The

ARRL
Antenna Book

Edited by
Gerald L. Hall, K1TD

Published by
THE AMERICAN RADIO RELAY LEAGUE, INC.
Newington, Connecticut
 Copyright © 1974 by

THE AMERICAN RADIO RELAY LEAGUE, INC.

Copyright secured under the Pan-American Convention

International Copyright secured

This work is Publication No. 15 of The Radio Amateur's

Library, published by the League. All rights reserved. No
part of this work may be reproduced in any form except by
written permission of the publisher. All rights of translation
are reserved. Printed in U.S.A.

Quedan reservados todos los derechos

Thirteenth Edition
Fifth Printing, 1980

Library of Congress Catalog Card Number: 55-8966

$5.00 in U.S.A.
$5.50 elsewhere
CONTENTS

Foreword.5
Chapter 1 Wave Propagation.7
Antenna Fundamentals .24
3 Transmission Lines.68
4 Multielement Directive Arrays .130
5 Long-Wire Antennas .165
6 Multiband Antennas .178
7 Antennas for 1.8 MHz.189
8 Antennas for 3.5 and 7 MHz.195
9 Antennas for 14, 21, and 28 MHz.202
10 HF Antennas for Restricted Space.211
11 VHF and UHF Antenna Systems .224
12 Antennas for Space Communications .250
13 Construction of Wire Antennas.264
14 Rotatable Antennas.277
15 Mobile Antennas.287
16 Specialized Antennas .293
17 Measurements .301
18 Finding Directions .323
Index.331
J
FOREWORD

Radio amateurs are people with diversified interests, ranging from

low frequencies to uhf and above, through radiotelephony, code
operation, and teletype to television, from the sociability of “rag
chewing” and the excitement of contests to handling traffic for the
public benefit. In all these activities there is one common element — the
antenna. It is fair to say that the ultimate success of the station is in
most cases determined more by the antenna than by any other single
category of equipment.
It is the purpose of this volume to assemble such of the available
information on antennas as may be useful to amateurs. The subject has
many ramifications, and consideration of the antenna alone is not
enough: To be able to choose an antenna intelligently one must know
something about how radio waves travel, because the energy radiated
must be properly directed if it is to do the most good; one must also
know something about transmission Unes, because power can be wasted
between the transmitter and antenna if the two are not properly
connected. Wave propagation and transmission lines therefore require
detailed treatment along with the antenna itself. There is no one “best”
antenna system for all purposes, but with the help of the information in
this book any amateur should be able to take the three steps that lead
to a successful installation — first, to determine the things the antenna
must do if it is to provide the best communication between two points;
second, to choose the type of antenna that best meets the re¬
quirements; and third, to select a suitable method of transferring power
from the transmitter to the antenna.
The book has three principal divisions. Chapters One through Five
deal with the principles of antennas and transmission lines, wave
propagation and its relationship to antenna design, and the performance
characteristics of directive antenna systems. Beginning with Six, there is
a series of chapters in which complete data are given on specific designs
for the various amateur bands, including those suitable for space
communications — satellites, moonbounce, meteor scatter, etc. The
remaining chapters deal with the highly important mechanical features
of construction and related subjects such as measurements and
determining geographical directions.
It is sincerely hoped that you will find the work helpful. If there are
some things left untreated that you wish to know more about, some
things you do not understand, we shall be grateful if you, the user of
the book, will give us your suggestions on how a future edition can be
made even more valuable to you.
Richard L. Baldwin, W1 RU
Newington, Connecticut General Manager, ARRL
 Chapter 1

Wave Propagation

Because radio communication is carried on by

means of electromagnetic waves traveling through
the earth’s atmosphere, it is desirable to know
something about the characteristics of waves and
the way in which their behavior is influenced by
the conditions they meet in their trip from the
radio transmitter to the receiver. While a detailed
knowledge of wave propagation is not at all
essential to the amateur who wants to put up an
effective antenna, a few facts must be understood
before the principles of antenna design can be
intelligently applied. An antenna may - and
usually does - radiate the power applied to it with
a high degree of efficiency, but if that power does
not travel to the desired receiving point but goes Fig. 1-1 — Representation of the magnetic and
somewhere else instead, the antenna is a failure. electric fields of a vertically polarized plane wave
Our purpose in this chapter, therefore, is to traveling along the ground. The arrows indicate
discuss those features of wave propagation that instantaneous directions of the fields for a wave
have some bearing on the design of an antenna traveling perpendicularly out of the page toward
system. In doing this we do not mean to imply that the reader. Reversal of the direction of one set of
lines reverses the direction of travel. There is no
there is nothing more that the wide-awake amateur change in direction when both sets are reversed.
will want or need to know about the subject. Such a dual reversal occurs, in fact, once each half
Effective radio communication results from a cycle.
combination of equipment, antenna system, and
operating skill, which must include that ability to since light also is an electromagnetic wave. In
anticipate “conditions” - a skill that is based on empty space this speed is 299,793,077 meters or
an understanding of the vagaries of wave propaga¬ 186,282.386 miles per second. In normal use for
tion. But the latter, fascinating though they are, calculations these figures may be rounded off to
are somewhat outside the scope of this book when 300,000,000 meters per second and 186,000 miles
they do not directly affect antenna design. per second, and these latter values will be used
throughout the remainder of this book. The path
of a ray from the source to any point on the
WAVE CHARACTERISTICS spherical surface always is a straight line — a radius
The reader who has some knowledge of elec¬ of the sphere.
tricity has already been introduced to the idea of It is obvious that in a remarkably short time a
electric and magnetic fields. A radio wave is a sphere growing outward from the center would be
special combination of both types of fields, with very large indeed. An observer on such a spherical
the energy divided equally between the two. If the surface would conclude, if he could “see” the wave
waves could originate at a point source in “free” in his vicinity, that it does not appear to be
space — empty space such as occurs, for all spherical at all but instead seems like a flat surface
practical purposes, in the interplanetary and inter¬ - just as the earth seems to human beings to be
stellar stretches of the universe — they would flat rather than spherical. A wave that is far enough
spread out in ever-growing spheres with the source from the source to appear flat is called a plane
as the center. The speed at which the spheres wave. The radio waves with which we deal in
expand would be the same as the speed of light, communication always meet this condition, at least

7
 8 Chapter 1

after they have traveled a short distance from the corresponding instants in each period will be
transmitting antenna. identical, and the fields caused by those identical
A typical representation of the lines of electric currents will also be identical. As the fields move
and magnetic force in a plane wave is given in Fig. outward they become more thinly spread over
1-1. The nature of wave propagation is such that larger and larger surfaces, so their amplitudes
the electric and magnetic lines always are mutually decrease with distance from the antenna. But they
perpendicular, as indicated in the drawing. The do not lose their identity with respect to the
plane containing the set of crossed lines represents instant of the period at which they were generated.
the wave front. The direction of wave travel always That is, the phase of the outwardly moving surface
is perpendicular to the wave front, but whether the remains constant. It follows, then, that at intervals
direction is “forward” or “backward” is deter¬ of 10 meters (in the example above) measured
mined by the relative directions of the electric and outward from the antenna the phase of the waves
magnetic forces. at any given instant is identical.
If the wave is traveling through anything other In this fact we have the means for defining
than empty space its speed is not 300,000,000 rather precisely both “wave front” and “wave¬
meters per second but is something less. Just how length.” The wave front is simply a surface in every
much less depends on the substance or medium part of which the wave is in the same phase. The
though which the wave is traveling. If the medium wavelength is simply the distance between two
is air instead of empty space, the reduction in wave fronts having identical phase at any given
speed is so small that it can be ignored in most instant. In the example, the wavelength is 10
calculations. In solid insulating materials the speed meters because the distance between two wave
is generally much slower; for example, in distilled fronts having the same phase is, as we found, 10
water (which is a good insulator) the waves travel meters. This distance, incidentally, always must be
only one-ninth as fast as they do in space. In good measured perpendicular to the wave fronts; in
conductors such as metals the speed is so low that other words, along the same line that represents
the opposing fields set up by currents induced in the direction in which the wave is traveling.
the conductor by the wave itself occupy practically Measurements made along any other line between
the same space as the original wave and thus almost the two wave fronts would lead to the erroneous
cancel it out. This is the reason for the skin effect conclusion that the wave is longer than it really is.
in conductors at high frequencies and also the Expressed in a formula, the length of a wave is
reason why thin metal enclosures form good
shields for electrical circuits at radio frequencies.

Phase and Wavelength

Because the speed at which radio waves travel is where X = wavelength
so great, we are likely to fall into the habit of V = velocity of wave
ignoring the time that elapses between the instant ! = frequency of current causing the wave
at which a wave leaves the transmitting antenna
and the instant at which it arrives at the receiving
antenna. It is true that it takes only one seventh of The wavelength will be in the same length units as
a second for a wave to travel around the world, and the velocity when the frequency is expressed in the
from a communication standpoint that is hardly same time units as the velocity. For waves traveling
worth worrying about. But there is another consid¬ in free space (and nearly enough for waves travel¬
eration that makes this factor of time extremely ing through air) the wavelength is
important.
The wave is brought into existence because an
alternating current flowing in a conductor (which X (meters) =
j(MHz)
is usually an antenna) sets up the necessary electric
and magnetic fields. The alternating currents used We shall continually be encountering this idea
in radio work may have frequencies anywhere from of phase in succeeding chapters in this book,
a few tens of thousands to several billion hertz. because it is fundamental to the operation of
Suppose that the frequency is 30 megahertz (MHz) antenna systems. It is essential, therefore, to have a
— that is, 30,000,000 hertz (Hz). One of the cycles clear understanding of what it means if antenna
or periods will be completed in 1/30,000,000 behavior is to be appriciated. Basically, “phase”
second, and since the wave is traveling at a’spee'd of means “time,” but when something goes through
approximately 300,000,000 meters per second it periodic variations with time in the way that an
will have moved only 10 meters during the time alternating current does, corresponding instants in
that the current is going through one complete succeeding periods are said to have the same phase
period of alternation. To put it another way, the even though the actual time difference is the
electromagnetic field 10 meters away from the duration of one period. In using the word phase in
antenna is caused by the current that was flowing this fashion we are inherently using the peroid as
in the antenna one period earlier in time; the field the unit of time measurement. Four o’clock
20 meters away is caused by the current that was yesterday afternoon corresponds to four o’clock
flowing two periods earlier, and so on. this afternoon in much the same way that an
Now if each period of current is simply a instant in an ac cycle corresponds to the identical
repetition of the one preceding it, the currents at instant in the preceding period.
Wave Propagation 9

earth. A wave “with its feet on the ground” as in

Fig. 1-1 is, as a matter of fact, usually vertically
polarized. This is because the ground acts as a
rather good conductor, particularly at frequencies
below about 10 MHz, and it is one of the laws of
nur
electromagnetic action that electric lines touching
«n««— the surface of a good conductor must do so
perpendicularly. Over partially conducting ground
Fig. 1-2 — The instantaneous amplitude of both there may be a forward tilt to the wave front; this
fields (electric and magnetic) varies sinusoidally tilt in the electric lines is greater as the energy loss
with time as shown in this graph. Since the fields in the ground becomes greater.
travel at constant velocity, the graph also rep¬ Waves traveling in contact with the surface of
resents the instantaneous distribution of field the earth are of little usefulness in amateur
intensity along the wave path. The distance communication because as the frequency is raised
between two points of equal phase, such as A-B the distance over which such a “surface” wave will
and B-C, is the length of the wave. travel without excessive loss of energy or attenua¬
tion becomes smaller and smaller. The surface wave
is of most utility at low frequencies and through
In Fig. 1-2 points A. B and C are all in the same the standard broadcast band. At high frequencies
phase because they are corresponding instants in the wave reaching the receiving antenna ordinarily
the same period.This is a cOnventionaldrawing of a has not had much contact with the earth and its
sine-wave alternating current with time progressing polarization is not necessarily vertical. If the
to the right. It also represents an instantaneous electric lines of force are horizontal, the wave is
“snapshot” of the distribution of intensity of the said to be horizontally polarized. However, the
traveling fields if distance is substituted for time in polarization can be anything between horizontal
the horizontal axis. In that case the distance and vertical. In many cases, the polarization is not
between A and B or between B and C represents fixed but continually rotates. When this occurs the
one wavelength. This shows that the field-intensity wave is said to be elliptically polarized.
distribution follows the sine curve, both as to
amplitude and polarity, to correspond exactly to Attenuation
the time variations in the current that produced In free space the field intensity of the wave
the fields. It must be remembered that this is an decreases directly with the distance from the
instantaneous picture; the actual wave travels along source. That is, if the field strength one mile from
just as a wave in water does. To an observer at any the source has a value, let us say, of 100 microvolts
fixed point along the wave path, the field intensity per meter, the field strength at two miles will be 50
goes through time variations corresponding to the microvolts per meter, at 100 miles will be 1
time variations of the current that initiated the microvolt per meter, and so on. This decrease in
wave. field strength is caused by the fact that the energy
in the wave has to spread out over larger and larger
Field Intensity spheres as the distance from the source is in¬
The strength of a wave is measured in terms of creased.
In actual communication by radio the attenua¬
the voltage between two points lying on an electric
tion of the wave may be much greater than this
line of force in the plane of the wave front. The
“inverse-distance” law would indicate. For one
unit of length is the meter, and since the voltage in
thing, the wave is not traveling in empty space. For
a wave usually is quite low, the measurement is
another, the receiving antenna seldom is situated so
made in microvolts per meter. The voltage so
measured goes through time variations just like that there is a clear “line of sight” between it and
the transmitting antenna. Since the earth is spheri¬
those of the original current that caused the wave,
cal and the waves do not penetrate its surface to
and so is measured like any other ac voltage - that
is, in terms of the effective value or, sometimes, any considerable extent, communication has to be
by some means that will bend the waves around
the peak value.
There are few, if any, occasions in amateur the curvature of the earth. These means exist, but
work where a measurement of actual field strength they usually involve additional energy losses that
is necessary. This is fortunate, because the equip¬ increase the attenuation of the wave with distance.
ment required is elaborate. It is comparatively
easy, however, to make measurements of relative Reflection, Refraction and Diffraction
field strength, and thus determine whether an It has been mentioned that radio waves and
adjustment to the antenna system has resulted in light waves are the same type of wave; the only
an improvement or not. difference is in the scale of wavelength. We are all
familiar with the reflection of light; radio waves are
Polarization reflected in much the same way. Frequently,
A wave such as is shown in Fig. 1-1 is said to be however, the reflecting surfaces are small (in terms
polarized in the direction of the electric lines of of wavelength) compared with the surfaces from
force. In this drawing the polarization is vertical which we see light waves reflected. An object the
because the electric lines are perpendicular to the size of an automobile, for instance, will not reflect
10
Chapter 1

much of the energy in an 80-meter wave. On the Most of the optical examples of refraction are
other hand, it may be a very good reflector of based on two homogeneous substances having a
waves only a meter or two in length. The thickness very definite common boundary, as between air
of the object is of some importance because the and glass. In that case the rays travel in straight
waves penetrate it to an extent depending on its lines inside either medium and the bending takes
characteristics. In a material of given conductivity, place at the common surface. In radio transmission
for example, longer waves will penetrate farther it is frequently the case that the boundary between
than shorter ones and so require a greater thickness the two areas of differing dielectric constant is not
for good reflection. Thin metal is a good reflector at all sharp; the dielectric constant simply changes
even at quite long wavelengths, but in poorer gradually over quite a distance along the wave
conductors such as the earth — which certainly path. This causes the wave bending also to be
meets the requirement of having a surface that is gradual, and the wave path becomes curved.
large compared with any radio wavelength - the A somewhat less familiar optical phenomenom
longer wavelengths may penetrate quite a few feet. that has its radio counterpart is diffraction. To the
Reflection may also take place from any eye, the shadows cast by a pin-point source of light
surface that represents a change in the dielectric appear to be quite sharp. However, close examina¬
constant of the medium in which the wave is tion shows that light bends around the edge of an
moving. A familiar example in optics is the object to some extent, depending on the thickness
reflection of light from the surface of a pane of of the edge. This effect becomes greater as the
glass that is itself quite transparent to light waves. wavelength is increased, and can be of some
When viewed from certain angles, it is practically importance at radio frequencies. For example, with
impossible to see through the pane of glass because waves traveling in a straight line one would expect
of the reflected light. that no signal could be heard behind a hill, but the
Another phenomenon that has a rather familiar bending caused by diffraction does produce a
counterpart in optics is refraction, or the bending signal in the “shadow area.” At high radio frequen¬
that takes place when the wave enters (at an angle) cies the diffracted signal is weak compared with
a medium having a different dielectric constant the direct ray, and frequently is masked by
than the medium it has just left. This bending is stronger signals reaching the same spot by other
caused by the fact that the wave travels at a means such as reflection or refraction in the
different speed when the dielectric constant is atmosphere.
changed. The part of the wave that enters the new Both reflection and refraction can take place in
medium first is either slowed down or speeded up, various parts of the atmosphere, and the mecha¬
depending on the relative dielectric constants, and nisms by which they occur are likewise varied. The
so tends to get ahead of or fall behind the sections result is that radio waves frequently are "scat¬
of the wave that enter later. The effect is to change tered,” just as light is scattered in the atmosphere.
the direction in which the wave is moving. The Such scattering accounts for the reception of
classic example in optics is the apparent sharp bend signals under conditions when they would not be
in a stick held partly in and partly out of a body of expected from the simplified pictures of wave
water at an angle. travel now to be discussed.

THE GROUND WAVE

Waves travel close to the earth in several wave that provides reception up to distances of
different ways, some of which involve relatively 160 kilometers or more in the standard broadcast
little contact with the ground itself. The selection band during the daytime. The attenuation of this
of proper nomenclature therefore becomes some¬
what confusing, but more or less by common
consent the term ground wave is applied to waves
that stay close to the earth and do not reach the
receiving point by reflection or refraction from the
much higher region of the atmosphere known as
the ionosphere. The ground wave therefore can
be a wave traveling in actual contact with the
ground, such as the wave pictured in Fig. 1-1 ; or it
can be a wave that goes directly from the transmit¬
ting antenna to the receiving antenna when the two
antennas are high enough so that they can “see”
each other. It can also be a wave that is refracted
or reflected in the atmosphere near the earth (the
troposphere).

THE SURFACE WAVE

A wave that travels in contact with the earth’s Fig. 1-3 — Typical hf ground-wave range as a
surface is called a surface wave. It is the type of function of frequency.
Wave Propagation 11

type of wave is rather high, so the intensity dies off

rapidly with distance from the transmitter. The
attenuation also increases rapidly with frequency,
with the result that the surface wave is of little
value in amateur communications, with the possi¬
ble exception of the 1.8- and 3.5-MHz bands. Fig.
1-3 shows typical ground-wave coverage as a
function of frequency throughout the hf spectrum. Fig. 1-4 — The ray traveling directly from the
transmitting antenna to the receiving antenna
As explained earlier, the surface wave must be combines with a ray reflected from the ground to
essentially vertically polarized. The transmitting form the space wave.
and receiving antennas therefore must generate and
receive vertically polarized waves, if the surface
wave is to be utilized to advantage. In general becomes increasingly important. It is the
terms, this means that both antennas must be dominating factor in ground-wave communication
vertical. at vhf and uhf.
The space-wave picture presented is a simplified
one and, as usual, there are practical complications
THE SPACE WAVE that modify it. There is some loss of energy when
The conditions that exist when the transmitting the ray strikes the ground, so the reflected ray does
antenna and receiving antenna are within Une of not arrive at the receiving antenna with the same
sight of each other are showm in Fig. 1-4. One ray intensity as the direct ray. Because of the ground
travels directly from the transmitter to the receiver loss the phase of the reflected ray is not shifted
and consequently is attenuated in about the same exactly 180 degrees. For both these reasons the
way as a wave in free space. However, the wave two waves never cancel completely at the receiving
from the transmitting antenna also strikes the antenna. Also, at frequencies in the uhf region it is
ground between the two antennas, and the ray that possible to form the wave into a beam, much like
does so at the proper angle to reach the receiving the light beam from a flashlight. Such a beam puts
antenna (the angle of incidence being equal to the most of the energy into the direct ray and reduces
angle of reflection, as in optics) combines with the the amount that can strike the ground, particularly
direct ray to produce the actual signal at the when the tranmitting and receiving antennas are
receiving antenna. both at high elevations. Thus the effect of the
In most practical cases where the communica¬ ground-reflected ray is minimized.
tion is between two stations on the ground (as Strictly speaking, the description above applies
contrasted with communication between the only to a horizontally polarized wave and perfectly
ground and an airplane, or between two airplanes) conducting earth. Practically, the polarization does
the angle at which the ground-reflected ray strikes not make much difference because the earth is
the earth and is reflected will be very small. That neither a perfect conductor nor a perfect dielectric.
is, the ray strikes the earth at almost grazing The overall result is that at frequencies below, say,
incidence. Now it happens that such a reflection 20 MHz, the space wave is inconsequential. But at
reverses the phase of the wave, so if the distance vhf it is readily possible to transmit to the horizon
traveled by the direct wave and the distance by means of the space wave.
traveled by the ground-reflected ray were exactly
the same, the two rays would arrive out of phase “Line-of-Sight” Propagation
and would cancel each other. Actually, the From inspection of Fig. 1-4 it appears that use
ground-reflected ray has to travel a little farther, of the space wave for communication between two
and so the phase difference between the two rays points depends on having a line of sight between
depends on the difference in path length as the two locations. This is not quite literally true.
measured in terms of wavelength. If the difference The structure of the atmosphere near the earth is
in the length of the two paths is 3 meters, for such that under “normal” conditions (a theoretical
example, the phase difference from this cause will normal, rather than an actual one; in many parts of
be only 3 degrees if the wave is 360 meters long. the world, at least, the “normal” is an average
This is only a negligible shift in phase from the 180 which is statistically useful but seldom represents
degrees caused by the reflection, and so the signal the actual condition of the atmosphere) the waves
strength would be small. On the other hand, if the are bent into a curved path that keeps them nearer
wavelength is 6 meters, the phase shift caused by
to the earth than true straight-line travel would.
the same difference in path length would be 180 This effect can be approximated by assuming that
degrees - enough to overcome completely the
the waves travel in straight lines but that the
180-degree reversal caused by the reflection, so the
earth’s radius is increased in dimension by one
two rays would add at the receiving antenna. In
third. On this assumption, the distance from the
short, the space wave is a negligible factor in
transmitting antenna to the horizon is given by the
communication at low frequencies, because the
difference in the distance traveled by the two rays following formula:
is always very small, when measured in terms of D(mi) = 1.415V/7 (ft)
wavelength. The space wave therefore is canceled
out at such frequencies. But as the frequency is or
raised (wavelength shortened) the space wave D (km) = 4.124x/^ (m)
 12 Chapter 1

In addition to the “normal” refraction or

bending, the waves also are diffracted around the
curvature of the earth, so that the actual distance
that can be covered does exceed the line-of-sight
distance. However, under ordinary conditions the
Fig. 1-5 - The distance, D, to the horizon from an amount of diffraction at vhf and uhf where the
antenna of height H is given by the formulas in the space wave is of chief importance is rather small,
text. The maximum line-of-sight distance between and the signal strength drops off very rapidly in a
two elevated antennas is equal to the sum of their short distance beyond the earth’s “shadow.”
distances to the horizon, as indicated in the To make maximum use of the ordinary space
drawing.
wave discussed here it is necessary that the antenna
be as high as possible above the surrounding
country. A hill that juts above the adjacent terrain
where H is the height of the transmitting antenna, is usually an excellent location. However, the peak
as shown in Fig. 1-5. The formula assumes that the of a hill is not necessarily the best spot, particular¬
earth is perfectly smooth out to the horizon; of ly if it is of the nature of a plateau. Arriving waves
course, any obstructions that rise along any given may have to be diffracted over the brow of the hill
path must be taken into consideration. The point to reach the antenna unless the latter is placed on a
at the horizon is assumed to be on the ground. If high pole or tower; in other words, the brow of the
the receiving antenna also is elevated, The maxi¬ hill may shield the antenna from waves arriving
mum line-of-sight distance between the two anten¬ from a desired direction. Also, it is advantageous to
nas is equal to D + Di ; that is, the sum of the have the ground drop off fairly sharply in front of
distance to the horizon from the transmitting the antenna, as this frequently prevents the
antenna and the distance to the horizon from the ground-reflected ray from approaching at such a
receiving antenna. The distances are given in graph flat angle as it would over level ground. Generally
form in Fig. 1-6. Two stations on a flat plain, one speaking, a location just below the peak of a hill is
having an antenna on a tower 60 feet high and the the optimum one for transmitting and receiving in
other having an antenna supported 40 feet in the a desired direction, as indicated in Fig. 1-7.
air, could be separated approximately 20 miles for
line-of-sight communication.

Fig. 1-7 — Propagation conditions are generally

best when the antenna is located slightly below the
top of a hill on the side that faces the distant
station. Communication is poor when there is a
sharp rise immediately in front of the antenna in
the direction of communication.

Since the space wave goes essentially in a

straight line from the transmitter to the receiver,
the antenna used for radiating it should concen¬
trate the energy toward the horizon. That is, the
antenna should be a “low-angle” radiator, because
energy radiated at angles above the horizon obvi¬
ously will pass over the receiving antenna. Similar¬
ly, the receiving antenna should be most responsive
to waves that arrive horizontally.
In general, the polarization of a space wave
remains constant during its travels. Therefore, the
receiving antenna should be designed to give
maximum response to the polarization set up at
the transmitting antenna. For vhf work both
horizontal and vertical polarization are used, the
former being more generally preferred. The princi¬
pal reason for this preference is that the chief
source of radio noise at vhf - that generated by
the spark in the ignition systems of automobiles -
Fig. 1-6 — Distance to the horizon from an antenna is predominantly vertically polarized. Thus hori¬
of given height. The solid curve includes the effect zontally polarized antennas tend to discriminate
of atmospheric refraction. The optical line-of-sight against such noise and thereby improve the signal-
distance is given by the broken curve. to-noise ratio.
 13
Wave Propagation

inversions are present practically continuously at

PROPAGATION IN THE TROPOSPHERE
heights of the order of a few hundred feet or less.
Weather conditions in the atmosphere at The boundary of the inversion is usually well
heights from a few thousand feet to a mile or two enough defined so that waves traveling horizontally
at times are responsible for bending waves down¬ are “trapped” by the refracting layer of air and
ward. This tropospheric refraction makes commu¬ continually bent back toward the earth. The air
nication possible over far greater distances than can layer and the earth form the upper and lower walls
be covered with the ordinary space wave. The of a “duct” in which waves are guided in much the
amount of the bending increases with frequency, same fashion as in a metallic wave guide. The waves
so tropospheric communication improves as the therefore follow the curvature of the earth for
frequency is raised. The bending is relatively distances far beyond the optical horizon of the
inconsequential at frequencies below 28 MHz, but transmitter (sometimes for hundreds of miles).
provides interesting communication possibilities at Because the height of an atmospheric duct is
50 MHz and above. relatively small, only waves smaller than a certain
Refraction in the troposphere takes place when limit will be trapped. If the refracting layer is only
masses of air become stratified into regions having a few feet above the surface the lowest usable
differing dielectric constants. If the boundary frequency may be as high as a few thousand
between the two masses of air is sharply defined, megahertz, so ultrahigh or superhigh frequencies
reflection as well as refraction may take place for must be used. Under some conditions, however,
waves striking the boundary at grazing angles. The the height and dielectric characteristics of the layer
most common cause of tropospheric refraction is may be such that waves in the medium vhf region
the temperature inversion. Normally, the tempera¬ will be transmitted. The line of distinction, if any,
ture of the lower atmosphere decreases at a between ducting and ordinary tropospheric propa¬
constant rate of approximately 3° F per 1000 feet gation is hard to draw in such a case.
of height. When this rate is decreased for any A feature of duct transmission is that the
reason, a temperature inversion is said to exist and antennas, both transmitting and receiving, must be
greater-than-normal wave bending takes place. inside the duct if communication is to be estab¬
Some of the types of temperature inversion are the lished. If the duct extends only a few feet above
dynamic inversion, resulting when a warm air mass the earth and the transmitting antenna is on a
overruns a colder mass; the subsidence inversion, tower or promontory above the duct, no signals
caused by the sinking of an air mass heated by will be heard at the receiving point. Likewise, a
compression; the nocturnal inversion, brought receiving antenna situated above the duct will not
about by the rapid cooling of surface air after pick up energy trapped nearer the earth.
sunset; and the cloud-layer inversion, caused by the Atmospheric ducts also are formed between
heating of air above a cloud layer by reflection of two layers of air having suitable characteristics. If
the sun’s rays from the upper surface of the clouds. the lower layer refracts the waves upward while the
Sharp transitions in the water-vapor content of the upper layer refracts them downward, waves will be
atmosphere may also bring about refraction and trapped between the two layers and again can
reflection of vhf waves. travel for great distances. In such a case antennas
Because the atmospheric conditions that pro¬ either below or above the duct will be ineffective.
duce tropospheric refraction are seldom stable over Ducts of this type have been observed from
any considerable period of time, the strength of airplanes, where good signals will be received with
the received signal usually varies or “fades” over a the plane at the optimum height but the signal
wide range. Hourly and seasonal variations are strength drops off rapidly ar either higher or lower
observed. Best conditions often occur in the altitudes.
evening and just before sunrise, and conditions are Much remains to be learned about the extent of
generally poorest at midday when the atmosphere duct transmission at amateur frequencies. There
is relatively stable. Tropospheric refraction is gen¬ appears to be no significant difference in the signal
erally greatest in the early summer and early fall. It strength with either horizontal or vertical polariza¬
is also more pronounced along the seacoasts. tion.
The tropospheric wave maintains essentially the
same polarization throughout its travel, so the
transmitting and receiving antennas should have Other Modes
the same type of polarization. Since waves that Although the transmission modes just discussed
enter the refracting region at anything other than are the most common ones at vhf, they are by no
practically grazing incidence are not bent enough means the only methods by which communication
to be useful for communication, the transmitting can be carried on at such frequencies. New modes
antenna should be designed for maximum radiation are continually being discovered as better antennas
horizontally. The receiving antenna likewise should and equipment are developed. However, it is
be a low-angle affair if the received signal is to be invariably true that best results in all types of vhf
most efficiently utilized. propagation are obtained when the transmitting
antenna system concentrates as much as possible of
Atmospheric Ducts its energy in the very low vertical angles, and when
In some parts of the world, particularly in the the receiving antenna likewise is most responsive to
tropics and over large bodies of water, temperature signals arriving horizontally.
 14
Chapter 1

THE SKY WAVE

At frequencies below 30 MHz practically all Layer Characteristics

amateur communication except for local work over
distances of a few miles (see Fig. 1-3) is carried on The ionized layers or regions are designated by
by means of the sky wave. This is the wave that, on letters. The lowest ionized layer known, at a height
leaving the transmitting antenna, would travel on of about 60 to 92 kilometers (37-57 miles), is
out into empty space if it were not for the fact called the D region. Because it is in a relatively
that under certain conditions it can be sufficiently dense part of the atmosphere the atoms broken up
reflected or refracted, high up in the earth’s into ions by sunlight quickly recombine, so the
atmosphere, to reach the earth again at distances amount of ionization depends directly on the
varying from zero to about 4000 kilometers (2500 amount of sunlight. Thus ZT-region ionization is
miles) from the transmitter. By successive re¬ maximum at local noon and disappears at sun¬
flections at the earth’s surface and in the upper down. When electrons in the D region are set into
atmosphere, communication can be established motion by a passing wave, the collisions between
over distances of thousands and thousands of particles are so frequent (because of the rather high
miles. On occasion, radio signals have been known air density) that a substantial proportion of the
to travel completely around the earth. wave energy may be used up as heat. The prob¬
ability of collisions depends on the distance an
THE IONOSPHERE electron can travel under the influence of the wave.
This distance depends on the frequency of the
The region of the earth’s upper atmosphere in wave, because during a long cycle (low frequency)
which the radio waves are bent back to the earth’s the electron has time to move farther, before the
surface is called the ionosphere. This is an area direction of the field reverses and sends it back
where the air pressure is so low that “free” again, than it does in a short cycle (high fre¬
electrons and ions can move for a long time quency). If the frequency is low enough the
without getting close enough to each other to be collisions between particles will be so frequent that
attracted together and thus recombine into a practically all energy in the wave will be absorbed
neutral atom. The ionosphere is sometimes referred in the D region. This usually happens at fre¬
to as the rarefied atmosphere.
A radio wave entering a region in which there quencies in the 3.5-4.0 MHz amateur band at the
are many free electrons will be affected in much time of maximum D-region ionization, particularly
the same way as one entering a region of differing for waves that enter the layer at the lower vertical
dielectric constant; that is, its direction of travel angles and thus have to travel a relatively long
will be altered. A simple explanation of the rather distance through it. At times of sunspot maxima
complex mechanism involved is given earlier in this even waves entering the layer directly upward will
chapter under the subheading of Reflection, Re¬ be almost wholly absorbed, in this frequency band,
fraction and Diffraction. around the middle of the day. The absorption is
Ultraviolet radiation from the sun is the pri¬ less in the 7-MHz band and is quite small at 14
mary cause of ionization in the upper atmosphere. MHz and higher frequencies. The D region is
The amount of ionization does not change uni¬ relatively ineffective in bending high-frequency
formly with height above the earth, as might be waves back to earth, and so plays no significant
part in amateur long-distance communication ex¬
expected at first thought. Instead, it is found that
cept as an absorber of energy. It is the principal
there ye relatively dense regions (or layers) of
reason why daytime communication on the lower
ionization, quite thick vertically, at rather
frequencies (3.5 and 7 MHz) is confined to
well-defined heights. However, the interregional relatively short distances.
troughs, as verified through rocket soundings dur¬
The lowest layer that affords long-distance
ing the International Geophysical Year (IGY), communication is in the E region, which appears
1958, are more shallow than had been thought for 100 to 115 kilometers (62-71 miles) above the
many years. The ionization level is not uniform earth. It is a region of fairly high atmospheric
within a specific region, but rather tapers off density and consequently the ionization varies with
gradually above and below the peak ionization the height of the sun (angle above the horizon).
level at the center of the region. Because the Ultraviolet radiation is not alone in affecting the
amount of ultraviolet and other types of radiation ionization of this region. A certain amount of
received from the sun at any given time over any ionization is also caused by solar photons, X rays,
given point above the earth’s surface will vary, the and meteoric activity within the region. The
height and ionization intensity of the various ionization level drops rapidly after sundown, when
regions will likewise vary. The condition, then, of
ions and electrons recombine in the absence of
the ionosphere over any given point will vary with
sunlight, and reaches a minimum level of ionization
the time of day (the received radiation from the
after midnight local time. It again increases
sun is maximum at solar noon), the season of the
rapidly at sunrise and reaches a maximum at about
year (because of the tilt of the earth’s axis and the
noon local time. As in the case of the D region, the
varying seasonal distance between earth and sun),
E region (or layer) absorbs energy from low-
and the so-called 11-year sunspot cycle (refer to’
later section on Ionosphere Variations). frequency waves during the time of maximum
ionization.
Wave Propagation 15

Fig. 1-8 — Behavior of waves

on encountering the
ionosphere. Waves entering the
ionized region at angles higher
than the critical angle are not
bent enough to be returned to
earth. Waves entering at angles
below the critical angle reach
the earth at increasingly
greater distances as the angle
approaches the horizontal.

Most long-distance communications result from Fig. 1-8 shows a condition that is frequently
ionization which takes place in the F region. Its typical of the way waves are bent in a single layer.
principal area is the F2 layer which varies con¬ (When several layers are involved, the paths are
siderably in ionization level and in height. It may naturally more complex, since the layers have
be anywhere from 210 to 420 kilometers (130-261 differing characteristics.) In this case the layer is
miles) above the earth, depending on season, capable of refracting waves that enter it at low
latitude, time of day, and the portion of the angles. However, as the angle at which the ray
sunspot cycle prevalent at a given time. At these strikes the layer is increased, a critical angle is
heights the atmosphere is very thin, and so the ions reached at which the ray just manages to be bent
and electrons are slow to recombine. Because of back to earth. Rays entering at still greater angles
this, the level of ionization is not so responsive to are not bent enough and pass through the layer
the height of the sun; it reaches a maximum into empty space. Since such rays are useless for
shortly after noon local time, but tapers off quite communication, it is obvious that energy radiated
gradually thereafter. The amount of ionization at angles above the critical angle is wasted.
continues to remain at a fairly high level through¬ Note also that the point at which a ray reaches
out the night, reaching a minimum just before the earth on its return journey from the ionosphere
sunrise. At sunrise it increases rapidly and attains depends on the angle at which it left the trans¬
the daytime level in the course of an hour or two. mitting antenna. The larger the angle with the
During the day the F region sometimes splits surface of the earth the shorter the distance from
into two layers. The lower and weaker one, the transmitter to the point at which the returning
occurring at a height of 160 kilometers (99 miles), ray arrives.
is designated the Ft layer. This layer plays only a
minor role in long-distance communications, acting Skip Distance
more like the E than the F2 layer in terms of daily When the critical angle is less than 90 degrees
life cycle and communications effectiveness. At the highest angle wave that can be bent back to
night the Ft layer disappears and the F2 layer earth will return at an appreciable distance from
height drops somewhat. the transmitter. For some distance, then, de¬
pending on the critical angle, there is a region
Refraction in the Ionosphere about the transmitter where the sky-wave signal
The amount by which the path of a wave is will not be heard. This “silent” region, extending
bent in an ionized layer depends on the intensity from the limit of the useful ground wave to the
of ionization and the wavelength. The greater the distant point where the sky-wave signal can first be
ionization, the more the bending at any given heard, is called the skip zone, because all signals
frequency. Or, to put it another way, for a given skip over it. The skip zone is indicated by the skip
degree of ionization the bending will be greater as distance in Fig. 1-8.
the frequency of the wave is lowered - in other The skip distance - the distance from the
words, as its wavelength is increased. transmitter to the point where the sky-wave signal
Two extremes thus become possible. If the is first heard — depends on the critical angle and
ionization is intense enough and the frequency is the layer height. The lower the critical angle the
low enough, a wave entering the ionized region farther the skip distance extends. Since higher
perpendicularly will be turned back to earth. But frequencies are, in general, bent less than lower
as the frequency is raised or the ionization is frequencies, the skip distance is greater the higher
decreased, a condition will eventually be reached the frequency. For a given critical angle, it is also
where the bending will not be sufficient to return greater the greater the height of the layer in which
the wave to earth, even though the wave leaves the the bending takes place. Thus for the same critical
transmitting antenna at the lowest possible angle angle, the skip distance with F2 -layer bending will
and thus requires the least bending in the iono¬ be greater than for waves returned to earth from
sphere. A typical “in-between” condition is illus¬ the E layer, because the F2 layer is higher.
trated in Fig. 1-8, a simplified illustration of the When waves at any and all angles are returned
paths taken by high-frequency waves and con¬ to earth from the layer, there is, of course, no skip
sidering only the effect of a single layer. zone. In such instances the sky wave frequently is
 16
Chapter 1

stronger than the ground wave, even as close as a come back. The highest frequency that is returned
few miles from the transmitter location. This is to earth is known as the critical frequency. As the
because the wave is attenuated less in its travel up frequency is further increased beyond the critical
to the layer and back again than it is in going a few frequency, the wave must enter the ionosphere at
miles over the ground, surprising as it may seem. progressively smaller angles in order for it to be
Single- and Multihop Propagation bent back to earth. By using very low angles,
long-distance transmission via the F2 layer is
Fig. 1-8 also shows two of the modes by which possible at frequencies up to about 3.5 times the
the signal can reach a distant receiving point. In critical frequency. Thus, the critical frequency isa
one case the wave is bent in the layer at a point measure of the reflecting ability of the ionosphere.
about midway between the transmitter and the Since the refracted wave acts as though it were
receiving point, B. The wave thus makes the trip in reflected from a mirror at the virtual height, it is
one “hop.” However, that is not the only possi¬ customary to use the terms “reflection” and
bility. A ray that is reflected midway between the ‘refraction” almost interchangeably in connection
transmitter and point A (which in turn is midway with ionospheric propagation. In most cases the
between the transmitter and B) will be reflected actual process is refraction. However, it is possible
when it strikes the earth at A and will go up to the for true reflection to occur if the boundary of the
layer again. Here it is once more reflected, re¬ layer is sharply defined and the wave strikes it at a
turning to earth, finally, at B. This is “two-hop” small enough angle.
transmission. More than two hops are readily Virtual heights, of course, depend on the height
possible. of the ionized region. The critical frequencies vary
Multihop propagation over long distances tends with the intensity of ionization in the layers, being
to become more complex than the simple geo¬ greater when the ionization increases. Since the
metrical picture given here would indicate, partly ionization is greatest at the peak of the sunspot
because different ionospheric conditions usually cycle, critical frequencies are highest in both the E
exist at each point of reflection. Observation and F2 layers during that period. Conversely, they
indicates, however, that the hops are well defined are lowest during a sunspot minimum. The /Mayer
over distances up to several thousand miles, and critical frequency ranges from about 1 to 4
antennas for commercial point-to-point circuits megahertz depending on the period in the sunspot
usually are designed to radiate and receive best at cycle and the time of day. The F2 critical
the vertical angles associated with the number of frequency varies with the time of day, the season,
hops between the transmitting and receiving and the sunspot cycle, ranging from a low of
points. The smallest possible number of hops is perhaps 2 to 3 MHz at night in a sunspot minimum
best, in general, since each additional hop intro¬ to a high of 12 or 13 MHz in daytime during a
duces additional attentuation because of losses at sunspot maximum. Whenever the critical frequency
the reflection points. Even more important, how¬ is above an amateur band, it is possible to
ever, is the effect of the ionosphere itself as communicate on that band over all distances from
described later. Ionospheric absorption may be so zero to the maximum that absorption will permit.
much less, at a given operating frequency, with say
three or four hops instead of two over a given path
that the received signal will be much stronger in Maximum Usable Frequency
spite of the additional reflection loss. Of more interest, from a practical standpoint,
Virtual Height and Critical Frequencies than the critical frequency is the frequency range
over which communication can be carried on via
By using a frequency low enough so that waves one or the other of the two reflecting layers. In
entering the ionosphere at the maximum angle of particular, it is useful to know the maximum
90 degrees (i.e., waves going vertically from the usable frequency (abbreviated muf) for a particular
transmitting antenna to the ionosphere) are re¬
turned to earth, it is possible to measure the height
of the ionosphere. This is done by measuring the
time taken by the wave to go up and back.
Knowing the time and velocity of propagation, the
distance can be readily calculated. The distance so
found is the virtual height, or the height from
which a pure reflection would give the same effect
as the refraction that actually takes place. The
method is illustrated in Fig. 1-9. Because a certain
amount of time is required for the wave to make
the turn at the top of its travel, the virtual height is
somewhat higher than the actual height, as the
illustration shows.
If the transmitting frequency is gradually in¬ Fig. 1-9 The “virtual" height of the refracting
creased while height measurements of this type are layer is measured by sending a wave vertically to
being made, eventually a frequency range will be the layer and measuring the time it takes for it to
encountered where the virtual height seems to come back to the receiver. The actual height is
increase rapidly, until finally the wave does not somewhat less because of the time required for the
wave to ' turn around" in the ionized region.
 17
Wave Propagation

distance at the time of day at which communi¬

cation is desired. It is always advantageous to use
the highest possible frequency because the ab¬
sorption is less the higher the frequency. Therefore
the muf always gives the greatest signal strength at
the receiving point for a given transmitting power.
The distance to be considered in determining the
muf is the length of one hop in multihop trans¬
mission.
The muf depends upon the critical frequency
and thus is subject to seasonal variations as well as
variations throughout the day. To employ the muf
for very long distances with the smallest number of
hops requires that the antenna system radiate well
at very low vertical angles.

Lowest Usable Frequency

As the frequency is decreased below the muf,

the signal strength also decreases because of greater

LIMITS OF F-2 REGION

absorption. Eventually, as the frequency continues
to be lowered, the signal will disappear in the
background noise that is always present. Thus
there is a low-frequency limit, under a given set of
ionospheric conditions, as well as a high-frequency
limit to the range of frequencies that can be used
for a given distance. The lowest usable frequency
(abbreviated luf) depends considerably on the
transmitter power available, since high power will SINGLE HOP DISTANCE
“push” the signal through the noise where low
power would fail. But when the frequency in use is Fig. 1-10 — Distance plotted against wave angle
near the muf, even low-power signals often will (one-hop transmission) for the nominal range of
give surprising signal strength at long distances. virtual heights for the E and F2 layers, and for the
In commercial communications it is considered layer.
good practice to operate on a frequency which is
about 15 percent below the muf. This allows for The maximum distance that can be covered by
day-to-day variations in the ionosphere. This some- a single-hop transmission is approximately 2000
what lower frequency is known as the optimum kilometers (1250 miles) when reflection is from
working frequency, abbreviated owf. The the E layer, and approximately 4000 kilometers
international abbreviation, fot, is generally used, (2500 miles) when reflected from the F2 layer.
from the initial letters of the French words for These distances are based on average virtual heights
optimum working frequency, “Frequence Opti¬ (refer to Fig. 1-9), and in both cases a wave angle
mum de Travail.” Since amateur stations work in of zero is required. The actual distance covered by
fixed bands of frequencies, it is not possible to single-hop transmissions is usually somewhat less,
choose either the muf or fot at will. Instead, the at least at frequencies below about 28 MHz,
time of day at which optimum conditions can be because of ground losses at wave angles below
expected for a given path and distance on a about 3 degrees. The wave angle required for any
particular band must be determined. single-hop distance may be determined by referring
to Fig. 1-10 - the shaded areas represent the
Transmission Distance and Layer Height ranges of possible layer heights. Transmission paths
Consideration of Fig. 1-8 shows that the dis¬ in excess of 4000 kilometers (the maximum
tance at which a particular ray returns to earth single-hop range) obviously require multiple hops.
depends upon the angle at which it enters the This will be discussed later in the chapter. One-hop
layer. This angle in turn is determined by the angle transmission, when possible, will provide the great¬
(called the wave angle) at which the ray leaves the est signal strength at the receiving point because
transmitting antenna. Not shown in the drawing, there is some energy loss at each reflection,
but inherent in the geometry of the situation, is whether in the ionosphere or at the earth. At the
the fact that the distance also depends on the layer longer distances, this requires a small wave angle,
height. As the layer height is increased, the or “low-angle radiation” from the antenna. High-
distance at which a ray leaving at a fixed wave angle radiation is most useful for covering short
angle returns to earth also increases. The same distances. It will be appreciated that “long dis¬
wave angle, therefore, will result in transmission tance” and “short distance” are relative terms
over a greater distance when the wave is reflected when it is remembered that the distance depends
by the F2 layer than when it is reflected by the E on the layer height as well as the wave angle. At
times when the frequency in use is reflected by the
layer.
18
Chapter 1

E layer the distance will be one thing, but at control points are located one fourth of the path
another time of day when the E layer is ineffective length from each terminal point. If the muf at the
and the F2 layer comes into play the same wave transmitter’s control point is, say 14 MHz, iono¬
angle from the same antenna will cover a much spheric propagation in the direction of the receiver
larger distance. That is one reason why it is is possible on that frequency. If the muf at the
possible to communicate over longer distances at receiver’s control point is 14 MHz or higher,
night on frequencies in the vicinity of 7 MHz than ionospheric propagation is possible over the com¬
it is in the daytime. plete path and the signal will be heard. On the
Long-Distance Transmission other hand, if the muf at the receiving control
point is only 10 MHz, a 14-MHz signal from the
From the discussion in the preceding section, it transmitter will not be heard. The transmitting
should be clear that transmission over distances frequency must then be lowered to 10 MHz before
greater than 4000 km (2500 miles) must involve communications will be possible. In other words,
multihop propagation, because 4000 km is the the lower of the mufs at the two control points is
maximum distance that can be covered by one hop the muf of the circuit. The muf values at control
via the highest layer. Since multihop transmission points in any part of the world can be determined
increases the energy loss, it is quite important, for in advance from charts described later in this
most effective long-distance transmission, that a chapter under the subheading of Prediction Charts.
frequency near the muf be used, and that the In theory, communication is possible at any
antenna concentrate the radiation at low angles so frequency below the circuit muf, while in practice
that the number of reflections will be small. the absorption becomes too great if the frequency
The propagation of waves over long paths is is lowered too much below the muf.
complicated by a number of factors. For example, The E layer may be effective at the control
at the particular frequency used the E layer may point at either end of the circuit. This will be the
reflect the waves along part or parts of the path
case if the frequency to be used is below the
while the Flayer does the reflecting at other parts. ¿■-layer muf at that particular time. This fact
This will depend on the time of day, whether the
should not be forgotten, because frequently it
path is generally north-south or east-west, the part happens that the F2 layer is controlling one end of
of the world over which the path lies - in short, on
the path and the E layer the other. It can also
the state of the ionosphere all along the path. It is
happen that the E layer controls both ends of the
also possible that a wave reflected downward from path. Under such circumstances the F2 muf may
the F2 layer will be reflected upward from the E
be so high at both ends of the circuit that high
layer instead of being reflected from the earth.
absorption would be expected, whereas the actual
However, all these possibilities have but little effect case is that rather good signals will be received
on the primary consideration in DX-antenna design because the operating frequency is near the ¿-layer
- that the antenna should concentrate the radia¬ muf at one or both ends.
tion at the lowest possible angle.
The control-point method of prediction does
Despite the complexity of long-distance propa¬ not take into account how the waves travel from
gation, there is a method of determining average the transmitter to the receiver. Its justification is
communications possibilities in advance. This is that it has been found to be a useful method, on
based on predictions of the muf at two “control the average, for predicting whether or not commu¬
points” in the ionosphere. For distances beyond nication will be possible at a given frequency, or
4000 km, the control points are located at 2000 for selecting a frequency that will give communica¬
km (1250 miles) from the transmitter and the tion between any two points.
receiver, respectively, along the great-circle path The vertical angle at which a wave arrives at the
between them. For shorter distances, the two receiving point in long-distance transmission has

TABLE 1-1
Freq. Angle below which Angle above which Angle above which
MHz signals arrived 99% signals arrived 50% signals arrived 99%,
of the time of the time of the time
7 35° 22° io°
14 17° 11° 6°
21 12° 7° 4°
28 9° 5° 3°

Table 1-1 — Measured vertical angles of arrival of

signals from England at receiving location in New
Jersey.
Wave Propagation 19

Fig. 1-11 — Typical

daytime propagation of
high frequencies (14 to
28 MHz). The waves are
partially bent in going
through the E and Ft
layers, but not enough
to be returned to earth.
The actual reflection is
from the Ft layer.

been found by measurement to vary over a kilometers (20,000 mi.) via long path are not
considerable range. For example, measurements on uncommon, particularly at sunrise and sunset local
a path from England to the New Jersey (U.S.A.) time.
coast indicate that on 7 MHz the wave angle of the Occasionally waves arrive from directions that
seem to bear no visible relationship to the direction
received signal at times is as high as 35° and on 14
MHz is at times as high as 17°. For 99 percent of in which the transmitting station lies. While there
the time it is below those figures on these two are well-authenticated cases of this, and reasonable
frequencies. On the other hand, the same measure¬ explanations have been worked out on the basis of
ments showed that for 99 percent of the time the known behavior of the ionosphere, it is probable
angle was above 10° on 7 MHz and above 6° on 14 that the apparent shift in direction frequently
observed by amateurs is a result of “scattering,”
MHz. For about half the time the angle was
between 22° and 35° on 7 MHz and between 11° described in a later section. It is also possible that a
and 17° on 14 MHz. Whether or not there is exact combination of the vertical angle at which the
reciprocity between the transmitting and receiving wave arrives and minor responses of the antenna
wave angles, these figures clearly indicate the system being used gives a false direction indication.
importance of keeping the wave angle low. They Accurate direction finding with the sky wave at
also show that the higher the frequency, the less high frequencies is extremely difficult, requiring
useful the higher wave angles become in terms of highly specialized design and construction of
transmission over long distances. The above men¬ equipment.
tioned information is presented in Table 1-1 for
MISCELLANEOUS FEATURES OF
easy reference.
SKY-WAVE PROPAGATION
Polarization and Direction of Travel Although not having any very direct bearing on
antenna design, there are several aspects of sky-
Because of the nature of refraction in the wave transmission that are of considerable interest
ionosphere, the polarization of the refracted wave from an operating standpoint. The ability to
usually is shifted from the direction it had on recognize and appraise unusual propagation effects
leaving the transmitting antenna. It is therefore not often will help to explain seeming inconsistencies
at all necessary to use antennas having the same that may wronglj' be blamed on faulty antenna
polarization at the receiving and transmitting
design.
points. At the frequencies for which sky waves are
useful, most amateurs use horizontal antennas. Ionosphere Variations
Depending on the type, such antennas may gener¬
ate either horizontally or elliptically polarized The daily and seasonal variations in the ionized
waves. layers that result from changes in the amount of
For the most part, a wave follows the most ultraviolet light received from the sun have already
direct path between the transmitter and receiver. been mentioned. Reference has also been made to
In other words, it follows the great circle connect¬ the 11-year sunspot cycle, which directly affects
ing the two points. Because of variations in the propagation conditions because there is a rather
ionosphere, the actual path may vary slightly, and direct correlation between sunspot activity and
shifts of as much as 5 degrees from the true ionization. The 11-year figure for the time between
great-circle path occur at times. successive peaks of sunspot activity is only an
There are always two great-circle paths connect¬ average; any given cycle may vary a few years
ing two points on the earth’s surface, one repre¬ either way. The peak in 1968 was of average
senting the shortest distance between them and the stature, having a maximum smoothed sunspot
other a path in exactly the opposite direction — number not much in excess of 110. By contrast,
around the world the other way. Most communica¬ the peak which occurred in 1957-1958 had a
tion is via the “short path.” However, “long-path” smoothed number of sunspots exceeding 200 and
communication is not uncommon, particularly was the highest ever recorded. On occasions the F2
when there is not too much difference between the muf rose well into the vhf portion of the spectrum.
two distances. At certain times of the day, when Daily sunspot counts are recorded, and monthly
the short path would be inoperative, the iono¬ and yearly averages determined. The smoothed
sphere may be able to support communication over sunspot number (also called the 12-month running
the long path. Contacts in excess of 32,000 average) for a given month is the mean for the
20 Chapter 1

preceding 6 months and the succeeding 6 months absorption changes with the height of the sun.
about the month in question. The sunspot number There is also the daily variation of the muf; when
is not the actual number of spots observed, but the muf drops below the frequency in use, the
rather a weighted figure which takes into account signal will “fade out.”
such factors as the number of groups of sunspots, In addition, the ionization at any part of the
the number of individual spots counted, and the layer is in a continual state of change; there is
equipment used to make the measurements. The turbulence in the ionosphere, just as there is some
result is known as the Wolf number, R (after the turbulence in the atmosphere even on quiet days
man who derived the system for standardizing the when the weather seems stable. The amount of
sunspot count), and has been in use since the absorption is continually varying; waves entering
mid-18th century. Because international records the ionosphere at slightly different angles will be
are kept in Zurich, Switzerland, Wolf numbers are refracted differently; the polarization is continu¬
also known as Zurich smoothed sunspot numbers. ally changing with refraction. The wave reaching
The smoothed sunspot number (SSN) has been the receiving antenna is usually made up of a group
used extensively as an index of solar activity. of rays each of which has been acted on a little
Although it is not entirely satisfactory as a measure differently by the ionosphere. Sometimes the rays
of this activity, it is the only index with a series of are more or less in phase when they strike the
observations long enough for prediction purposes. receiving antenna; at other times some of the rays
Recently, it has been determined that the 10.7-cm may be out of phase with others. The result is a
solar radio noise flux (0) is generally a more continual variation in signal strength that may
accurate measure of solar activity. However, the occur at rates varying from several times a second
available series of observations is still too limited to once every few minutes.
for present prediction techniques. Since the two When transmission conditions are not alike for
indices are highly correlated, either index may be waves of slightly different frequency, the sidebands
used. in voice transmission may have a different fading
At a sunspot minimum there is a period of a pattern than the carrier or than each other. This is
year or two when the F2 muf does not get as high known as selective fading. It causes severe distor¬
as 28 MHz in the temperate latitudes. This mainly tion of the modulation, especially when the carrier
affects east-west communications. Transequatorial of an a-m or fm signal fades down while the
openings are not uncommon even with smoothed sidebands do not. The distortion is, in general,
sunspot counts below 5. worse with frequency modulation than with ampli¬
A small, but regular, variation in sunspot tude modulation, and is least with single-sideband
activity occurs over a period of 27 or 28 days. This transmission. Selective fading is more serious at the
is the time required for the sun to make one lower frequencies, such as 4 MHz, where the
rotation on its axis. The consequent rise and fall of sideband frequencies represent a larger percentage
the muf makes a noticeable change in propagation of the carrier frequency than they do at a
conditions at frequencies from 14 to 28 MHz. frequency such as 28 MHz.
Fading may be entirely different at two re¬
"One-Way” Skip ceiving points only a short distance apart. By the
use of antennas separated by a wavelength or two,
In long-distance hf communications via the feeding separate receivers, it is possible to over¬
ionosphere the relationship between receiving and come the effects of amplitude fading, but not of
transmitting capability over a given circuit may not selective fading. Similar use of inputs from anten¬
be exactly reciprocal. Investigations of this phe¬ nas of differing polarization will often serve the
nomenon are, thus far, inconclusive. Several theo¬ same purpose. Such receiving arrangements are
ries are currently popular. For whatever the reason known as “diversity” systems.
or reasons, the propagated waves do not always
take exactly the same paths from the two terminal
points, and so show considerable variation in Ionospheric Storms
alternate transmission and reception.
This condition of “one-way” skip has been Unusual eruptions on the sun cause distur¬
thought to exist for any or all of the following bances in the ionosphere called ionospheric storms.
reasons; tilting of the ionosphere, different muf These are accompanied by disturbances in the
conditions at each end of the path, the presence of earth’s magnetic field called magnetic storms.
E layer at one end of the path and not the other, Storms of this type are most frequent during the
and higher O-region absorption at one end of the sunspot cycle peak. They have a pronounced effect
path than at the other. When “one-way” skip on radio communication. The most prominent
conditions occur, one end of the path is generally features of these storms are the reduction in the F2
in the darkened hemisphere of the earth and the critical frequencies and the increase in D region
other in daylight. absorption. The practical consequence of this
lowering of the muf and raising of the luf is a
Fading narrowing of the usable frequency spectrum. On
high frequencies, communication frequently
Variations in the strength of a received signal becomes impossible, as though the refracting layers
are classified under the general term, fading. had disappeared. The storms vary in intensity and
Long-period variations are to be expected through duration. They may last from one to several days.
the day, on any given frequency, because the Ionospheric storms tend to recur at approxi-
Wave Propagation 21

mately 28-day intervals since they are associated begin some 18 to 36 hours after the occurrence of
with particular sunspots or groups of sunspots, and the solar flare, produce a variety of adverse effects
these tend to maintain a more or less fixed position in the ionosphere. For example, it sometimes
on the sun’s surface as it rotates. The period of causes the muf to drop as low as half its pre-storm
rotation, as mentioned before, is about 28 days. value while at the same time the luf rises greatly.
The Fi layer seems to disappear at times, or may
Aurora appear to split into several layers - causing severe
rapid fading and echoes. At this time, ionospheric
During magnetic storms auroral activity be¬
propagation conditions are poor. Ionospheric
comes more pronounced and extends farther from
storms may last from 2 to 5 or more days. During
the polar regions than is normally the case. During
the peak of the sunspot cycle they are more
abnormal auroral activity a peculiar form of wave
intense, but are of shorter duration than during
propagation is frequently observed, in which the
sunspot minimum periods. During the storms, only
auroral curtain acts as a reflector. Waves directed
line-of-sight and ground-wave communications re¬
toward the polar regions will be reflected back and
can be used for communication on frequencies and main reliable.
over distances that normally would be skipped
over. When this condition prevails, it is necessary, Scatter Signals
when directive antennas are used, that both the When a skip zone exists it might be expected,
transmitting and receiving antennas be directed from the simplified explanations of propagation
toward the polar regions rather than along the given earlier, that no signals at all would be heard
great-circle path between the two stations. While from stations too near to be reached by the sky
most of the hf bands are rendered useless for wave and too far away for the ground wave to be
long-distance communication because of the in¬ heard. Actually, however, signals from these sta¬
creased absorption and instability of the tions usually can be detected. The strength is low
ionosphere, the higher frequencies, beginning with and sometimes the signals have a “fluttery” or
about 28 MHz, are enhanced for beyond-local “warbly” fade which is very characteristic. This is
communications. East-west paths of up to about the result of the signals having been scattered so
1300 miles are possible using this form of propaga¬ much that they arrive at the receiving point from
tion. In the southern hemisphere one would beam random directions and in random phase relation¬
southward, since the aurora is a geomagnetic polar ships.
phenomenon. Several different forms of scatter propagation
Characteristically, aurora imparts a rapid flutter are known and are of use to amateurs. These
to signals. This flutter, which is not constant, may include backscatter and sidescatter, tropospheric
be anywhere from 100 to 2000 Hz and makes cw and ionospheric scatter, and transequatorial scat¬
the preferred mode for communications, as most ter. Tropospheric and transequatorial types are
phone signals (even ssb) are difficult to read unless considered to be forward-scatter modes of propaga¬
signal levels are extremely high. The higher one tion.
goes in frequency to work via an auroral opening, The troposphere, which lies between the earth
the stronger phone signals must be in order to be and the ionosphere upward to a variable height of
readable. 8 to 20 km, is capable of supporting medium¬
Propagation via auroral-curtain reflection distance communication with weak but very reli¬
occurs more often near the poles than at midlati¬ able signals regardless of the condition of the
tudes. It is nonexistent in the tropics. The auroral ionosphere. It is especially useful on vhf, where
display need not be visible at a given location in modest power and antennas may be employed to
order to take advantage of its properties; auroral sustain 200-mile reliable communications paths.
propagation may continue night and day for Higher power and bigger antennas can push the
several days but the display would not be visible maximum path length for this mode to close to
during the daylight hours in most latitudes. In 500 miles in the amateur service. Tropospheric (or
December and January, auroral propagation is least tropo) scatter signals show rapidly decreasing signal
likely. Its peaks occur in March and September, strength with increase in distance from their
but it is apt to happen at any time. source, but signal strength decreases very slowly as
a function of increase in frequency.
The D region, E and F layers are all useful for
Sudden Ionospheric Disturbances ionospheric scatter. Above the muf, where for the
At times, without warning, sky-wave hf com¬ most part signals pass through, rather than being
munication is severely disrupted by sudden iono¬ returned by the ionosphere, some signal does get
spheric disturbances (SID). Solar flares, heavily returned to earth by means of being scattered off
laden with ultraviolet and X rays, affect the D irregularities in the layers. These scattered signals
region in such a manner as to render communica¬ may continue forward along the great-circle path
tions all but impossible. D-rcgion absorption in¬ or may be returned to the earth in any random
creases markedly, because of sharply increased direction. As with tropo-scattered signals, iono¬
ionization. High-frequency signals suddenly fade spheric scatter signals are quite weak. But because
out and gradually return some minutes to some communication via this mode is possible from the
hours later, eventually reaching pre-SID levels. higher layers of the ionosphere, the distances
Solar flares are frequently followed by iono¬ covered are also greater — up to 2000 km via E
spheric storms. These storms, which generally layer is possible. If the ionospheric layers are intact
22 Chapter 1

(no SID or ionospheric storm in progress), iono¬ Sporadic-E Propagation

spheric scatter may be used. Such communication
requires higher power and bigger antennas than At about the height of the E layer, highly
does tropospheric scatter, however. ionized clouds are randomly and sporadically
formed. They are small and last for only a few
Sidescatter and backscatter are offshoot forms hours at a time. Because of their transient nature
of ionospheric propagation modes. Simply stated, and their altitude, they are called sporadic-E
it is possible to communicate with some distant clouds. To differentiate them from other E-layer
point by routes other than direct (great-circle phenomena these sporadic E clouds are designated
path), by “bouncing” a signal from an area of the E&.
earth which is reachable via the ionosphere from Eg clouds vary in intensity and move rapidly
both ends of the path. For example, stations in from southeast to northwest, at midlatitudes in the
each other’s skip zone could communicate via northern hemisphere. In the northern hemisphere
backscatter during a “solid” band opening to a they occur predominantly from May through
more distant area. If both stations are using August, with a minor peak in midwinter. The
rotatable directive arrays, optimum results will be seasonal months are reversed in the southern
obtained with beams directed toward the area for hemisphere.
which the band is open, rather than toward each The mechanism for the formation of Es is
other. The characteristics of backscatter propaga¬ believed to be wind shear. This explains ambient
tion are very weak but quite stable signals with ionization being redistributed and compressed into
essentially no fading. If the geographic area which a ledge of high density without the need for
is open to both terminals is very large, the signals production of extra ionization. Neutral winds of
may have a “hollow” or echo-effect sound, result¬ high velocity, flowing in opposite directions at
ing from different propagation delays of the signals slightly different altitudes, produce shears and, in
being scattered back and consequent differences in the presence of the geomagnetic field, the ions are
received phase relationships of the signal compo¬ collected at a particular altitude, forming a thin
nents. These effects are more pronounced if signals over-dense layer. Data from rockets penetrating Es
are scattered back over more than one hop, as may layers confirm the electron density, wind veloc¬
occasionally occur with high-gain directional anten¬ ities, and height parameters.
nas and transmitters running at powers near the Since midlatitude Es is directly associated with
legal amateur limit. Backscatter signals may be terrestrial or meteorological rather than solar phe¬
observed with dipole antennas and power levels nomena, it is not surprising that the occurrence of
below 100 watts if the frequency of operation is intense Es does not show direct correlation with
just below the muf. sunspot activity.
Sidescatter propagation is closely related to The occurrence of intense Es is markedly
that of backscatter, the only difference being that reduced at middle latitudes when the earth’s
the signals which are bounced from the earth are magnetic field is disturbed.
scattered sideways, rather than back in the direc¬ The muf of intense Es clouds is a function of
tion of the transmitter. For example, stations their ionization density. The highest frequency
which are 5000 or 6000 km apart might establish which can be propagated obliquely is not known,
usable communications on a frequency somewhat but propagation to approximately 200 MHz has
above the prevailing muf on the great-circle path been reported a number of times. The probability
between them, if there is a band opening from each of a 144-MHz opening is about 3 to 4 percent of
terminal point to some common area off the side the occurrence of 50-MHz openings. However,
of the great-circle route. Under such conditions, signal strength observed over the vhf range does
signals will arrive at each terminal from a direction not appear to be frequency dependent.
which may be dozens of degrees off the great-circle Normal one-hop, single-cloud Es DX is limited
path. As with backscatter, optimum results will be to about 1250 miles, but during the summer season
otained if rotatable beam antennas are pointed multiple clouds are common, and most propaga¬
toward the area for which the band is open. All tion is via more than one cloud. With the right
else being equal, sidescatter signals are considerably distribution, distances up to 2500 miles or more
stronger than backscatter signals. are possible over land.
Transequatorial scatter (TE) is primarily useful
for vhf work. The ionosphere over the equator is Meteor Trails
higher, thicker, and more densely ionized than it is Meteors entering the upper atmosphere travel at
elsewhere. Because of this, transequatorial iono¬ such high speed that a large amount of energy is
spheric propagation is better than in other direc¬ released when the meteor is slowed down by
tions. This is also true of transequatorial scatter, friction with the air. Part of this is used in ionizing
which, at its extremes, can be useful up to about the atmosphere along the path followed by the
8000 kilometers (5000 miles). The TE-mode muf is meteor. Even a very small meteor can ionize a
approximately 1.5 times the muf observed that day region 50 or more feet in diameter and a mile or so
over the same path. Thus, if the daytime muf were long. Such a region is large enough to refract the
to reach 35 MHz, then the TE-mode muf would shorter wavelengths back to earth. The ions quick¬
reach about 52 MHz during the day. This mode is ly recombine, however, so the effect of a meteor
most useful on the amateur six-meter band. Two- usually lasts only a short time — from a fraction of
meter propagation via this mode is considered a second to a few seconds, in the average case. It is
impossible. long enough, though, to produce a “burst” of
Wave Propagation 23

signal from stations not normally heard, oi heard

only weakly by scatter propagation. Bursts caused
by meteors can be observed at amateur frequencies
from 14 MHz through 220 MHz, with contacts
having been made through 144 MHz.
In its movement through space, the earth rather
frequently encounters swarms of meteors, or
“meteor showers,” which ionize the E region.
During meteor showers the bursts are so frequent
that it is sometimes possible to carry on con¬
tinuous communication on 28 and 50 MHz by that
means. Just as in the case of sporadic-E patches,
the ionized meteor trail must be midway between
two stations if a burst is to be heard, and the two
stations must be separated by enough distance so
that the wave angle will be low enough to be Fig. 1-12 — With these four volumes, frequency
refracted. During some of the stronger bursts of estimations for sky-wave signals can be made
manually for any month of any year. The only
the more intense showers, it is not unusual to have other information required is the Zurich smoothed
a signal enhanced by as much as 40 dB. Because relative sunspot number for the period of interest.
the meteor trail is so short lived, communication
relies on several (or many) parallel trails at one
time. The result is a Doppler shift of as much as 2 spheric Predictions, OT-TRER 13, and may be
kHz. obtained from the Superintendent of Documents,
A table of meteor showers appears in Chapter 2 U.S. Government Printing Office, Washington, DC
of The Radio Amateur’s VHF Manual, published 20402, for $9.30. Vol. 1 is available separately for
by the ARRL. 30 cents, and Vols. 2, 3, or 4 are available
Meteor-scatter communication, being basically separately for $3.00 each. Vol. 4 is not required
a single-hop mode, provides a maximum working unless estimations are being made for smoothed
distance of only about 1300 miles, the same as for sunspot numbers greater than 110.
any other E-layer single-hop mode of communi¬ The Institute for Telecommunication Sciences
cation. (ITS), Boulder, Colo., issues weekly forecasts
which contain effective solar activity indices
(12-month moving-average Zurich sunspot num¬
PREDICTION CHARTS bers) for use with these volumes. For many
Studies of the ionosphere and correlation of amateurs, a more likely source of this information
theories with observed signals from various dis¬ may be the Propagation Forecast Bulletins trans¬
tances over different paths have made it possible to mitted by W1AW on a regularly scheduled basis.
estimate the median maximum usable frequencies These bulletins are revised weekly, containing a
(mufs) to be expected over any part of the earth summary of the information from the ITS fore¬
for a particular smoothed sunspot number. Charts casts. W1AW transmission schedules are published
in the form of world maps with frequency con¬ monthly in QST, or may be obtained by writing
tours are available for making frequency estima¬ ARRL Headquarters, 225 Main St., Newington, CT
tions manually. The maps and instructions for their 06111.
use are contained in four volumes, pictured in Fig.
1-12.
With these four volumes, estimations of mufs Bibliography
may be made in advance for any month of any
year, if the smoothed sunspot number (or a Source material and more extended discussion
predicted value) is known. The use of the maps is of topics covered in this chapter can be found in
explained in Vol. 1. Vols. 2, 3, and 4 each contain the references given below.
maps for predicted Zurich smoothed relative sun¬ Davies, National Bureau of Standards Monograph
spot numbers of 10, 110, and 160, respectively. 80, Ionospheric Radio Propagation, Supt. of
Linear interpolation of data from two of the latter Documents, Washington 25, D.C.
three volumes must be made for periods of solar Hall, “High-Frequency Propagation Estimations for
the Radio Amateur,” QST, March, 1972.
activity at intermediate levels. The maps provide Jacobs and Leinwoll, “Short Wave Radio and the
data for the E and E2 layers, with the Ft layer also Ionosphere,” CQ, November, 1969.
being taken into account. Jacobs and Leinwoll, “VHF Ionospheric Propa¬
It has been the experience of amateurs using gation,” CQ, November, 1969.
these volumes that quite reliable frequency estima¬ Schultz, “Don’t Be Afraid of the Big Bad Black¬
tions are obtainable. Any observed departures tend out,” CQ, November, 1969.
Terman, Radio Engineering,, McGraw-Hill Book
to indicate that the estimated values are sometimes Co., New York. N. Y.
on the conservative side — that actual mufs for a The Radio Amateur’s VHF Manual, ARRL,
given time over a given path may be slightly higher Newington, Conn.
than estimated. Wilson, “Midlatitude Intense Sporadie-E Pro¬
The four-volume set is called OT Telecommuni¬ pagation,” QST, December, 1970 (Part I) and
cations Research and Engineering Report lono- March, 1971 (Part II).
 Chapter 2

Antenna Fundamentals

An antenna is an electric circuit of a special uted around the circuit wiring and coil conductors.
kind. In the ordinary type of circuit the dimen¬ Such circuits are said to have lumped constants. In
sions of coils, capacitors and connections usually an antenna, on the other hand, the inductance,
are small compared with the wavelength that capacitance, and resistance are distributed along
corresponds to the frequency in use. When this is the wire. Such a circuit is said to have distributed
the case most of the electromagnetic energy stays constants. Circuits with distributed constants are
in the circuit itself and either is used up in so frequently straight-line conductors that they are
performing useful work or is converted into heat. customarily called linear circuits.
But when the dimensions of wiring or components
become appreciable compared with the wave¬ RESONANCE IN LINEAR CIRCUITS
length, some of the energy escapes by radiation in
the form of electromagnetic waves. When the The shortest length of wire that will resonate to
circuit is intentionally designed so that the major a given frequency is one just long enough to permit
portion of the energy is radiated, we have an an electric charge to travel from one end to the
antenna. other and then back again in the time of one rf
Usually, the antenna is a straight section of cycle. If the speed at which the charge travels is
conductor, or a combination of such conductors. equal to the velocity of light, approximately
Very frequently the conductor is a wire, although 300,000,000 meters per second, the distance it will
rods and tubing also are used. In this chapter we cover in one cycle or period will be equal to this
shall use the term “wire” to mean any type of velocity divided by the frequency in hertz, or
conductor having a cross section that is small
compared with its length. X. _ 300,000,000
7-
The strength of the electromagnetic field radi¬
ated from a section of wire carrying radio-frequen¬
cy current depends on the length of the wire and in which X is the wavelength in meters. Since the
the amount of current flowing.! Other things being charge traverses the wire twice, the length of wire
equal, the field strength will be directly propor¬ needed to permit the charge to travel a distance X
tional to the current. It is therefore desirable to in one cycle is x/2, or one-half wavelength.
make the current as large as possible, considering Therefore the shortest resonant wire will be
a half wavelength long.
the power available. In any circuit that contains
The reason for this length can be made clear by
both resistance and reactance, the largest current a simple example. Imagine a trough with barriers at
will flow (for a given amount of power) when the each end. If an elastic ball is started along the trough
reactance is “tuned out” — in other words, when from one end, it will strike the far barrier, bounce
the circuit is made resonant at the operating back, travel along to the near barrier, bounce again,
frequency. So it is with the common type of and continue until the energy imparted to it
antenna; the current in it will be largest, and the originally is all dissipated. If, however, whenever it
radiation therefore greatest, when the antenna is returns to the near barrier it is given a new push
resonant. just as it starts away, its back-and-forth motion can
In an ordinary circuit the inductance is usually be kept up indefinitely. The impulses, however,
concentrated in a coil, the capacitance in a must be timed properly; in other words, the rate or
capacitor, and the resistance is principally concen¬ frequency of the impulses must be adjusted to the
trated in resistors, although some may be distrib- length of travel and the rate of travel. Or, if the
+ It would also be true to say that the field timing of the impulses and the speed of the ball are
strength depends on the voltage across the section
of wire, but it is generally more convenient to fixed, the length of the trough must be adjusted to
measure current. The electromagnetic field consists “fit.”
of both magnetic and electric energy, with the In the case of the antenna, the speed is
total energy equally divided between the two. One essentially constant, so we have the alternatives of
cannot exist without the other in an electro¬
magnetic wave, and the voltage in an antenna is adjusting the frequency to a given length of wire,
just as much a measure of the field intensity as the or the length of wire to a given operating frequen¬
current. cy. The latter is usually the practical condition.

24
Antenna Fundamentals 25

By changing the units in the equation just given

and dividing by 2, the formula

I - 492
AMHz)

is obtained. In this case Iis the length in feet of a

half wavelength for a frequency f given in mega¬
hertz, when the wave travels with the velocity of Fig. 2-1 — Current and voltage distribution on a
light. This formula is the basis upon which several half-wave wire. In this conventional representation
significant lengths in antenna work arc developed. the distance at any point (X, for instance) from the
wire, represented by the heavy line, to the curve
It represents the length of a half wavelength in gives the relative intensity of current or voltage at
space, when no factors that modify the speed of that point. The relative direction of current flow
propagation exist. To determine a half wavelength (or polarity of voltage) is indicated by drawing the
in meters, the relationship is curve either above or below the line that represents
the antenna. The curve above, for example, shows
/ = iso that the instantaneous polarity of the voltage in
/(MHz) one half of the antenna is opposite to that in the
other half.

Current and Voltage Distribution

neous value of current at any point varies sinusoi¬
If the wire in the first illustration had been dally at the applied frequency, but its amplitude is
infinitely long the charge (voltage) and the current different at every point along the wire as shown by
(an electric current is simply a charge in motion) the curve. The standing-wave curve itself has the
would both slowly decrease in amplitude with shape of a half sine wave, at least to a good
distance from the source. The slow decrease would approximation.
result from dissipation of energy in the form of The voltage along the wire will behave differ¬
radio waves and in heating the wire because of its ently; it is obviously greatest at the end since at
resistance. However, when the wire is short the this point we have two practically equal charges
charge is reflected when it reaches the far end, just adding. As we move back along the wire, however,
as the ball bounced back from the barrier. With the outgoing and returning charges are not equal
radio-frequency excitation of a half-wave antenna, and their sum is smaller. At the quarter-wave point
there is of course not just a single charge but a the returning charge is of equal magnitude but of
continuous supply of energy, varying in voltage opposite sign to the outgoing charge, since at this
according to a sine-wave cycle. We might consider time the polarity of the voltage wave from the
this as a series of charges, each of slightly different source has reversed (one-half cycle). The two
amplitude than the preceding one. When a charge voltages therefore cancel each other and the
reaches the end of the antenna and is reflected, the resultant voltage is zero. Beyond the quarter-wave
direction of current flow reverses, since the charge point, away from the end of the wire, the voltage
is now traveling in the opposite direction. How¬ again increases, but this time with the opposite
ever, the next charge is just reaching the end of the polarity.
antenna, so we have two currents of practically the It will be observed, therefore, that the voltage is
same amplitude flowing in opposite directions. The maximum at every point where the current is
resultant current at the end of the antenna minimum, and vice versa. The polarity of the
therefore is zero. As we move farther back from current or voltage reverses every half wavelength
the end of the antenna the magnitudes of the along the wire, but the reversals do not occur at
outgoing and returning currents are no longer the the same points for both current and voltage; the
same because the charges causing them have been respective reversals occur, in fact, at points a
supplied to the antenna at different parts of the rf quarter wave apart.
cycle. There is less cancellation, therefore, and a A maximum point on a standing wave is called
measurable current exists. The greatest difference a loop (or antinode); a minimum point is called a
- that is, the largest resultant current - will be node.
found to exist a quarter wavelength away from the
end of the antenna. As we move back still farther Harmonic Operation
from this point the current will decrease until, a
half wavelength away from the end of the antenna, If there is reflection from the end of a wire, the
it will reach zero again. Thus, in a half-wave number of standing waves on the wire will be equal
antenna the current is zero at the ends and to the length of the wire divided by a half
maximum at the center. wavelength. Thus, if the wire is two half waves long
This current distribution along a half-wave wire there will be two standing waves; if three half
is shown in Fig. 2-1. The distance measured waves long, three standing waves, and so on. These
vertically from the antenna wire to the curve longer wires, each multiples of a half wave in
marked “current,” at any point along the wire, length, will also be resonant, therefore, at the same
represents the relative amplitude of the current as frequency as the single half-wave wire. When an
measured by an ammeter at that point. This is antenna is two or more half waves in length at the
called a standing wave of current. The instanta- operating frequency it is said to be harmonically
26 Chapter 2

near end in exactly the time of one rf cycle. Then

it will be found that the physical length of the
circuit with the lower velocity of propagation is
shorter than the physical length of the other. The
electrical lengths, however, are identical, each
being a half wavelength.
In alternating-current circuits the instantaneous
Fig. 2-2 — Harmonic operation of along wire. The values of current or voltage are determined by the
wire is long enough to contain several half waves. instant during the cycle at which the measurement
The current and voltage curves cross the heavy line is made (assuming, of course, that such a measure¬
representing the wire to indicate that there is ment could be made rapidly enough). If the
reversal in the direction of the current, and a
reversal in the polarity of the voltage, at intervals current and voltage follow a sine curve — which is
of a half wavelength. The reversals of current and the usual case — the time, for any instantaneous
voltage do not coincide, but occur at points a value, can be specified in terms of an angle. The
quarter wavelength apart. sine of the angle gives the instantaneous value
when multiplied by the peak value of the current
or voltage. A complete sine curve occupies the 360
resonant, or to operate at a harmonic. The number
degrees of a circle and represents one cycle of ac
of the harmonic is the number of standing waves
current or voltage. Thus a half cycle is equal to 180
on the wire. For example, a wire two half waves
degrees, a quarter cycle to 90 degrees, and so on.
long is said to be operating on its second harmonic;
It is often convenient to use this same form of
one three half waves long on its third harmonic,
representation for linear circuits. When the electri¬
and so on.
cal length of a circuit is such that a charge traveling
Harmonic operation is often utilized in antenna in one direction takes the time of one cycle or
work because it permits operating the same anten¬ period to traverse it, the length of the circuit is said
na on several harmonically related amateur bands. to be 360 degrees. This corresponds to one
It is also an important principle in the operation of wavelength. On a wire a half wave in electrical
certain types of directive antennas. length, the charge completes a one-way journey in
one half cycle and its length is said to be 180
Electrical Length degrees. The angular method of measurement is
The electrical length of a linear circuit such as quite useful for lengths that are not easily
an antenna wire is not necessarily the same as its remembered fractions or simple multiples of such
physical length in wavelengths or fractions of a fractions. A chart for converting fractions of a
wavelength. Rather, the electrical length is mea¬ wavelength to degrees is given in Fig. 2-3.
sured by the time taken for the completion of a
specified phenomenon. Velocity of Propagation
For instance, we might imagine two linear The velocity at which electromagnetic waves
circuits having such different characteristics that travel through a medium depends upon the dielec¬
the speed at which a charge travels is not the same tric constant of the medium. At rf the dielectric
in both. Suppose we wish to make both circuits constant of air is practically unity, so the waves
resonant at the same frequency, and for that travel at essentially the same velocity as light in a
purpose adjust the physical length of each until a vacuum. This is also the velocity, very closely, of
charge started at one end travels to the far end, is the charge traveling along a wire.
reflected and completes its return journey to the If the dielectric constant is greater than 1, the
velocity of propagation is lowered. Thus the
introduction, in appreciable quantity, of insulating
material having a dielectric constant greater than 1
will cause the wave to slow down. This effect is
encountered in practice in connection with both
antennas and transmission lines. It causes the
electrical length of the line or antenna to be greater
than actual physical length.

Length of a “Half-Wave” Antenna

Even if the antenna could be supported by
insulators that did not cause the electromagnetic
fields traveling along the wire to slow down, the
physical length of a practical antenna always is
somewhat less that its electrical length. That is, a
“half-wave” antenna is not one having the same
length as a half wavelength in space. It is one
having an electrical length equal to 180 degrees.
Or, to put it another way, it is one with a length
Fig. 2-3 — Chart for converting electrical degrees to which has been adjusted to “tune out” any
fractions of a wavelength. reactance, so it is a resonant antenna.
Antenna Fundamentals 27

The antenna length required to resonate at a

given frequency (independently of any dielectric
effects) depends on the ratio of the length of the
conductor to its diameter. The smaller this ratio
(or the “thicker” the wire), the shorter the antenna
for a given electrical length. This effect is shown in
Fig. 2-4 as a factor (Â) by which a free-space half
wavelength must be multiplied to find the resonant
length, as a function of the ratio of the free-space
half wavelength to conductor diameter, known as
the length/diameter ratio. The curve is based on
theoretical considerations and is useful as a guide
to the probable antenna length for a given frequen¬
cy. It applies only to conductors of uniform
diameter (tapered elements such as are used in
some types of beam antennas will generally be Fig. 2-4 — The solid curve shows the factor, K, by
longer, for the same frequency) and does not which the length of a half wave in free space
should be multiplied to obtain the physical length
include any effects introduced by the method of of a resonant half-wave antenna having the length/
supporting the conductor. diameter ratio shown along the horizontal axis.
A length/diameter ratio of 10,000 is roughly The broken curve shows how the radiation resis¬
average for wire antennas (actually, it is approxi¬ tance of a half-wave antenna varies with the
mately the ratio for a 7-MHz half-wave antenna length/diameter ratio.
made of No. 12 wire). In this region K changes
rather slowly and a half-wave antenna made of wire
is about 2 percent shorter than a half wavelength in end effect, as there is some current flowing into
space. the end capacitance.
The shortening effect is most pronounced when End effect increases with frequency and varies
the length/diameter ratio is 100 or less. An antenna slightly with different installations. However, at
constructed of 1-inch diameter tubing for use on frequencies up to 30 MHz (the frequency range
144 MHz for example, would have a length/diam¬ over which wire antennas are most likely to be
eter ratio of about 40 and would be almost 5 used) experience shows that the length of a
percent shorter than a free-space half wavelength. half-wave antenna is of the order of 5 percent less
If the antenna is made of rod or tubing and is than the length of a half wave in space. As an
not supported near the ends by insulators, the average, then, the physical length of a resonant
following formula will give the required physical half-wave antenna may be taken to be
length of a half-wave antenna based on Fig. 2-4.
/ (feet) = X 0'95 =
468
( ' /(MHz) /(MHz)
Length (feet) y (MHz)
or
. n r ^ \ - 5905 X Similarly
Length (inches) y(MHz)

or /(meters) y (MHz)
Length (meters) - /(MHz) -
These formulas are reasonably accurate for
finding the physical length of a half-wave antenna
where K is taken from Fig. 2-4 for the particular for a given frequency, but do not apply to
length/diameter ratio of the conductor used. antennas longer than a half wave in length. In the
practical case, if the antenna length must be
End Effect adjusted to exact frequency (not all antenna
systems require it) the length should be “pruned”
If the formulas of the preceding section are
to resonance.
used to determine the length of a wire antenna, the
antenna will resonate at a somewhat lower frequen¬
cy than is desired. The reason for this is that there
ANTENNA IMPEDANCE
is an additional “loading” effect caused by the
insulators that must be used at the ends of the wire In the simplified description given earlier of
for suspending it. These insulators and the wire voltage and current distribution along an antenna it
loop that ties the insulator to the antenna add a was stated that the voltage was zero at the center
small amount of capacitance to the system. This of a half-wave antenna (or at any current loop
capacitance helps to tune the antenna to a slightly along a longer antenna). It is more accurate to say
lower frequency, in much the same way that that the voltage reaches a minimum rather than
additional capacitance in any tuned circuit will zero. Zero voltage with a finite value of current
lower the resonant frequency. In an antenna it is would imply that the circuit is entirely without
called end effect The current at the ends of the resistance. It would also imply that no energy is
antenna does not quite reach zero because of the radiated by the antenna, since a circuit without
28 Chapter 2

—s—
--1_- that this current is always in phase with the applied
voltage, regardless of the antenna length. For the
sake of simplicity only the current flowing in one
leg of the antenna is considered; conditions in the
other leg are similar.
Power Sourct In the case of the resonant antenna, the current
travels out to the end and back to the driving point
Fig. 2-5 — The center-fed antenna discussed in the in one half cycle, since one leg of the antenna is 90
text. It is assumed that the leads from the source degrees long and the total path out and back is
of power to the antenna have zero length. therefore 180 degrees. This would make the phase
of the reflected component of current differ from
that of the outgoing current by 180 degrees, since
resistance would take no real power from the the latter current has gone through a half cycle in
driving source. the meantime. However, it will be remembered
Actually, of course, an antenna, like any other that there is a phase shift of 180 degrees at the end
circuit, consumes power. The current that flows in of the antenna, because the direction of current
it therefore must be supplied at a finite value of reverses at the end. The total phase shift between
voltage. The impedance of the antenna is simply the outgoing and reflected currents, therefore, is
equal to the voltage applied to its terminals divided 360 degrees. In other words, the reflected com¬
by the current flowing into those terminals. If the ponent arrives at the driving point exactly in phase
current and voltage are exactly in phase the with the outgoing component. The reflected com¬
impedance is purely resistive. This is the case when ponent, shown at C, adds to the outgoing com¬
the antenna is resonant. If the antenna is not ponent to form the resultant or total current at the
exactly resonant the current will be somewhat out driving point The resultant current is shown at D,
of phase with the applied voltage and the antenna and in the case of the resonant antenna it is easily
shows reactance along with resistance. seen that the resultant is exactly in phase with the
Most amateur transmitting antennas are op¬ applied voltage. This being the case, the load seen
erated at or quite close to resonance so that by the source of power is a pure resistance.
reactive effects are in general comparatively small. Now consider the antenna that is too short to
They are nevertheless present, and must be taken be resonant. The outgoing component of current is
into account whenever an antenna is operated at still in phase with the applied voltage, as shown at
other than the exact design frequency. B. The reflected component, however, gets back to
In the following discussion it is assumed that the driving point too soon, because it travels over a
power is applied to the antenna by opening the path less than 180 degrees, out and back. This
conductor at the center and applying the driving means that the maximum value of the reflected
voltage across the gap. This is shown in Fig. 2-5.
While it is possible to supply power to the antenna
by other methods, the selection of different driving
points leads to different values of impedance; this
can be appreciated after study of Fig. 2-1, which
shows that the ratio of voltage to current (which is,
by definition, the impedance) is different at every
point along the antenna. To avoid confusion it is
desirable to use the conditions at the center of the
antenna as a basis.

The Antenna as a Circuit

If the frequency applied at the center of a
half-wave antenna is varied above and below the
resonant frequency, the antenna will exhibit much
the same characteristics as a conventional series-
resonant circuit. Exactly at resonance the current
at the input terminals will be in phase with the
applied voltage. If the frequency is on the low side
of resonance the phase of the current will lead the
voltage; that is, the reactance of the antenna is
capacitive. When the frequency is on the high side
of resonance the opposite occurs; the current lags
the applied voltage and the antenna exhibits
inductive reactance. Fig. 2-6 — Current flow in resonant and off-
It is not hard to see why this is so. Consider the resonant antennas. The initial current flow, B,
antennas shown in Fig. 2-6, one resonant, one too caused by the source of power, is in phase with the
long for the applied frequency, and one too short. applied voltage, A. This is the outgoing current
discussed in the text. The reflected current, C,
In each case the applied voltage is shown at A, and combines with the outgoing current to form the
the instantaneous current going into the antenna resultant current, D, at the input terminals of the
because of the applied voltage is shown at B. Note antenna.
Antenna Fundamentals 29

component occurs at the driving point ahead of (in Resistance

time) the maximum value of the outgoing com¬ The energy supplied to an antenna is dissipated
ponent, since that particular charge took less than in the form of radio waves and in heat losses in the
a half cycle to get back. Including the 180-degree wire and nearby dielectrics. The radiated energy is
reversal at the end of the antenna, the total phase the useful part, but so far as the antenna is
shift is therefore less than 360 degrees. This is concerned it represents a loss just as much as the
shown at C, and the resultant current is the energy used in heating the wire is a loss. In either
combination of the outgoing and reflected com¬ case the dissipated power is equal to PR: in the
ponents as given at D. It can be seen that the case of heat losses, R is a real resistance, but in the
resultant current leads the applied voltage, so the case of radiation R is an assumed resistance, which,
antenna looks like a resistance in series with a if present, would dissipate the power that is
capacitance. The shorter the antenna, the greater actually radiated from the antenna. This fictitious
the phase shift between voltage and current; that resistance is called the radiation resistance. The
is, the capacitive reactance increases as the antenna total power loss in the antenna is therefore equal to
is shortened. P (Ra + R), where Ro is the radiation resistance
When the antenna is too long for the applied and R the real resistance, or ohmic resistance.
frequency the reflected component of current In the ordinary half-wave antenna operated at
arrives too late to be exactly in phase with the amateur frequencies the power lost as heat in the
outgoing component, because it must travel over a conductor does not exceed a few percent of the
path more than 180 degrees long. The maximum total power supplied to the antenna. This is
value of the reflected component therefore occurs because the rf resistance of copper wire even as
later (in time) than the maximum value of the small as No. 14 is very low compared with the
outgoing component, as shown at C. The resultant radiation resistance of an antenna that is reason¬
current at the antenna input terminals therefore ably clear of surrounding objects and is not too
lags behind the applied voltage. The phase lag close to the ground. Therefore it can be assumed
increases as the antenna is made longer. That is, an that the ohmic loss in a reasonably well-located
antenna that is too long shows inductive reactance antenna is negligible, and that all of the resistance
along with resistance, and this reactance increases shown by the antenna is radiation resistance. As a
with an increase in antenna length over the length radiator of electromagnetic waves such an antenna
required for resonance. is a highly efficient device.
If the antenna length is increased to 180 The value of radiation resistance, as measured
degrees on each leg, the go-and-return path length at the center of a half-wave antenna, depends on a
for the current becomes 360 degrees. This, plus the number of factors. One is the location of the
180-degree reversal at the end, makes the total antenna with respect to other objects, particularly
phase shift 540 degrees, which is the same as a the earth. Another is the length/diameter ratio of
180-degree shift. In this case the reflected current the conductor used. In “free space” - with the
arrives at the input terminals exactly out of phase antenna remote from everything else — the radia¬
with the outgoing component, so the resultant tion resistance of a resonant antenna made of an
current is very small. The resultant is in phase with infinitely thin conductor is approximately 73
the applied voltage, so the antenna impedance is ohms. The concept of a free-space antenna forms a
again purely resistive. The resistance under these convenient basis for calculation because the modi¬
conditions is very high, and the antenna has the fying effect of the ground can be taken into
characteristics of a parallel-resonant circuit. A account separately. If the antenna is at least several
voltage loop, instead of a current loop, appears at wavelengths away from ground and other objects,
the input terminals when each leg of the antenna is it can be considered to be in free space insofar as
180 degrees long. its own electrical properties are concerned. This
The amplitude of the reflected component is condition can be met with antennas in the vhf and
less than that of the component of current going uhf range.
into the antenna. This is the result of energy loss The way in which the free-space radiation
by radiation as the current travels along the wire. It resistance varies with the length/diameter ratio of a
is perhaps easier to understand if, instead of half-wave antenna is shown by the broken curve in
thinking of the electromagnetic fields as being Fig. 2-4. As the antenna is made thicker the
brought into being by the current flow, we adopt radiation resistance decreases. For most wire anten¬
the more fundamental viewpoint that current flow nas it is close to 65 ohms. It will usually lie
along a conductor is caused by a moving electro¬ between 55 and 60 ohms for antennas constructed
magnetic field. When some of the energy escapes of rod or tubing.
from the system because the field travels out into The actual value of the radiation resistance - at
space, it is not hard to understand why the current least so long as it is 50 ohms or more - has no
becomes less as it travels farther. There is simply appreciable effect on the radiation efficiency of
less energy left to cause it. The difference between the antenna. This is because the ohmic resistance is
the outgoing- and reflected-current amplitudes only of the order of 1 ohm with the conductors
accounts for the fact that the current does not go used for thick antennas. The ohmic resistance does
to zero at a voltage loop, and a similar difference not become important until the radiation resis¬
between the applied- and reflected-voltage compo¬ tance drops to very low values — say less than 10
nents explains why the voltage does not go to zero ohms — as may be the case when several antennas
at a current loop. are coupled together to form an array of elements.
30 Chapter 2

for three values of length/diameter ratio are shown;

L represents the length of a half wave in space,
approximately, and D is the diameter of the
conductor in the same units as the length. The
point where each curve crosses the zero axis
(indicated by an arrow in each case) is the length at
which an antenna of that particular length/diam¬
eter ratio is resonant. The effect of L/D on the
resonant length also is illustrated by these curves:
the smaller the ratio, the shorter the length at
which the reactance is zero.
It will be observed that the reactance changes
about twice as rapidly in the antenna with the
LENGTH Of ANTENNA IN WAVELENGTHS smallest diameter (A) as it does in the antenna with
the largest diameter (C). With still larger diameters
the rate at which the reactance changes would be
Fig. 2-7 — Resistance and reactance at input
terminals of a center-fed antenna as a function of even smaller. As a practical matter it is advanta¬
its length near a half wavelength. As shown by geous to keep the reactance change with a given
curves A, B and C, the reactance is affected more change in length as small as possible. This means
by the length/diameter ratio of the conductor than that when the antenna is operated over a small
is the radiation resistance. band of frequencies centered on its resonant
frequency, the reactance is comparatively low, and
that the impedance change with frequency is small.
The radiation resistance of a resonant antenna This simplifies the problem of supplying power to
is the “load” for the transmitter or for the rf the antenna when it must be worked at frequencies
transmission line connecting the transmitter and somewhat different from its resonant frequency.
antenna. Its value is important, therefore, in At lengths considerably different from the
determining the way in which the antenna and resonant length the reactance changes more rapidly
transmitter or Une are coupled together. than in the curves shown in Fig. 2-7. As in the case
The resistance of an antenna varies with its of the resistance change, the change is most rapid
length as well as with the ratio of its length to its when the length exceeds 135 degrees (3/8 wave¬
diameter. When the antenna is approximately a length) and approaches 180 degrees (1/2 wave¬
half wave long, the resistance changes rather slowly length) on a side. In this case the reactance is
with length. This is shown by the curves of Fig. inductive and reaches a maximum at a length
2-7, where the change in resistance as the length is somewhat less that 180 degrees. Between this
varied a few percent on either side of resonance is maximum point and 180 degrees of electrical
shown by the broken curves. The resistance de¬ length the reactance decreases very rapidly, be¬
creases somewhat when the antenna is slightly coming zero when the length is such as to be
short, and increases when it is sUghtly long. These parallel resonant.
curves also illustrate the effect of changing the Very short antennas have a large capacitive
frequency applied to an antenna of fixed length, reactance. It was pointed out in the preceding
since increasing the frequency above resonance is section that with antennas shorter than 90 degrees
the same thing as having an antenna that is too on a side the resistance decreases at a fairly
long, and vice versa. uniform rate, but this is not true of the reactance.
The range covered by the curves in Fig. 2-7 is It increases rather rapidly when the length of a side
representative of the frequency range over which a is shortened below about 45 degrees.
fixed antenna is operated between the limits of an The behavior of antennas with different
amateur band. At greater departures from the length/diameter ratios corresponds to the behavior
resonant length the resistance continues to de¬ of ordinary resonant circuits having different 0s.
crease somewhat uniformly as the antenna is When the Q of a circuit is low, the reactance is
shortened, but tends to increase rapidly as the small and changes rather slowly as the applied
antenna is made longer. The resistance increases frequency is varied on either side of the resonant
very rapidly when the length of a leg exceeds about frequency. If the Q is high, the converse is true.
135 degrees, or about 3/8 wavelength, and reaches The response curve of the low-0 circuit is “broad”;
a maximum when the length of one side is 180 that of the high-0 circuit “sharp.” So it is with
degrees. This is considered in more detail in later antennas; a thick antenna works well over a
sections of this chapter. comparatively wide band of frequencies while a
thin antenna is rather sharp in tuning. The 0 of the
Reactance thick antenna is low; the 0 of the thin antenna is
high, assuming essentially the same value of radia¬
The rate at which the reactance of the antenna tion resistance in both cases.
increases as the length is varied from resonance
depends on the length/diameter ratio of the con¬
ductor. The thicker the conductor the smaller the Coupled Antennas
reactance change for a given change in length. This A conventional tuned circuit far enough away
is shown by the reactance curves in Fig. 2-7. Curves from all other circuits so that no external coupling
Antenna Fundamentals 31

exists can be likened to an antenna in free space, in

one sense. That is, its characteristics are unaffected
by its surroundings. It will have a Q and resonant
impedance determined by the inductance, capaci¬
tance and resistance of which it is composed, and
those quantities alone. But as soon as it is coupled
to another circuit its Q and impedance will change,
depending on the characteristics of the other
circuit and the degree of coupling.
A similar situation arises when two or more
“elementary” antennas - half-wave antennas, fre¬
quently called half-wave dipoles - are coupled
together. This coupling takes place merely by
having the two antennas in proximity to each
other. The sharpness of resonance and the radia¬
tion resistance of each “element” of the system are
affected by the mutual interchange of energy
between the coupled elements. The exact effect
depends on the degree of coupling (that is, how
close the antennas are to each other in terms of
wavelength, and whether the wires arc parallel or
not) and the tuning condition (whether tuned to
resonance or slightly off resonance) of each ele¬
ment. Analysis is extremely difficult and even then
has to be based on some simplifying assumptions
that may not be true in practice. Only a few
relatively simple cases have been analyzed. Such
data as are available for even moderately compli¬
cated systems of coupled antennas are confined to
a few types and are based on experimental mea¬
surements. They are, therefore, subject to the Fig. 2-8 — How the feed point makes a difference
inaccuracies that accompany any measurements in in current distribution along the antenna. With
a field where measurement is difficult at best. center feed, increasing the length of each side of
Antenna systems consisting of coupled ele¬ the antenna keeps the current flowing in the same
ments will be taken up in later chapters. At this direction in the two halves, up to the point where
point it is sufficient to appreciate that the free- each side is a half-wavelength long. For harmonic
space values that have been discussed in this operation, the antenna must be fed in such a way
chapter may be modified drastically when more that the current direction reverses in alternate
half-wavelength sections. Suitable methods are
than one antenna element is involved in the shown at D and E.
system. It has already been pointed out that the
presence of the ground, as well as nearby conduc¬
tors or dielectrics, also will modify the free-space HARMONICALLY OPERATED ANTENNAS
values. The free-space characteristics of the elemen¬ An antenna operated at a harmonic of its
tary half-wave dipole are only the point of depar¬ fundamental frequency has considerably different
ture for a practical antenna system. In other words, properties than the half-wave dipole previously
they give the basis for understanding antenna discussed. It must first be emphasized that har¬
principles but cannot be applied too literally in the monic operation implies that there is a reversal of
practical case. the direction of current flow in alternate half-wave
It is of interest to note that the comparison sections of the antenna, as shown in Fig. 2-2 and
between an isolated tuned circuit and an antenna again at A in Fig. 2-8. In the latter figure, the curve
in free space is likewise not to be taken too shows the standing wave of current intensity along
literally. In one sense the comparison is wholly the wire; the curve is above the line to indicate
misleading. The tuned circuit is usually so small, current flow in one direction (assumed to be to the
physically, in comparison with the wavelength, right, in the direction of the arrow) and below the
that practically no energy escapes from it by line to indicate current flow in the opposite
radiation. An antenna, to be worthy of the name, direction in the other half-wave section. (During
is always so large in comparison with the wave¬ the next radio-frequency half cycle the current
length that practically all the energy supplied to it flow in the left-hand half-wave section would be
escapes by radiation. Thus, the antenna can be said toward the left, and in the right-hand half-wave
to be very tightly coupled to space, while the section to the right; this alternation in direction
tuned circuit is not coupled to anything. This very takes place in each succeeding half cycle. However,
fundamental difference is one reason why antenna the direction of current flow in adjacent half-wave
systems cannot be analyzed as readily, and with as sections is at all times opposite.) The antenna in
satisfactory results in the shape of simple formulas, this drawing is one wavelength long and is opera¬
as ordinary electrical circuits. ting on its second harmonic.
32 Chapter 2

Now consider the half-wave antenna shown at Because the number of half waves varies with
Fig. 2-8 B. It is opened in the center and fed by a the harmonic on which the antenna is operated,
source of rf power through leads that are assumed consideration of the formulas together with that
to have zero length. Since one terminal of the for the half-wave antenna (the fundamental fre¬
generator is positive at the same instant that the quency) will show that the relationship between
other terminal is negative, current flows into one the antenna fundamental frequency and its har¬
side of the generator while it is flowing out at the monics is not exactly integral. That is, the "sec¬
other terminal. Consequently the current flows in ond-harmonic” frequency to which a given length
the same direction in both sections of the half¬ of wire is resonant is not exactly twice its
wave antenna. It has the amplitude distribution fundamental frequency; the “third-harmonic” res¬
shown by the curve over the antenna wire. onance is not exactly three times it fundamental,
If we now increase the length of the wire on and so on. The actual resonant frequency on a
each side of the generator in Fig. 2-8B to one half harmonic is always a little higher than the exact
wavelength, we have the situation shown in Fig. multiple of the fundamental. A full-wave
2-8C. At the instant shown, current flows into the (second-harmonic) antenna, for example, must be a
generator from the left-hand half-wave section, and little longer than twice the length of a half-wave
out of the generator into the right-hand half-wave antenna.
section. Thus the currents in the two sections are Frequently it is desired to determine the
in the same direction, just as they were in Fig. electrical length of a harmonically operated wire
2-8B. The current distribution in this case obvious¬ antenna of fixed physical length for a given
ly is not the same as in Fig. 2-8A. Although the frequency. With a rearrangement of the terms of
over-all lengths of the antennas shown at A and C the above formulas, the following equation is
are the same, the antenna at A is operating on a useful for making these determinations:
harmonic but the one in C is not.
For true harmonic operation it is necessary that = fL (feet) ¿ZJmeters)
025 = 025
the power be fed into the antenna at the proper A 984 300 ’
point. Two methods that result in the proper
current distribution are shown at D and E in Fig. where X is the length of the wire in wavelengths at
2-8. If the source of power is connected to the the frequency, f, in megahertz, and L is the
antenna at one end, as in D, the direction of physical length of the wire.
current flow will be reversed in alternate half-wave
sections. Or if the power is inserted at the center of Impedance of Harmonic Antennas
a half-wave section, as in E, there will be a similar A harmonic antenna can be looked upon as a
reversal of current in the next half-wave section. series of half-wave sections placed end to end
For harmonic operation, therefore, the antenna (collinear) and supplied with power in such a way
should be fed either at the end or at a current loop. that the currents in alternate sections are out of
If the feed point is at a current node the current phase. There is a certain amount of coupling
distribution will not be that expected on a harmon¬ between adjacent half-wave sections. Because of
ic antenna. this coupling, as well as the effect of radiation
from the additional sections, the impedance as
Length of a Harmonic Wire measured at a current loop in a half-wave section is
The physical length of a harmonic antenna is not the same as the impedance at the center of a
not exactly the same as its electrical length, for the half-wave antenna.
same reasons discussed earlier in connection with Just as in the case of a half-wave antenna, the
the half-wave antenna. The physical length is impedance consists of two main components,
somewhat shorter than the length of the same radiation resistance and reactance. The ohmic or
number of half waves in space because of the loss resistance is low enough to be ignored in the
length/diameter ratio of the conductor and the end practical case. If the antenna is exactly resonant
effects. Since the latter are appreciable only where there will be no reactance at the input terminals
insulators introduce additional capacitance at a and the impedance consists only of the radiation
high-voltage point along the wire, and since a resistance. The value of the radiation resistance
harmonic antenna usually has such insulation only depends on the number of half waves on the wire
at the ends, the end-effect shortening affects only and, as in the case of the half-wave antenna, is
the half-wave sections at each end of the antenna. modified by the presence of nearby conductors
It has been found that the following formulas for and dielectrics, particularly the earth. As a point of
the length of a harmonic antenna of the usual wire departure, however, it is of interest to know the
sizes work out very well in practice: order of magnitude of the radiation resistance of a
theoretical harmonic antenna consisting of an
W^feet) = -° 5) infinitely thin conductor in free space, with its
J(MriZ) length adjusted to exact harmonic resonance. The
or radiation resistance of such an antenna having a
length of one wavelength is approximately 90
ohms, and as the antenna length is increased the
resistance also increases. At ten wavelengths it is
where N is the number of half waves on the approximately 160 ohms, for example. The way in
antenna. which the radiation resistance of a theoretical
Antenna Fundamentals 33

harmonic wire varies with length is shown by curve

A in Fig. 2-20. It is to be understood that the (A) M®-
radiation resistance is always measured at a current
loop.
When the antenna is operated at a frequency
slightly off its exact resonant frequency, reactance (B)
as well as resistance will appear at its input
terminals. In a general way, the reactance varies
with applied frequency in much the same fashion
as in the case of the half-wave antenna already
described. However, the reactance varies at a more
rapid rate as the applied frequency is varied; on a
harmonic antenna a given percentage change in
applied frequency causes a greater change in the
phase of the reflected current as related to the
outgoing current than is the case with a half-wave
antenna. This is because, in traveling the greater
length of wire in a harmonic antenna, the reflected
current gains the same amount of time in each
half-wave section, if the antenna is too short for
resonance, and these gains add up as the current Fig. 2-9 — The percentage frequency change from
travels back to the driving point. When the antenna one high-order harmonic to the next (for example,
between the 10th and 11th harmonics shown at C
is too long, the reverse occurs and the reflected and D) is much smaller than between the funda¬
current progressively drops behind in phase as it mental and second harmonic (A and B). This
travels back to the point at which the voltage is makes impedance variations more rapid as the wire
applied. This effect increases with the length of the becomes longer in terms of wavelength.
antenna, and the change of phase can be quite
rapid when the frequency applied to an antenna
operated on a high-order harmonic is varied. wire is an odd number of quarter wavelengths long,
Another way of looking at it is this: Consider then rises again with increasing frequency until it
the antenna of Fig. 2-9A, driven at the end by a reaches a new maximum when the frequency is 2/
source of power having a frequency 1 /2 f, where f This behavior of reactance and resistance is
is the fundamental or half-wave resonant frequency shown in Fig. 2-10. A similar change in reactance
of the antenna. When the frequency 1/2 / is and resistance occurs when the frequency is moved
applied, there is one quarter wavelength on the from any harmonic to the next adjacent one, as
wire, with the current distribution as shown. At well as between the fundamental and second
this frequency the antenna is resonant, and it harmonic shown in the drawing. That is, the
appears as a pure resistance of low value to the impedance goes through a cycle, starting with a
source of power because the current is large and high value of pure resistance at f then becoming
the voltage is small at the feed point. capacitive and decreasing, passing through a low
If the frequency is now increased slightly the value of pure resistance, and then becoming induc¬
antenna will be too long and the resultant current tive and increasing until it again reaches a high
at the input terminals will lag behind the applied value of pure resistance at 2/. This cycle occurs as
voltage (as explained by Fig. 2-6), and the antenna the frequency is continuously varied from any
will have inductive reactance along with resistance. harmonic to the next higher one.
As we continue to raise the frequency, the value of Looking now at Fig. 2-9D, the frequency has
reactance increases to a maximum and then de¬ been increased to 10/, ten times its original value,
creases, reaching zero when the frequency is f, such so the antenna is operated on its tenth harmonic.
that the wire is a half wavelength long, shown at B. Raising the frequency to 11/, the eleventh harmon¬
On further increasing the frequency, the reactance ic, causes the impedance of the antenna to go
becomes capacitive, increases to a maximum, and through the complete cycle described above. But
then decreases, reaching zero when the wire is an 11/ is only 10% higher than 10/ so a 10% change
odd number of quarter wavelengths long. As the in frequency has caused a complete impedance
frequency is increased still further, the reactance cycle. In contrast, changing from / to 2/isa 100%
again becomes inductive, reaches a maximum, and increase in frequency - for the same impedance
again goes to zero at frequency 2/. At this point cycle. The impedance changes ten times as fast
there are two complete standing waves of current when the frequency is varied about the 10th
(two half waves or one wavelength) and the wire is harmonic as it does when the frequency is varied
exactly resonant on its second harmonic. This last the same percentage about the fundamental.
condition is shown in Fig. 2-9C. To offset this, the actual impedance change —
In varying the frequency from 1/2 f to 2/ the that is, the ratio of the maximum to the minimum
resistance seen by the source of power also varies. impedance through the impedance cycle - is not as
This resistance increases as the frequency is raised great at the higher harmonics as it is near the
above 1/2 f and reaches a maximum when the wire fundamental. This is because the radiation resis¬
is a half wavelength long, decreases as the frequen¬ tance increases with the order of the harmonic,
cy is raised above f, reaching a minimum when the raising the minimum point on the resistance curve
34 Chapter 2

polarization as that of any one of the elements. A

system composed of a group of horizontal dipoles,
for example, will be horizontally polarized. If both
horizontal and vertical elements are used and
radiate in phase, the polarization will be the
resultant of the contributions made by each set of
elements to the total electromagnetic field at a
distance. In such a case the resultant polarization
will be linear, tilted between horizontal and verti¬
cal. If vertical and horizontal elements are fed out
of phase, where the beginning of the rf period
applied to the feed point of the vertical elements is
not in time coincidence with that applied to the
horizontal, the resultant polarization will be either
elliptical or circular. With circular polarization, the
wave front, as it appears in passing to a fixed
observer, rotates every quarter period between
vertical and horizontal, making a complete
360-degree rotation once every period. Field inten¬
Fig. 2-10 — This drawing shows qualitatively the sities are equal at all instantaneous polarizations.
way in which the reactance and resistance of an Circular polarization is frequently used for space
end-fed antenna vary as the frequency is increased communications.
from one half the fundamental (1/2 /) to the Harmonic antennas also are polarized in the
second harmonic (2/). Relative resistance values are
shown on the horizontal axis, and reactance values direction of the wire axis. However, in some
on the vertical axis; increasing frequency appears in combinations of harmonic wires such as the V and
a clockwise direction on the curve. Actual values of rhombic antennas described in a later chapter, the
resistance and reactance and the frequencies at polarization becomes elliptical in most directions
which the reactances are maximum will depend on with respect to the antenna.
the size of the conductor and the height of the As pointed out in Chapter One, sky-wave
antenna above ground. transmission usually changes the polarization of
the traveling waves. The polarization of receiving
and transmitting antennas in the 3- to 30-MHz
and also lowering the maximum point. If the curve
of Fig. 2-10 were continued through higher order range, where almost all communication is by means
harmonics, it would thus spiral inward, toward a of the sky wave (except for distances of a few
central point. This comes about because the miles), therefore need not be the same at both ends
reflected current returning to the input end of a of a communication circuit. In this range the
long harmonic wire is not as great as the outgoing choice of polarization for the antenna is usually
current since energy has been lost by radiation; this determined by factors such as the height of
is not taken into account in the theoretical pictures available antenna supports, the polarization of
of current distribution so far discussed. man-made rf noise from nearby sources, probable
The overall result, then, is that the magnitude energy losses in nearby houses, wiring, etc., and the
of the impedance variations becomes less as the likelihood of interfering with neighborhood broad¬
wire is operated at increasingly higher harmonics. cast or TV reception.
Nevertheless, the impedance reaches a maximum at
each adjacent harmonic and a minimum halfway
between, independently of the actual values of Reciprocity in Receiving
impedance. and Transmitting
The basic conditions existing when an antenna
is used for radiating power are not the same as
OTHER ANTENNA PROPERTIES when it is used for receiving a distant signal. In the

Polarization VERTICAL HORIZONTAL

POLARIZATION POLARIZATION
The polarization of a half-wave dipole is the
same as the direction of its axis. That is, at Dipole
distances far enough from the antenna for the
waves to be considered as plane waves (see Chapter
Direction of , /
One) the direction of the electric component of Electric Field '
the field is the same as the direction of the antenna
wire. Vertical and horizontal polarization, the two
most commonly used for antennas, are indicated in EARTH

Fig. 2-1 1.
Antennas composed of a number of half-wave Fig. 2-11 — Vertical and horizontal polarization of
elements with all arranged so that their axes lie in a dipole. The direction of polarization is the
the same or parallel directions will have the same direction of the electric field with respect to earth.
Antenna Fundamentals 35

former case the electromagnetic field originates tion than a horizontal antenna. On the average, an
with the antenna and the waves are not plane- antenna that transmits well in a certain direction
polarized in the immediate vicinity. In the re¬ will give favorable reception from the same direc¬
ceiving case the antenna is always far enough away tion, despite ionosphere variations.
from the transmitter so that the waves which the
antenna intercepts are plane-polarized. This causes
Pickup Efficiency
the current distribution in a receiving antenna to
be different than in a transmitting antenna except Although the transmitting and receiving proper¬
in a few special cases. These special cases, however, ties of an antenna are, in general, reciprocal, there
are those of most interest in amateur practice, is another fundamental difference between the two
since they occur when the antenna is resonant and cases that is of very great practical importance. In
is delivering power to a receiver. the transmitting case all the power supplied to the
For all practical purposes, then, the properties antenna is radiated (assuming negligible ohmic
of a resonant antenna used for reception are the resistance) regardless of the physical size of the
same as its properties in transmission. It has the antenna system. For example, a 300-MHz half¬
same directive pattern in both cases, and so will wave radiator, which is only about 50 cm (19
deliver maximum signal to the receiver when the inches') long, radiates every bit as efficiently as a
signal comes from the direction in which the 3.5-MHz half-wave antenna, which is about 41
antenna transmits best. The impedance of the meters (134 feet) long. But in receiving, the
antenna is the same, at the same point of measure¬ 300-MHz antenna does not abstract anything like
ment, in receiving as in transmitting. the amount of energy from passing waves that the
In the receiving case, the antenna is to be 3.5-MHz antenna does.
considered as the source of power delivered to the This is because the section of wave front from
receiver, rather than as the load for a source of which the antenna can draw energy extends only
power as in transmitting. Maximum output from about a quarter wavelength from the conductor. At
the receiving antenna is secured when the load to 3.5 MHz this represents an area roughly 1/2
which the antenna is connected is matched to the wavelength or 41 meters in diameter, but at 300
impedance of the antenna. Under these conditions MHz the diameter of the area is only about 1/2
half of the total power picked up by the antenna meter. Since the energy is evenly distributed
from the passing waves is delivered to the receiver throughout the wave front regardless of the wave¬
and half is reradiated into space. length, the effective area that the receiving antenna
“Impedance matching” in the case of a receiv¬ can utilize varies directly with the square of the
ing antenna does not have quite the same meaning wavelength. A 3.5-MHz half-wave antenna there¬
as in the transmitting case. This is considered in fore picks up something like 7000 times as much
later chapters. energy as a 300-MHz half-wave antenna, the field
The power gain (defined later in this chapter) in strength being the same in both cases.
receiving is the same as the gain in transmitting, The higher the frequency, consequently, the
assuming that certain conditions are met. One such less energy a receiving antenna has to work with.
condition is that both the antenna under test and This, it should be noted, does not affect the gain of
the comparison antenna (usually a half-wave anten¬ the antenna. In making gain measurements, both
na) work into load impedances matched to their the antenna under test and the comparison antenna
own impedances so that maximum power is de¬ are working at the same frequency. Both therefore
livered in both cases. In addition, the comparison are under the same handicap with respect to the
antenna should be oriented so that it gives max¬ amount of energy that can be intercepted. Thus
imum response to the signal used in the test; that the effective area of an antenna at a given
wavelength is directly proportional to its gain.
is, it should have the same polarization as the
incoming signal and should be placed so that its Although the pickup efficiency decreases rapidly
with increasing frequency, the smaller dimensions
direction of maximum gain is toward the signal
of antenna systems in the vhf and uhf regions make
source. it relatively easy to obtain high gain. This helps to
In long-distance transmission and reception via
the ionosphere the relationship between receiving overcome the loss of received signal energy.
and transmitting may not be exactly reciprocal.
This is because the waves do not take exactly the The Induction Field
same paths at all times and so may show consider¬ Throughout this chapter the fields we have
able variation in alternate transmission and recep¬ been discussing are those forming the traveling
tion. Also, when more than one layer is involved in electromagnetic waves - the waves that go long
the wave travel it is sometimes possible for distances from the antenna. These are the radiation
transmission to be good in one direction and fields. They are distinguished by the fact that their
reception to be poor in the other, over the same intensity is inversely proportional to the distance
path. In addition, the polarization of the waves is and that the electric and magnetic components,
shifted in the ionosphere, as pointed out in although perpendicular to each other in the wave
Chapter One. The tendency is for the arriving wave front, ere in phase in time. Several wavelengths
to be elliptically polarized, regardless of the polari¬ from the antenna, these are the only fields that
zation of the transmitting antenna, and a vertically need to be considered.
polarized antenna can be expected to show no Close to the antenna, however, the situation is
more difference between transmission and recep¬ much more complicated. In an ordinary electric
36 Chapter 2

circuit containing inductance or capacitance the the two types of field have equal intensity.
magnetic field is a quarter cycle out of phase (in Although the induction field is of no impor¬
time) with the electric field. The intensity of these tance insofar as effects at a distance are concerned,
fields decreases in a complex way with distance it is important when antenna elements are coupled
from the source. These are the induction fields. together, particularly when the spacing between
The induction field exists about an antenna along elements is small. Also, its existence must be kept
with the radiation field, but dies away with much in mind in making field-strength measurements
greater rapidity as the distance from the antenna is about an antenna. Error may occur if the measur¬
increased. At a distance equal to the wavelength ing equipment is set up too close to the antenna
divided by 2rr, or slightly less than 1/6 wavelength. system.

RADIATION PATTERNS AND DIRECTIVITY

A graph showing the actual or relative intensity, The solid pattern of an isotropic radiator,
at a fixed distance, as a function of the direction therefore, would be a sphere, since the field
from the antenna system is called a radiation strength is the same in all directions. In any plane
pattern. At the outset it must be realized that such containing the isotropic antenna (which may be
a pattern isa three-dimensional affair and therefore considered as a point in space, or a “point source”)
cannot be represented in a plane drawing. The the pattern is a circle with the antenna at its
“solid” radiation pattern of an antenna in free center. The isotropic antenna has the simplest
space would be found by measuring the field possible directive pattern; that is, it has no direc¬
strength at every point on the surface of an tivity at all.
imaginary sphere having the antenna at its center. An infinite variety of pattern shapes, some
The information so obtained is then used to quite complicated, is possible with actual antenna
construct a solid figure such that the distance from systems.
a fixed point (representing the antenna) to the
surface, in any direction, is proportional to the
field strength from the antenna in that direction. RADIATION FROM DIPOLES
In the analysis of antenna systems it is conve¬
THE ISOTROPIC RADIATOR
nient to make use of another fictitious type of
The radiation from a practical antenna never antenna called an elementary doublet or elementa¬
has the same intensity in all directions. The ry dipole. This is just a very short length of
intensity may even be zero in some directions from conductor, so short that it can be assumed that the
the antenna; in others it may be greater than one current is the same throughout its length. (In an
would expect from an antenna that did radiate actual antenna, it will be remembered, the current
equally well in all directions. But even though no is different all along its length.) The radiation
actual antenna radiates with equal intensity in all intensity from an elementary doublet is greatest at
directions, it is neverthless useful to assume that right angles to the line of the conductor, and
such an antenna exists. It can be used as a decreases as the direction becomes more nearly in
“measuring stick” for comparing the properties of line with the conductor until, right off the ends,
actual antenna systems. Such a hypothetical an¬ the intensity is zero. The directive pattern in a
tenna is called an isotropic radiator. single plane, one containing the conductor, is

Fig. 2-12 — At A, directive diagram of an elementary doublet in the plane

containing the wire axis. The length of each dash-line arrow represents the relative
field strength in that direction. At B, the solid pattern of the same antenna. These
same diagrams apply to any antenna considerably less than a half wavelength long.
Antenna Fundamentals 37

appearing in the pattern off the ends of the wire as

the antenna approaches a full wavelength. (The
antenna is assumed to be driven at the center, as in
Fig. 2-8B and 2-8C.) The solid pattern from a
half-wave wire is formed, just as in the case of the
doublet, by rotating the plane diagram shown in
Fig. 2-13 about the wire as an axis.
The single-plane diagrams just discussed are
actually cross sections of the solid pattern, cut by
planes in which the axis of the antenna lies. If the
solid pattern is cut by any other plane the diagram
will be different. For instance, imagine a plane
passing through the center of the wire at right
angles to the axis. The cross section of the pattern
for either the elementary doublet or the half-wave
antenna will simply be a circle in that case. This is
shown In Fig. 2-14 where the dot at the center
represents the antenna as viewed “end on,” as if
one were looking into the top of the doughnut of
Fig. 2-12B. In other words, the antenna is perpen¬
dicular to the page. This means that in any
direction in a plane at right angles to the Wire, the
Fig. 2-13 — Plane directive diagram {E plane) of a field intensity is exactly the same at the same
half-wave antenna. The solid line shows the direc¬ distance from the antenna. At right angles to the
tion of the wire, although the antenna itself is wire, then, an antenna a half wave or less in length
considered to be merely a point at the center of is nondirectional. Also, at every point on such a
the diagram. As explained in the text, a diagram circle the field is in the same phase.
such as this is simply a cross section of the solid
figure that describes the relative radiation in all
possible directions. The radial scale is purely
arbitrary and is porportional to the field strength E- AND H-PLANE PATTERNS
(voltage). This is also true of the diagram in Fig. The solid pattern of an antenna cannot ade¬
2-16. quately be shown with field-strength data on a flat
sheet of paper. For this purpose, cross-sectional or
plane diagrams are very useful. Two such diagrams,
shown in Fig. 2-12A. It consists of two tangent one in the plane containing the axis of the antenna
circles. The solid pattern is the doughnut-shaped and one in the plane perpendicular to the axis, can
figure which results when the plane shown in the give a great deal of information. The former is
drawing is rotated about the conductor as an axis, called the ¿’-plane pattern and the latter the
Fig. 2-12B. //-plane pattern. These designations are used be¬
The radiation from an elementary doublet is cause they represent the planes in which the
not uniform in all directions because there is a electric (symbol E), and the magnetic (symbol H)
definite direction to the current flow along the
lines of force lie, respectively. The E lines are taken
conductor. It will be recalled that a similar
to represent the polarization of the antenna,
condition exists in the ordinary electric and mag¬
consistent with the description of antenna polariza¬
netic fields set up when current flows along any
tion given earlier. The electromagnetic field pic¬
conductor; the field strength near a coil, for
tured in Fig. 1-1, as an example, is the field that
example, is greatest at the ends and least on the
would be radiated from a vertically polarized
outside of the coil near the middle of its length.
There is nothing strange, therefore, in the idea that
the field strength should depend on the direction
in which it is measured from the radiator.
When the antenna has appreciable length, so
that the current in every part is not the same at
any given instant, the shape of the radiation
pattern changes. In this case the pattern is the
summation of the fields from each elementary
doublet of which the antenna may be assumed to
consist, strung together in chain fashion. If the
antenna is short compared with a half wavelength
there is very little change in the pattern, but at a
half wavelength the pattern takes the shape shown
in cross section in Fig. 2-13. The intensity de¬
creases somewhat more rapidly, as the angle with Fig. 2-14 — Directive diagram of a doublet or
the wire is made smaller, than in the case of the dipole in the plane perpendicular to the wire axis
elementary doublet. If the wire length is increased (H plane). The direction of the wire is into or out
further, this tendency continues, with a wider null of the page.
38 Chapter 2

wavelengths does have some radiation at right

angles to its axis.
The greater the number of half waves in a
harmonic antenna, the larger the number of lobes
into which the pattern splits. A feature of all such
patterns is the fact that the “main” lobe — the one
that gives the largest field strength at a given
distance - always is the one that makes the
smallest angle with the antenna wire. Furthermore,
this angle becomes smaller as the length of the
antenna is increased. Fig. 2-15 shows how the angle
which the main lobe makes with the axis of the
antenna varies with the antenna length in
wavelengths. The angle shown by the solid curve is
the maximum point of the lobe; that is, the
Fig. 2-15 — Angle at which the field intensity from direction in which the field strength is greatest.
the main lobe of a harmonic antenna is maximum, The broken curve shows the angle at which the
as a function of the wire length in wavelengths. first null (the one that occurs at the smallest angle
The curve labeled "First Null" locates the angle at
which the intensity of the main lobe decreases to with the wire) appears. There is also a null in the
zero. The null marking the other boundary of the direction of the wire itself (0 degrees) and so the
main lobe always is at zero degrees with the wire total width of the main lobe is the angle between
axis. the wire and the first null. It can be seen from Fig.
2-15 that the width of the lobe decreases as the
wire becomes longer. At 1 wavelength, for exam¬
antenna; that is, an antenna in which the conduc¬ ple, it has a width of 90 degrees, but at 8
tor is mounted vertically over the earth. wavelengths the width is slightly less than 30
After a little practice, and with the exercise of a degrees.
little imagination, the complete solid pattern can A plane diagram of the radiation pattern of a
be visualized with fair accuracy from inspection of 1-wavelength harmonic wire is shown in Fig. 2-16.
just the two diagrams. Plane diagrams are plotted This is a free-space diagram in the plane containing
on polar-coordinate paper - graph paper with the wire axis (E plane), corresponding to the
radial lines marking the 360 degrees of a circle, and diagrams for the elementary doublet and half-wave
having a linear scale along the radii for marking dipole shown in Figs. 2-12 and 2-13. It is based on
amplitudes. The points on the pattern where the an infinitely thin antenna conductor with ideal
radiation is zero are called nulls, and the curved current distribution, and in a practical antenna
section from one null to the next on the plane system will be modified by the presence of the
diagram, or the corresponding section on the solid earth and other effects that will be considered
pattern, is called a lobe or ear. later.

HARMONIC-ANTENNA PATTERNS

In view of the change in radiation pattern as the

length of the antenna is increased from the
elementary doublet to the half-wave dipole, it is to
be expected that further pattern changes will occur
as the antenna is made still longer. We find, as a
matter of fact, that the patterns of harmonic
antennas differ very considerably from the pattern
of the half-wave dipole.
As explained earlier in this chapter, a harmonic
antenna consists of a series of half-wave sections
with the currents in adjacent sections always
flowing in opposite directions. This type of current
flow causes the pattern to be split up into a
number of lobes. If there is an even number of
half-waves in the harmonic antenna there is always
a null in the plane at right angles to the wire; this is
because the radiation from one half-wave section
cancels the radiation from the next, in that
particular direction. If there is an odd number of
half-waves in the antenna, the radiation from all
but one of the sections cancels itself out in the
plane perpendicular to the wire. The “left-over” Fig. 2-16 — Free-space directive diagram of a
section radiates like a half-wave dipole, and so a 1-wavelength harmonic antenna in the plane con¬
harmonic antenna with an odd number of half taining the wire axis.
Antenna Fundamentals 39

HOW PATTERNS ARE FORMED

The radiation pattern of the half-wave dipole is
found by summing up, al every point on the
surface of a sphere with the antenna at its center,
the field contributions of all the elementary
dipoles that can be imagined to make up the
full-size dipole. Antenna systems often are com¬
posed of a group of half-wave dipoles arranged in
various ways, in which case each half-wave dipole is
called an antenna element. An antenna having two
or more such dipoles is called a multielement
antenna. (A harmonic antenna can be considered
to be constructed of a number of such elements
connected in series and fed power appropriately, as
described earlier, but is not usually classed as a
multielement antenna.)
In a multielement antenna system the overall
radiation pattern is determined by the way in
which the fields at a distant point from the
separate antenna elements combine. With two
antenna elements, for example, the field strength
at a given point depends on the amplitudes and
phase relationship of the fields from each antenna.
A requirement in working out a radiation pattern is
that the field strength be measured or calculated at Fig. 2-17 — Interference between waves from two
a distant point — distant enough so that, if the separated radiators causes the resultant directional
elements carry equal currents, the field strength effects to differ from those of either radiator
from each is exactly the same even though the size alone. The two radiators shown here are separated
of the antenna system may be such that one one-half wavelength. The radiation fields of the
two cancel along the line XY but, at distances
antenna element is a little nearer the measuring which are large compared with the separation
point than another. On the other hand, this slight between the radiators, add together along line AB.
difference in distance, even though it may be only The resultant field decreases uniformly as the line
a small fraction of a wavelength in many miles, is is swung through intermediate positions from AB
very important in determining the phase relation¬ toXY.
ships between the fields from the various elements.
The principle on which the radiated fields
combine to produce the directive pattern, in the with the field denoted by the solid circles, and the
case of multielement antennas, is illustrated in the distance between the solid and broken circles is
simple example shown in Fig 2-17. In this case it is therefore one-half wavelength.
assumed that there are two antenna elements, each Similarly, the solid circles centered on the
having a circular directive pattern. The two ele¬ lower antenna locate all points at which the field
ments therefore could be half-wave dipoles ori¬ intensity from that antenna is maximum and has
ented perpendicular to the page (which gives the the same direction as the solid circles about the
plane pattern shown in Fig. 2-14). The separation upper antenna. In other words, these circles repre¬
between the two elements is assumed to be a half sent points in the same phase as the solid circles
wavelength and the currents in them are assumed around the upper antenna. The broken circle
to be equal. Furthermore, the two currents are in having the lower antenna at its center likewise
phase; that is, they reach their maximum values in locates the points of opposite phase.
the same polarity at the same instant. Considering now the fields from both antennas,
Under these conditions the fields from the two it can be seen that along the line AB the fields
antennas will be in the same phase at any point from the two always are exactly in phase, because
that is equally distant from both antenna elements. every point along AB is equally distant from both
At the instant of time selected for the drawing of antenna elements. However, along the line XY the
Fig. 2-17 the solid circles having the upper antenna field from one antenna always is out of phase with
at their centers represent, let us say, the location of the other, because every point along XY is a
all points at which the field intensity is maximum half wavelength nearer one element than the other.
and has the direction indicated by the arrowheads. It takes one-half cycle longer, therefore, for the
The distance between each pair of concentric solid field from the more distant element to reach the
circles, measured along a radius, is equal to one same point as the field from the nearer antenna,
wavelength because, as described earlier in this and thus the one field arrives 180 degrees out of
chapter, it is only at intervals of this distance that phase with the other. Since we have assumed that
the fields are in phase. The broken circle locates the points considered are sufficiently distant so
the points at which the field intensity is the same that the amplitudes of the fields from the two
as in the case of the solid circles, but is oppositely antennas are the same, the resultant field at any
directed. It is therefore 180 degrees out of phase point along XY is zero and the antenna combina-
 40
Chapter 2

tion shown will have a null in that direction.

However, the two fields add together along AB and
the field strength in that direction will be twice the
amplitude of the field from either antenna alone.
The drawing of Fig. 2-17 is not quite accurate
because it cannot be made large enough. Actually,
the two fields along AB do not have exactly the
same direction until the distance to the measuring
point is large enough, compared with the dimen¬
sions of the antenna system, so that the waves
become plane. In a drawing of limited size the
waves are necessarily represented as circles - that
is, as representations of a spherical wave. The
reader, therefore, should imagine Fig. 2-17 as being
so much enlarged that the circles crossing AB are
substantially straight lines in the region under
discussion.

Pattern Construction Fig. 2-19 — Graphical construction in the example

discussed in the text.
The drawing of Fig. 2-17 does not tell us much
about what happens to the field strength at points
that do not lie on either AB or XY although we printed page make it necessary to use the imagina¬
could make the reasonable guess that the field tion, because we assume that P is far enough from
strength at intermediate points probably would A and B so that the lines AP and BP are, for all
decrease as the point was moved along the arc of a practical purposes, parallel. When this is so, the
circle farther away from AB and nearer to XY. To distance d, between B and a perpendicular dropped
construct an actual pattern it is necessary to use a to BP from A, will be equal to the difference in
different method. It is simple in principle and can length between the distance from A to P and the
be done with a ruler, protractor and pencil, or by distance from B to P. The distance d thus measures
trigonometry. the difference in the distance the waves from A
In Fig. 2-18 the two antennas, A and B, are and B have to cover to reach P; dis also, therefore,
assumed to have circular radiation patterns, and to a measure of the difference in the time of arrival or
carry equal currents in the same phase. (In other phase of the waves at P.
words, the conditions are the same as in Fig. 2-17.) Under the assumed conditions, the relative field
The relative field strength at a distant point Pis to strengths easily can be combined graphically. The
be determined. Here again the limitations of the phase angle in degrees between the two fields at P
is equal to

~ X 360
A

where X is the wavelength and d is found by

constructing a figure similar to that shown in Fig.
2-18 for P in any desired direction. The angle 8is
the angle between a line to P and the line drawn
between the two antenna elements, and is used
simply to identify the direction of P from the
antenna system; X and d must be expressed in the
same length units.
For example, let us assume that e is 40 degrees.
We then arbitrarily choose a scale such that four
inches is equal to one wavelength — a scale large
enough for reasonable accuracy but not too large
to be unwieldy. Since the two antenna elements
are assumed to be a half-wavelength apart, we start
the drawing by placing two points two inches apart
and connecting them by a line, as shown in Fig.
2-19. Then, using B as a center and employing the
protractor, we lay off an angle of 40 degrees and
Fig. 2-18 — Graphical construction to determine
the relative phase, at a distant point, of waves draw the fine BC. The next step is to drop a
originating at two antennas, A and B. The phase is perpendicular from A to BC; this may be done
determined by the additional distance, d, that the with the 90-degree mark on the protractor but the
wave from B has to travel to reach the distant corner of an ordinary sheet of paper will do just
point. This distance will vary with the angle that about as well The distance BD is then measured,
the direction to P makes with the axis of the preferably with a ruler graduated in tenths of
antenna system, 8. inches rather than the more usual eighths. By
Antenna Fundamentals 41

actual measurement distance BD is found to be For two elements fed out of phase but having
1.53 inches. The phase difference is therefore d/x equal currents, the relationship
X 360 = 1.53/4 X 360 = 138 degrees.
The relative field strength in the direction given
by e(40 degrees in this example) is found by Field strength = 2 cos^y - ^cosdj
arbitrarily selecting a line length to represent the
strength of the field from each antenna, and then
combining them “vectorially.” One inch is a may be used, where a is the phase difference
convenient length to select. XY, Fig. 2-19, is such a between the two fed elements. Simple trigonomet¬
line, representing the strength of the field from ric equations are insufficient for determining array
antenna element A. We then measure off an angle patterns when the currents in the elements are
of 138 degrees from XY, using Yas a center, and unequal or when there are more than two ele¬
draw YZ one inch long to represent the strength ments.
and phase of the field from antenna clement B. It should be noted that the simple methods
The angle is measured off clockwise from XY described above for determining pattern shapes do
because the field from B lags behind that from A. not take mutual coupling between elements into
The distance from X to Z then represents the account, i.e., the fact that current flowing in one
relative field strength resulting from the combina¬ element will induce a voltage and therefore a
tion of the separate fields from the two antennas, resultant current into the other, and vice versa.
and measurement shows it to be approximately When mutual coupling is taken into account the
0.72 inch. In the direction 6, therefore, the field shape of the pattern remains the same for a given
strength is 72% as great as the field from either condition of element spacing and phasing, but the
antenna alone. Using trigonometry, the determina¬ magnitudes of the resultant vectors used in plotting
tion may be made by using the equation, points for various values of 0 are altered by a fixed
factor. The “fixed” factor will vary with changes in
spacing and phasing of the elements. Therefore, a
Field strength = 2 cos ( $ cos 0) direct comparison of the sizes of different patterns
obtained by these simple procedures cannot be
used for determining, say, the gain of one antenna
where 5 = spacing between elements in electrical
system over another, even though both patterns
degrees. were derived by using the same scale. Mutual
By selecting different values for eand proceed¬
coupling is covered in more detail in Chapter Four.
ing as above in each case, the complete pattern can
be determined. When 0 is 90 degrees, the phase
difference is zero and YZ and XY are simply
end-to-end along the same line. The maximum field DIRECTIVITY AND GAIN
strength is therefore twice that of either antenna
alone. When eis zero, YZ lies on top of XT (phase It has been stated that all antennas, even the
difference 180 degrees) and the distance XZ is simplest types, exhibit directive effects in that the
therefore zero; in other words the radiation from B intensity of radiation is not the same in all
cancels that from A at such an angle. directions from the antenna. This property of
The patterns of more complex antenna systems radiating more strongly in some directions than in
can readily be worked out by this method, others is called the directivity of the antenna. It
although more labor is required if the number of can be expressed quantitatively by comparing the
elements is increased. But whether or not actual solid pattern of the antenna under consideration
patterns are worked out, an understanding of the with the solid pattern of the isotropic antenna. The
method will do much to make it plain why certain field strength (and thus power per unit area or
combinations of antenna elements result in specific “power density”) will be the same everywhere on
directive patterns. the surface of an imaginary sphere having a radius
The illustration above is a very simple case, of many wavelengths and having an isotropic
but it is only a short step to systems in which the antenna at its center. At the surface of the same
antenna elements do not carry equal currents or imaginary sphere around an actual antenna radia¬
currents in the same phase. A difference in current ting the same total power, the directive pattern will
amplitude is easily handled by making the lengths result in greater power density at some points and
of lines XY and YZ proportional to the current in less at others. The ratio of the maximum power
the respective elements; if the current in B is one density to the average power density taken over
half that in A, for example, YZ would be drawn the entire sphere (the latter is the same as from the
one half as long as XY. If B's current leads the isotropic antenna under the specified conditions) is
current in A by 25 degrees, then after the angle the numerical measure of the directivity of the
determined by the distance d is found the line YZ antenna. That is,
is simply rotated 25 degrees in the counterclock¬
wise direction before measuring the distance XZ.
The rotation would be clockwise for any line
representing a lagging current. The lead or lag of
current always has to be referred to the current in where D is the directivity,/* is the power density at
one element of the system, but any desired its maximum point on the surface of the sphere,
element can be chosen as the reference. and Pav is the average power density.
42
Chapter 2

Gain the efficiency is quite high (the loss amounts only

The gain of an antenna is closely related to its to a few percent of the total), and in such cases the
directivity. Since directivity is based solely on the gain is essentially equal to the directivity.
shape of the directive pattern, it does not take into The more the directive diagram is compressed —
account any power losses that may occur in an or, in common terminology, the “sharper” the
actual antenna system. In order to determine gain lobes - the greater the power gain of the antenna.
these losses must be subtracted from the power This is a natural consequence of the fact that as
supplied to the antenna. The loss is normally a power is taken away from a larger and larger
constant percentage of the power input, so the portion of the sphere surrounding the radiator it is
antenna gain is added to the smaller and smaller volume repre¬
sented by the lobes. The power is therefore
concentrated in some directions at the expense of
others. In a general way, the smaller the volume of
the solid radiation pattern, compared with the
volume of a sphere having the same radius as the
where G is the gain expressed as a power ratio, kis
length of the largest lobe in the actual pattern, the
the efficiency (power radiated divided by power greater the power gain.
input) of the antenna, and P and Pav are as above.
Gain referred to an isotropic radiator is neces¬
For many of the antenna systems used by amateurs
sarily theoretical; that is, it has to be calculated

TABLE 2-1
Power Ratio to Decibel Conversion
Decimal Increments
Rutin ■<> 1 -2 .3 .4 .5 .6 .7 .8 .9
1 0.00 0.41 0.79 1.14 1.46 1.76 2.04 2.30 2.SS 2.79
2 3.01 3.22 3.42 3.62 3.80 3.98 4.15 4.31 4.47 4 62
3 4.77 4.91 5.05 5.19 5.32 5.44 4.56 5.68 5.80 5.91
4 6.02 6.13 6.23 6.34 6.44 6.S3 6.63 6.72 6.81 6 90
5 6.99 7.08 7.16 7.24 7.32 7.40 7.48 7.56 7.63 7.71
6 7.78 7.85 7.92 7.99 8.06 8.13 8.20 8.26 8.33 8.39
8.45 8.51 8.57 8.63 8.69 8.75 8.81 8.86 8.92 8.98
8 9.03 9.08 9.14 9.19 9.24 9.29 9.34 9.40 9.44 9 49
9 9.54 9.59 9.64 9.68 9.73 9.78 9.82 9.87 9.91 9^96
10 1000 10.04 10.09 10.13 10.17 10.21 10.2S 10.29 10.33 10 37
XIO +10
X100 +20
X1000 +30
X 10,000 +40
X 100.000 +50

Voltage Ratio to Decibel Conversion

Decimal Increments
Ratio .0 .1 .2 .3 .4 .5 .6 .7 .8 .9

1 0.00 0.83 1.58 2.28 2.92 3.52 4.08 4.61 5.11 5.58
2 6.02 6.44 6.8S 7.23 7.60 7.96 8.30 8.63 8.94 9.25
3 9.54 9.83 10.10 10.37 10.63 10.88 11.13 11.36 11.60 11.82
4 12.04 12.26 12.46 12.67 12.87 13.06 13.26 13.44 13.62 13.80
5 13.98 14.15 14.32 14.49 14.65 14.81 14.96 15.12 15.27 1S 42
6 15.56 15.71 15.85 15.99 16.12 16.26 16.39 16.52 16^65 16 78
16.90 17.03 17.15 17.27 17.38 17.50 17.62 17.73 17.84 17'95
8 18.06 18.17 18.28 18.38 18.49 18.59 18.69 18.79 18.89 18 99
9 19.08 19.18 19.28 19.37 19.46 I9.S5 19.65 19.74 19.82 19 91
10 20.00 20.09 20.17 20.26 20.34 20.42 20.51 20 59 20 67 20 75
xIO +20
xlOO +40
XIOOO +60
X 10.000 +80
X 100.000 +100

At the top, power ratio to decibel conversion, and at the bottom, voltage ratio to
decibel conversion. The decibel value is read from the body of the table for the
desired ratio, including decimal increment. For example, a power ratio of 2.6 is
equivalent to 4.15 dB. A voltage ratio of 4.3 (voltages measured across like
impedances) is equivalent to 12.67 dB. Values from the table may be extended, as
indicated at the lower left in each section. For example, a power ratio of 17, which is
the same as 10 x 1.7, is equivalent to 10 + 2.30 = 12.30 dB. Similarly, a power ratio
of 170 (100 X 1.7) = 20 + 2.30 = 22.30 dB.
Antenna Fundamentals 43

rather than measured because the isotropic radiator dB = 20 log

has no existence. In practice, measurements on the
antenna being tested usually are compared with
measurements made on a half-wave dipole. The When a voltage ratio is used both voltages must be
latter should be at the same height and have the measured in the same value of impedance. Unless
same polarization as the antenna under test, and this is done the decibel figure is meaningless,
the reference field — that from the half-wave because it is fundamentally a measure of a power
dipole comparison antenna — should be measured ratio.
Table 2-1 shows the number of decibels corres¬
in the most favored direction of the dipole. The
ponding to various power and voltage ratios. One
data can be secured either by measuring the field
advantage of the decibel is that successive power
strengths produced at the same distance from both
antennas when the same power is supplied to each, gains expressed in decibels may simply be added
together. Thus a gain of 3 dB followed by a gain of
or by measuring the power required in each
antenna to produce the same field strength at the 6 dB gives a total gain of 9 dB. In ordinary power
ratios, the ratios would have to be multiplied
same distance. together to find the total gain. Furthermore, a
A half-wave dipole has a theoretical gain of
2.14 dB over an isotropic radiator. Thus the gain of reduction in power is handled simply by subtract¬
an actual antenna over a half-wave dipole can be ing the requisite number of decibels. Thus reducing
referred to isotropic by adding 2.14 dB to the the power to 1/2 is the same as subtracting 3
measured gain, or if the gain is expressed over an decibels. We might, for example, have a power gain
Isotropic antenna it can be referred to a half-wave of 4 in one part of a system and a reduction of 1/2
dipole by subtracting 2.14 dB. in another part, so that the total power gain is 4 X
It should be noted that the field strength 1/2 = 2. In decibels, this would be 6 - 3 = 3 dB. A
(voltage) produced by an antenna at a given point power reduction or “loss” is simply indicated by
is proportional to the square root of the power. putting a negative sign in front of the appropriate
That is, when the two are expressed as ratios (the number of decibels.
usual case), pt ,—p~
Power Gains of Harmonic Antennas
Antenna patterns often are plotted in terms of In splitting off into a series of lobes, the solid
relative field strength, and if these are to be radiation pattern of a harmonic antenna is com¬
interpreted in power the field-strength ratio must pressed into a smaller volume as compared with the
be squared. For example, if the field strength is single-lobed pattern of the half-wave dipole. This
doubled, the power ratio is 2 2, or 4. means that there is a concentration of power in
certain directions with a harmonic antenna, partic¬
The Decibel ularly in the main lobe. The result is that a
harmonic antenna will produce an increase in field
As a convenience the power gain of an antenna strength, in its most favored direction, over a
system is usually expressed in decibels. The decibel half-wave dipole in its most favored direction,
is an excellent practical unit for measuring power when both antennas are supplied with the same
ratios because it is more closely related to the amount of power.
actual effect produced than the power ratio itself. The power gain from harmonic operation is
One decibel represents a just-detectable change in small when the antenna is small in terms of
signal strength, regardless of the actual value of the wavelengths, but is quite appreciable when the
signal voltage. A 20-decibel (20-dB) increase in antenna is fairly long. The theoretical power gain
signal, for example, represents 20 observable of harmonic antennas or “long wires” is shown by
“steps” in increased signal. The power ratio (100 to curve B in Fig. 2-20, using the half-wave dipole as a
1) corresponding to 20 dB would give an entirely base. A 1-wavelength or “second harmonic” anten¬
exaggerated idea of the improvement in communi¬ na has only a slight power gain, but an antenna 9
cation to be expected. The number of decibels wavelengths long will show a power increase of 5
corresponding to any power ratio is equal to 10 times over the dipole. This gain is secured in one
times the common logarithm of the power ratio, or direction by reducing or eliminating the power
P, radiated in other directions; thus the longer the
dB = 10 log p-
wire the more directive the antenna becomes.
If the voltage ratio is given, the number of decibels Curve A in Fig. 2-20 shows how the radiation
is equal to 20 times the common logarithm of the resistance, as measured at a current loop, varies
ratio. That is, with the length of a harmonic antenna.

GROUND EFFECTS
those waves that are radiated from the antenna at
REFLECTION FROM THE GROUND angles lower than the horizon. These downcoming
The performance of an antenna, particularly waves strike the surface and are reflected by a
with respect to its directive properties, is considera¬ process very similar to that by which light waves
bly modified by the presence of the earth under¬ are reflected from a mirror. As in the case of light
neath it. The earth acts like a huge reflector for waves, the angle of reflection is the same as the
44 Chapter 2

were a perfect conductor for electric currents,

reflection would take place without a change in
phase when the waves are vertically polarized.
Under similar conditions there would be a com¬
plete reversal ( 180 degrees) of phase when a
horizontally polarized wave is reflected. The actual
earth is, of course, not a perfect conductor, but is
usually assumed to be one for purposes of calculat¬
ing the vertical pattern of an antenna. The error is
small except at very low vertical angles.
As an example, when the path of the reflected
ray is exactly a half wave longer than the path of
the direct ray, the two waves will arrive out of
phase if the polarization is vertical. This corre¬
sponds to the condition illustrated in Fig. 2-17
along the line XY. However, if the path of the
reflected ray is just a wavelength longer than that
of the direct ray, the two rays arrive in phase.

Image Antennas
It is often convenient to use the concept of an
image antenna to show the effect of reflection. As
Fig. 2-21 shows, the reflected ray has the same
path length (AD equals BD) that it would if it
Fig. 2-20 — The variation in radiation resistance originated at a second antenna of the same
and power in the major lobe of long-wire antennas.
Curve A shows the change in radiation resistance
with antenna length, as measured at a current loop,
while Curve B shows the power in the lobes of
maximum radiation for long-wire antennas as a
ratio to the maximum of a half-wave antenna.

angle of incidence, so that a wave striking the

surface at an angle of, for instance, 15 degrees, is
reflected upward from the surface at the same
angle.
The reflected waves combine with the direct
waves (those radiated at angles above the horizon¬ Fig. 2-21 — At any distant point, P, the field
tal) in various ways, depending upon the orienta¬ strength will be the resultant of two rays, one
tion of the antenna with respect to earth, the direct from the antenna, the other reflected from
height of the antenna, its length, and the character¬ the ground. The reflected ray travels farther than
istics of the ground. At some vertical angles above the direct ray by the distance BC, where the
the horizontal the direct and reflected waves may reflected ray is considered to originate at the
be exactly in phase - that is, the maximum field "image" antenna.
strengths of both waves are reached at the same
time at the same spot, and the directions of the characteristics as the real antenna, but situated
fields are the same. In such a case the resultant below the ground just as far as the actual antenna
field strength is simply equal to the sum of the is above it. Like an image in a mirror, this image
two. At other vertical angles the two waves may be antenna is “in reverse,” as shown in Fig. 2-22.
completely out of phase - that is, the fields are If the real antenna is horizontal, and is instan¬
maximum at the same instant and the directions taneously charged so that one end is positive and
are opposite, at the same spot. The resultant field the other negative, then the image antenna, also
strength in that case is the difference between the horizontal, is oppositely poled; the end under the
two. At still other angles the resultant field will positively charged end of the real antenna is
have intermediate values. Thus the effect of the negative, and vice versa. Likewise, if the lower end
ground is to increase the intensity of radiation at of a half-wave vertical antenna is instantaneously
some vertical angles and to decrease it at others. positive, the end of the vertical image antenna
The effect of reflection from the ground is nearest the surface is negative. Now if we look at
shown graphically in Fig. 2-21. At a sufficiently the antenna and its image from a remote point on
large distance, two rays converging at the distant the surface of the ground, it will be obvious that
point can be considered to be parallel. However, the currents in the horizontal antenna and its
the reflected ray travels a greater distance in image are flowing in opposite directions, or are 180
reaching P than the direct ray does, and this degrees out of phase, but the currents in the
difference in path length accounts for the effect vertical antenna and its image are flowing in the
described in the preceding paragraph. If the ground same direction, or are in phase. The effect of
Antenna Fundamentals
45

to 2-41 apply to vertical antennas an odd number

of half waves long. Comparing the two sets, it is
seen that the positions of nulls (multiplying factor
zero) and maxima (multiplying factor 2) are
interchanged for the two sets of conditions.
It must be remembered that these graphs are
not plots of vertical patterns of antennas, but
represent simply multiplying factors representing
the result of reflection from the ground. With the
distinction between vertical and horizontal anten¬
Fig. 2-22 — Horizontal and vertical half-wave nas noted, the graphs apply equally well to all
antennas and their images. antennas. Also, it should be understood that they
apply at vertical angles only. The ground, if of
ground reflection, or the image antenna, is there¬ uniform characteristics, makes no distinction be¬
fore different for horizontal and vertical half-wave tween geographical directions — that is, horizontal
antennas. The physical reason for this difference is angles from the antenna - in reflecting waves.
the fact that vertically polarized waves are Fig. 2-42 shows the angles at which nulls and
reflected from a perfectly conducting earth with maxima occur as a function of the height of the
no change in phase, but that horizontally polarized antenna. This chart gives a rough idea of the
waves have their phase shifted by 180 degrees on ground-reflection pattern for heights intermediate
reflection. to those shown in detail in Figs 2-24 to 2-41. It
By using techniques similar to those discussed also facilitates picking the right height for any
earlier for determining patterns of antennas con¬ desired angle of radiation.
taining more than one element, it is possible to
derive the vertical pattern which will result from an
antenna and its image beneath the surface of the
earth. The resultant pattern is a modification of GROUND CHARACTERISTICS
the free-space pattern of the antenna. Fig. 2-23
shows how such modification takes place for a As already explained, the charts of Figs. 2-24 to
horizontal half-wave antenna. The patterns at the 2-41 are based on the assumption that the earth is
left show the relative radiation when one views the a perfect conductor. The actual ground is far from
antenna from the side; those at the right show the being “perfect” as a conductor of electricity.
radiation pattern when one looks at the end of the Actually, its behavior depends considerably on the
antenna. Changing the height from one-fourth to transmitted frequency. At low frequencies
one-half wavelength makes quite a difference in the through the standard broadcast band, for example
upward radiation — that is, the radiation at high - most types of ground do act very much like a
angles. (The radiation angle or wave angle is good conductor. At these frequencies the waves
measured from the ground up.) can penetrate for quite a distance and thus find a

Reflection Factor
The effect of reflection can be
expressed as a factor which, when
multiplied by the free-space figure
for relative intensity of radiation at a
given vertical angle from an antenna,
gives the resultant relative radiation
intensity at that same angle. The
limiting conditions are those repre¬
sented by the direct ray and reflected
ray being exactly in phase and exact¬
ly out of phase when both, assuming
there are no ground losses, have
exactly equal amplitudes. Thus the
resultant field strength at a distant
point may be either twice the field
strength from the antenna alone, or
zero.
The way in which the reflection
factor (based on perfectly conducting
ground) varies with antenna height is
shown in the series of graphs. Figs. Fig. 2-23 — Effect of the ground on the radiation
2-24 to 2-41. Figs. 2-24 to 2-35 apply from a horizontal half-wave antenna, for heights of
to horizontal antennas of any length, one-fourth and one-half wavelength. Dashed lines
and to vertical antennas an even show what the pattern would be if there were no
number of half waves long. Figs. 2-36 reflection from the ground.
46 Chapter 2

Factors by which the free-space radiation pattern of a horizontal antenna should be multiplied to include
the effect of reflection from perfectly conducting ground. These factors affect only the vertical angle of
radiation (wave angle).

large cross section in which to cause current flow shown by the charts, actually is a small fraction.
along their paths. The resistance of even a moder¬ Thus the apparent advantage of the vertical anten¬
ately good conductor will be low if its cross section na for very low-angle radiation is not realized in
is large enough. The ground acts as a fairly good practice in this frequency range.
conductor even at frequencies as high as the The “effective reflecting plane” of the ground
3.5-MHz band, and so the charts give a rather good - that is, the surface from which the reflection is
approximation of the effect of the ground at this considered to take place at the heights given in the
frequency. charts - seldom coincides with the actual surface
In the higher frequency region the penetration of the ground. Usually it will be found that this
decreases and the ground may even take on the plane appears to be a few feet below the surface; in
characteristics of a lossy dielectric rather than a other words, the height of the antenna taken for
good conductor. The chief effect of this change is purposes of estimating reflection is a few feet more
to absorb most of the energy radiated at the very than the actual height of the antenna. A great deal
low angles, in the frequency region from about 7 depends upon the character of the ground, and in
to 21 MHz. In general, the reflection factor will be some cases the reflecting plane may be “buried” a
lower than given by the charts at angles of less than surprising distance. Thus in some instances the
about 10 degrees, and it is generally considered charts will not give an accurate indication of the
that the radiation below about 3 degrees is very effect of reflection. On the average, however, they
small compared with the radiation at higher angles. will give a reasonably satisfactory representation of
This applies to both vertical and horizontal anten¬ reflection effects, with the qualifications with re¬
nas, so that the “zero-angle” reflection factor with spect to high frequencies and low angles mentioned
a vertical half-wave antenna, theoretically 2 as above.
Antenna Fundamentals 47

Factors by which the free-space radiation pattern of a horizontal antenna should be multiplied to include
the effect of reflection from perfectly conducting ground. These factors affect only the vertical angle of
radiation (wave angle).

In general, the effects of placing the antenna this frequency region. Also, there is little practical
over real earth, rather than a perfect conductor, are difference between horizontal and vertical polariza¬
to decrease the magnitude of the lobes of the tion.
pattern and to fill in the nulls. These effects are
shown in Fig. 2-43.
In the vhf and uhf region (starting in the GROUND REFLECTION AND
vicinity of the 28-MHz band) a different situa¬ RADIATION RESISTANCE
tion exists. At these frequencies little, if any, use is
made of the part of the wave that travels in contact Waves radiated from the antenna directly down¬
with the ground. The antennas, both transmitting ward reflect vertically from the ground and, in
and receiving, usually are rather high in terms of passing the antenna on their upward journey,
wavelength. The wave that is actually used - at induce a current in it. The magnitude and phase of
least for line-of-sight communication - is in most this induced current depends upon the height of
cases several wavelengths above the surface of the the antenna above the reflecting surface.
ground. At such a height there is not consequential The total current in the antenna thus consists
loss of energy; the direct ray travels from the of two components. The amplitude of the first is
transmitter to the receiver with only the normal determined by the power supplied by the transmit¬
attenuation caused by spreading, as explained in ter and the free-space radiation resistance of the
Chapter One. The loss of energy in the reflected antenna. The second component is induced in the
ray is beneficial rather than otherwise, as also antenna by the wave reflected from the ground.
explained in that chapter. The net result is that The second component, while considerably smaller
radiation at very low angles is quite practicable in than the first at most useful antenna heights, is by
 48 Chapter 2

no means inappreciable. At some heights the two conducting ground. Over actual ground the varia¬
components will be more or less in phase, so the tions will be smaller, and tend to become negligible
total current is larger than would be expected from as the height approaches a half wavelength. The
the free-space radiation resistance. At other heights antenna on which this chart is based is assumed to
the two components are out of phase, and at such have an infinitely thin conductor, and thus has a
heights the total current is the difference between somewhat higher free-space value of radiation
the two components. resistance (73 ohms) than an antenna constructed
Thus merely changing the height of the antenna of wire or tubing. (See Fig. 2-7.)
above ground will change the amount of current
flow, assuming that the power input to the antenna
is held constant. A higher current at the same value Ground Screens
of power means that the effective resistance of the The effect of a perfectly conducting ground can
antenna is lower, and vice versa. In other words, be simulated, under the antenna, by installing a
the radiation resistance of the antenna is affected metal screen or mesh (such as chicken wire) near or
by the height of the antenna above ground. Fig. on the surface of the ground. The screen preferably
2-44 shows the way in which the radiation resis¬ should extend at least a half wavelength in every
tance of a horizontal half-wave antenna varies with direction from the antenna. Such a screen will
height, in terms of wavelengths, over perfectly rather effectively establish the height of the anten-

Factors by which the free-space radiation pattern of ahalf-wave vertical antenna should be multiplied to
include the effect of reflection from perfectly conducting ground. These factors affect only the vertical
angle of radiation (wave angle).

Fig. 2-37 — Vertical dipole antenna with center 3/8 Fig. 2-40 — Vertical dipole antenna with center 1
wavelength high. wavelength high.

1-1/2 wavelengths high.

Antenna Fundamentals 49

Fig. 2-42 — Angles at which nulls and maxima (factor = 2) in the ground-reflection factor appear for
antenna heights up to two wavelengths. The solid lines are maxima, dashed lines nulls, for all horizontal
antennas and for vertical antennas having a length equal to an even multiple of one-half wavelength. For
vertical antennas an odd number of half waves long, the dashed lines are maxima and the solid lines nulls.
For example, if it is desired to have the ground reflection give maximum reinforcement of the direct ray
from a horizontal antenna at a 20-degree wave angle (angle of radiation) the antenna height should be
0.75 wavelength. The same height will give a null at 42 degrees and a second maximum at 90 degrees.
Values may also be determined from the trigonometric relationship 0 = arc sin C4/4/i), where 0 is the
wave angle and his the antenna height expressed in wavelengths. For the first maxima (horizontal
antennas), Ahas a value of 1; for the first null A has a value of 2, for the second maxima 3, for the second
null 4, and so on.

na insofar as radiation resistance is concerned, quarter wavelength or more, losses in the ground
since it substitutes for the actual earth underneath are much less important. For these types the
the antenna. considerable constructional problems are not justi¬
For vertical quarter-wave antennas the screen fied by the possible improvement.
also reduces losses in the ground near the antenna, Ground screens will affect only the very high-
since if the screen conductors are solidly bonded to angle rays from horizontal antennas, and will not
each other the resistance is much lower than the appreciably modify the effect of the earth itself at
resistance of the ground itself. With other types of the lower radiation angles which ordinarily are
antennas - e.g., horizontal — at heights of a used for long-distance communication.

DIRECTIVE DIAGRAMS AND THE WAVE ANGLE

In the discussion of radiation patterns or assumed to be merely a point at the center of the
directive diagrams of antennas it was brought out pattern, so the plane must pass through this point.)
that such patterns always are three-dimensional Now with two exceptions - surface waves at
affairs, but that it is difficult to show, on a plane low frequencies and space waves at vhf and higher
sheet of paper, more than a cross section of the — the wave angle used for communication is not
solid pattern at a time. The cross sections usually zero. In ionospheric transmission, waves sent
selected are those cut by a plane that contains the directly upward can be reflected back to earth, if
wire axis, and those cut by a plane perpendicular the frequency is low enough. On the other hand, as
to the wire axis. pointed out earlier in this chapter, in most of the
If the antenna is horizontal, the former pattern frequency range useful for ionospheric communica¬
(the cross section cut by a plane containing the tion, waves leaving at an angle of less than about 3
wire) represents the radiation pattern of the degrees are largely absorbed by ground losses. What
antenna when the wave leaves the antenna (or we are interested in at these frequencies, then, is
arrives at it) at zero angle of elevation above the the directive pattern of the antenna at a wave angle
earth. The angle of elevation - the “vertical” angle that is of value in communication.
referred to in the discussion of ground effects - is EFFECTIVE DIRECTIVE DIAGRAMS
usually called the wave angle. In the case of the
vertical antenna the radiation pattern at zero wave The directive diagram for a wave angle of zero
angle is given by the cross section cut by the plane elevation (purely horizontal radiation) does not
perpendicular to the wire axis. (In the latter case it give an accurate indication of the directive proper¬
must be remembered that the antenna itself is ties of a horizontal antenna at wave angles above
 50 Chapter 2

Fig. 2-43 — At A, a typical free-space vertical radiation pattern of a horizontal multielement array,
such as a Yagi antenna. The forward direction of the array is to the right. At B, the free-space pattern
as modified by placing the antenna one wavelength above a perfect conductor (multiplying factors
from Fig. 2-31). Because the earth is not a perfect conductor, a typical measured radiation pattern for
a 1-X-high Yagi will more nearly resemble that shown at C, calculated by assuming that only portions
of the energy striking the earth are reflected.

(C)

zero. For example, consider the half-wave dipole antenna with a cross section of its free-space
pattern in Fig. 2-13. It shows that there is no radiation pattern, cut by a plane that is vertical
radiation directly in line with the antenna itself, with respect to the earth and which contains the
and this is true at zero wave angle. However, if the axis of the antenna conductor. In other words, the
antenna is horizontal and some wave angle other view is looking broadside at the antenna wire. (For
than zero is considered, it is not true at all. the moment, reflections from the ground are
The reason why will become clear on inspection neglected.) The Unes OA, OB and OC all point in
of Fig. 2-45, which shows a horizontal half-wave the same geographical direction (the direction in
which the wire itself points), but make different
angles with the antenna in the vertical plane. In
other words, they correspond to different wave
angles or angles of radiation, with all three rays
aimed along the same line on the earth’s surface.
So far as compass directions are concerned, all
three waves are leaving the end of the antenna.
The purely horizontal wave OA has zero ampli¬
tude, but at a somewhat higher angle correspond¬
ing to the line OB the field strength is appreciable.
At a still higher angle corresponding to the line OC
the field strength is still greater. In this particular
pattern, the higher the wave angle the greater the
field strength in the compass direction OA. It
should be obvious that it is necessary, in plotting a
directive diagram that purports to show the behav¬
ior of the antenna in different compass directions,
to specify the angle of radiation for which the
diagram applies. When the antenna is horizontal
the shape of the diagram will be altered considera¬
bly as the wave angle is changed.
As described in Chapter One, the wave angles
Fig. 2-44 — Variation in radiation resistance of a that are useful depend on two things — the
horizontal half-wave antenna with height above distance over which communication is to be carried
perfectly conducting ground.
on, and the height of the ionospheric layer that
Antenna Fundamentals 51

Fig. 2-45 — The effective directive pattern of the

antenna depends upon the angle of radiation
considered. As shown by the arrows, the field
strength in a given compass direction will be quite
different at different vertical angles.

does the reflecting. Whether the E or Ft layer (or a

combination of the two) will be used depends on
the operating frequency, the time of day, season,
and the sunspot cycle. The same half-wave
antenna, operating on the same frequency, may be
almost nondirectional for distances of a few
hundred miles but will give substantially better
results broadside than off its ends at distances of
the order of 1000-1200 miles during the day when
transmission is by the E layer. In the evening, when
the Flayer takes over, the directivity may be fairly
well marked at long distances and not at all Fig. 2-46 — Directive patterns of a horizontal
pronounced at 1000 miles or less. From this it half-wave antenna at three radiation angles, 9, 15
might seem that it would be impossible to predict (solid line), and 30 degrees. The direction of the
the directivity of an antenna without all sorts of antenna itself is shown by the arrow. These
patterns are plotted to a 30-dB scale, which is
qualifications. However, it is possible to get a very about proportional to signal strength as determined
good idea of the directivity by choosing a few by ear. If 30 dB represents an S-9 signal, 0 on the
angles that, on the average, are representative for scale will be about S-3. All three patterns are
different types of work. With patterns for such plotted to the same maximum, but the actual
angles available it is fairly simple to interpolate for amplitudes at the various angles will depend upon
intermediate angles. Combined with some know¬ the antenna height, as described in the text. The
ledge of the behavior of the ionosphere, a fairly patterns shown here indicate only the shape of the
good estimate of the directive characteristics of a directive diagram as the angle is varied.
particular antenna can be made for the particular
time of day and distance of interest.
In the directive patterns given in Figs. 2-46 to
2-49, inclusive, the wave angles considered are 9,
15 and 30 degrees. These represent, respectively,
the median values of a range of angles that have
been found to be effective for communication at
28, 21, 14 and 7 MHz. Because of the variable
nature of ionospheric propagation, the patterns
should not be considered to be more than general
guides to the sort of directivity to be expected.
In the directive patterns of Figs. 2-46 to 2-49,
the relative field strength has been plotted in
decibels. This makes the patterns more representa¬
tive of the effect produced than is the case when
the relative intensity is plotted in either voltage or
power. Since one S unit on the signal-strength scale
is roughly 5 or 6 dB, it is easy to get an
approximate idea of the operation of the antenna.
For example, off the ends of a half-wave antenna
the signal can be expected to be “down” between
2 and 3 S units compared with its strength at right
angles or broadside to the antenna, at a wave angle
of 15 degrees. This would be fairly representative Fig. 2-47 — The horizontal patterns for a one¬
of its performance on 14 MHz at distances of 500 wavelength antenna at vertical angles of 9, 15 and
miles or more. With a wave angle of 30 degrees the 30 degrees.
52 Chapter 2

angle of 15 degrees the radiation in any direction is

3.5 dB above the radiation at 9 degrees, and at a
wave angle of 30 degrees it is 6 dB above. To put it
another way, at a wave angle of 30 degrees the
antenna is about one S unit better than it is at 9
degrees. There is about a half S-unit difference
between 9 and 15 degrees, and between 15 and 30
degrees. If we wished, we could add 3.5 dB to
every point on the 15-degree graph in Fig. 2-46,
and 6 dB to every point on the 30-degree graph,
and thus show graphically the comparison in
amplitude as well as shape of the directive pattern
at the three angles. This has been done in the graph
of Fig. 2-50. However, it is generally unnecessary
to take the trouble to draw separate graphs because
it is so easy to add or subtract the requisite number
of decibels as based on the appropriate ground¬
reflection factor.
It should be emphasized again that the patterns
are based on idealized conditions not realized over
actual ground. Nevertheless, they are useful in
Fig. 2-48 — The horizontal patterns for a 1-1/2- indicating about what order of effect to expect.
wavelength antenna at vertical angles of 9, 15 and
30 degrees. Using the Patterns
Directive patterns can be of considerable help
signal off the ends would be down only 1 to 2 S in solving practical problems in the choice and
units, while at an angle of 9 degrees it would be location of antennas, particularly in cases where a
down 2-1/2 to 3-1/2 S units. Since high wave angles simple type of antenna (such as the antennas
become less useful as the frequency is increased, discussed in this chapter) must be used. While it is
this illustrates the importance of running the not to be expected that antenna performance will
antenna wire in the proper direction if best results conform to the theoretical with the same exactness
are wanted in a particular direction at the higher that you would expect Ohm’s Law calculations to
frequencies. work out, the results averaged over a period of
time will be sufficiently close to the predictions to
Height Above Ground make a little preliminary estimating worthwhile.
The shapes of the directive patterns given in
Figs. 2-46 to 2-49 are not affected by the height of
the antenna above the ground. However, the
amplitude relationships between the patterns of a
given antenna for various wave angles are modified
by the height. In the figures as given, the scale is
such that the same field intensity is assumed in the
direction of maximum radiation, regardless of the
wave angle. To make best use of the patterns the
effect of the ground-reflection factor should be
included.
Take the horizontal half-wave antenna shown in
Fig. 2-46 as an example, and assume that the
antenna is a half wavelength above perfectly
conducting ground. The graph of the ground¬
reflection factor for this height is given in Fig.
2-27. For angles of 9, 15, and 30 degrees the values
of the factor as read from the curve are 1.0, 1.5
and 2.0, respectively. These factors are applied to
field strength. For convenience, take the 9-degree
angle as a reference. Then at a wave angle of 15
degrees the field strength will be 1.5 times the field
strength at 9 degrees, in any compass direction, and
at a wave angle of 30 degrees will be 2.0 times the Fig. 2-49 — The horizontal patterns for a 2-wave-
field strength at 9 degrees, in any compass direc¬ length antenna at vertical angles of 9, 15, and 30
tion. Since the factors apply to field strength degrees. As the antenna length is made still longer
(voltage), the ratios just obtained may be convert¬ the major lobes move closer to the wire, and the
ed to decibels by using the voltage section of Table number of minor lobes increases. The total number
2-1. A voltage ratio of 1.5 corresponds to 3.5 dB, of lobes on each side of the wire is numerically
equal to twice the number of wavelengths of the
and a ratio of 2.0 corresponds to 6 dB. Hence at an wire.
Antenna Fundamentals 53

Here is one example: Suppose that a clear space included for the sake of completeness. Arranging
of about 70 feet is available between two supports the information in table form gives:
that will hold the antenna about 35 feet above
ground. The operating frequency is to be 28 MHz, Antenna Length in
and the position of the supports is such that the Wavelengths
antenna will run west of north by 10 degrees. The 0.5 1 1.5 2
principal direction of transmission is to be 35 Relative intensity at 45
degrees east of north, but there is another area in degrees, dB 27 30 28 28
the general direction 15 degrees south of west that Gain from harmonic op¬
is also hoped to be covered as well as possible. The eration, dB 0 0.4 0.8 1.1
situation is shown in Fig. 2-51 (in this figure the Total dB 27 30.4 28.8 29.1
last direction is shown with reference to the
north-south line). Relative intensity at 85
In the available space, it is possible to erect degrees, dB 30 17 27 20
antennas 1/2, 1, 1-1/2 or 2 wavelengths long. Since Gain from harmonic op¬
the supports are at fixed height, the ground¬ eration, dB 0 0.4 0.8 1.1
reflection factor will be the same for all the Total dB 30 17.4 27.8 21.1
possible antennas and so may be left out of the
estimates. The principal direction is 45 degrees off It is seen that either a 1-wavelength or 2-wave-
the line of the antenna and the secondary direction length antenna will give the best results in the
is 85 degrees off. For simple antennas such as these principal direction, but that neither is as good as a
the directive patterns are symmetrical about the half-wave antenna in the secondary direction.
wire axis and so we do not have to worry about Perhaps an acceptable compromise in results could
whether the angles lie east or west of the antenna. be obtained with the 1-1/2-wavelength antenna.
However, in a case such as this, the best results
would be obtained by using two antennas since
there is room to string them end to end. A good
combination, for example, would be a 1-wave¬
length and l/2 Jwavelength antenna, arranged with
a little space between the ends so the coupling
between the pair of antennas is substantially
reduced.
Another example: Space is available to erect a
2-wavelength antenna at a height of 30 feet, for
operation on 28 MHz, and it is possible to orient
the antenna so that its major lobe will point in the
direction of transmission desired. Alternatively, a
self-supporting half-wave dipole could be erected at
a height of 45 feet and oriented so that its
maximum radiation would be in the desired direc¬
tion. Which antenna is likely to be the better one?

Fig. 2-50 — These diagrams compare the ampli¬

tudes of radiation at wave angles of 9, 15 and 30
degrees from a horizontal half-wave antenna when
the height is 1/2 wavelength.

Since the frequency is 28 MHz, the 9-degree

patterns should be used. From Fig. 2-46 we see
that the relative amplitudes at 45 and 85 degrees
are 27 and 30 dB, respectively, for the half-wave
antenna. From Fig. 2-47 the corresponding ampli¬
tudes are 30 and 17 dB; from Fig. 2-48 the
amplitudes are 28 and 27 dB; and from Fig. 2-49
the amplitudes are 28 and 20 dB. To these
amplitudes we should add the gains realized by
harmonic operation as given in Fig. 2-20. These
are, for the 1/2-, 1-, 1-1/2- and 2-wavelength
antennas, respectively, 1, 1.1, 1.2 and 1.3 in
power. Converted to decibels by using Table 2-1
they are 0, 0.4, 0.8 and 1.1 dB respectively. They
are small enough to be less important than the
probable error in reading the charts, but will be Fig. 2-51 — Example discussed in the text.
 54 Chapter 2

From Fig. 2-20, the power gain of a 2-wave-

length antenna over a dipole is 1.3, and from Table
2-1 this power ratio corresponds to a gain of 1.1
dB. At 28 MHz the length of one wavelength in ah¬
is 984/28 « 35 feet. At 30 feet the height in
wavelengths is 30/35 = 0.86k, and at 45 feet it is
45/35 = 1.3k. (It is not necessary to carry out the
calculations to more than two significant figures
because the height above the effective ground
plane is not known. As explained earlier, the
“effective” height would tend to be higher than
the actual height.) A height of 0.86k is near
enough to 7/8k to permit us to use Fig. 2-30, in Fig. 2-52 — Gain or loss in decibels because of
which the ground-reflection factor is shown as 1.5 change in antenna current with radiation resis¬
at a wave angle of 9 degrees. From Table 2-1 the tance, for fixed power input. Perfectly conducting
corresponding figure is 3.5 dB. A height of 1.3k is ground is assumed.
slightly over 1-1/4k, and inspecting Figs. 2-32 and
2-33 shows that the reflection factor therefore will
he in the vicinity of 1.9, or about 5.6 dB. Putting from the actual case. What it means is that with an
these figures into table form, we have: antenna of given dimensions, a change that reduces
the radiation resistance in the right way will be
Dipole 2k Antenna accompanied by a change in the directive pattern
that in turn will increase the field strength in some
Relative intensity of main directions at the expense of reduced field strength
lobe, dB 30 30 in other directions. This principle is used in certain
Gain from harmonic op¬ types of directive systems described in detail in
eration, dB 0 1.1 Chapter Four.
Ground-reflection factor, The shape of the directional pattern in the
dB 5.6 3.5 vertical plane is, as previously described, modified
Total dB 35.6 34.6 by the height of the antenna above ground. The
effect of height on radiation resistance has been
shown in Fig. 2-44 for the horizontal half-wave
The difference, 1 dB, is in favor of the dipole dipole. The plots of ground-reflection factors
antenna, but for all practical purposes the two shown in Figs. 2-24 to 2-35, inclusive, show the
antennas could be considered equally good in the actual shape of the pattern of such a half-wave
desired direction; the additional height of the dipole in the vertical plane at right angles to the
dipole overcomes the gain of the harmonic anten¬ wire. That is, they show the variation in intensity
na. The choice therefore could be based purely on with wave angle in the direction broadside to the
other considerations such as convenience in erec¬ antenna. In an approximate way, the radiation
tion, the fact that the dipole antenna will be resistance is smaller as the area of the pattern is
effective over a wider horizontal angle than the less, as can be seen by comparing the ground
harmonic antenna, or perhaps the desire to mini¬ reflection patterns with the curve of Fig. 2-44.
mize the effectiveness in unwanted directions. Varying the height of a horizontal half-wave
Once again it must be emphasized that calcula¬ antenna while the power input is held constant will
tions such as these should not be taken too cause the current in the antenna to vary as its
literally. Too many factors, particularly the be¬ radiation resistance changes. Under the idealized
havior of the ground, are unknown. The calcula¬ conditions represented in Fig. 2-44 (an infinitely
tions are useful principally as a guide to determin¬ thin conductor over perfectly conducting ground)
ing the type of antenna that, in all probability, will the field intensity at the optimum wave angle for
best meet the required working conditions.

RADIATION RESISTANCE AND GAIN

The field strength produced at a distant point

by a given antenna system is directly proportional
to the current flowing in the antenna. In turn, the
amount of current that will flow, when a fixed
amount of power is applied, will be inversely
proportional to the square root of the radiation
resistance. Lowering the radiation resistance will
increase the field strength and raising the radiation
resistance will decrease it This is not to be Fig. 2-53 — Solid curve, relative intensity vs. height
interpreted broadly as meaning that a low value of at a wave angle of 15 degrees, because of reflection
from perfect ground. Broken curve, height and
radiation resistance is good and a high value is bad, effect of change in radiation resistance (Fig. 2-52)
regardless of circumstances, because that is far combined.
Antenna Fundamentals 55

each height will vary as shown in Fig. 2-52. In this wavelength that is very nearly as good as the next
figure the relative field intensity is expressed in maximum at a height of 1 to 1-1/4 wavelength. The
decibels, using the field when the radiation resis¬ change in radiation resistance also has the effect of
tance is 73 ohms as a reference (0 dB). From this steepening the curve at the lower heights and
cause alone, there is a gain of about 1 dB when the flattening it in the optimum region. Thus it would
antenna height is 5/8 wavelength as compared with be expected that, for this wave angle, increasing
either 1/2 or 3/4 wavelength. the height of a half-wave dipole is very much
The gain or loss from the change in radiation worthwhile up to about 5/8 wavelength, but that
resistance should be combined with the reflection further increases would not result in any material
factor for the particular wave angle and antenna improvement. At 14 MHz, where a 15-degree wave
height considered, in judging the overall effect of angle is taken to be average, 5/8 wavelength is
height on performance. For example, Fig. 2-53 about 45 feet.
shows the reflection factor, plotted in decibels, for There is, of course, some difficulty in applying
a wave angle of 15 degrees (solid curve). This curve the information obtained in this fashion because of
is based on data from Figs. 2-24 to 2-35, inclusive. the uncertainty as to just where the ground plane
Taken alone, it would indicate that a height of is. One possibility, if the antenna can be raised and
slightly less than 1 wavelength is optimum for this lowered conveniently, is to measure the current in
wave angle. However, when the values taken from it while changing its height, keeping the power
the curve of Fig. 2-52 are added, the broken curve input constant. Starting with low heights, the
results. Because of the change in radiation resis¬ current should first go through a minimum (at a
tance, there is a maximum near a height of 5/8 theoretical height of about 3/8 wavelength) and

wavelength. wavelength.

Fig. 2-58 — In direction of wire; height 3/4

wavelength.
56 Chapter 2

Vertical-plane radiation patterns of horizontal half-wave antennas above perfectly conducting ground.

wavelength.

4*3 I I 0 I 2 34
Fig. 2-62 — In direction of wire; height 1-1/4 Fig. 2-63 — At right angles to wire; height 1-1/4
wavelength. wavelength.

Fig. 2-64 — In direction of wire; height 1-1/2

wavelength. wavelength.

then increase to a maximum as the height is antenna height for this wave angle is 1/4 wave¬
increased. The height at which this maximum is length. However, it is in the region below 1/4
obtained is the optimum. wavelength that the radiation resistance decreases
It should be kept in mind that no one wave most rapidly. At a height of 1/8 wavelength there
angle does all the work. Designing for optimum is a gain of 3.5 dB over a height of 1/4 wavelength
results under average conditions does not mean (Fig. 2-52) because of lowered radiation resistance.
that best results will be secured for all types of To offset this, the ground-reflection factor for a
work and under all conditions. For long-distance wave angle of 60 degrees is about 1.25 at 1/8
work, for example, it is best to try for the lowest wavelength (Fig. 2-24) as compared with 2.0 for
possible angle - 10 degrees or less is better for 1/4 wavelength; this is a loss of 4 dB. There is thus
multihop propagation at 14 MHz, for example. a difference of only 1-1/2 dB, which is not
However, an antenna that radiates well at such low observable, between 1/8 and 1/4 wavelength. At
angles may not be as good for work over shorter 3.5 MHz, this is a considerable difference in actual
distances as one having a broader lobe in the height, since 1/8X is about 35 feet and 1/4X is
vertical plane. about 70 feet. For short-distance work the cost of
The effect of radiation resistance is somewhat the supports required for the greater height would
more marked at the lower frequencies. To cover a not be justified.
distance of 200 miles at night (F-layer propaga¬ Information on the variation in radiation resis¬
tion) requires a wave angle of 60 degrees. As shown tance with height for antenna types other than the
by the patterns of Figs. 2-24 to 2-26, optimum half-wave dipole is not readily available. A harmon-
Antenna Fundamentals 57

ic antenna can be expected to show such variations, A series of such patterns for a horizontal
but in general an antenna system that tends to half-wave dipole at different heights is given in
minimize the radiation directly toward the ground Figs. 2-54 to 2-65, inclusive. It will be noted that
under the antenna can be expected to have a lesser the patterns of Figs. 2-54 through 2-57 are the
order of variation in radiation resistance with same as those shown in Fig. 2-23, while those
height than is the case with the half-wave dipole. shown at right angles to the wire are the same (for
the appropriate antenna height) as Figs. 2-24 to
VERTICAL DIRECTIVITY PATTERNS 2-35, inclusive. The scale here is simply an arbi¬
trary one in which the length of a radius drawn
It was explained in the preceding section that
from the origin to any point on the graph is
the directive patterns of Fig. 2-46 to 2-50, inclu¬
proportional to field strength (voltage). The reduc¬
sive, show the relative intensity of radiation in
tion in field strength off the ends of the wire at the
different compass directions for each wave angle
lower angles, as compared with the broadside field
selected, but do not attempt to show the ampli¬
strength, is quite apparent. It is also apparent that,
tude relationship between the wave angles. This is
at some heights, the high-angle radiation off the
because the intensity at different wave angles varies
with the height of the antenna above the ground, ends is nearly as great.as the broadside radiation.
and an extremely large number of diagrams would In vertical planes making some angle intermedi¬
be needed to represent the range of heights and ate between 0 and 90 degrees with the wire axis,
lengths of antennas encountered in practice. The the pattern will have a shape intermediate between
information on relative intensity at different wave the two planes shown. By visualizing a smooth
angles is easily secured from the ground-reflection transition from the end-on pattern to the broadside
charts. pattern as the horizontal angle Is varied from 0 to
However, it is helpful in forming a picture of 90 degrees a fairly good mental picture of the
the operation of antennas to use a form of actual solid pattern can be formed.
representation in which the vertical directional In the case of a vertical half-wave dipole, the
characteristic is shown for different heights. Inas¬ horizontal directional pattern is simply a circle at
much as we are still confronted by a three- any wave angle (although the actual field strength
dimensional pattern, it is only possible to do this will vary, at the different wave angles, with the
for selected vertical planes oriented in various height above ground). Hence one vertical pattern is
directions with respect to the antenna axis. In the sufficient to give complete information, for a
case of the horizontal half-wave dipole a plane selected antenna height, about the antenna in any
running in a direction along the axis and another direction with respect to the wire. A series of such
broadside to the antenna will give a good deal of patterns is given in Figs. 2-66 to 2-69, inclusive.
information. These patterns are formed by multiplying one lobe

Vertical-plane radiation patterns of vertical half-wave antennas above perfectly conducting ground. The
height is that of the center of the antenna. Dotted lines indicate approximate effect of attenuation of the
very low-angle radiation because of ground losses.

Fig. 2-67 — Height 1/2 wavelength. Fig. 2-69 — Height 1 wavelength.

58 Chapter 2

of the free-space pattern of a half-wave dipole by

the ground-reflection factor that applies at each
wave angle for the antenna height selected, to
obtain the resultant relative field strength at each
wave angle. The solid pattern in each case is
formed by rotating the plane pattern about the
90-degree axis of the graph.
The effect of ground losses at high frequencies
is simulated by the broken curves at the very low
wave angles. In other respects the curves are based
on the assumption that the antenna is erected over
perfectly conducting ground.

SOME PRACTICAL CONSIDERATIONS

At the risk of being repetitious, we must state
again that the results from a practical antenna
cannot be expected to be exactly according to the
theoretical performance outlined in this chapter. Fig. 2-70 — The effect of feeding an antenna at the
The theory that leads to the impedances, radiation end is to cause a tilt to the directional pattern, as
patterns, and power-gain figures discussed is neces¬ shown by these experimentally determined pat¬
terns. A, half wave; B, 1 wavelength; C, 1-1/2
sarily based on idealized assumptions that cannot wavelengths; D, 2 wavelengths. Solid patterns are
be exactly realized, although they may be ap¬ theoretical, dotted patterns experimental. In each
proached, in practice. case the antenna is fed from the left-hand end.
The effect of imperfectly conducting earth has
been mentioned several times. It will cause the does not have the perfect symmetry indicated in
actual radiation resistance of an antenna to differ the drawings in this chapter. The lobes pointing
somewhat from the theoretical figure at a given away from the end at which the antenna is fed are
height. In addition, there is the effect of the tilted somewhat toward the direction of the
length/diameter ratio of the conductor to be antenna wire, and the lobes pointing toward the
considered. Nevertheless, the theoretical figure will fed end are tilted away from the wire. The latter
approximate the actual radiation resistance closely also have less amplitude than the former. Typical
enough for most practical work. The value of measured patterns are shown in Fig. 2-70. There is
radiation resistance is of principal importance in even a tilt to the pattern of a half-wave antenna
determining the proper method for feeding power when it is fed at one end; however, when such an
to the antenna through a transmission line, and a antenna is fed at the center the pattern is symmet¬
variation of 10 or even 20 percent will not be rical.
serious. Adjustments can easily be made to com¬ Finally, the effect of nearby conductors and
pensate for the discrepancy between practice and dielectrics cannot, of necessity, be included in the
theory. theoretical patterns. Conductors such as power and
So far as radiation patterns are concerned, the telephone Unes, house wiring, piping, etc., close to
effect of imperfect earth is to decrease the ampli¬ the antenna can cause considerable distortion of
tude of the reflected ray and to introduce some the pattern if currents of appreciable magnitude
phase shift on reflection. The phase shift is are induced in them. Under similar conditions they
generally small with horizontal polarization. Both can also have a marked effect on the radiation
effects combine to make the maximum reflection resistance. Poor dielectrics such as green foliage
factor somewhat less than 2, and to prevent near the antenna can introduce loss, and may make
complete cancellation of radiation in the nulls in a noticeable difference between summer and win¬
the theoretical patterns, as shown in Fig. 2-43C. ter performance.
There may also be a slight change in the wave angle The directional effects of an antenna will
at which maximum reinforcement occurs, as a conform more closely to theory if the antenna is
result of the phase shift. The effect of ground located in clear space, at least a half wavelength
losses on very low angles already has been empha¬ from anything that might affect its properties. In
sized.
Aside from ground effects, the theoretical
Fig. 2-71 — The
patterns of the antennas discussed are developed half-wave anten¬
on the basis of sinusoidal distribution of current na and its
along the antenna, and on the assumption (in grounded quar¬
harmonic antennas) that the value of the current is ter-wave coun¬
the same at every current loop. Neither is strictly terpart. The
true. In particular, the current in a long harmonic missing quarter
antenna is not the same at every loop because some wavelength can
energy is lost all along the antenna by radiation. be considered to
be supplied by
This affects both the current going out and the the image in
current returning after reflection at the far end of ground of good
the antenna. The result is that the radiation pattern conductivity.
Antenna Fundamentals 59

cities, it may be difficult to find such a space at masts and guy wires usually have relatively little
low frequencies. The worst condition arises when effect on the performance of horizontal antennas
nearby wires or piping happen to be resonant, or because, being vertical or nearly so, they do not
nearly so, at the operating frequency. Such reso¬ pick up much energy from a horizontally polarized
nances often can be destroyed by bonding pipes or wave. In considering nearby conductors, too, the
BX coverings at trial points, checking with a transmission line that feeds the antenna should not
diode-detector wavemeter to determine the mea¬ be overlooked. Under some conditions that are
sures necessary to reduce the induced current. rather typical with amateur antennas, currents will
Metal masts or guy wires can cause distortion of be induced in the line by the antenna, leading to
the pattern unless detuned by grounding or by some undesirable effects. This is considered in
breaking up the wires with insulators. However, Chapter Three.

SPECIAL ANTENNA TYPES

The underlying principles of antenna operation current and voltage distibution are shown in Fig.
have been discussed in this chapter in terms of the 2-7 2A. When the antenna is shorter than a quarter
half-wave dipole, which is the elementary form wave but is loaded to resonance, the current and
from which more elaborate antenna systems are voltage distribution are part sine waves along the
built. However, there are other types of antennas antenna wire. If the loading coil is substantially
that find some application in amateur work, free from distributed capacitance, the voltage
particularly when space limitations do not permit across it will increase uniformly from minimum at
using a full-sized dipole. These include, principally, the ground, as shown at B and C, while the current
grounded antennas and loops. will be the same throughout.
Extremely short antennas are used, of neces¬
THE GROUNDED ANTENNA sity, in mobile work on the lower frequencies such
as the 3.5-MHz band. These may be “base loaded”
In cases where vertical polarization is required
as shown at B and C in Fig. 2-72, but there is a
— for example, when a low wave angle is desired at
small advantage to be realized by placing the
frequencies below 4 MHz - the antenna must be
loading coil at the center of the antenna. In neither
vertical. At low frequencies the height of a vertical
case, however, is the current uniform throughout
half-wave antenna would be beyond the construc¬
the coil, since the inductance required is so large
tional reach of most amateurs. A 3.5-MHz vertical
that the coil tends to act like a Unear circuit rather
half wave would be 133 feet high, for instance.
than Uke a “lumped” inductance.
However, if the lower end of the antenna is If the antenna height is greater than a quarter
grounded it need be only a quarter wave high to wavelength the antenna shows inductive reactance
resonate at the same frequency as an ungrounded at its terminals and can be tuned to resonance by
half-wave antenna. The operation can be under¬ means of a capacitance of the proper value. This
stood when it is remembered that ground having
high conductivity acts as an electrical mirror, and is shown in Fig. 2-73A. As the length is increased
progressively from 1/4 to 1/2 wavelength the
the missing half of the antenna is supplied by the current loop moves up the antenna, always being at
mirror image. This is shown in Fig. 2-71. a point 1/4 wavelength from the top. When the
The directional characteristic of a grounded height is 1/2 wavelength the current distribution is
quarter-wave antenna will be the same as that of a as shown at B in Fig. 2-73. There is a voltage loop
half-wave antenna in free space. Thus a vertical (current node) at the base, and power can be
grounded quarter-wave antenna will have a circular
applied to the antenna through a parallel-tuned
radiation pattern in the horizontal plane. In the
circuit, resonant at the same frequency as the
vertical plane the radiation will decrease from
antenna, as shown in the figure.
maximum along the ground to zero directly over¬
head. Up to a little more than 1/2 wavelength,
The grounded antenna may be
much smaller than a quarter wave¬
length and still be made resonant by
“loading” it with inductance at the
base, as in Fig. 2-72 at B and C. By
adjusting the inductance of the load¬
ing coil even very short wires can be
tuned to resonance.
The current along a grounded
quarter-wave vertical wire varies prac¬
tically sinusoidally, as is the case with
a half-wave wire, and is highest at the
ground connection. The rf voltage, Fig. 2-72 — Current and voltage distribution on a grounded
however, is highest at the open end quarter-wave antenna (A) and on successively shorter anten¬
and minimum at the ground. The nas loaded to resonate at the same frequency.
60 Chapter 2

The antenna is approximately self-resonant at a

height of 90 degrees (1/4 wavelength). The actual
resonant length will be somewhat less because of
the length/diameter ratio mentioned earlier.
In the range of heights covered by Fig. 2-74 the
radiation resistance is practically independent of
the length/diameter ratio. At greater heights the
length/diameter ratio is important in determining
the actual value of radiation resistance. At a height
of 1/2 wavelength the radiation resistance may be
as high as several thousand ohms.
The variation in radiation resistance with
Fig. 2-73 — Current and voltage distribution on heights below 60 degrees is shown in Fig. 2-75. The
grounded antennas longer than 1/4 wavelength.
(A) between 1/4 and 3/8 wave, approximately; (B) values in this range are very low.
half wave.

increasing the height compresses the directive

pattern in the vertical plane; this results in an
increase in field strength for a given power input
at the very low radiation angles. The theoretical
improvement is about 2 dB for a half-wave antenna
as compared with a quarter-wave antenna.

Radiation Resistance
The radiation resistance of a grounded vertical
antenna, as measured between the base of the
antenna and ground, varies as shown in Fig. 2-74 as
a function of the antenna height. The word
“height” as used in this connection has the same
meaning as “length” as applied to a horizontal
antenna. This curve is for an antenna based on (but
not directly connected to) ground of perfect
conductivity. The height is given in electrical
degrees, the 60-135 degree range shown corre¬ Fig. 2-75 — Same as Fig. 2-74, for heights below 60
degrees.
sponding to heights from 1/6 to 3/8 wavelength.
A very approximate curve of reactance vs.
height is given in Fig. 2-76. The actual reactance
will depend on the length/diameter ratio, so this
curve should be used only as a rough guide. It is
based on a ratio of about 1000 to 1. Thicker
antennas can be expected to show lower reactance
at a given height, and thinner antennas should
show more. At heights below and above the range
covered by the curve, larger reactance values will
be encountered, except for heights in the vicinity
of 1/2 wavelength. In this region the reactance
decreases, becoming zero when the antenna is
resonant.
Efficiency
The efficiency of the antenna is the ratio of the
radiation resistance to the total resistance of the
system. The total resistance includes radiation
resistance, resistance in conductors and dielectrics,
including the resistance of loading coils if used, and
the resistance of the grounding system, usually
referred to as “ground resistance.”
It was stated earlier in this chapter that a
half-wave antenna operates at very high efficiency
Fig. 2-74 — Radiation resistance vs. free-space because the conductor resistance is negligible com¬
antenna height in electrical degrees for a vertical
antenna over perfectly conducting ground, or over pared with the radiation resistance. In the case of
a highly conducting ground plane. This curve also the grounded antenna the ground resistance usually
may be used for center-fed antennas (in free space) is not negligible, and if the antenna is short
by multiplying the radiation resistance by two; the (compared with a quarter wavelength) the resis¬
height in this case is half the actual antenna length. tance of the necessary loading coil may become
Antenna Fundamentals 61

preferable to a few long radials. The conductor size

is relatively unimportant, No. 12 to No. 28 copper
wire being suitable.
The measurement of ground resistance at the
operating frequency is difficult. The power loss in
the ground depends on the current concentration
near the base of the antenna, and this depends on
the antenna height. Typical values for small radial
systems (15 or less) have been measured to be from
about 5 to 30 ohms, for antenna heights from 1/16
to 1/4 wavelength.
Additional information on ground losses is
contained in Chapter Seven.
Top Loading
Because of the difficulty of obtaining a really
Fig. 2-76 — Approximate reactance of a vertical low-resistance ground system, it is always desirable
antenna over perfectly conducting ground and to make a grounded vertical antenna as high as
having a length/diameter ratio of about 1000. possible, since this increases the radiation resis¬
Actual values will vary considerably with length/ tance. (There is no point in going beyond a half
diameter ratio. The remarks under Fig. 2-74 also wavelength, however, as the radiation resistance
apply to this curve. decreases with further increases in height.) At the
low frequencies where a grounded antenna is
appreciable. To attain an efficiency comparable generally used, the heights required for the realiza¬
with that of a half-wave antenna, in a grounded tion of high radiation resistance usually are imprac¬
antenna having a height of 1/4 wavelength or less, ticable for amateur work. The objective of design
great care must be used to reduce both ground of vertical grounded antennas which are necessarily
resistance and the resistance of any required 1/4 wavelength or less high is to make the current
loading inductors. Without a fairly elaborate loop come near the top of the antenna, and to
grounding system, the efficiency is not likely to keep the current as large as possible throughout the
exceed 50 percent and may be much less, particu¬ length of the vertical wire. This requires “top
larly at heights below 1/4 wavelength. loading,” which means replacing the missing height
by some form of electrical circuit having the same
Grounding Systems characteristics as the missing part of the antenna,
The ideal grounding system for a vertical so far as energy traveling up to the end of the
grounded antenna would consist of about 120 antenna is concerned.
wires, each at least 1/2 wavelength long, extending One method of top loading is shown in Fig.
radially from the base of the antenna and spaced 2-77. The vertical section of the antenna termi¬
equally around a circle. Such a system is the nates in a “flat-top” which supplies a capacitance
practical equivalent of perfectly conducting ground at the top into which current can flow. The simple
and has negligible resistance. The wires can either single-conductor system shown at A is more readily
be laid directly on the surface of the ground or visualized as a continuation of the antenna so that
buried a few inches below. the dimension % is essentially the overall length of
Such a system would be practical in an amateur the antenna. If this dimension is a half wavelength,
installation in very few cases. Unfortunately, the the resistance at the antenna terminals (indicated
resistance increases rapidly when the number of by the small circles, one being grounded) will be
radials is reduced, and if at all possible at least 15 high. A disadvantage of this system is that the
radials should be used. Experimental measurements horizontal portion also radiates to some extent,
have shown that even with this number the although there is cancellation of radiation in the
resistance is such as to decrease the antenna direction at right angles to the wire direction, since
efficiency to about 50 percent at a height of 1/4 the currents in the two portions are flowing in
wavelength. opposite directions.
It has also been found that as the number of
radials is reduced the length required for optimum
results with a particular number of radials also
decreases; in other words, if only a small number
of radials can be used there is no point in
extending them out a half wavelength. With 15
radials, for example, a length of 1/8 wavelength is
sufficient. With as few as two radials the length is
almost unimportant, but the efficiency of a
quarter-wave antenna with such a grounding Fig. 2-77 — Simple top loading of a vertical
system is only about 25 percent. It is considerably antenna. The antenna terminals, indicated by the
lower with shorter antennas. small circles, are the base of the antenna and
In general, a large number of radials, even ground, and should not be taken to include the
though some or all of them have to be short, is length of any lead-ins or connecting wires.
62 Chapter 2

be calculated. The inductance may be the usual

Fig. 2-78 — Top loading with type of rf coil, with suitable protection from the
lumped constants. The induc¬ weather.
tance, L, should be adjusted to
give maximum field strength The minimum value of capactive reactance
with constant power input to required depends principally upon the ground
the antenna. A parallel-tuned resistance. Fig. 2-79 is a set of curves giving the
circuit, independently re¬ reactances required under representative condi¬
Antenna
sonant at the operating fre¬ tions. These curves are based on obtaining 75
quency, is required for coup¬ percent of the maximum possible increase in field
ling to the transmitter when strength over an antenna of the same height
the top loading is adjusted to without top loading, and apply with sufficient
bring a current node at the
lower end of the antenna. accuracy to all antenna heights. An inductance coil
of reasonably low-loss construction is assumed.
The general rule is to use as large a capacitance
(low capacitive reactance) as the circumstances will
permit, since an increase in capacitance will cause
an improvement in the field strength. It is particu¬
A multiwire system such as is shown at B will larly important to do this when, as is usually the
have more capacitance than the single-conductor case, the ground resistance is not known and
arrangement and thus will not need to be as long, cannot be measured.
for resonating at a given frequency, but requires
extra supports for the additional wires. Ideally, an
arrangement of this sort should be in the form of a
cross, but parallel wires separated by several feet
will give a considerable increase in capacitance over
a single wire. With either system shown in Fig.
2-77, dimension X, the length from the base of the
antenna along one conductor to the end, should be
not more than one-half wavelength nor less than
one-fourth wavelength.
Instead of a flat top, it is possible to use a
simple vertical wire with concentrated capacitance
and inductance at its top to simulate the effect of
the missing length. The capacitance used is not the
usual type of capacitor, which would be ineffective
since the connection is one-sided, but consists of a
metallic structure large enough to have the neces¬
sary self-capacitance. Practically any sufficiently Fig. 2-80 — Capacitance of sphere, disk and
large metallic structure can be used for the cylinder as a function of their diameters. The
purpose, but simple geometric forms such as the cylinder length is assumed equal to its diameter.
sphere, cylinder and disk are preferred because of
the relative ease with which their capacitance can The capacitance of three geometric forms is
shown by the curves of Fig. 2-80 as a function of
their size. For the cylinder, the length is specified
equal to the diameter. The sphere, disk and
cylinder can be constructed from sheet metal, if
such construction is feasible, but the capacitance
will be practically the same in each case if a
“skeleton” type of construction with screening or
networks of wire or aluminum tubing is used.

Ground-Plane Antennas
Instead of being actually grounded, a 1/4-wave
antenna can work against a simulated ground called
a ground plane. Such a simulated ground can be
formed from wires 1/4 wavelength long radiating
from the base of the antenna, as shown in Fig.
2-81. It is obvious that with 1/4-wave radials the
antenna and any one radial have a total length of
1/2 wavelength and therefore will be a resonant
Fig. 2-79 — Inductive and capacitive reac¬ system. However, with only one radial the directive
tance required for top loading a grounded antenna pattern would be that of a half-wave antenna bent
by the method shown in Fig. 2-78. The reactance into a right angle at the center; if one section is
values should be converted to inductance and vertical and the other horizontal this would result
capacitance, using the ordinary formulas, at the in equal components of horizontal and vertical
operating frequency. polarization and a nonuniform pattern in the
Antenna Fundamentals 63

horizontal plane. This can be overcome by using a

ground plane in the shape of a disk with a radius of
1/4 wavelength. The effect of the disk can be
simulated, with simpler construction, by using at
least four straight radials equally spaced around the
circle, as indicated in the drawing.
The ground-plane antenna is widely used at vlif,
for the purpose of establishing a “ground” for a
vertical antenna mounted many wavelengths above
actual ground. This prevents a metallic antenna
support from carrying currents that tend to turn
the system into the equivalent of a vertical
long-wire antenna and thus raising the wave angle.
At frequencies of the order of 14 to 30 MHz, a
ground plane of the type shown in Fig. 2-81 will
permit using a quarter-wave vertical antenna (for
nondirectional low-angle operation) at a height Fig. 2-81 — The ground-plane antenna. Power is
that will let the antenna be clear of its surround¬ applied between the base of the antenna and the
ings. Such short antennas mounted on the ground center of the ground plane, as indicated in the
itself are frequently so surrounded by energy¬ drawing.
absorbing structures and trees as to be rather
ineffective. Since the quarter-wave radials are problem, because it can easily be “tuned out” and
physically short at these frequencies, it is quite the antenna system thereby resonated, even though
practical to mount the entire system on a roof top the antenna by itself is not resonant. The purpose
or pole. A ground plane at a height of a half of resonating either the antenna or the system as a
wavelength or more closely approximates perfectly whole is simply to facilitate feeding power to the
conducting earth, and the resistance curve of Fig. antenna, and such resonating or tuning does not
2-74 applies with reasonable accuracy. The antenna affect the antenna’s radiating properties.
itself may be any desirable height and does not Physical conditions frequently make it neces¬
have to be exactly a quarter wavelength. The sary to use antennas shorter than a half wave¬
radials, however, preferably should be quite close length. The directive pattern of a short antenna
to a quarter wave in length. does not differ greatly from that of a half-wave
A ground plane also can be of considerable antenna, and in the limit the pattern approaches
benefit at still lower frequencies provided the that shown in Fig. 2-12. The difference in the field
radials and the base of the antenna can be a strength that this shift in pattern shape causes is
moderate height above the ground. In order to take negligible. The most important difference is the
over the function of the actual ground connection, decrease in radiation resistance and its effect on
the ground plane must be so disposed that the field the efficiency of the antenna. This has been
of the antenna prefers to travel along the ground¬ discussed in the preceding section on grounded
plane wires instead of in the ground itself, thus antennas. The curve of Figs. 2-74 and 2-75 can be
confining the current to the highly conducting used for any center-fed nongrounded antenna by
wires rather than letting it flow in lossy earth. This using half the actual antenna length and multiply¬
is particularly necessary where the current is ing the corresponding radiation resistance by two.
greatest, i.e., close to the antenna. If the ground For example, a center-fed dipole antenna having an
plane has to be near the earth, the number of wires actual length of 120 degrees (1/3 wavelength) has a
should be increased, using as many as is practicable half length of 60 degrees and a radiation resistance
and spacing them as evenly as possible in a circle of about 13 ohms per side or 26 ohms for the
around the antenna. If the construction of a whole antenna. The reactance, which will be
multiwire ground plane is impracticable, a better capacitive and of the order of 400 ohms (Fig. 2-76,
plan is to bury as many radials as possible in the same technique as above), can be tuned out by a
ground as described earlier. loading coil or coils. As described earlier, low-
resistance coils must be used if the antenna
efficiency is to be kept reasonably high. However,
SHORT ANTENNAS IN GENERAL the ground resistance loss can be neglected in a
horizontal center-fed antenna of this type if the
Although the discussions in this chapter have height is a quarter wavelength or more.
principally been in terms of self-resonant antennas, A short antenna should not be made shorter
particularly those a half wavelength long, it would than the physical circumstances require, because
be a mistake to assume that there is anything the efficiency decreases rapidly as the antenna is
peculiarly sacred about resonance. The resonant made shorter. For example, a center-fed antenna
length happens to be one that is convenient to having an overall length of 1/4 wavelength (half
analyze. At other lengths the directive properties length 45 degrees) wdl have a radiation resistance
will be different, the radiation resistance will be of 2 X 7 = 14 ohms, as shown by Fig. 2-75.
different, and the impedance looking into the Depending on the length/diameter ratio, the react¬
terminals of the antenna will contain reactance as ance will be 500 to 1000 ohms. A loading coil of
well as resistance. The reactance presents no the same reactance probably will have a resistance
64 Chapter 2

of 3 to 6 ohms, so the probable efficiency will be maximum in the plane of the loop and in the
70 to 80 percent, or a loss of 1 to 1.5 dB. While direction looking from the low-current side to the
this is not too bad, further shortening not only high-current side. If the side opposite the terminals
further decreases the radiation resistance but enters is opened at the center as shown at B (strictly
a length region where the reactance increases very speaking, it is then no longer a loop because it is no
rapidly, so that the coil resistance quickly becomes longer a closed circuit) the direction of current
larger than the radiation resistance. Where the flow remains unchanged but the maximum current
antenna must be short, a small length/diameter flow occurs at the terminals. This reverses the
ratio (thick antenna) is definitely desirable as a direction of maximum radiation.
means of keeping down the reactance and thus The radiation resistance at a current antinode
reducing the size of loading inductance required. (which is also the resistance at X-Y in Fig. 2-82B)
is of the order of 50 ohms. The impedance at the
LOOP ANTENNAS terminals in A is a few thousand ohms. This can be
A loop antenna is a closed-circuit antenna - reduced by using two identical loops side by side
that is, one in which a conductor is formed into with a few inches spacing between them and
one or more turns so that its two ends are close applying power between terminal X on one loop
together. Loops can be divided into two general and terminal Yon the other.
classes, those in which both the total conductor
length and the maximum linear dimension of a
turn are very small compared with the wave¬
length, and those in which both the conductor
length and the loop dimensions begin to be
comparable with the wavelength.
A “small” loop can be considered to be simply
a rather large coil, and the current distribution in
such a loop is the same as in a coil. That is, the
current is in the same phase and has the same
amplitude in every part of the loop. To meet this
condition the total length of conductor in the loop
must not exceed about 0.8 wavelength. Small loops
are discussed further in the chapter on Specialized
Antennas.
A “large” loop is one in which the current is
not the same either in amplitude or phase in every
part of the loop. This change in current distribu¬
tion gives rise to entirely different properties as
compared with a small loop.
Fig. 2-83 — Inductive loading in the sides of a
half-wave loop to increase the directivity and gain.
Half-Wave Loops Maximum radiation or response is in the plane of
the loop in the direction shown by the arrow.
The smallest size of “large” loop generally used
is one having a conductor length of 1/2 wave¬
length. The conductor is usually formed into a Unlike a half-wave dipole or a small loop, there
square, as shown in Fig. 2-82, making each side 1/8 is no direction in which the radiation from a loop
wavelength long. When fed at the center of one of the type shown in Fig. 2-82 is zero. There is
side the current flows in a closed loop as shown at appreciable radiation in the direction perpendicular
A. The current distribution is approximately the to the plane of the loop, as well as to the “rear” -
same as on a half-wave wire, and so is maximum at the opposite direction to the arrows shown. The
the center of the side opposite the terminals X-Y, front-to-back ratio is of the order of 4 to 6 dB. The
and minimum at the terminals themselves. This small size and the shape of the directive pattern
current distribution causes the field strength to be result in a loss of about 1 dB when the field
strength in the optimum direction from such a
loop is compared with the field from a half-wave
dipole in its optimum direction.
The ratio of the forward radiation to the
backward radiation can be increased and the field
strength likewise increased at the same time to give
a gain of about 1 dB over a dipole, by using
inductive reactances to “load” the sides joining the
front and back of the loop. This is shown in Fig.
2-83. The reactances, which should have a value of
approximately 360 ohms, decrease the current in
the sides in which they are inserted and increase it
in the side having the terminals. This increases the
Fig. 2-82 — Half-wave loops, consisting of a single directivity and thus increases the efficiency of the
turn having a total length of 1/2 wavelength. loop as a radiator.
Antenna Fundamentals 65

Fig. 2-84 — At A and B,

loops having sides 1/4
wavelength long, and at
C having sides 1/3
wavelength long (total
conductor length one
wavelength). The po¬
larization depends on
the orientation of the
loop and the point at
which the terminals
X- Yare located.

One-Wavelength Loops The one-wavelength loop is more frequently

Loops in which the conductor length is one used as an element of a directive antenna array (the
wavelength have different characteristics than half¬ quad and delta-loop antennas described in later
wave loops. Three forms of one-wavelength loops chapters) than singly, although there is no reason
are shown in Fig. 2-84. At A and B the sides of the why it cannot be used alone. In the quad and delta
squares are equal to 1/4 wavelength, the difference loop, it is nearly always driven so that the
being in the point at which the terminals are polarization is horizontal.
inserted. At C the sides of the triangle are equal to
1/3 wavelength. The relative direction of current FOLDED DIPOLES
flow is as shown in the drawings. This direction In the diagram shown in Fig. 2-85, suppose for
reverses halfway around the perimeter of the loop, the moment that the upper conductor between
since such reversals always occur at the junction of points B and C is disconnected and removed. The
each half-wave section of wire. system is then a simple center-fed dipole, and the
The directional characteristics of loops of this direction of current flow along the antenna and
type are opposite in sense to those of a small loop. line at a given instant is as given by the arrows.
That is, the radiation is maximum perpendicular to Then if the upper conductor between B and C is
the plane of the loop and is minimum in any restored, the current in it will flow away from B
direction in the plane containing the loop. If the and toward C, in accordance with the rule for
three loops shown in Fig. 2-84 are mounted in a reversal of direction in alternate half-wave sections
vertical plane with the terminals at the bottom, the along a wire. However, the fact that the second
radiation is horizontally polarized. When the ter¬ wire is “folded” makes the currents in the two
minals are moved to the center of one vertical side conductors of the antenna flow in the same
in A, or to a side corner in B, the radiation is direction. Although the antenna physically resem¬
vertically polarized. If the terminals are moved to a bles a transmission line (see Chapter Three) it is
side corner in C, the polarization will be diagonal, not actually a line but is merely two conductors in
containing both vertical and horizontal compo¬ parallel. The connections at the ends of the two are
nents. assumed to be of negligible length.
In contrast to straight-wire antennas, the elec¬ A half-wave dipole formed in this way will have
trical length of the circumference of a one-wave¬ the same directional properties and total radiation
length loop is shorter than the actual length. For resistance as an ordinary dipole. However, the
loops made of wire and operating at frequencies transmission line is connected to only one of the
below 30 MHz or so, where the ratio of conductor conductors. It is therefore to be expected that the
length to wire diameter is large, the loop will be antenna will “look” different, in respect to its
close to resonance when input impedance, as viewed by the line.
r _ 1005 The effect on the impedance at the antenna
input terminals can be visualized quite readily. The
center impedance of the dipole as a whole is the
or same as the impedance of a single-conductor dipole
— that is, approximately 70 ohms. A given amount
Leng,h(m}= !^)
of power will therefore cause a definite value of

The radiation resistance of a resonant one-wave-

length loop is approximately 100 ohms, when the
ratio of conductor length to diameter is large. As
the loop dimensions are comparable with those of
a half-wave dipole, the radiation efficiency is high.
In the direction of maximum radiation (that is,
broadside to the plane of the loop, regardless of
the point at which it is fed) the one-wavelength
loop will show a small gain over a half-wave dipole.
Theoretically, this gain is about 2 dB, and measure¬ Fig. 2-85 — Direction of current flow in a folded
ments have confirmed that it is of this order. dipole and associated transmission line.
66 Chapter 2

Fig. 2-86 — The folded dipole.

current, I. In the ordinary half-wave dipole this

current flows at the junction of the line and
antenna. In the folded dipole the same total
current also flows, but is equally divided between
two conductors in parallel. The current in each
conductor is therefore Ij2. Consequently, the line
“sees” a higher impedance because it is delivering
the same power at only half the current. It is easy
to show that the new value of impedance is equal
to four times the impedance of a simple dipole. If
more wires are added in parallel the current Fig. 2-88 — Impedance step-up ratio for the
three-conductor folded dipole. The conductors
that are not directly driven must have the same
diameter, but this diameter need not be the same
as that of the driven conductor. Dimensions are
indicated in the inset.

continues to divide between them and the terminal

impedance is raised still more. This explanation is a
simplified one based on the assumption that the
conductors are close together and have the same
diameter.
The two-wire system in Fig. 2-86 is an espe¬
cially useful one because the input impedance is so
close to 300 ohms that it can be fed directly with
300-ohm Twin-Lead or open line without any
other matching arrangement.
The folded dipole has a somewhat “flatter”
impedance-vs.-frequency characteristic than a sim¬
ple dipole. That is, the reactance varies less rapidly,
as the frequency is varied on either side of
resonance, than with a single-wire antenna.

Harmonic Operation
A folded dipole will not accept power at twice
the fundamental frequency, or any even multiples
of the fundamental. At such multiples the folded
section simply acts like a continuation of the
transmission line. No other current distribution is
possible if the currents in the two conductors of
Fig. 2-87 — Impedance step-up ratio for the the actual transmission lines are to flow in opposite
two-conductor folded dipole, as a function of directions.
conductor diameters and spacing. Dimensions di, On the third and other odd multiples of the
d2 and S are shown in the inset drawing. The fundamental the current distribution is correct for
step-up ratio, r, may also be determined from: operation of the system as a folded antenna. Since
2 the radiation resistance of a 3/2-wave antenna is
not greatly different from that of a half-wave
antenna, a folded dipole can be operated on its
third harmonic.
Antenna Fundamentals 67

Multi- and Unequal-Conductor Folded Dipoles “effective diameter” of the folded dipole will lie
somewhere between the actual conductor diameter
Larger impedance ratios than 4 to 1 are and the maximum distance between conductors.
frequently desirable when the folded dipole is used The relatively large effective thickness of the
as the driven element in a directive array because antenna reduces the rate of change of reactance
the radiation resistance is frequently quite low. A with frequency, so the tuning becomes relatively
wide choice of impedance step-up ratios is available broad and the antenna length is not too critical for
by varying the relative size and spacing of the a given frequency.
conductors, and by using more than two. Fig. 2-87 Further information on the folded dipole, as
gives design information of this nature for two- pertains to feeding and matching, is contained in
conductor folded dipoles and Fig. 2-88 is a similar Chapter Three in the section, Coupling the Line to
chart for three-conductor dipoles. Fig. 2-88 the Antenna.
assumes that the three conductors are in the same
plane, and that the two that are not directly
connected to the transmission line are equally
spaced from the driven conductor.
In computing the length of a folded dipole OTHER TYPES OF ANTENNAS
using thick conductors - i.e., tubing such as is used The half-wave dipole and the few special types
in rotary beam antennas — it should be remem¬ of antennas described in this chapter form the basis
bered that the resonant length may be appreciably for practically all antenna systems in amateur use
less than that of a single-wire antenna cut for the at frequencies from the vhf region down. Other
same frequency. Besides the shortening required fundamental types of radiators are applicable at
with thick conductors, as discussed earlier in this microwaves, but they are not used at lower
chapter, the parallel conductors tend to act like the frequencies because the dimensions are such as to
boundaries of a conducting sheet of the same be wholly impracticable when the wavelength is
width as the spacing between the conductors. The measured in meters rather than centimeters.

Bibliography

Source material and more extended discussion Textbooks on Antennas

of topics covered in this chapter can be found in and Transmission Lines
the references given below.
Jasik, Antenna Engineering Handbook, McGraw-
Brown, Lewis and Epstein, “Ground Systems as a Hill Book Co., New York, N.Y.
Factor in Antenna Efficiency,” Proc. I.R.E., Johnson, Transmission Lines and Networks,
June, 1937. McGraw-Hill Book Co., New York, N.Y.
Carter, Hansell and Lindenblad, “Development of Jordan, Electromagnetic Waves and Radiation
Directive Transmitting Antennas by R.C.A. Systems, Prentice-Hall, Inc., Englewood Cliffs,
Communications, ” Proc. I.R.E., October, 1931. New Jersey.
Dome, “Increased Radiating Efficiency for Short King, The Theory of Linear Antennas, Harvard
Antennas,” QST, September, 1934. University Press, Cambridge, Mass.
Grammer, “The All-Around Radiation Characteris¬ Laport, Radio Antenna Engineering. McGraw-Hill
tics of Horizontal Antennas,” QST, November, Book Co., New York, N.Y.
1936. Kraus, Antennas, McGraw-Hill Book Co., New
Grammer, “More on the Directivity of Horizontal York, N.Y.
Antennas,” QST, March, 1937. Schelkunoff, Advanced Antenna Theory, John
King, Mimno and Wing, Transmission Lines, Anten¬ Wiley & Sons, Inc., New York, N.Y.
nas and Wave Guides, McGraw-Hill Book Co., Schelkunoff and Friis, Antenna Theory and Prac¬
New York, N.Y. tice, John Wiley & Sons, Inc., New York, N.Y.
Lindsay, “Quads and Yagis,” QST, May, 1968. Skilling, Electric Transmission Lines, McGraw-Hill
Reinartz, “Half-Wave Loop Antennas,” QST, Oc¬ Book Co., New York, N.Y.
tober, 1937. Slurzburg and Osterheld, Essentials of Radio,
Terman, Radio Engineering, McGraw-Hill Book McGraw-Hill Book Co., New York, N.Y.
Co., New York, N.Y. Southworth, Principles and Applications of Wave¬
Williams, “Radiating Characteristics of Short-Wave guide Transmission, D. Van Nostrand Co., Inc.,
Loop Aerials,” Proc. I.R.E., October, 1940. New York, N.Y.
 Chapter 3

Transmission Lines

The desirability of installing an antenna in a at every point in space only when the two
clear space, not too near buildings or power and conductors occupy the same space — an obviously
telephone lines, has been emphasized in the preced¬ impossible condition if they are to remain separate
ing chapter. On the other hand, the transmitter conductors.
that generates the rf power for driving the antenna The best that can be done is to make the two
is usually, as a matter of necessity, located at some fields cancel each other as completely as possible.
distance from the antenna terminals. The connect¬ This can be accomplished by making the distance,
ing link between the two is the rf transmission line d, between the two conductors small enough so
or feeder. Its sole purpose is to carry rf power from that the time interval during which the field from
one place to another, and to do it as efficiently as X is moving to Y is a very small part of a cycle.
possible. That is, the ratio of the power transferred When this is the case the phase difference between
by the line to the power lost in it should be as large the two fields at any given point will be so close to
as the circumstances will permit. 180 degrees that the cancellation is practically
At radio frequencies every conductor that has complete.
appreciable length compared with the wavelength
in use will radiate power. That is, every conductor
becomes an antenna. Special care must be used,
therefore, to minimize radiation from the conduc¬
tors used in rf transmission lines. Without such
care, the power radiated by the line may be much
larger than that which is lost in the resistance of
conductors and dielectrics. Power loss in resistance
is inescapable, at least to a degree, but loss by
radiation is largely avoidable.

Preventing Radiation
Radiation loss from transmission lines can be
prevented by using two conductors so arranged and
operated that the electromagnetic field from one is
balanced everywhere by an equal and opposite
field from the other. In such a case the resultant
field is zero everywhere in space; in other words,
there is no radiation.
For example, Fig. 3-1A shows two parallel Fig. 3-1 — The two basic types of transmission
conductors having currents /I and /2 flowing in lines.
opposite directions. If the current/I at point Yon
the upper conductor has the same amplitude as the Practicable values of d, the separation between
current 12 at the corresponding point X on the the two conductors, are determined by the physi¬
lower conductor, the fields set up by the two cal limitations of line construction. A separation
currents will be equal in magnitude. Because the that meets the condition of being “very small” at
two currents are flowing in opposite directions, the one frequency may be quite large at another. For
field from /I at T will be 180 degrees out of phase example, if dis six inches, the phase difference
with the field from 12 at X. However, it takes a between the two fields at Ywill be only a fraction
measurable interval of time for the field from X to of a degree if the frequency is 3500 kHz. This is
travel to Y. I f /1 and /2 are alternating currents, because a distance of six inches is such a small
the phase of the field from /I at Y will have fraction of a wavelength (one wavelength = 360
changed in such a time interval, and so at the degrees) at 3500 kHz. But at 144 MHz the phase
instant the field from X reaches Ythe two fields at difference would be 26 degrees, and at 420 MHz it
Y are not exactly 180 degrees out of phase. The would be 73 degrees. In neither of these cases
two fields will be exactly 180 degrees out of phase could the two fields be considered to “cancel”

68
Transmission Lines 69

each other. The separation must be very small in Fig. 3-3 — A current
comparison with the wavelength used; it should pulse traveling along
never exceed 1 percent of the wavelength, and a transmission line
smaller separations are desirable. at the speed of light
Transmission lines consisting of two parallel would reach the suc¬
cessive positions
conductors as in Fig. 3-1A are called parallel¬ shown at intervals
conductor lines, or open-wire lines, or two-wire of 0.1 microsecond.
lines.
A second general type of line construction is
shown in Fig. 3-1B. In this case one of the ■*-30 METERS-*
conductors is tube-shaped and encloses the other *— 0.1 y SEC.—*
conductor. This is called a coaxial line (“coax”) or
concentric line. The current flowing on the inner
conductor is balanced by an equal current flowing
in the opposite direction on the inside surface of
the outer conductor. Because of skin effect the
current on the inner surface of the tube does not 60 METERS
penetrate far enough to appear on the outer -0.2 y SEC-
surface. In fact, the total electromagnetic field
outside the coaxial line, as a result of currents
flowing on the conductors inside, always is zero
because the tube acts as a shield at radio frequen¬
cies. The separation between the inner conductor
and the outer conductor is therefore unimportant 90 METERS

from the standpoint of reducing radiation. ■0.3 ^ses-

CU RR ENT FLOW IN LONG LINES

In Fig. 3-2, imagine that the connection be¬ has reached in its travel; at this point it is present
tween the battery and the two wires is made in both wires, with the electrons moving in one
instantaneously and then broken. During the time direction in one wire and in the other direction in
the wires are in contact with the battery terminals, the other wire. If the line is infinitely long and has
electrons in wire No. 1 will be attracted toward the no resistance or other cause of energy loss, the
positive battery terminal and an equal number of pulse will travel undiminished forever.
electrons in wire No. 2 will be repelled away from Extending the example of Fig. 3-3, it is not
the negative terminal. This happens only near the hard to see that if instead of one pulse we started a
battery terminals at first, since electrical effects do whole series of them on the line at equal time
not travel at infinite speed, so some time will intervals, they would travel along the line with the
elapse before the currents become evident at more same time and distance spacing between them,
extreme parts of the wires. By ordinary standards each pulse independent of the others. In fact, each
the elapsed time is very short, since the speed of pulse could have a different amplitude, if the
travel along the wires may be almost 300,000,000 battery voltage were varied to that end. Further¬
meters per second, so it becomes necessary to more, the pulses could be so closely spaced that
measure time in millionths of a second (micro¬ they touched each other, in which case we should
seconds) rather than in more familiar time units. have current present everywhere along the line
simultaneously.

Wavelength
It follows from this that an alternating voltage
applied to the line would give rise to the sort of
current flow shown in Fig. 3-4. If the frequency of
the ac voltage is 10,000,000 hertz (cycles per
second) or 10 MHz, each cycle will occupy 0.1
microsecond, so a complete cycle of current will be
For example, suppose that the contact with the present along each 30 meters of line. This is a
battery is so short that it can be measured in a very distance of one wavelength. Any currents observed
small fraction of a microsecond. Then the “pulse” at B and D occur just one cycle later in time than
of current that flows at the battery terminals the currents at A and C. To put it another way, the
during this time can be represented by the vertical currents initiated at A and C do not appear at B
line in Fig. 3-3. At the speed of light, this pulse will and D, one wavelength away, until the applied
travel 30 meters along the line in 0.1 microsecond; voltage has had time to go through a complete
30 meters more, making a total of 60 meters, in cycle.
0.2 microsecond; a total of 90 meters in 0.3 Since the applied voltage is always changing,
microsecond, and so on for as far as the line the currents at A and C are changing in proportion.
reaches. The current does not exist all along the The current a short distance away from A and C —
wires but is only present at the point that the pulse for instance, at X and Y — is not the same as the
70 Chapter 3

current at A and C because the current at X and Y

was caused by a value of voltage that occurred a; Bi c;
slightly earlier in the cycle. This is true all along < I I —-
the line; at any instant the current anywhere along
the line from A to B and C to D is different from Tl T T T T
the current at every other point in that same
distance. The series of drawings shows how the
instantaneous currents might be distributed if we
could take snapshots of them at intervals of Fig. 3-5 — Equivalent of a transmission line in
one-quarter cycle. The current travels out from the terms of ordinary circuit constants. The values of L
input end of the line in waves. and C depend on the line construction.
At any selected point on the line the current
goes through its complete range of ac values in the
time of one cycle just as it does at the input end. light. The actual velocity is very close to that of
light in a line in which the insulation between
Therefore (if there are no losses) an ammeter
conductors is solely air. The presence of dielectrics
inserted in either conductor would read exactly the
other than air reduces the velocity, since electro¬
same current at any point along the line, because magnetic waves travel more slowly in dielectrics
the ammeter averages the current over a whole than they do in a vacuum. Because of this the
cycle. The phases of the currents at any two wavelength as measured along the line will depend
separated points would be different, but an am¬ on the velocity factor that applies in the case of
meter cannot show phase.
the particular type of line in use. (See later section
in this chapter for actual figures.) The wavelength
Velocity of Propagation in a practical line is always shorter than the
In the example above it was assumed that wavelength in free space.
energy traveled along the line with the velocity of
CHARACTERISTIC IMPEDANCE
If this is a “perfect” line — one without
resistance - a question immediately comes up:
What is the amplitude of the current in the pulse?
Will a larger voltage result in a larger current, or is
the current theoretically infinite for any applied
voltage, as we would expect from applying Ohm’s
Law to a circuit without resistance? The answer is
that the current does depend directly on the
voltage, just as though resistance were present.
The reason for this is that the current flowing
in the line is something like the charging current
that flows when a battery is connected to a
capacitor. That is, the line has capacitance. How¬
ever, it also has inductance. Both of these are
“distributed” properties. We may think of the line
as being composed of a whole series of small
inductors and capacitors connected as in Fig. 3-5,
where each coil is the inductance of an extremely
small section of wire and the capacitance is that
existing between the same two sections. Each
inductance limits the rate at which each immedi¬
ately following capacitor can be charged, and the
effect of the chain is to establish a definite
relationship between current and voltage. Thus the
line has an apparent “resistance,” called its charac¬
teristic resistance — or, a more general term, its
characteristic impedance or surge impedance. The
conventional symbol for characteristic impedance
is Zo.

TERMINATED LINES
The value of the characteristic impedance is
equal to-jL/C in a perfect line - i.e., one in which
the conductors have no resistance and there is no
leakage between them - where L and C are the
Fig. 3-4 — Instantaneous current along a transmis¬ inductance and capacitance, respectively, per unit
sion line at successive time intervals. The frequency length of line. The inductance decreases with
is such that the time of one cycle is 0.1 micro¬ increasing conductor diameter, and the capacitance
second. decreases with increasing spacing between the
Transmission Lines 71

conductors. Hence a line with large conductors

closely spaced will have relatively low characteris¬
tic impedance, while one with thin conductors
widely spaced will have high impedance. Practica¬
ble values of Zo for parallel-conductor lines range
from about 200 to 800 ohms and for typical
coaxial lines from 50 to 100 ohms.

Matched Lines
In this picture of current traveling along a
transmission line we have assumed that the line was
infinitely long. Practical lines have a definite
length, and they are connected to or terminated in Fig. 3-7 — Mismatched lines. A — termination not
a load at the “output” end, or end to which the equal to Zo; B — short-circuited line; C —
power is delivered. If the load is a pure resistance open-circuited line.
of a value equal to the characteristic impedance of
the line, Fig. 3-6, the current traveling along the Mismatched Lines
line to the load does not find conditions changed Now take the case where the terminating
in the least when it meets the load; in fact, the load resistance, R, is not equal to Zo, as in Fig. 3-7. The
just “looks like” still more transmission line of the load R no longer “looks like” more line to the
same characteristic impedance. section of line immediately adjacent. Such a line is
said to be mismatched. The more R differs from
Zo, the greater the mismatch. The power reaching
R is not totally absorbed, as it was when R was
equal to Zo, because R requires a different voltage-
to-current ratio than the one at which the power is
traveling along the line. The result is that R absorbs
Fig. 3-6 — A transmission line terminated in a only part of the power reaching it (the incident
resistive load equal to the characteristic impedance power); the remainder acts as though it had
of the line. bounced off a wall and starts back along the line
toward the source. This is reflected power and the
greater the mismatch the larger the percentage of
The reason for this can perhaps be made a little
the incident power that is reflected. In the extreme
clearer by considering it from another viewpoint.
case where R is zero (a short circuit) or infinity (an
In flowing along a transmission line, the power is
open circuit) all of the power reaching the end of
handed from one of the elementary sections in Fig.
3-5 to the next. When the line is infinitely long this the line is reflected.
Whenever there is a mismatch, power is travel¬
power transfer always goes on in one direction —
ing in both directions along the line. The voltage-
away from the source of power. From the stand¬ to-current ratio is the same for the reflected power
point of section B, Fig. 3-5, for instance, the power as for the incident power, since this ratio is
it has handed over to section C has simply determined by the Zo of the line. The voltage and
disappeared in C. So far as section B is concerned, current travel along the line in both directions in
it makes no difference whether C has absorbed the the same sort of wave motion shown in Fig. 3-4.
power itself or has in turn handed it along to more When the source of power is an ac generator, the
line. Consequently, if we substitute something for outgoing or incident voltage and the returning or
section C that has the same electrical characteris¬ reflected voltage are simultaneously present all
tics, section B will not know the difference. A pure along the line, so the actual voltage at any point
resistance equal to the characteristic impedance of along the line is the sum of the two components,
C, which is also the characteristic impedance of the
taking phase into account. The same thing is true
line, meets this condition. It absorbs all the power
just as the infinitely long line absorbs all the power of the current.
The effect of the incident and reflected compo¬
transferred by section B. nents on the behavior of the line can be under¬
A line terminated in a purely resistive load
stood more readily by considering first the two
equal to the characteristic impedance is said to be
matched. In a matched transmission line, power limiting cases - the short-circuited line and the
travels outward along the line from the source until open-circuited line. If the line is short-circuited as
it reaches the load, where it is completely ab¬ in Fig. 3-7B, the voltage at the end must be zero.
sorbed. Thus with either the infinitely long line or Thus the incident voltage must disappear suddenly
its matched counterpart the impedance presented at the short. It can do this only if the reflected
to the source of power (the line-input impedance) voltage is opposite in phase and of the same
is the same regardless of the line length. 11 is simply amplitude. This is shown by the vectors in Fig. 3-8.
equal to the characteristic impedance of the line. The current, however, does not disappear in the
The current in such a line is equal to the applied short circuit; in fact, the incident current flows
voltage divided by the characteristic impedance, through the short and there is in addition the
and the power put into it is E2 /Zo or PZo, by reflected component in phase with it and of the
same amplitude. The reflected voltage and current
Ohm’s Law.
72 Chapter 3

(A) (B) (C) the same amplitude. Vectors for the open-circuited
case are shown in Fig. 3-9.
I Where there is a finite value of resistance at the
end of the line, Fig. 3-7A, only part of the power
reaching the end of the line is reflected. That is,
the reflected voltage and current are smaller than
the incident voltage and current. If R is less than
A Zo the reflected and incident voltages are 180
degrees out of phase, just as in the case of the
short-circuited line, but the amplitudes are not
E nn/
J small equal because all of the voltage does not disappear
at R. Similarly, if R is greater than Zo the reflected
and incident currents are 180 degrees out of phase,
as they were in the open-circuited line, but all of
the current does not disappear in R so the
amplitudes of the two components are not equal.
These two cases are shown in Fig. 3-10. Note that
the resultant current and voltage are in phase in R,
since R is a pure resistance.
Fig. 3-8 — Voltage and current at the short circuit
on a short-circuited line. These vectors show how
the outgoing voltage and current (A) combine with
the reflected voltage and current (B) to result in
high current and very low voltage in the short-
circuit (C). (A)

must have the same amplitudes as the incident

voltage and current because no power is used up in
the short circuit; all the power starts back toward
the source. Reversing the phase of either the
current or voltage (but not both) will reverse the
direction of power flow; in the short-circuited case
the phase of the voltage is reversed on reflection
but the phase of the current is not.
If the line is open-circuited (Fig. 3-7C) the (B)
current must be zero at the end of the line. In this
case the reflected current is 180 degrees out of
phase with the incident current and has the same
amplitude. By reasoning similar to that used in the
short-circuited case, the reflected voltage must be
in phase with the incident voltage, and must have
Fig. 3-10 — Incident and reflected components of
voltage and current when the line is terminated in a
pure resistance not equal toZo. In the case shown,
£ the reflected components have half the amplitude
of the incident components. A - R less than Zo; B
— R greater than Zo.

Reflection Coefficient
The ratio of the reflected voltage to the
incident voltage is called the reflection coefficient
Thus

(A)
P " Et
where pis the reflection coefficient, Et is the
reflection voltage, and Ef is the incident or forward
voltage. The reflection coefficient is determined by
the relationship between the line Zo and the actual
(C) load at the terminated end of the line. For any
given line and load it is a constant if the line has
Fig. 3-9 — Voltage and current at the end of an negligible loss in itself. The coefficient can never be
open-circuited line. A — outgoing voltage and larger than 1 (which indicates that all the incident
current; B — reflected voltage and current; C — power is reflected) nor smaller than zero (indica¬
resultant. ting that the line is perfectly matched by the load).
Transmission Lines
73

If the load is purely resistive, the reflection against position along the line, graphs like those of
coefficient can be found from Fig. 3-12 will result. If we could go along the line
with a voltmeter and ammeter plotting the current
_ R — Zo and voltage at each point, we should find that the
p R + Zo data collected gave curves like these. In contrast, if
where R is the resistance of the load terminating the load matched the Zo of the line, similar
the line. In this expression pis positive if R is measurements along the the line would show that
larger than Zo and negative if R is smaller than Zo. the voltage is the same everywhere (and similarly
The change in signs accompanies the change in for the current). The mismatch between load and
phase of the reflected voltage described above. line is responsible for the variations in amplitude
which, because of their wave-like appearance, are
called standing waves.
STANDING WAVES From the earlier discussion it should be clear
As might be expected, reflection cannot occur that when R is greater than Zo, the voltage will be
at the load without some effect on the voltages and largest and the current smallest at the load. This is
currents all along the line. A detailed description just the reverse of the case shown in Fig. 3-12. In
tends to become complicated, and what happens is such case the curve labeled E would become the I
most simply shown by vector diagrams. Fig. 3-11 is (current) curve, while the current curve would
an example in the case where R is less than Zo. The become the voltage curve.
voltage and current vectors at the load, R, are Some general conclusions can be drawn from
shown in the reference position; they correspond inspection of the standing-wave curves: Al a
with the vectors in Fig. 3-10A. Going back along position 180 degrees (1/2 wavelength) from the
the line from R towards the power source, the load, the voltage and current have the same values

Fig. 3-11 — Incident and re¬

flected components at various
positions along the line, to¬
gether with resultant voltages
and currents at the same posi¬
tions. The case shown is for R
less than Zo.

incident vectors, El and fl, lead the vectors at the they do at the load. At a position 90 degrees from
load according to their position along the line the load the voltage and current are “inverted.”
measured in electrical degrees. (The corresponding That is, if the voltage is lowest and the current
distances in fractions of a wavelength also are highest at the load (R less than Zo), then 90
shown.) The vectors representing reflected voltage degrees or 1/4 wavelength from the load the
and current, El and 72, successively lag the same voltage reaches its highest value and the current
vectors at the load. This lag and lead is the natural reaches its lowest value. In the case where R is
consequence of the direction in which the incident greater than Zo, so that the voltage is highest and
and reflected components are traveling, together the current lowest at the load, the voltage has its
with the fact that it takes time for the power to lowest value and the current its highest value at a
travel along the line. The resultant voltage, E, and point 90 degrees from the load.
current, I, at each of these positions are shown Note that the conditions existing at the 90-
dotted. Although the incident and reflected com¬ degree point also are duplicated at the 270-degree
ponents maintain their respective amplitudes (the point (3/4 wavelength). If the graph were contin¬
reflected component is shown at half the incident¬ ued on toward the source of power it would be
component amplitude in this drawing) their phase found that this duplication occurs at every point
relationships vary with position along the line. The that is an odd multiple of 90 degrees (odd multiple
phase shifts cause both the amplitude and phase of of a quarter wavelength) from the load. Similarly,
the resultants to vary with position on the line. the voltage and current are the same at every point
If the amplitude variations (disregarding phase) that is a multiple of 180 degrees (any multiple of
of the resultant voltage and current are plotted one-half wavelength) as they are at the load.
74
Chapter 3

when R is less than Zo. The smaller quantity is

always used in the denominator of the fraction so
the SWR will be a number larger than 1.
This relationship shows that the greater the
mismatch — that is, the greater the difference
between Zo and R - the larger the SWR. In the
case of open- and short-circuited lines the SWR
becomes infinite. On such lines the voltage and
current become zero at the minimum points (£ m j n
and /m jn) since total reflection occurs at the end
of the line and the incident and reflected compo¬
nents have equal amplitudes.
Fig. 3-12 - Standing waves of current and voltage
along the line, for R less than Zo.
INPUT IMPEDANCE
The relationship between the voltage and cur¬
Standing-Wave Ratio rent at any point along the line (including the
The ratio of the maximum voltage along the effects of both the incident and reflected compo¬
line to the minimum voltage - that is, the ratio of nents) becomes more clear when only the resultant
^max ^min * n Fig- 3-12 — is called the voltage voltage and current are shown, as in Fig. 3-13.
standing-wave ratio (abbreviated VSWR) or simply Note that the voltage and current are in phase not
the standing-wave ratio (SWR). The ratio of the only at the load but also at the 90-degree point,
maximum current to the minimum current the 180-degree point, and the 270-degree point.
^max/^min) is the same as the VSWR, so either This is also true at every point that is a multiple of
current or voltage can be measured to determine 90 degrees from the load.
the standing-wave ratio. Suppose the line were cut at one of these points
The standing-wave ratio is an index of many of and the generator or source of power were con¬
the properties of a mismatched line. It can be nected to that portion terminated in R. Then the
measured with good accuracy with fairly simple generator would “see” a pure resistance, just as it
equipment, and so is a convenient quantity to use would if it were connected directly to R. However,
in making calculations on line performance. If the the value of the resistance it sees would depend on
load contains no reactance, the SWR is numerically the line length. If the length is 90 degrees, or an
equal to the ratio between the load resistance, R, odd multiple of 90 degrees, where the voltage is
and the characteristic impedance of the line; that high and the current low, the resistance seen by the
is, generator would be greater than R. If the length is
180 degrees or a multiple of 180 degrees, the
SWR = -§- voltage and current relationships are the same as in
Zo R, and therefore the generator “sees” a resistance
when R is greater than Zo, and equal to the actual load resistance at these line
lengths.
SWR = The current and voltage are exactly in phase
/\ only at points that are multiples of 90 degrees

tro* » K)’ c 5* 0* 4 5* —- a’

■
1*^4
i ‘I 1

eJ 1
■*- INOUC TIVE -►+•-CAPAC TIVE-►+
-•- INOUC TIVE-*4

Fig. 3-13 — Resultant voltages

and currents along a mis¬
matched line. Above - R less
than Zo; below — R greater
than Zo.
Transmission Lines
75

from the load. At all other points the current

either leads or lags the voltage, and so the load seen
by the generator when the line length is not an
exact multiple of 90 degrees is not a pure
resistance. The input impedance of the Une - that
is, the impedance seen by the generator connected
to the line — in such a case has both resistive and
reactive components. When the current lags behind
the voltage the reactance is inductive; when it leads
the voltage the reactance is capacitive. The upper
drawing in Fig. 3-13 shows that when the line is Fig. 3-14 — Series circuits which may be used to
terminated in a resistance smaller than Zo the represent the input impedance of a length of
reactance is inductive in the first 90 degrees of line transmission line.
moving from the load toward the generator, is
capacitive in the second 90 degrees, inductive in cuits are important from the standpoint of design¬
the third 90 degrees, and so on every 90 degrees or ing coupling networks so that the desired amount
quarter wavelength. The lower drawing illustrates of power will be taken from the source. Their
the case where R is greater than Zo. The voltage practical application is described later in this
and current vectors are merely interchanged, since, chapter.
as explained in connection with Fig. 3-11, in this
case the vector for the reflected current is the one REACTIVE TERMINATIONS
that is reversed in phase on reflection. The reac¬ So far the only type of load considered has
tance becomes capacitive in the first 90 degrees, been a pure resistance. In general, the actual load
inductive in the second, and so on. usually will be fairly close to being a pure
resistance, since most transmission lines used by
Factors Determining the Input Impedance amateurs are connected to resonant antenna
The magnitude and phase angle of the input systems, which are principally resistive in nature.
impedance depend on the SWR, the line length, Consequently, the resistive load is an important
and the Zo of the line. If the SWR is small, the practical case.
input impedance is principally resistive at all line However, an antenna system is purely resistive
lengths; if the SWR is high, the reactive component only at one frequency, and when it is operated
may be relatively large. The input impedance of over a band of frequencies without readjustment -
the line can be represented by a series circuit of the usual condition - its impedance will contain a
resistance and reactance, as shown in Fig. 3-14
where Rs is the resistive component and X s is the
reactive component. Frequently the “s” subscripts
are omitted, and the series-equivalent impedance
denoted as R + jX. The / is an operator function,
used to indicate that the values for R and X cannot
be added directly, but that vector addition must be
used if the overall impedance is to be determined. WWW-
(This is analogous to solving a right triangle for the
length of its hypotenuse, where R and X represent
the length of its two sides. The length of the
hypotenuse represents Z, the overall impedance.) B o-lWLW-
By convention, a plus sign is assigned to/ when the *s
SERIES EQUIVALENT
reactance is inductive (R + jX), and a minus sign is
used when the reactance is capacitive (R - jX).

Equivalent Circuits for the Input Impedance

The series circuits shown in Fig. 3-14 are
equivalent to the actual input impedance of the
line because they have the same total impedance * x s

and the same phase angle. It is also possible to x s

form a circuit with resistance and reactance in
parallel that will have the same total impedance
and phase angle as the line. This equivalence is
shown in Fig. 3-15. The individual values in the Fig. 3-15 — Input impedance of a line terminated
parallel circuit are not the same as those in the in a pure resistance. The input impedance can be
series circuit (although the overall result is the represented either by a resistance and reactance in
same) but are related to the series-circuit values by series or a resistance and reactance in parallel. The
the equations shown in the drawing. relationship between the R and X values in the
Either of the two equivalent circuits may be series and parallel equivalents is given by the
used, depending on which happens to be more formulas. X may be either inductive or capacitive,
convenient for the particular purpose. These cir- depending on the line length, Zo, and the load.
 76
Chapter 3

certain amount of reactance along with resistance. ically, with the aid of a very useful device, the
The effect of such a combination is to increase the Smith Chart. If the terminating impedance is
standing-wave ratio — that is, as between two known, it is a simple matter to determine the input
loads, one having only resistance of, say, 100 ohms impedance of the line for any length by means of
as compared with a reactive load having the same the Chart. Conversely, with a given line length and
total impedance, 100 ohms, the SWR will be higher a known (or measured) input impedance, the load
with the reactive load than with the purely resistive impedance may be determined by means of the
load. Also as between two loads containing the Chart — a convenient method of remotely deter¬
same value of resistance but one being without mining an antenna impedance, for example.
reactance while the other has a reactive component Named after its inventor, Phillip H. Smith, the
in addition, the SWR will be higher with the one Smith Chart was originally described in Electronics
having the reactance. for January, 1939. Smith charts may be obtained
The effect of reactance in the load is to shift at most university book stores. They may be
the phase of the current with respect to the voltage ordered (50 for $2.50, postpaid when remittance is
both in the load itself and in the reflected enclosed) from Phillip H. Smith, Analog Instru¬
components of voltage and current. This in turn ments Co., P. O. Box 808, New Providence, NJ
causes a shift in the phase of the resultant current 07974. For 8-1/2 X 11-inch paper charts with
with respect to the resultant voltage. The net result normalized coordinates, request Form 82-BSPR.
is to shift the points along the line at which the Smith charts with 50-ohm coordinates (Form
various effects already described will occur. With a 5301-7569) are available at the same price from
load having inductive reactance the point of General Radio Co., West Concord, MA 01781.
maximum voltage and minimum current is shifted Although its appearance may at first seem
toward the load. The reverse occurs when the somewhat formidable, the Smith Chart is really
reactance in the load is capacitive. nothing more than a specialized type of graph,
with curved, rather than rectangular, coordinate
lines. The coordinate system consists simply of two
families of circles — the resistance family and the
SMITH-CHART TRANSMISSION-LINE reactance family. The resistance circles (Fig. 3-16)
are centered on the resistance axis (the only
CALCULATIONS
straight line on the Chart), and are tangent to the
outer circle at the bottom of the Chart. Each circle
It has already been stated that the input is assigned a value of resistance, which is indicated
impedance, or the impedance seen when “looking at the point where the circle crosses the resistance
into” a length of line, is dependent upon the SWR, axis. All points along any one circle have the same
the length of the line, and the Zo of the line. The resistance value.
SWR, in turn, is dependent upon the load which The values assigned to these circles vary from
terminates the line. There are complex mathema¬ zero at the top of the chart to infinity at the
tical relationships which may be used to calculate bottom, and actually represent a ratio with respect
the various values of impedances, voltages, cur¬ to the impedance value assigned to the center point
rents, and SWR values which exist in the operation of the Chart, indicated 1.0. This center point is
of a particular transmission line. However, it is called prime center. If prime center is assigned a
much easier to determine such parameters graph- value of 100 ohms, then 200 ohms resistance is
represented by the 2.0 circle, 50 ohms by the 0.5
circle, 20 ohms by the 0.2 circle, and so on. If a
value of 50 is assigned to prime center, the 2.0
circle now represents 100 ohms, the 0.5 circle 25
ohms, and the 0.2 circle 10 ohms. In each case, it
may be seen that the value on the Chart is
determined by dividing the actual resistance by the
number assigned to prime center. This process is
called normalizing. Conversely, values from the
Chart are converted back to actual resistance values
by multiplying the Chart value times the value
assigned to prime center. This feature permits the
use of the Smith Chart for any impedance values,
and therefore with any type of uniform transmis¬
sion line, whatever its impedance may be. As
mentioned above, specialized versions of the Smith
Chart may be obtained with a value of 50 ohms at
prime center. These are intended for use with
50-ohm lines.
Now consider the reactance circles (Fig. 3-17)
which appear as curved lines on the Chart because
only segments of the complete circles are drawn.
Fig. 3-16 — Resistance circles of the Smith Chart These circles are tangent to the resistance axis,
coordinate system. which itself is a member of the reactance family
Transmission Lines 77

(with a radius of infinity). The centers are dis¬

placed to the right or left on a line tangent to the
bottom of the chart. The large outer circle bound¬
ing the coordinate portion of the Chart is the
reactance axis.
Each reactance circle segment is assigned a
value of reactance, indicated near the point where
the circle touches the reactance axis. All points
along any one segment have the same reactance
value. As with the resistance circles, the values
assigned to each reactance circle are normalized
with respect to the value assigned to prime center.
Values to the right of the resistance axis are
positive (inductive), and those to the left of the
reactance axis are negative (capacitive).
When the resistance family and the reactance
family of circles are combined, the coordinate
system of the Smith Chart results, as shown in Fig.
3-18. Complex series impedances can be plotted on
this coordinate system.

Impedance Plotting Fig. 3-17 — Reactance circles (segments) of the

Smith Chart coordinate system.
Suppose we have an impedance consisting of 50
ohms resistance and 100 ohms inductive reactance
(Z = 50 + /TOO). If we assign a value of 100 ohms bottom of the Chart, at the intersection of the
to prime center, we normalize the above imped¬ resistance and reactance axes. These two special
ance by dividing each component of the impedance cases are sometimes used in matching stubs, des¬
by 100. The normalized impedance would then be cribed later in this chapter.
= 0.5 + /T.O. This impedance would be Standing-Wave-Ratio Circles
plotted on the Smith Chart at the intersection of
Members of a third family of circles, which are
the 0.5 resistance circle and the +1.0 reactance
not printed on the chart but which are added
circle, as indicated in Fig. 3-18. If a value of 50
during the process of solving problems, are stand-
ohms had been assigned to prime center, as for
ing-wave-ratio or SWR circles. See Fig. 3-19. This
50-ohm coaxial line, the same impedance would be
family is centered on prime center, and appears as
plotted at the intersection of the = 1.0 concentric circles inside the reactance axis. During
calculations, one or more of these circles may be
resistance circle, and the = 2.0 positive added with a drawing compass. Each circle repre-
reactance circle, or at 1 + /2 (also indicated in Fig.
3-18).
From these examples, it may be seen that the
same impedance may be plotted at different points
on the Chart, depending upon the value assigned to
prime center. It is customary when solving trans¬
mission-line problems to assign to prime center a
value equal to the characteristic impedance, or Zo,
of the line being used. This value should always be
recorded at the start of calculations, to avoid
possible confusion later. In using the specialized
charts with the value of 50 at prime center, it is, of
course, not necessary to normalize impedances
when working with 50-ohm line. The resistance
and reactance values may be plotted directly.

Short and Open Circuits

On the subject of plotting impedances, two
special cases deserve consideration. These are short
circuits and open circuits. A true short circuit has
zero resistance and zero reactance, or 0 + /0. This
impedance would be plotted at the top of the
Chart, at the intersection of the resistance and the Fig. 3-18 — The complete coordinate system of the
reactance axes. An open circuit has infinite resis¬ Smith Chart. For simplicity, only a few divisions
tance, and would therefore be plotted at the are shown for the resistance and reactance values.
78 Chapter 3

scales of the chart, by drawing lines

tangent to the circle. Or, the radius
of the SWR circle may be simply
transferred to the external scale by
placing the point of a drawing com¬
pass at the center, or 0, line and
inscribing a short arc across the
appropriate scale. It will be noted
that when this is done in Fig. 3-20,
the external STANDING-WAVE
voltage-ratio scale indicates
the SWR to be 3.0 (at A) — our
condition for initially drawing the
circle on the Chart (and the same as
the SWR reading on the resistance
axis).

Solving Problems with the

Smith Chart
Suppose we have a transmission
Fig. 3-19 — Smith Chart with SWR circles added. line with a characteristic impedance
of 50 ohms, and an electrical length of 0.3
wavelength. Also, suppose we terminate this line
sents a value of SWR, every point on a given circle with an impedance having a resistive component of
25 ohms and an inductive reactance of 25 ohms (Z
representing the same SWR. The SWR value for a
given circle may be determined directly from the = 25 + /25), and desire to determine the input
chart coordinate system, by reading the resistance impedance to the line. Because the line is not
value where the SWR circle crosses the resistance terminated in its characteristic impedance, we
axis, below prime center. (The reading where the know that standing waves will be present on the
circle crosses the resistance axis above prime center line, and that, therefore, the input impedance to
indicates the inverse ratio.) the line will not be exactly 50 ohms. We proceed as
Consider the situation where a load mismatch follows: First, normalize the load impedance by
in a length of line causes a 3-to-l standing-wave dividing both the resistive and reactive components
ratio to exist. If we temporarily disregard line by 50 (Zo of the line being used). The normalized
losses, we may state that the SWR remains con¬ impedance in this case is 0.5 + /O.5. This is plotted
stant throughout the entire length of this line. This on the Chart at the intersection of the 0.5
is represented on the Smith Chart by drawing a 3:1 resistance and the +0.5 reactance circles, as in Fig.
constant-SWR circle (a circle with a radius of 3 on 3-21. Then draw a constant-SWR circle passing
the resistance axis), as in Fig. 3-20. The design of through this plotted point. The radius of this circle
the Chart is such that any impedance encountered may then be transferred to the external scales with
the drawing compass. From the external s.w.V.R.
anywhere along the length of this mismatched line
scale, it may be seen (at A), that the voltage ratio
will fall on the SWR circle, and may be read from
of 2.6 exists for this radius, indicating that our line
the coordinates merely by progressing around the
is operating with an SWR of 2.6 to 1. This figure is
SWR circle by an amount corresponding to the
converted to decibels in the adjacent scale, where
length of the line involved.
8.4 dB may be read (at B), indicating that the ratio
This brings into use the wavelength scales, of the voltage maximum to the voltage minimum
which appear, in Fig. 3-20, near the outer perime¬ along the line is 8.4 dB.
ter of the Smith Chart. These scales are calibrated Next, with a straightedge, draw a radial line
in terms of portions of an electrical wavelength from prime center through the plotted point to
along a transmission line. One scale, running intersect the wavelengths scale, and read a value
counterclockwise, starts at the generator or input from the wavelengths scale. Because we are starting
end of the line and progresses toward the load, from the load, we use the TOWARD-GENERA-
while the other scale starts at the load and TOR or outermost calibration, and read 0.088
proceeds toward the generator in a clockwise wavelength (at C). To obtain the line input
direction. The complete circle represents one half impedance, we merely find the point on the SWR
wavelength. Progressing once around the perimeter circle which is 0.3 wavelength toward the generator
of these scales corresponds to progressing along a from the plotted load impedance. This is accom¬
transmission line for a half wavelength. Because plished by adding 0.3 (the length of the line in
impedances will repeat themselves every half wave¬ wavelengths) to the reference or starting point,
length along a piece of line, the Chart may be used 0.088; 0.3 + 0.088 = 0.388. Locate 0.388 on the
for any length of line by disregarding or subtract¬ TOWARD-GENERATOR scale (at D), and draw a
ing from the line’s total length an integral, or second radial line from this point to prime center.
whole number, of half wavelengths. The intersection of the new radial line with the
Also shown in Fig. 3-20 is a means of transfer¬ SWR circle represents the line input impedance, in
ring the radius of the SWR circle to the external this case 0.6 — /0.65. To find the actual line input
Transmission Lines 79

Fig. 3-20 — Example discussed in text.

impedance, multiply by 50 — the value assigned to reflected. (The amount of reflected power is
prime center, which equals 30 —/32.5, or 30 ohms proportional to the square of the reflected volt¬
resistance and 32.5 ohms capacitive reactance. This age.)
is the impedance which a transmitter must match if
such a system were a combination of antenna and Admittance Coordinates
transmission line, or is the impedance which would
be measured on an impedance bridge if the Quite often it is desirable to convert impedance
measurement were taken at the line input. information to admittance data — conductance and
In addition to the Une input impedance and the susceptance. Working with admittances greatly
SWR, the Chart reveals several other operating simplifies determining the resultant when two
characteristics of the above system of line and complex impedances are connected in parallel, as
load, if a closer look is desired. For example, the in stub matching. The conductance values may be
voltage reflection coefficient, both magnitude and added directly, as may be the susceptance values,
phase angle, for this particular load is given. The to arrive at the overall admittance for the parallel
phase angle is read under the radial line drawn combination. This admittance may then by con¬
through the plot of the load impedance where the verted back to impedance data.
line intersects the ANGLE-OF-REFLECTION- On the Smith Chart, the necessary conversion
COEFFICIENT scale. This scale is not included in may be made very simply. The equivalent admit¬
Fig. 3-21, but will be found on the Smith Chart, tance of a plotted impedance value lies diametri¬
just inside the wavelengths scales. In this example, cally opposite the impedance point on the Chart.
the reading would be about 116.5 degrees. This In the foregoing example, where the normalized
indicates the angle by which the reflected voltage line input impedance is 0.6 —/0.65, the equivalent
wave lags the incident wave at the load. It will be admittance lies at the intersection of the SWR
noted that angles on the left half, or capacitive- circle and the extension of the straight line passing
reactance side, of the Chart are negative angles, a from point D to prime center. Although not shown
“negative” lag indicating that the reflected voltage in Fig. 3-21, the normalized admittance value may
wave actually leads the incident wave. be read as 0.76 + /0.84 if the Une is extended.
The magnitude of the voltage-reflection- (Capacitance is considered to be a positive suscep¬
coefficient may be read from the external tance and inductance a negative susceptance.) The
REFLECTION-COEFFICIENT-VOLTAGE scale, admittance in mhos is determined by dividing the
and is seen to be approximately 0.44 (at E) for this normalized values by the Zo of the line. For this
example, meaning 44 percent of the incident example the admittance would be 0.76/50 +
voltage is reflected. Adjacent to this scale on the /0.84/50 = 0.0152 +/0.0168 mho.
POWER calibration, it is noted (at F) that the Of course admittance coordinates may be con¬
power reflection coefficient is 0.20, indicating that verted to impedance coordinates just as easily — by
20 percent of the incident power would be locating the point on the Smith Chart which is
 80
Chapter 3

Fig. 3-21 — Example discussed in text.

diametrically opposite the point representing the The SWR of 1.7 may be read from the S.W.V.R.
admittance coordinates, on the same SWR circle. scale (at A). Now draw a radial line from prime
Determining Actual Antenna Impedances center through this plotted point to the wave¬
lengths scale, and read a reference value, which is
To determine an actual antenna impedance 0.195 (at B), on the TOWARD-LOAD scale.
from the Smith Chart, the procedure is similar to Remember, we are starting at the generator end of
the previous example. The electrical length of the the transmission line.
feed line must be known and the impedance value To locate the load impedance on the SWR
at the input end of the line must be determined circle, we add the line length, 2.35 wavelengths to
through measurement, such as with an impedance¬ the reference value from the wavelengths scale, and
measuring bridge like the Macromatcher, described locate the new value on the TOWARD-LOAD
in the Measurements chapter. In this case, the scale; 2.35 + 0.195 = 2.545. However, the calibra¬
antenna is connected to the far end of the line and tions extend only from 0 to 0.5, so we must
becomes the load for the line. Whether the antenna subtract a whole number of half wavelengths from
is intended purely for transmission of energy, or this value and use only the remaining value. In this
purely for reception makes no difference; the situation, the largest integral number of half
antenna is still the terminating or load impedance wavelengths that can be subtracted is 5, or 2.5
on the line as far as these measurements are wavelengths. Thus, 2.545 - 2.5 = 0.045, and the
concerned. The input or generator end of the line 0.045 value is located on the toward-load
would be that end connected to the device for scale (at C). A radial line is then drawn from this
measurement of the impedance. In this type of value to prime center, and the coordinates at the
problem, the measured impedance is plotted on the intersection of the second radial line and the SWR
Chart, and the TOWARD-LOAD wavelengths scale circle represent the load impedance. To read this
is used in conjunction with the electrical line value closely, some interpolation between the
length to determine the actual antenna impedance. printed coordinate lines must be made, and the
For example, assume we have a measured input value of 0.62 — /0.18 is read. Multiplying by 50,
impedance to a 50-ohm line of 70 -/25 ohms. The the actual load or antenna impedance is 31 - /9
line is 2.35 wavelengths long, and is terminated in ohms, or 31 ohms resistance with 9 ohms capaci¬
an antenna. We desire to determine the actual tive reactance.
antenna impedance. Normalize the input imped¬ Problems may be entered on the chart in yet
ance with respect to 50 ohms, which comes out 1.4 another manner. Suppose we have a length of
- /0.5, and plot this value on the Chart. See Fig. 50-ohm line feeding a base-loaded resonant vertical
3-22. Draw a constant-SWR circle through the ground-plane antenna which is shorter than a
point, and transfer the radius to the external scales. quarter wave. Further, suppose we have an SWR
Transmission Lines 81

Fig. 3-22 — Example discussed in text.

monitor in the line, and that it indicates an SWR of Determination of Line Length
1.7 to 1. The line is known to be 0.95 wavelength In the example problems given so far in this
long. We desire to know both the input and the section, the line length has conveniently been
antenna impedances. stated in wavelengths. The electrical length of a
From the data given, we have no impedances to piece of line depends upon its physical length, the
enter onto the chart. We may, however, draw a radio frequency under consideration, and the
circle representing the 1.7 SWR. We also know, velocity of propagation in the line. If an imped¬
from the definition of resonance, that the antenna ance-measurement bridge is capable of quite reli¬
presents a purely resistive load to the line; Le., no able readings at high line-SWR values, the line
reactive component. Thus, the antenna impedance length may be determined through line input¬
must lie on the resistance axis. If we were to draw impedance measurements with short- or open¬
such an SWR circle and observe the Chart with circuit terminations. A more direct method is to
only the circle drawn, we would see two points measure the line’s physical length and apply the
which satisfy the resonance requirement for the value to a formula. The formula is:
load. These points are 0.59 + /0 and 1.7 + /0.
Multiplying by 50, these values represent 29.5 and
85 ohms resistance. This may sound familiar, 984Æ
because, as was discussed earlier, when a line is
terminated in a pure resistance, the SWR in the where
line equals ZjJZo or Zo/Z^, where Zr - load N = Number of electrical wavelengths in the
resistance and Zo = line impedance. Une,
If we consider antenna fundamentals described L = Line length in feet,
in Chapter Two, we know that the theoretical F = Frequency in megahertz, and
impedance of a quarter-wave ground-plane antenna K = Velocity or propagation factor of the line.
is approximately 36 ohms. We therefore can quite The factor K may be obtained from transmission¬
logically discard the 85-ohm impedance figure in line data tables which appear later in this chapter.
favor of the 29.5-ohm value. This is then taken as
the actual load-impedance value for the Smith ATTENUATION
Chart calculations. The line input impedance is
found to be 0.64 — /0.21, or 32 — /10.5 ohms, The discussion in the preceding part of this
after subtracting 0.5 wavelength from 0.95, and chapter applies to all types of transmission lines,
finding 0.45 wavelength on the TOWARD- regardless of their physical construction. It is,
GENERATOR scale. (The wavelength reference in however, based on the assumption that there is no
this case is 0.) power loss in the line. Every actual line will have
 82 Chapter 3

logarithmically with the length. That is, if 10% of

the input power is lost in a section of line of
certain length, 10% of the remaining power will be
lost in the next section of the same length, and so
on. For this reason it is customary to express line
losses in terms of decibels per unit length, since the
decibel is a logarithmic unit. Calculations are very
simple because the total loss in a line is found by
multiplying the dB loss per unit length by the total
length of the line. Line loss is usually expressed in
decibels per 100 feet. It is necessary to specify the
frequency for which the loss applies, since the loss
varies with frequency.
Conductor loss and dielectric loss both increase
as the operating frequency is increased, but not in
the same way. This, together with the fact that the
relative amount of each type of loss depends on
the actual construction of the line, makes it
impossible to give a specific relationship between
loss and frequency that will apply to all types of
lines. Each line has to be considered individually.
UNE LOSS IN DB WHEN MATCHED Actual loss values are given in a later section.

Fig. 3-23 — Increase in line loss because of standing Effect of SWR

waves (SWR measured at the load). To determine The power lost in a line is least when the line is
the total loss in decibels in a line having an SWR terminated in a resistance equal to its characteristic
greater than 1, first determine the loss for the
particular type of line, length and frequency, on impedance, and increases with an increase in the
the assumption that the line is perfectly matched standing-wave ratio. This is because the effective
(Table 3-1). Locate this point on the horizontal values of both current and voltage become larger as
axis and move up to the curve corresponding to the the SWR becomes greater. The increase in effective
actual SWR. The corresponding value on the current raises the ohmic losses in the conductors,
vertical axis gives the additional loss in decibels and the increase in effective voltage increases the
caused by the standing waves. losses in the dielectric.
The increased loss caused by an SWR greater
some inherent loss, partly because of the resistance than 1 may or may not be serious. If the SWR at
of the conductors, partly, because of the fact that the load is not greater than 2, the additional loss
caused by the standing waves, as compared with
power is consumed in every dielectric used for
insulating the conductors, and partly because in the loss when the line is perfectly matched, does
many cases a small amount of power escapes from not amount to more than about 1/2 dB even on
the line by radiation. very long lines. Since 1/2 dB is an undetectable
Losses in a line modify its characteristic imped¬ change in signal strength, it can be said that from a
ance slightly, but usually not to a sufficient extent practical standpoint an SWR of 2 or less is, so far
to be significant. They will also affect the input as losses are concerned, every bit as good as a
perfect match.
impedance; in this case the theoretical values will
be modified only slightly if the line is short and has 1 he effect of SWR on line loss is shown in Fig.
only a small loss, but may be changed considerably 3-23. The horizontal axis is the attenuation, in
if an appreciable proportion of the power input to decibels, of the line when perfectly matched. The
the line is dissipated by the line itself. A large loss vertical axis gives the additional attenuation, in
may exist because the line is long, because it has decibels, caused by standing waves. For example, if
inherently high loss per unit length, because the the loss in a certain line is 4 dB when perfectly
standing-wave ratio is high, or because of a matched, an SWR of 3 on that same line will cause
combination of two or all three of these factors. an additional loss of 1.1 dB, approximately. The
The reflected power returning to the input total loss on the poorly matched Une is therefore 4
terminals of the line is less when the line has losses + 1.1 = 5.1 dB. If the SWR were 10 instead of 3,
than it would be if there were none. The overall the additional loss would be 4.3 dB, and the total
effect is that the SWR changes along the line, being loss 4 + 4.3 = 8.3 dB.
highest at the load and smallest at the input It is important to note that the curves in Fig.
terminals. A long, high-loss line therefore tends to 3-23 which represent SWR are the values which
act, so far as its input impedance is concerned, as exist at the load. In most cases of amateur
though the impedance match at the load end were operation, this will be at the antenna end of a
better than is actually the case. length of transmission line. The SWR as measured
at the input or transmitter end of the line will be
Line Losses less, depending on the line attenuation. It is not
always convenient to measure SWR directly at the
The power lost in a transmission line is not antenna. However, by using the graph shown in
directly proportional to the line length but varies Eig. 3-24, the SWR at the load can be obtained by
Transmission Lines 83

measuring it at the input to the transmission line necessary to draw a spiral inward and clockwise
and using the known (or estimated) loss of the from the load impedance toward the generator.
transmission line. (See later section on testing The rate at which the curve spirals toward prime
coaxial cable.) For example, if the SWR at the center is related to the attenuation in the line.
transmitter end of a line is measured as 3 to 1 and Rather than drawing spiral curves, a simpler
the line is known to have a total attenuation method is used in solving line-loss problems, by
(under matched conditions) of 1 dB, the SWR at means of the external scale TRANSMISSION¬
the load end of the line will be 4.5 dB. From Fig. LOSS, 1-DB STEPS in Fig. 3-25. Because this is
3-23, the additional loss is nearly 1 dB because of only a relative scale, the dB steps are not num¬
the presence of the SWR. The total line loss in this bered.
case is 2 dB. If we start at the top end of this external scale
It is of interest to note that when the line loss is and proceed in the direction indicated toward
high with perfect matching, the additional loss in generator, the first dB step is seen to occur at a
dB caused by the SWR tends to be constant radius from center corresponding to an SWR of
regardless of the matched Une loss. The reason for about 9 (at A); the second dB step falls at an SWR
this is that the amount of power available to be of about 4.5 (at B), the third at 3.0 (at C), and so
reflected from the load is reduced, because rela¬ forth, until the 15th dB step falls at an SWR of
tively little power reaches the load in the first about 1.05 to 1. This means that a Une terminated
place. For example, if the line loss with perfect in a short or open circuit (infinite SWR) and having
matching is 6 dB, only 25% of the power originally an attenuation of 15 dB, would exhibit an SWR of
put into the line reaches the load. If the mismatch only 1.05 at its input. It wiU be noted that the dB
at the load (the SWR at the load) is 4 to 1, 36% of steps near the lower end of the scale are very close
the power reaching the load will be reflected. Of together, and a Une attenuation of 1 or 2 dB in this
the power originally put into the line, then, 0.25 X area wiU have only slight effect on the SWR. But
0.36 = 0.09 or 9% will be reflected. This in turn near the upper end of the scale, 1- or 2-dB loss has
will be attenuated 6 dB in traveling back to the considerable effect on the SWR.
input end of the line, so that only 0.09 X 0.25 = In solving a problem utilizing Une-loss informa¬
0.0225 or slightly over 2% of the original power tion, it is necessary only to modify the radius of
actually gets back to the input terminals. With such the SWR circle by an amount indicated on the
a small proportion of power returning to the input TRANSMISSION-LOSS, 1-DB-STEPS scale. This is
terminals the SWR measured at the input end of accompUshed by drawing a second SWR circle, of
the Une would be only about 1.35 to 1 - although
it is 4 to 1 at the load. In the presence of line losses
the SWR always decreases along the line going
from the load to the input end.
On lines having low losses when perfectly
matched, a high standing-wave ratio may increase
the power loss by a large factor. However, in this
case the total loss may still be inconsequential in
comparison with the power delivered to the load.
An SWR of 10 on a line having only 0.3 dB loss
when perfectly matched will cause an additional
loss of 1 dB, as shown by the curves. This loss
would produce a just-detectable difference in signal
strength.

LINE-LOSS CONSIDERATIONS
WITH THE SMITH CHART
The problems presented earlier ignored attenua¬
tion, or fine losses. Quite frequently it is not even
necessary to consider losses when making calcula¬
tions; any difference in readings obtained would be
almost imperceptible on the Smith Chart. When
the Une losses become appreciable, as described
above, loss considerations may be warranted in
making Smith Chart calculations. This involves
only one simple step, in addition to the procedures
previously presented.
Because of line losses, the SWR does not remain
constant throughout the length of the line, as just SWR AT TRANSMITTER
discussed. As a result, there is a decrease in SWR as
one progresses away from the load. To truly Fig. 3-24 — SWR at input end of transmission line
represent this situation on the Smith Chart, instead vs. SWR at load end for various values of matched-
of drawing a constant-SWR circle, it would be line loss.
 84 Chapter 3

either greater or lesser radius than the first, as the by 50, the actual load impedance is 32.5 - ;'3O
case may be. ohms. The SWR in this problem was seen to
For example, assume that we have a 50-ohm increase from 1.9 at the line input to 2.3 (at I) at
line 0.282 wavelength long, with 1-dB inherent the load, with the 1-dB line loss taken into
attenuation. The line input impedance is measured consideration.
as 60 + /35 ohms. We desire to know the SWR at In the example above, values were chosen to
the input and at the load, and the load impedance. fall conveniently on or very near the “tick marks”
As before, we normalize the 60 + ;35-ohm imped¬ on the 1-DB, scale. Actually, it is a simple matter
ance, plot it on the Chart, and draw a constant- to interpolate between these marks when making a
SWR circle and a radial line through the point. In radius correction. When this is necessary, the
this case, the normalized impedance is 1.2 + /O.7. relative distance between marks for each dB step
From Fig. 3-25, the SWR at the line input is seen should be maintained while counting off the
to be 1.9 (at D), and the radial line is seen to cross proper number of steps.
the TOWARD-LOAD scale at 0.328 (at E). To the Adjacent to the 1-DB-STEPS scale lies a LOSS¬
0.328 we add the line length, 0.282, and arrive at a COEFFICIENT scale. This scale provides a factor
value of 0.610. To locate this point on the by which the matched-line loss in dB should be
TOWARD-LOAD scale, first subtract 0.500, and multiplied to account for the increased losses in
locate 0.110 (at F); then draw a radial line from the line when standing waves are present. These
this point to prime center. added losses do not affect the standing-wave ratio
To account for line losses, transfer the radius of or impedance calculations; they are merely the
the SWR circle to the external 1-DB-STEPS scale. additional dielectric and copper losses of the line
This radius will cross the external scale at G, the caused by the fact that the line conducts more
fifth dB mark from the top. Since the line loss was average current and must withstand more average
given as 1 dB, we strike a new radius (at H), one voltage in the presence of standing waves. In the
“tick mark” higher (toward load) on the same above example and in Fig. 3-25, the loss coefficient
scale. (This will be the fourth dB tick mark from at the input end is seen to be 1.21 (at J), and 1.39
the top of the scale.) Now transfer this new radius (at K) at the load. As a good approximation, the
back to the main chart, and scribe a new SWR loss coefficient may be averaged over the length of
circle of this radius. This new radius represents the line under consideration; in this case, the average is
SWR at the load, and is read as about 2.3 on the 1.3. This means that the total losses in the line are
external S.W.V.R. scale. At the intersection of the 1.3 times the matched loss of the line (1 dB), or
new circle and the load radial line, we read 0.65 - 1.3 dB, the same result that may be obtained from
/0.6 as the normalized load impedance. Multiplying Fig. 3-23 for the data of the above example.

Fig. 3-25 — Example of Smith Chart calculations

taking line losses into account.
Transmission Lines 85

Smith Chart Procedure Summary

To summarize briefly, any calculations made on
the Smith Chart are performed in four basic steps,
although not necessarily in the order listed.
1) Normalize and plot a line input (or load)
impedance, and construct a constant-SWR circle.
2) Apply the line length to the wavelengths
scales.
3) Determine attenuation or loss, if required,
by means of a second SWR circle.
4) Read normalized load (or input) impedance,
and convert to impedance in ohms.
The Smith Chart may be used for many types
of problems other than those presented as ex¬
amples here. The transformer action of a length of
line - to transform a high impedance (with
perhaps high reactance) to a purely resistive imped¬
ance of low value - was not mentioned. This is
known as “tuning the line,” for which the Chart is
very helpful, eliminating the need for cut-and-try Fig. 3-26 — Increase in maximum value of current
procedures. The Chart may also be used to or voltage on a line with standing waves, as referred
calculate lengths for shorted or open matching to the current or voltage on a perfectly matched
stubs in a system, described later in this chapter. In line, for the same power delivered to the load.
fact, in any application where a transmission line is Voltage and current at minimum points are given
by the reciprocals of the values along the vertical
not perfectly matched, the Smith Chart can be of axis. The curve is plotted from the relationship,
value. current (or voltage) ratio = \/SWR.

VOLTAGES AND CURRENTS ON LINES voltage and current are both equal to 81.8% of the
The power reflected from a mismatched load incident voltage and current. The reflected power
does not represent an actual loss, except as it is is proportional to the square of either the current
attenuated in traveling back to the input end of the or voltage, and so is equal to (0.818)’ = 0.67 times
line. It merely represents power returned, and the the incident power, or 67 watts. Since we have
actual effect is to reduce the power taken from the assumed that the line has no losses, this amount of
power arrives back at the input terminals and
source. That is, it reduces the coupling between the
subtracts from the original 100 watts, leaving only
power source and the line. This is easily overcome
33 watts as the amount of power actually taken
by readjusting the coupling until the actual power
from the source.
put into the line is the same as it would be with a
In order to put 100 watts into the 60-ohm load
matched load. In doing this, of course, the voltages
the coupling to the source must be increased so
and currents at loops along the line are increased.
that the incident power minus the reflected power
As an example, suppose that a line having a
equals 100 watts, and since the power absorbed by
characteristic impedance of 600 ohms is matched
the load is only 33% of that reaching it, the
by a resistive load of 600 ohms and that 100 watts
incident power must equal 100/0.33 = 303 watts.
of power goes into the input terminals. The line
In a perfectly matched line, the current and voltage
simply looks like a 600-ohm resistance to the
source of power. By Ohm’s Law the current and with 303 watts input would be 0.71 ampere and
426 volts, respectively. The reflected current and
voltage in such a matched line are
voltage are 0.818 times these values, or 0.581
/ = ^P/R ampere and 348 volts. At current maxima or loops
the current will therefore be 0.71 + 0.58 = 1.29 A,
E = ^PR and at a minimum point will be 0.71 - 0.58 = 0.13
Substituting 100 watts for P and 600 ohms for/?, A. The voltage maxima and minima will be 426 +
the current is 0.408 ampere and the voltage is 245 348 = 774 volts and 426 - 348 = 78 volts.
volts. Assuming for the moment that the line has (Because of rounding off figures in the calculation
no losses, all the power will reach the load so the process, the SWR does not work out to be exactly
voltage and current at the load will be the same as 10 in either the voltage or current case, but the
at the input terminals. error is very small.)
Now suppose that the load is 60 ohms instead In the interests of simplicity this example has
of 600 ohms. The SWR is 10, therefore. The been based on a line with no losses, but the
reflection coefficient, or ratio of the reflected approximate effect of line attenuation could be
voltage or current to the voltage or current arriving included without much difficulty. If the matched-
line loss were 3 dB, for instance, only half the
at the load, is SWR - 1
input power would reach the load, so new values of
p “ SWR + 1 current and voltage at the load would be computed
In this case the reflection coefficient is (10 - accordingly. The reflected power would then be
l)/(10 + 1) = 9/11 = 0.818, so that the reflected based on the attenuated figure, and then itself
86 Chapter 3

attenuated 3 dB to find the power arriving back at simply repeats, at its input or sending end, what¬
the input terminals. The overall result would be, as ever impedance exists at its output or receiving
stated before, a reduction in the SWR at the input end; it does not matter whether the impedance at
terminals as compared with that at the load, along the receiving end is resistive, reactive, or a com¬
with less actual power delivered to the load for the bination of both. Sections of line having such
same power input to the fine. length can be cut in or out without changing any
Fig. 3-26 shows the ratio of current or voltage of the operating conditions, at least when the
at a loop, in the presence of standing waves, to the losses in the line itself are negligible.
current or voltage that would exist with the same
power in a perfectly matched line. Strictly speak¬ Impedance Transformation with
ing, the curve applies only near the load in the case Quarter-Wave Lines
of lines with appreciable losses. However, the curve
shows the maximum possible value of current or The input impedance of a line an odd multiple
voltage that can exist along the line whether there of a quarter wavelength long is
are line losses or not, and so is useful in determin¬
ing whether or not a particular line can operate
safely with a given SWR.
where Zs is the input impedance and ZR is the
load impedance. If ZR is a pure resistance, Zs also
SPECIAL CASES will be a pure resistance. Rearranging this equation
gives
Besides the primary purpose of transporting
power from one point to another, transmission Zo = x/ZsZR
lines have properties that are useful in a variety of This means that if we have two values of imped¬
ways. One such special case is a line an exact ance that we wish to “match,” we can do so if we
multiple of one-quarter wavelength (90 degrees) connect them together by a quarter-wave transmis¬
long. As shown earlier, such a line will have a sion line having a characteristic impedance equal to
purely resistive input impedance when the termina¬ the square root of their product.
tion is a pure resistance. Also, unterminated - i.e., A quarter-wave line is, in effect, a transformer.
short-circuited or open-circuited - lines can be It is frequently used as such in antenna work when
used in place of conventional inductors and capaci¬ it is desired, for example, to transform the imped¬
tors since such lines have an input impedance that ance of an antenna to a new value that will match a
is substantially a pure reactance when the line given transmission line. This subject is considered
losses are low. in greater detail in a later section of this chapter.
The Half-Wavelength Line Lines as Circuit Elements
When the line length is an even multiple of 90 An open- or short-circuited line does not deliver
degrees (that is, a multiple of a half wavelength), any power to a load, and for that reason is not,
the input resistance is equal to the load resistance. strictly speaking, a “transmission” line. However,
As a matter of fact, a line an exact multiple of a the fact that a line of the proper length has
half wave in length (disregarding line losses) inductive reactance makes it possible to substitute

CIRCUIT EQUIVALENT
AT TERMINALS A-B

Fig. 3-27 — Lumped-constant

circuit equivalents of open-
and short-circuited transmis¬
sion lines.
Transmission Lines 87

the line for a coil in an ordinary circuit.

Likewise another line of appropriate
length having capacitive reactance can be
substituted for a capacitor.
Sections of lines used as circuit ele¬
ments are usually a quarter wavelength
or less long. The desired type of reac¬
tance (inductive or capacitive) or the
desired type of resonance (series or
parallel) is obtained by shorting or open¬
ing the far end of the line. The circuit
equivalents of various types of line sec¬
tions are shown in Fig. 3-27.
When a line section is used as a
reactance, the amount of reactance is
determined by the characteristic imped¬
ance and the electrical length of the line.
The type of reactance exhibited at the
input terminals of a line of given length
depends on whether it is open- or short-
circuited at the far end.
The equivalent “lumped” value for any “induc¬
tor” or “capacitor” may be determined with the
aid of the Smith Chart. Line losses may be taken
into account if desired, as explained in an earlier Fig. 3-28 — Smith Chart determination of input
section. See Fig. 3-28. Remember that the right impedances for short- and open-circuited line
sections, disregarding line losses.
half of the Smith Chart coordinate system is used
for impedances containing inductive reactances,
and the left half for capacitive reactances. For either case is numerically equal to the characteris¬
example, a section of 600-ohm line 3/16-wave- tic impedance of the tine. In using the Smith Chart
length long (0.1875 X) and short-circuited at the it should be kept in mind that the electrical length
far end is represented by LI, drawn around a of a line section depends on the frequency and
portion of the perimeter of the Chart. The “load” velocity of propagation as well as on the actual
is a short-circuit, 0 + /0 ohms, and the TOWARD physical length.
GENERATOR wavelengths scale is used for mark¬ In the case of a line having no losses, and to a
ing off the Une length. At A in Fig. 3-28 may be close approximation when the losses are small, the
read the normalized impedance as seen looking inductive reactance of a short-circuited line less
into the length of Une, 0 + /2.4. The reactance is than a quarter wave in length is
therefore inductive, equal to 600 X 2.4 = 1440 XL (ohms) = Zo tan /
ohms. The same Une open-circuited (termination
where Iis the length of the line in electrical degrees
impedance = ~, the point at the bottom of the
and Zo is the characteristic impedance of the line.
Chart) is represented by L2 in Fig. 3-28. At B the
The capacitive reactance of an open-circuited line
normaUzed Une-input impedance may be read as 0
less than a quarter wave in length is
- /0.41 ; the reactance in this case is capacitive, 600
X 0.41 = 246 ohms. (Line losses are disregarded in Xc (ohms) = Zo cot I
these examples.) From Fig. 3-28 it is easy to At lengths of line that are exact multiples of a
visuaUze that if LI were to be extended by a quarter wavelength, such lines have the properties
quarter wavelength, represented by L3, the line- of resonant circuits. At lengths where the input
input impedance would be identical to that ob¬ reactance passes through zero at the top of the
tained in the case represented by L2 alone. In the Smith Chart, the line acts like a series-resonant
case of L2, the line is open-circuited at the far end, circuit, as shown at D of Fig. 3-27. At lengths for
but in the case of L3 the Une is terminated in a which the reactances theoretically pass from “posi¬
short. The added quarter wavelength of Une for L3 tive” to “negative” infinity at the bottom of the
provides the “transformer action” discussed in the Smith Chart, the line simulates a parallel-resonant
previous section. circuit, as at C of Fig. 3-27. The effective Q of such
The equivalent inductance and capacitance as linear resonant circuits is very high if the tine
determined above can be found by substituting losses, both in resistance and by radiation, are kept
these values in the formulas relating inductance down. This can be done without much difficulty,
and capacitance to reactance, or by using the particularly in coaxial lines, if air insulation is used
various charts and calculators available. The fre¬ between the conductors. Air-insulated open-wire
quency corresponding to the tine length in degrees lines are likewise very good at frequencies for
must be used, of course. In this example, if the which the conductor spacing is very small in terms
frequency is 14 MHz the equivalent inductance and of wavelength.
capacitance in the two cases are 16.4 mH and 46.2 Applications of line sections as circuit elements
pF, respectively. Note that when the tine length is in connection with antenna and transmission-line
45 degrees (0.125 wavelength) the reactance in systems are discussed later in this chapter.
 88 Chapter 3

LINE CONSTRUCTION AND OPERATING CHARACTERISTICS

The two basic types of transmission lines, Although once in universal use, such lines have
parallel-conductor and coaxial, can be constructed now been largely superseded by prefabricated lines.
in a variety of forms. Both types can be divided Prefabricated open-wire lines (sold principally
into two classes: those in which the majority of the for television receiving applications) are available in
insulation between the conductors is air, only the nominal characteristic impedances of 450 and 300
minimum of solid dielectric necessary for me¬ ohms. The spacers, of low-loss material such as
chanical support being used; and those in which polystyrene, are molded on the conductors at
the conductors are imbedded in and separated by a relatively small intervals so there is no tendency for
solid dielectric. The former class (air¬
insulated) has the lowest loss per unit
length because there is no power loss in
dry air so long as the voltage between
conductors is below the value at which Wire
size
corona forms. At the maximum power
permitted in amateur transmitters it is
seldom necessary to consider corona
unless the SWR on the line is very high.

AIR-INSULATED LINES
A typical type of construction used
for parallel-conductor or “two-wire” air¬
insulated transmission lines is shown in
Fig. 3-29. The two line wires are sup¬
ported a fixed distance apart by means
of insulating rods called spacers. Spacers
may be made from insulating material,
such as phenolic, or can be purchased
ready-made. Materials commonly used in
manufactured spacers are isolantite,
Lucite, and polystyrene. The spacers
used vary from two to six inches, the
smaller spacings being desirable at the
higher frequencies (28 MHz) so that
radiation will be minimized. It is neces¬
sary to use the spacers at small enough
intervals along the line to prevent the
two wires from swinging appreciably 8 2 3 4 S 6 78 910 <2 IS
with respect to each other in a wind. For amateur SPACING (S) INCHES, CENTER. TO CENTER
purposes, lines using this construction ordinarily
have No. 12 or No. 14 conductors, and the Fig. 3-30 — Characteristic impedance vs. conductor
characteristic impedance is from 500 to 600 ohms. size and spacing for parallel-conductor lines.

NanorU
the conductors to swing with respect to each other.
ï A conductor spacing of one inch is used in the
“450-ohm” line and 1/2 inch in the “300-ohm”
line. The conductor size is usually about No. 18.
The impedances of such lines are somewhat lower
than given by Fig. 3-30 for the same conductor size
and spacing, because of the effect of the dielectric
constant of the numerous spacers used. The attenu¬
ation is quite low and lines of this type are entirely
satisfactory for transmitting applications at ama¬
teur powers.
When an air-insulated line having still lower
Fig. 3-29 — Typical open-wire line construction. characteristic impedance is needed, metal tubing
Commercial spacers are usually provided with having a diameter from 1/4 to 1/2 inch is frequent¬
grooved ends for the line conductors. The conduc¬ ly used. With the larger conductor diameter and
tor is held in place by a tie wire anchored in a hole relatively close spacing it is possible to build a line
near the groove. having a characteristic impedance as low as about
Transmission Lines 89

200 ohms. This type of construction is used

principally for quarter-wave matching transformers
at the higher frequencies.

Characteristic Impedance
The characteristic impedance of an air-insulated
parallel-conductor line, neglecting the effect of the
insulating spacers, is given by:
Zo = 276 log I

where Zo = Characteristic impedance

b= Center-to-center distance between Fig. 3-31 — Construction of air-insulated coaxial
conductors lines.
a= Radius of conductor (in same units as
b)
It does not matter what units are used fora and b construction, provided the air inside the line can be
so long as they are the same units. Both quantities kept dry. This usually means that air-tight seals
may be measured in centimeters, inches, etc. Since must be used at the ends of the line and at every
it is necessary to have a table of common loga¬ joint.
rithms to solve practical problems, the solution is The characteristic impedance of an air-insulated
given in graphical form in Fig. 3-30 for a number coaxial line is given by the formula
of common conductor sizes.

Zo = 138 log £
Four-Wire Lines
Another type of parallel-conductor line that is where Zo = Characteristic impedance
useful in some special applications is the four-wire b= Inside diameter of outer conductors
line. In cross-section, the conductors of the four- a= Outside diameter of inner conductor
wire line are at the corners of a square, the spacings (in same units as b)
being of the same order as those used in two-wire Again it does not matter what units are used for b
lines. The conductors at opposite corners of the and a, so long as they are the same. Curves for
square are connected together to operate in paral¬ typical conductor sizes are given in Fig. 3-32.
lel This type of line has a lower characteristic The formula and curves for coaxial lines are
impedance than the simple two-wire type. Also, approximately correct for lines in which bead
because of the more symmetrical construction it is spacers are used, provided the beads are not too
better balanced, electrically, to ground and other closely spaced.
objects that may be close to the line. The spacers
for a four-wire line may be disks of insulating
material, X-shaped members, etc.

Coaxial Lines
In coaxial lines of the air-insulated type a
considerable proportion of the insulation between
conductors may actually be a solid dielectric,
because of the necessity for maintaining constant
separation between the inner and outer conduc¬
tors. This is particularly likely to be true in
small-diameter lines, typical construction of which
is shown in Fig. 3-31. The inner conductor, usually
a solid copper wire, is supported by insulating
beads at the center of the copper-tubing outer
conductor. The beads usually are isolantite and the
wire is generally crimped on each side of each bead
to prevent the beads from sliding. The material of
which the beads are made, and the number of them
per unit length of line, will affect the characteristic
impedance of the line. The greater the number of
beads in a given length, the lower the characteristic
impedance compared with the value that would be
obtained with air insulation only. The presence of
the solid dielectric also increases the losses in the
line. On the whole, however, a coaxial line of this
type tends to have lower actual loss, at frequencies Fig. 3-32 — Characteristic impedance of typical
up to about 100 MHz, than any other line air-insulated coaxial lines.
90 Chapter 3

FLEXIBLE LINES tors molded diametrically opposite each other in

the walls. This increases the leakage path across the
Transmission lines in which the conductors are dielectric surface. Also, much of the electric field
separated by a flexible dielectric have a number of between the conductors is in the hollow center of
advantages over the air-insulated type. They are the tube, which can be protected from dirt and
less bulky, weigh less in comparable types, main¬ weather by closing the exposed end of the tube to
tain more uniform spacing between conductors, are make it watertight. This type of line is fairly
generally easier to install, and are neater in impervious to weather effects. Care should be used
appearance. Both parallel-conductor and coaxial when installing it, however, to make sure that any
lines are available with this type of insulation. moisture that condenses on the inside with changes
The chief disadvantage of such lines is that the in temperature and humidity can drain out the
power loss per unit length is greater than in bottom end of the tube and not be trapped in one
air-insulated lines. The power loss causes heating of section. This type of line is made in two conductor
the dielectric, and if the heating is great enough — sizes (with different tube diameters), one for
as it may be with high power and a high standing¬ receiving applications and the other for transmit¬
wave ratio - the line may break down both ting.
mechanically and electrically. The transmitting-type 75-ohm Twin-Lead uses
stranded conductors about equivalent to solid No.
300-Ohm 12 wire, with quite close spacing between conduc¬
Twin-Lead tors. Because of the close spacing most of the field
is confined to the solid dielectric, very little
existing in the surrounding air. This makes the
is-Ohm 75-ohm line much less susceptible to weather
Twin-Lead effects than the 300-ohm ribbon type.
In addition to the parallel-conductor types
Polyethylene insulation. described above, there are also lightweight two-
Solid wire lines of 150 and 75 ohms. These are useful for
Coax receiving antennas, but are not heavy enough to
Vinyl outer coveriny carry very much power.
Braid outer conductor
-Inner conductor Coaxial Lines
Flexible “coax” is available in a rather large
Fig. 3-33 — Construction of parallel-conductor and
coaxial lines with solid dielectric. number of different types. However, the basic
construction is the same in all, and is typified in
the drawing in Fig. 3-33. The overall diameter
Parallel-Conductor Lines varies from a little less than 1/4 inch to somewhat
over an inch, depending chiefly on the power
The construction of a number of types of requirements for which the cable was designed. In
flexible lines is shown in Fig. 3-33. In the most some cables the inner conductor is stranded; in
common 300-ohm type (“Twin-Lead” is one trade others, solid wire is used. In some the outer
name) the conductors are stranded wire equivalent conductor is a single braid; in others it is double.
to No. 20 in cross-sectional area and are molded in The outer jacket, usually vinyl plastic, plays no
the edges of a polyethylene ribbon about a half part in the electrical performance of the cable, but
inch wide. The effective dielectric is partly solid is simply a waterproof covering.
and partly air. The presence of the solid dielectric The dielectric material used in these cables is
lowers the characteristic impedance of the line as polyethylene, a flexible plastic having low losses at
compared with the same conductors in air, the radio frequencies. In many types of cable (particu¬
result being that the impedance is approximately larly the RG series) the dielectric is solid poly¬
300 ohms. The fact that part of the field between ethylene. The losses in the solid dielectric are
the conductors exists outside the solid dielectric relatively low, if the operating frequency is not too
leads to an operating disadvantage in that dirt or high. A few types of cable (RG-62/U is an
moisture on the surface of the ribbon tends to example) use a polyethylene thread wound around
change the characteristic impedance. The operation the inner conductor; this makes the insulation
of the line is therefore affected by weather partly air, effecting a reduction in both loss and
conditions. The effect will not be very serious in a capacitance per un length. A popular type of
line terminated in its characteristic impedance, but dielectric in the more recently developed cables is
if there is a considerable standing-wave ratio a “foamed” polyethylene, which has a cellular struc¬
small change in Zo may cause wide fluctuations of ture similar to a honeycomb so that a large part of
the input impedance. Weather effects can be the dielectric is air. The loss per unit length in such
minimized by cleaning the line occasionally and cables is appreciably less than in those using solid
giving it a thin coating of a water-repellent material dielectric. The foam-polythylene dielectric is sus¬
such as silicone grease or automobile wax. ceptible to moisture penetration, which will cause
To overcome the effects of weather on the an increase in the attenuation.
characteristic impedance and attenuation of Solid coaxial cables are available in two princi¬
ribbon-type line, another type of Twin-Lead is pal characteristic impedances, approximately 50
made using a polyethylene tube with the conduc¬ and 75 ohms.
 Transmission Lines
TABLE 3-1 - TRANSMISSION LINES

1 2 3 4 5 6 7 8 9 10 11 12

Inner Cap. Power Rating, Watts

Nominal RG/U Outside Max. rms Connector
Mfrs. Jacket Cond. Velocity Up to
Type of Line Imp., Dia Dielectric per ft. Series
Type No. Size Factor Voltage
ohms inches pF 30 MHz 400 MHz
Flexible Coaxial 52 8 .405 7/21 SP 29.5 .66 5000 1720 465 UHF, N
Medium 52 8A .405 HA 7/21 SP 29.5 .66 5000 1720 465 UHF, N
50 621-111 .405 7/19 FP 24.5 .80 1500 UHF
50 T-4-50 .407 10 FP UHF
75 11 .405 7/26 SP 20.5 .66 5000 1400 340 UHF, N
75 11A .405 IIA 7/26 SP 20.5 .66 5000 1400 340 UHF, N
75 621-100 .405 I 14 FP 16.5 .80 3000 UHF
75 JT-204 .407 14 FP UHF

Small 53.5 58 .195 I 20 SP 28.5 .66 1900 580 135 UHF, BNC, N
50 58A .195 I 19/.0071 SP 30 .66 1900 550 105 UHF, BNC, N
53.5 58B .195 IIIA 20 SP 28.5 .66 1900 580 135 UHF, BNC, N
50 58C .195 IIA 19/.0071 SP 30 .66 1900 550 105 UHF, BNC, N
73 59 .242 I 22 cw SP 21.5 .66 2300 720 185 UHF, BNC, N
75 59B .242 IIA .023 cw SP 21 .66 2300 720 185 UHF. BNC. N
73 621-186 .242 P 20 cw FP 17.3 .80 1000 UHF. BNC
93 62 .242 I 22 cw SSP 13.5 .84 750 850 230 UHF. BNC. N

Parallel Conductor 75 214-023 7/21 SP 20 .71 1000

Flat or 300 214-056 7/28 SP 5.8 .82
Oval 300 214-022 16 cw SP 3.0 .82

Tubular 300 214-271 7/28 PA .82 500

300 214-076 7/26 PA 3.9 .82 1000
300 214-103 7/28 FP*

Column 3: T-4-50 and JT-204 are manufac¬ Column 6: Conductors are copper unless fol¬ closure solid polyethylene. Type 214-103 is intend¬
tured by Times Wire & Cable, Wallingford, Conn. lowed by CW (copper-weld). Decimal numbers give ed for use under adverse moisture and salt-spray
Other numbers are types made by Amphenol, wire diameter in inches; others are standard conditions. PA - Polyethylene tube with air core.
Chicago, Ill. copper-wire gauge except when preceding a virgule, Column 9: Open parallel-conductor line has a
Column 5: I Polyvinyl chloride (PVC), black. when the figure indicates number of strands: e.g., velocity factor of 0.95 to 0.975, depending on
HA Noncontaminating PVC, black or gray. H1A 7/21 means 7 strands of No. 21 copper wire. number of spacers and dielectric material of which
Polyethylene, black. Noncontaminating and Column 7: SP - Solid polyethylene. SSP - they are made. Polyethylene spacers used in types
abrasion-resistant. Recommended when cable is to Polyethylene strand wound around inner conduc¬ listed.
be buried underground. P - Polyethylene. X — tor; enclosed in solid tube of same material. FP — Column 12: Only connectors in common use
Xelon. Foamed polyethylene. FP* — Foamed poly¬ by amateurs are included.
ethylene surrounding each conductor; outer en¬
92 Chapter 3

If the entire field between the two conductors of the radiating antenna system to a greater or
of a line is in a solid dielectric, as in the case of lesser extent.
solid coaxial lines, the characteristic impedance of
the line is reduced by the factor 1/^/Fas compared ELECTRICAL LENGTH
with the impedance of an air-insulated line having
Whenever reference is made to a Une as being so
the same conductor size and spacing. The quantity
many wavelengths (such as a “half wavelength” or
kis the effective dielectric constant of the insulat¬
“quarter wavelength”) long, it is to be understood
ing material. In ribbon or tubular type parallel¬
that the electrical length of the line is meant. The
conductor Unes and in the special coaxial types physical length corresponding to an electrical
mentioned above, the field is partly in air and wavelength is given by
partly in the dielectric, so the redwtion factor
above cannot be applied directly. 004
Length (feet) = V
The attenuation and other characteristics of the
various types of lines commonly used by amateurs where / = Frequency in megahertz
are shown in Table 3-1 and Fig. 3-34. V = Velocity factor
A whole series of fittings for making detachable The velocity factor is the ratio of the actual
connections to flexible coaxial cable is available. velocity along the line to the velocity in free space.
These include general-purpose connectors, some of Values of V for several common types of lines are
which are quite inexpensive, and “constant¬ given in Table 3-1.
impedance” units especially designed so that Because a quarter-wavelength line is frequently
lengths of cable can be sphced together or ter¬ used as an impedance transformer, it is convenient
minated without causing a change in the charac¬ to calculate the length of a quarter-wave Une
teristic impedance. Such impedance “bumps” directly. The formula is
along a Une correspond in a general way to having a 246
load that is not matched to the line; that is, they Length (feet) = -y V
wiU cause some of the outgoing power to be
reflected back toward the input end. In most LINE INSTALLATION
amateur appUcations it is not necessary to worry
about such impedance discontinuities when using One great advantage of coaxial line, particularly
ordinary connectors because their effect at fre¬ the flexible dielectric type, is that it can be
quencies below 300 or 400 MHz is too small to be instaUed with almost no regard for its surround¬
of practical consequence. ings. It requires no insulation, can be run on or in
the ground or in piping, can be bent around
corners with a reasonable radius, and can be
SINGLE-WIRE LINE “snaked” through places such as the space between
There is one type of Une, in addition to those waUs where it would be impracticable to use other
already described, that deserves some mention types of lines. However, coax Unes always should
since it is still used to a limited extent. This is the be operated in systems that permit a low standing¬
single-wire line, consisting simply of a single wave ratio, and precautions must be taken to
conductor running from the transmitter to the prevent rf currents from flowing on the outside of
antenna. The “return” circuit for such a Une is the the Une. This point is discussed later in this
earth; in fact, the second conductor of the line can chapter.
be considered to be the image of the actual
conductor in the same way that an antenna strung Coaxial Fittings
above the earth has an image (see Chapter Two). There is a wide variety of fittings and connec¬
The characteristic impedance of the single-wire line tors designed to go with various sizes and types of
depends on the conductor size and the height of solid-dielectric coaxial line. The “UHF” series of
the wire above ground, ranging from 500 to 600 fittings is by far the most widely used type in the
ohms for No. 12 or No. 14 conductors at heights amateur field, largely because they have been
of 10 to 30 feet. By connecting the Une to the available for a long time and have been quite
antenna at a point that represents a resistive inexpensive on the surplus market. These fittings,
impedance of 500 to 600 ohms the line can be typified by the PL-259 plug and SO-239 chassis
matched and will operate without standing waves. fitting (Armed Services numbers) are quite ade¬
Although the single-wire line is very simple to quate for vhf and lower frequency applications,
instaU, it has at least two outstanding disadvan¬ but are not weatherproof.
tages. Since the return circuit is through the earth, The “N” series fittings are designed to maintain
the behavior of the system depends on the kind of constant impedance at cable joints. They are a bit
ground over which the antenna and transmission harder to assemble than the “UHF” type, but are
line are erected. In practice, it may not be possible better for frequencies above 300 MHz or so. These
to get the necessary good connection to actual fittings are weatherproof.
ground that is required at the transmitter. Second, The “BNC” fittings are for small cable such as
the line always radiates since there is no nearby RG-58/U, RG-59/U and RG-62/U. They feature a
second conductor to cancel the fields. The radia¬ bayonet-locking arrangement for quick connect
tion will be minimum when the Une is properly and disconnect, and are weatherproof.
terminated because the line current is lowest under Methods of assembling the connectors to the
those conditions. However, the Une is always a part cable are shown on accompanying pages.
Transmission Lines 93

Fig. 3-34 - Nominal attenuation in decibels per 100 feet of various types of transmission line. Total
attenuation is directly proportional to length. Attenuation will vary somewhat in actual cable samples,
and generally increases with age in coaxial cables having a type I jacket. Cables grouped together in thé
above chart have approximately the same attenuation. Types having foam polyethylene dielectric have
slightly lower loss than equivalent solid dielectric types, when not specifically shown above. The curve
for RG-58/U also applies to RG-58B/U. For RG-58A/U and RG-58C/U add 10 percent to the loss in
decibels given by the curve. RG-8A/U has the same loss as RG-8/U; RG-59B/U has the same loss as
RG-59.

Parallel-Wire Lines
point where the coupling occurs. Reflections take
In installing a parallel-wire line, care must be place from this coupled “load,” raising the stand¬
used to prevent it from being affected by moisture, ing-wave ratio. The effect is at its worst when one
snow and ice. In home construction, only spacers line wire is closer than the other to the external
that are impervious to moisture and are unaffected conductor. In such a case one wire carries a heavier
by sunlight and the weather should be used on load than the other, with the result that the line
air-insulated lines. Steatite spacers meet this re¬ currents are no longer equal. The line then be¬
quirement adequately, although they are somewhat comes “unbalanced.”
heavy. The wider the line spacing the longer the Solid-dielectric two-wire lines have a relatively
leakage path across the spacers, but this cannot be small external field because of the small spacing
carried too far without running into line radiation, and can be mounted within a few inches of other
particularly at the higher frequencies. Where an conductors without much danger of coupling
open-wire line must be anchored to a building or between the line and such conductors. Standoff
other structure, standoff insulators of a height insulators are available for supporting lines of this
comparable with the line spacing should be used if type when run along walls or similar structures.
mounted in a spot that is open to the weather. Sharp bends should be avoided in any type of
Lead-in bushings for bringing the line into a transmission line, because such bends cause a
building also should have a long leakage path. change in the characteristic impedance. The result
The line should be kept away from other is that reflections take place from each bend. This
conductors, including downspouting, metal win¬ is of less importance when the SWR is high than
dow frames, flashing, etc., by a distance equal to when an attempt is being made to match the load
two or three times the line spacing. Conductors to the line’s characteristic impedance. It may be
that are very close to the line will be coupled to it impossible to get the SWR down to a desired figure
in greater or lesser degree, and the effect is that of until the necessary bends in the line are made more
placing an additional load across the line at the gradual.
94
Chapter 3

BNC (UG-88/U) CONNECTORS 2) Bare 5/8" of

Connectors bearing suffix letters (UG-88C/U, etc.) center conductor
differ slightly in internal construction; assembly — don't nick con¬
and dimensions must be varied accordingly. ductor. Trim
braided shield
9/16" and tin.
1) Cut end square Slide coupling ring
and trim jacket 5/16" on cable.
for RG-58/U. 3) Screw the plug
assembly on cable.
Solder plug as¬
2) Fray shield and sembly to braid
strip inner dielectric through solder
1/8". Tin center con¬ SOLDI* HOLL holes. Solder con¬
ductor. ductor to contact
sleeve. Screw
coupling ring on
3) Taper braid and assembly.
slide nut (A), washer N (UG-21/U) CONNECTORS
(B), gasket (C>, and
clamp (D), over braid. 1) Remove 9/16"
Clamp is inserted so of vinyl jacket.
that its inner shoulder When using
fits squarely against d ouble-shielded
end of cable jacket. cable rem ove
5/8".
2) Comb out cop¬
4) With clamp in per braid as
place, comb out shown. Cut off di-
braid, fold back electric 7/32"
smooth as shown, and from end. Tin cen¬
trim 3/32" from end. ter conductor.

3) Taper braid as
5) Tin center conduc¬ shown. Slide nut,
tor of cable. Slip fe¬ washer and gasket
male contact in place over vinyl jacket.
and solder. Remove Slide clamp over
excess solder. Be sure cable dielectric is not heated braid with internal shoulder of clamp flush against
excessively and swollen so as to prevent dielectric end of vinyl jacket. When assembling connectors
entering body. with gland, be sure knife-edge is toward end of
cable and groove in gasket is toward the gland.

6) Push into body as

far as it will go. Slide 4) Smooth braid
nut into body and back over clamp
screw into place with and trim. Soft-
wrench until tight. solder contact to
Hold cable and shell center conductor.
rigidly and rotate nut. Avoid use of excessive heat and solder. See that
end of dielectric is clean. Contact must be flush
against dielectric. Outside of contact must be free
of solder. Female contact is shown; procedure is
similar for male contact.
7) This assembly pro¬
cedure applies to BNC
jacks. The assembly
for plugs is the same 5) Slide body into
except for the use of place carefully so
male contacts and a that contact enters
plug body. hole in insulator
(male contact
shown). Face of dielectric must be flush against
insulator. Slide completed assembly into body by
83-1SP (PL-259) PLUG pushing nut. When nut is in place, tighten with
wrenches. In connectors with gland, knife edge
1) Cut end of cable should cut gasket in half by tightening sufficiently.
even. Remove vinyl
jacket 1-1/8" - don't From Armed Services Index of R.F. Transmission
nick braid. Lines and Pittings, ASESA 49-2B.
Transmission Lines 95

83 SERIES (SO-239) WITH HOODS

1 ) Cut end of cable 3) Remove braid to
even. Remove vinyl expose dielectric to
jacket to dimension appropriate dimen¬
appropriate for type sion. Tin center conductor. Soldering and assembly
of hood. Tin exposed depends on the hood used, as illustrated.
braid.
2) Remove braid and
dielectric to expose
center conductor. Do
not nick conductor.

4) Slide hood over braid. Solder conductor to

contact. Slide hood flush against receptacle and bolt
both to chassis. Solder hood to braid as illustrated.
Tape this junction if necessary. (For UG-177/U.)
SOLDER HOOD TO BRAID
SOLDER CONDUCTOR
TO CONTACT SLEEI'E

5) Slide hood over braid. Bring receptacle flush

against hood. Solder hood to braid and conductor to
contact sleeve through solder holes as illustrated.
Tape junction if necessary. (For UG-372/U.)

6) Slide hood over braid and force under vinyl.

Place inner conductor in contact sleeve and solder.
Push hood flush against receptacle. Solder hood to
braid through solder holes. Tape junction if neces¬
sary. (For UG-106/U.)

83-1SP (PL-259) PLUG WITH ADAPTERS (UG-176/U OR UG-175/U)

1 ) Cut end of cable even. Remove vinyl jacket 3/4"

— don't nick braid. Slide coupling ring and adapter
on cable.

2) Fan braid slightly and fold back over cable.

3) Position adapter to dimension shown. Press braid

down over body of adapter and trim to 3/8". Bare
5/8" of conductor. Tin exposed center conductor.

4) Screw the plug assembly on adapter. Solde- braid

to shell through solder holes. Solder conductor to
contact sleeve.

5) Screw coupling ring on plug assembly.

 96 Chapter 3

TESTING COAXIAL CABLE

Unknown coaxial cable or cable that has been

exposed to the weather may have losses above the
published figures for the cable type. A simple
method for checking the losses in a cable is to use
an rf ammeter (mounted in a metal box with coax
fittings). Connect one end of the cable to a
nonreactive dummy load of the same impedance as
the coax. At the other end of the Une insert the rf
ammeter and connect it to a transmitter. Tune up
the transmitter and make a note of the exact
amount of current. Without touching the transmit¬
ter tuning, move the ammeter to the other end of
the line, at the dummy load, and note the meter
reading. Using Fig. 3-35, compare the readings to
determine the decibel loss that is present in the
line. Keep in mind that the cable must be
AMPERES AT LINE INPUT terminated in its characteristic impedance (SWR of
1); otherwise, the data from Fig. 3-35 will not be
Fig. 3-35 — Graph for determining losses in accurate.
transmission lines with an SWR of 1.

COUPLING THE TRANSMITTER TO THE LINE

In any system using a transmission line to feed equivalent to the line input-impedance range asso¬
the antenna the load that the transmitter “sees” is ciated with a 2-to-l or perhaps as high as 3-to-l
the input impedance of the line. As shown earlier, SWR in the line. This does not mean that a coaxial
this impedance is completely determined by the line must be used to feed the antenna', any type of
line length, the Zo of the line, and the impedance line can be used.
of the load (the antenna) at the output end of the
line. The line length and Zo are generally matters The Matching System
of choice regardless of the type of antenna used. The basic system then is as shown in Fig. 3-36.
The antenna impedance, which may or may not be Assuming that the transmitter is capable of deliver¬
known accurately, is (with Zo) the factor that ing its rated power into a load of the order of 50 to
determines the standing-wave ratio. 75 ohms, the coupling problem is one of designing
The SWR can be measured with relative ease, a matching circuit that will transform the actual
and from it the limits of variation in the line input line input impedance into a resistance of 50 or 75
impedance can be determined with little difficulty. ohms. This resistance will be unbalanced - that is,
It may be said, therefore, that the problem of one side will be grounded - but the actual line to
transferring power from the transmitter to the line the antenna probably will be balanced (parallel¬
can be approached purely on the basis of the conductor line) when a matching circuit is re¬
known Zo of the line and the maximum SWR that quired. Unbalanced (coaxial) line should not re¬
may be encountered. quire a matching circuit, since it should be used
So far as the transmitter itself is concerned, the only when it is sufficiently well matched by the
requirements of present-day amateur operation antenna to operate at a low SWR - one well within
almost invariably include complete shielding and the coupling capabilities of the transmitter itself.
provision for the use of low-pass filters to prevent Several types of matching circuits are available.
harmonic interference with television reception. In With some, such as an £ network using fixed-value
almost all cases this means that the use of coaxial components, it is necessary to know the actual
cable at the output of the transmitter is mandatory line-input impedance with fair accuracy in order to
because it is inherently shielded. Thus the transmit¬ arrive at a proper design. This information is not
ter output circuit should be designed to deliver full essential with the circuits described in the next
power into 50-75 ohms, with enough leeway to section, since they are capable of adjustment over a
take care of reasonable variations in load — wide range.
UrtHHA

Fig. 3-36 — Essentials of coup¬

ling system between transmit¬
ter and transmission line.
Transmission Lines 97
the value of Q is equivalent to tightening the
coupling, and vice versa.
If L2 does not have the optimum value, the
match may still be obtained by adjusting kand Q,
but one or the other - or both - must have a
larger value than is needed when X L2 is equal to
R2. In general, it is desirable to use as low a value
of Q as is practicable, since low Q values mean that
the circuit requires little or no readjustment when
shifting frequency within a band (provided RI
does not vary appreciably with frequency).

Circuit Q
In Fig. 3-37A, Q is equal to RI in ohms divided
by the reactance of Cl in ohms, assuming LI-Cl is
tuned to the operating frequency. This circuit is
suitable for comparatively high values of RI -
from several hundred to several thousand ohms. In
Fig. 3-37C, Q is equal to the reactance of Cl
divided by the resistance of RI, Ll-Cl again being
tuned to the operating frequency. This circuit is
Fig. 3-37 — Circuit arrangements for inductively suitable for low values of RI — from a few ohms
coupled impedance-matching circuit. A and B use a up to a hundred or so ohms. In Fig. 3-37B the Q
parallel-tuned coupling tank; B is equivalent to A depends on the placement of the taps on LI as well
when the taps are at the ends of L1. The as on the reactance of Cl. This circuit is suitable
series-tuned circuit at C is useful for very low
values of load resistance, R1. for matching all values of RI likely to be encoun¬
tered in practice.
Note that to change Q in either A or C, Fig.
3-37, it is necessary to change the reactance of Cl.
MATCHING WITH INDUCTIVE COUPLING Since the circuit is tuned essentially to resonance
Inductively coupled matching circuits are at the operating frequency, this means that the L/C
shown in basic form in Fig. 3-37. RI is the actual ratio must be varied in order to change Q. In Fig.
load resistance to which the power is to be 3-37B a fixed L/C ratio may be used, since Q can
delivered, and R2 is the resistance seen by the be varied by changing the tap positions. The Q will
power source. R2 depends on the circuit design increase as the taps are moved closer together, and
and adjustment; in general, the objective is to make will decrease as they are moved farther apart on
it equal to 50 or 75 ohms. LI and Cl form a LI.
resonant circuit capable of being tuned to the
operating frequency. The coupling between LI and
L2 is adjustable.
The circuit formed by Cl, LI and L2 is
equivalent to a transformer having a pnmary-to-
secondary impedance ratio adjustable over wide
Umits. The resistance “coupled into” L2 from LI
depends on the effective Q of the circuit
L1-C1-R1, the reactance of L2 at the operating
frequency, and the coefficient of coupling, k,
between the two coils. The approximate relation¬
ship is (assuming Cl is properly tuned)

R2=* ,Xl2 <?

where %l2 > s the reactance of L2 at the operating

frequency. The value of L2 is optimum when XL2
= R2, in which case the desired value of R2 is
obtained when

This means that the desired value of R2 may be

obtained by adjusting either the coupling, k, Fig. 3-38 — Line input impedances containing both
between the two coils, or by changing the Q of the resistance and reactance can be represented as
circuit L1-C1-R1 - or, if necessary, by doing both. shown enclosed in dotted lines, for capacitive
If the coupling is fixed, as is often the case, Q must reactance. If the reactance is inductive, a coil is
be adjusted to attain a match. Note that increasing substituted for the capacitance C.
98 Chapter 3

Reactive Loads - Series onance by readjusting Cl, as compared with the

and Parallel Coupling setting it would have with a purely resistive load.
More often than not, the load represented by Such a situation may occur when the load
the input impedance of the transmission line is reactance is low compared with the resistance in
reactive as well as resistive. In such a case the load the parallel-equivalent circuit, or when the
cannot be represented by a simple resistance, such reactance is high compared with the resistance in
as RI in Fig. 3-37. As stated earlier in this chapter, the series-equivalent circuit. The very considerable
we have the option of considering the load to be a detuning of the circuit that results is often ac¬
resistance in parallel with a reactance, or as a companied by an increase in Q, sometimes to
resistance in series with a reactance (Fig. 3-15). In values that lead to excessively high circulating
Figs. 3-37A and B it is convenient to use the currents in the circuit. This causes the efficiency to
parallel equivalent of the line input impedance. suffer. (Ordinarily the power loss in matching
The series equivalent is more suitable for Fig. circuits of this type is inconsequential, if the Q is
3-37C. below 10 and a good coil is used.) An unfavorable
Thus in Fig. 3-38 at A and B the load might be ratio of reactance to resistance in the input
represented by RI in parallel with the capacitive impedance of the line can exist if the SWR is high
reactance C, and in Fig. 3-3 8C by RI in series with and the line length is near an odd multiple of
a capacitive reactance C. In A, the capacitance Cis one-eighth wavelength (45 degrees).
in parallel with Cl and so the total capacitance is
the sum of the two. This is the effective capaci¬ Q of Line Input Impedance
tance that, with LI, tunes to the operating The ratio between reactance and resistance in
frequency. Obviously the setting of Cl will be at a the equivalent input circuit - that is, the Q of the
lower value of capacitance with such a load than it Une input impedance - is a function of line length
would with a purely resistive load such as is shown and SWR. There is no specific value of this Q of
in Fig. 3-37A. which it can be said that lower values are satis¬
In Fig. 3-38B the capacitance of C also in¬
creases the total capacitance effective in tuning the factory while higher values are not. In part, the
circuit. However, in this case the increase in maximum tolerable value depends on the tuning
effective tuning capacitance depends on the range available in the matching circuit. If the
positions of the taps; if the taps are close together tuning range is restricted (as it will be if the
the effect of C on the tuning is relatively small, but variable capacitor has relatively low maximum
it increases as the taps are moved farther apart. capacitance), compensating for the line input
reactance by absorbing it in the matching circuit -
In Fig. 3-38C, the capacitance C is in series with
that is, by retuning Cl in Fig. 3-38 - may not be
Cl and so the total capacitance is less than either.
possible. Also, if the Q of the matching circuit is
Hence the capacitance of Cl has to be increased in
low the effect of the line input reactance will be
order to resonate the circuit, as compared with the
purely resistive load shown in Fig. 3-37C.
If the reactive component of the load im¬
pedance is inductive, similar considerations apply.
In such case an inductance would be substituted
for the capacitance C shown in Fig. 3-38. The
effect in Figs. 3-38A and B would be to decrease
the effective inductance in the circuit, so Cl would
have to have a larger value of capacitance in order
to resonate the circuit to the operating frequency.
In Fig. 3-38C the effective inductance would be
increased, thus making it necessary to set Cl at a
lower value of capacitance for resonating the
circuit.

Effect of Line Reactance on Circuit Q

The presence of reactance in the line input
impedance can affect the Q of the matching
circuit. If the reactance is capacitive, the Q will not
change if resonance can be maintained by ad¬
justment of Cl without changing either the value
of LI or the position of the taps in Fig. 3-38B (as
compared with the Q when the load is purely
resistive and has the same value of resistance, RI).
If the load reactance is inductive the L/C ratio
changes because the effective inductance in the
circuit is changed and, in the ordinary case, LI is
not adjustable. This increases the Q in all three
circuits of Fig. 3-38. Fig. 3-39 — Compensating for react¬
When the load has appreciable reactance it is ance present in the line input imped¬
not always possible to adjust the circuit to res¬ ance.
Transmission Lines 99

greater than it will when the matching-circuit Q is

high.
As stated earlier, the optimum matching-circuit
design is one in which the Q is low, Le., a low
reactance-to-resistance ratio.

Compensating for Input Reactance Fig. 3-40 — Using a variable capacitance, C2, as an
When the reactance/resistance ratio in the line alternative to variable mutual inductance between
input impedance is unfavorable it is advisable to L1 and L2.
take special steps to compensate for it. This can be
done as shown in Fig. 3-39. Compensation consists is equal to the line Zo. To utilize a capacitor of
of supplying external reactance of the same more convenient size - maximum capacitance of
numerical value as the line reactance, but of the perhaps 250-300 pF - a value of inductance may
opposite kind. Thus in A, where the line input be used for L2 that will resonate at the lowest
impedance is represented by resistance and capa¬ frequency with the maximum capacitance
citance in parallel, an inductance L having the same available.
numerical value of reactance as C can be connected On the higher frequency bands the problem of
across the line terminals to “cancel out” the line variable capacitors does not arise since a reactance
reactance. (This is actually the same thing as tuning of 50 to 75 ohms is within the range of con¬
the line to resonance at the operating frequency.) ventional components.
Since the parallel combination of L and C is
equivalent to an extremely high resistance at Circuit Balance
resonance, the input impedance of the line Fig. 3-40 shows Cl as a balanced or split-stator
becomes a pure resistance having essentially the capacitor. This type of capacitor is desirable in a
same resistance as RI alone. practical matching circuit to be used with a
The case of an inductive line is shown at B. In balanced line, since the two sections are sym¬
this case the external reactance required is capa¬ metrical. In the ordinary single-section capacitor
citive, of the same numerical value as the reactance there is more capacitance to ground (or metal
of L. objects, such as a chassis, in the vicinity) from the
Where the series equivalent of the line input frame and rotor assembly than from the stator
impedance is used the external reactance is con¬ assembly. The rotor assembly of the balanced
nected in series, as shown at C and D in Fig. 3-39. capacitor may be grounded, if desired, or it may be
In general, these methods need not be used left “floating” and the center of LI may be
unless the matching circuit does not have sufficient grounded; or both may “float.” Which method to
range of adjustment to provide compensation for use depends on considerations discussed later in
the line reactance as described earlier, or when connection with antenna currents on transmission
such a large readjustment is required that the lines.
matching-circuit Q becomes undesirably high. The As an alternative to using a split-stator type of
latter condition usually is accompanied by heating capacitor, a single-section capacitor may be used,
of the coil used in the matching network. and a balun employed to couple the matching
network to the load, Baluns are discussed in a later
Methods for Variable Coupling section of this chapter.
The coupling between LI and L2, Figs. 3-37
and 3-38, preferably should be adjustable. If the A UNIVERSAL MATCHING NETWORK
coupling is fixed, such as with a fixed-position link, The circuit shown at A in Fig. 3-41, named the
the placement of the taps on LI for proper Ultimate Transmatch, may be used as a “universal”
matching becomes rather critical. The additional matching network. With reasonable component
matching adjustment afforded by adjustable values it will match impedances from quite low
coupling between the coils facilitates the matching values, an ohm or two, to quite high values, several
procedure considerably. L2 should be coupled to thousand ohms, whether or not reactances are
the center of LI for the sake of maintaining present in the load. Construction details for this
balance, since the circuit is used with balanced type of circuit are given in Chapter Six.
lines. Depending upon the settings of C1, C2, and L1,
If adjustable inductive coupling such as a the circuit may have a number of equivalents. A
swinging link is not feasible for mechanical reasons, basic equivalent is shown at B of Fig. 3-41, where
an alternative is to use a variable capacitor in series R and Cl, together, represent Cl alone in the
with L2. This is shown in Fig. 3-40. Varying C2 drawing at A. LI and Cl form a parallel-resonant
changes the total reactance of the circuit formed circuit; if tuned to resonance at the operating
by L2-C2, with much the same effect as varying frequency, they may be represented by a pure
the actual mutual inductance between LI and L2. resistance. Cl, a differential capacitor, provides the
The capacitance of C2 should be such as to capability of a continuously adjustable tap across
resonate with L2 at the lowest frequency in the the parallel combination of Cl and LI, represented
band of operation. This calls for a fairly large value by R at B.
of capacitance at low frequencies (about 1000 pF When the terminating impedance is low, op¬
at 3.5 MHz for 50-ohm line) if the reactance of L2 timum settings for a match will result when the
100 Chapter 3

adjusted for a “match.” As shown at C, the load

impedance is transformed up to a somewhat higher
value in the equivalent of an ¿ network, with the
setting of Cl providing the proper “tap” point for
the coax-termination impedance to be in the
50-ohm range.
With high-impedance resistive loads, correct
matching will be obtained with LI and Cl
resonated near the frequency of operation and
with C2 adjusted for maximum capacitance. If the
reactance of C2 at the operating frequency is low,
less than the total load impedance, the effects of
C2 in the circuit may, for the moment, be
disregarded. The resulting equivalent circuit is
shown at D of Fig. 3-41. The Q, and thus the
impedance of the parallel-resonant circuit, Ll-Cl,
HIGH ZO- -o may be adjusted by altering the L/C ratio of the
circuit while maintaining resonance. At the proper
INTERMEOIATE- LOW OR settings of LI and Cl, the impedance of the loaded
Z (50 OHMS) '-H INTERMEDIATE Z
parallel-resonant circuit will equal that of the
terminating impedance, Z, while the effective tap
C0MM0NO- -o point of differential capacitor Cl will provide a
50-ohm termination for the coax line. Reactances
in the load, including the practical effects of having
C2 in the circuit, may be tuned out with ap¬
propriate settings of LI and Cl to maintain overall
system resonance.
From the above paragraphs, it is easy to
visualize that with proper settings of LI, Cl, and
C2, the circuit of 3-41A will provide a match over
Fig. 3-41 — At A, the Ultimate Transmatch, a the continuous range of impedances between low
universal matching network. At B, C, and D, the and high.
various circuit-equivalent configurations which the
circuit at A may take, depending on the adjust¬
ment settings.
ADJUSTMENT OF MATCHING CIRCUITS
Adjustment of the matching circuit consists in
Ll-Cl combination is tuned for resonance at a
finding the proper settings for the taps on L1 (Fig.
frequency below that of the operating frequency.
3-40) or for the total inductance (Fig. 3-41), and
When so tuned, the combination will be inductive
at the operating frequency, represented by a coil in the proper settings of Cl and C2 (both circuits) so
C of Fig. 3-41. The circuit thus takes the that the line input impedance is transformed into a
equivalent form of an Lnetwork with the input 50- or 75-ohm load to match the Zo of the
being tapped down on the coil, as shown. (In particular type of coaxial cable used between the
¿-network matching, the higher of the two transmitter and matching circuit. The surest and by
impedances to be matched is connected to the far the simplest way to arrive at the correct
shunt-arm side of the network, and the lower adjustments is to use an SWR indicator designed
for the type of coax used.
impedance to the series-arm side.) Reactances in
the load, either inductive or capacitive, may be The setup using the SWR bridge is shown in
tuned out with C2. If the load itself contains Fig. 3-42. The matching circuit may be that of Fig.
capacitive reactance, less capacitive reactance 3-40 or 3-41. Adjustments may be made at any
(greater capacitance) will be required at C2 than if convenient power level within the power-handling
capabilities of the type of bridge used. Adjust the
the load were purely resistive. Conversely, less
transmitter output for a “forward” reading on the
capacitance will be required for the final setting of
bridge indicator well up on the indicator scale.
C2 if the load contains inductive reactance.
Then switch to read reflected power (or voltage,
For intermediate values of impedance, such as depending on the instrument calibration).
50 or 75 ohms, a matching network would not
ordinarily be required for most modern trans¬
mitters. However, it would be a nuisance in a Adjustment Procedure
multiband-antenna and matching-network system To adjust the circuit of Fig. 3-40, set the taps at
if it were necessary to disconnect and remove the trial positions equidistant from the center of the
matching network for one or two of the bands of coil, then vary Cl for minimum reading on the
operation and reinstall it for the others. If the bridge indicator. Next adjust C2 for the lowest
combination of antenna impedance and feed-line possible reading, then touch up Cl again with the
characteristics should happen to present a load to same objective. Continue until the indicator reads
the matching network in this intermediate¬ zero, or as close to it as possible. If a good null is
impedance range, the circuit of 3-41 may still be not obtained, try the taps at another position and
Transmission Lines 101

MATCHIH« CIRCUIT

Fig. 3-42 — Adjust¬

ment setup using
SWR indicator.

go through the procedure again. In most cases, it indicator will also result in the most efficient
will not be necessary to try many tap positions; in power transfer to the transmission line. However, it
fact, it is usually found that the tap positions are is possible that a good match will be accompanied
not at all critical. If C2 is not used but the coupling by excessive loss in the matching circuit. This is
between LI and L2 can be varied, the coupling unlikely to happen if the steps described for
adjustment takes the place of varying C2 in the obtaining a low Q are taken. If the settings are
above procedure. If neither C2 nor adjustable highly critical and/or it is impossible to obtain a
coupling between the coils is used, the tap posi¬ match, the use of additional reactance compensa¬
tions become rather criticai After the initial tion as described earlier is indicated.
adjustment of the circuit, try moving the tap Rf ammeters are useful for showing the com¬
positions out toward the ends of LI until it is just parative output obtained with various matching¬
possible to obtain a match by means of Cl and C2 network settings, and also for showing the im¬
at all frequencies within the band. This will result provement in output resulting from the use of
in the lowest possible operating Q and thus reactance compensation when it seems to be
minimize the necessity for readjustment of the required. Providing no basic circuit changes (such
circuit when shifting frequency. as grounding or ungrounding some part of the
A similar procedure is followed for adjustment matching circuit) are made during such compari¬
of the circuit of Fig. 341A. First set Cl at its sons, the current shown by the ammeters will
midposition and C2 at maximum capacitance. increase whenever the power put into the line is
Then vary LI. At some point, a sharp dip will be increased. Thus, the highest reading indicates the
noted in the reflected power indication. Once this greatest transfer efficiency, assuming that the
point is found, readjust Cl and C2, with possibly a power input to the transmitter is kept constant.
slight readjustment of LI, for zero reflected power. If the line Zo is matched by the antenna, the
For most loads, the settings will be found to be current can be used to determine the actual power
quite broad for a perfect match. The settings which input to the line. The power at the input terminals
will result in the lowest operating Q are those of is then equal to T’Zo, where / is the current and Zo
minimum inductance for LI and maximum capaci¬ is the characteristic impedance of the line. If there
tance for C2. are standing waves on the line this relationship
During the course of adjustments of either of does not hold. In such a case the current that will
the above circuits, switch the SWR indicator back flow into the line is determined by the line length,
to “forward” occasionally to make sure that this SWR, and whether the antenna impedance is higher
reading is staying well up on the scale. It some¬ or lower than the line impedance. Fig. 3-26 shows
times happens that an adjustment which apparent¬ how the maximum current to be expected will vary
ly reduces the reflected reading to zero or near with the standing-wave ratio. This information can
zero is simply detuning the circuit and the forward be used in selecting the proper ammeter range.
reading becomes quite low. The objective is to get Two ammeters, one in each line conductor, are
a zero reflected reading at matching-circuit settings shown in Fig. 3-42. The use of two instruments
that also give a high forward reading. gives a check on the Une balance, since the currents
With low Q, these circuits will work over an should be the same. However, a single meter can be
entire band without readjustment if the load is switched from one conductor to the other. If only
constant over the same frequency range. The load one instrument is used it is preferably left out of
seldom stays constant, however, since the input the circuit except when adjustments are being
impedance of the line changes with frequency with made, since it will add capacitance to the side in
most antennas. Readjustment becomes necessary which it is inserted and thus cause some unbalance.
whenever the input impedance changes enough to This is particularly important when the instrument
result in poor operation. Evidence of this is either is mounted on a metal panel.
the inability to adjust the transmitter output Since the resistive component of the input
circuit for proper loading of the final amplifier, or impedance of a line operating with an appreciable
such a high SWR in the coax that it shows signs of SWR is seldom known accurately, the rf current is
heating from the power lost in it. of little value as a check on power input to such a
Une. However, it shows in a relative way the
efficiency of the system as a whole. The set of
Measurement of Line Input Current coupling adjustments that results in the largest Une
The rf ammeters shown in Fig. 3-42 are not current with the least final-amplifier plate current
essential to the adjustment procedure but they, or is the one that delivers the greatest power to the
some other form of output indicator, are useful antenna with the lowest plate-power input.
accessories. In most cases the circuit adjustments For adjustment purposes, it is possible to
that lead to a match as shown by the SWR substitute small flashlight lamps, shunted across a
102 Chapter 3

few inches of the line wires, for the rf ammeters.

Their relative brightness shows when the current
increases or decreases. They have the advantage of
being inexpensive and of such small physical size
that they do not unbalance the circuit.

Coaxial-Line Feed
As mentioned earlier, a matching circuit should
not be necessary when coaxial line is used to feed
the antenna, since the SWR on such lines already
should be low enough to permit satisfactory
adjustment of the coupling by the normal transmit¬
ter controls. However, there are cases where the
additional frequency selectivity provided by the
matching circuit is desirable. For example, a
coax-fed multiband antenna system will not dis¬
criminate against transmitter harmonics, since the
system is designed to accept harmonically related
amateur frequencies without individual tuning ad¬ Fig. 3-44 — Air-core balun and toroidal trans¬
justments. former balun. The enclosure for the air-core
version measures five inches high by nine inches
The circuit of Fig. 3-40 may be altered slightly square; this balun is rated to handle 250 watts. The
for use with coaxial lines. Cl may be made a toroidal balun is built into a 2 X 3 X 3-1/2-inch
single-section capacitor, and the lower tap on L1 is enclosure, and will handle a full kilowatt. An
connected to the lower junction of L1 and Cl. 11 is Amidon T-200-2 core is used in this toroidal balun;
this junction where the shield of the coaxial line is a piece of phenolic insulating board is epoxy
connected, with the center conductor of the line cemented between the transformer and the enclo¬
connected to the upper, moveable tap. sure to prevent short-circuiting. (Amidon Associ¬
The circuit of Fig. 3-41A may be used without ates, 12033 Otsego St., N. Hollywood, CA 91601.)
the balun for coaxial lines. In either circuit,
adjustment is made in the same way as described
for the case of a balanced line. end of the winding, the inductance of the winding
considered as a plain coil is unimportant, since the
currents in the two conductors will be equal and
BALUN COILS opposite and there is substantially complete can¬
cellation of the external field just as in a normal
In Fig. 3-43, LI is a bifilar winding with an air
transmission line. The parallel conductors have a
core. When considered as a pair of parallel conduc¬ characteristic impedance dependent on their di¬
tors, it is equivalent to a transmission line. To a ameter and spacing. L2 is an identical winding.
voltage applied between the two terminals at one
If the two windings are connected as shown at
A, the two transmission lines are in series at the
right-hand end and in parallel at the left-hand end.
If each has a characteristic impedance equal to half
the input resistance of the balanced transmission
line to the antenna, their Zo will be matched by
the line input resistance, and the resistance looking
into the left-hand end of each will be equal to its
own Zo. Since they are in parallel at the left-hand
end, the total resistance looking into the windings
toward the antenna is equal to half the Zo of either
line, or one-fourth RI.
This arrangement therefore acts as an imped¬
ance transformer having a fixed ratio of 4 to 1,
balanced on the high-impedance side and un¬
balanced on the low-impedance side. Since a ratio
of 4 to 1 happens also to be the ratio between
matched 300-ohm twin line and 75-ohm coaxial
Une, such a balancing circuit or balun is useful for
matching a flat 300-ohm line to 75-ohm coax.
For parallel or “antenna” currents on the
Fig. 3-43 — Air-core balun coils as an impedance¬ transmission line to the antenna (see later section
matching device for coupling between balanced
and unbalanced lines. A properly designed set of
in this chapter) the two pairs of windings act as
coils will work over the 3.5-30-MHz range without normal inductances, since such currents are in
adjustment. However, only the fixed impedance phase in the two line wires. They thus tend to
ratios shown are available and the input impedance choke off parallel currents. This choking action
of the transmission line must be a pure resistance also is essential to keep the right-hand end of the
that will match the Zo of the coils. lower conductor of L2 properly above ground for a
Transmission Lines 103

VARIABLE IMPEDANCE
BALUN
4.1 BALUN TRANSFORMER
4ÍI TO IO.I
HIGH IMPEDANCE BALANCED

(A) (C)
Fig- 3-45 — Basic broad-band balun transformers.

balanced output. Thus the coils must be sufficient¬ At C of Fig. 3-45 is shown how the basic 4:1
ly large to give good isolation between the bal¬ balun may be altered for impedance matching in
anced and unbalanced ends at the lowest frequency the range between 4:1 and 10:1. This is accom¬
to be used. The upper frequency limit is that at plished by tapping the unbalanced input down on
which the winding, considered as an inductance one of the two windings. In the formula associated
rather than a transmission line, begins to show with Fig. 3-45C for determining the impedance
distributed-capacitance effects. In practice, a single ratio, kequals the ratio of the number of tapped
set of coils can be designed to work over the 3.5- turns to the total number of turns in the tapped
to 30-MHz range. Design is complicated because winding.
there is mutual coupling between turns, which Bifilar windings for balun transformers should
modifies the characteristic impedance. However, be from six to ten turns, depending on the
suitable units are available commercially (B&W ferrite-core permeability. A suitable ferrite material
3975). is Ferramic QI (Indiana General), with a perme¬
ability of 125. Very small size cores may be used
Balun Transformers for receiving and low-power applications. For
Because of their smaller physical size for full-power use, 2-1/2-inch OD Ferramic QI core
comparable power-handling capability, bifilar- with 1/2-inch cross section wound with No. 14
wound baluns on a toroidal core are preferred in Formex copper wire, seven turns per winding, is
practice over air-core baluns. The two types are recommended.
pictured in Fig. 3-44. Typically, ferrite toroidal- Balun Terminations
core baluns have bandwidths of 10 to 1, such as for
the frequency range from 3 to 30 MHz, and may The principles on which baluns operate should
be constructed sufficiently large to handle the full make it obvious that the termination must be
legal amateur power. Toroidal or transformer essentially a pure resistance in order for the proper
baluns are shown in Fig. 3-45. At A is a basic 4:1 impedance transformation to take place. If the
transformer balun, with the high impedance bal¬ termination is not resistive, the input impedance of
anced. Only two windings are required for this each bifilar winding will depend on its electrical
type of balun. characteristics and the input impedance of the
Transformer baluns having a 1:1 impedance main transmission line; in other words, the imped¬
ratio are usually trifilar wound — three windings ance will vary just as it does with any transmission
interlaced on the core. The 1:1 balun shown in B line, and the transformation ratio likewise will vary
of Fig. 3-45 has been modified slightly from that over wide limits.
design in that the third winding on the core has Baluns alone are convenient as matching devices
been separated from the bifilar winding. This when the above condition can be met, since they
modification results in improved balance at the require no adjustment. When used with a matching
higher frequencies with no change in other charac¬ network as described earlier, however, the
teristics. The third winding is a core-magnetizing impedance-transformation ratio of a balun be¬
winding which is effective only in extending the comes of only secondary importance, and loads
low-frequency range of the balun. The third containing reactance may be tolerated so long as
winding may be omitted entirely if operation is the losses in the balun itself do not become
confined to frequencies above about 10 MHz. excessive.
 104 Chapter 3

"TUNED" AND "UNTUNED" LINES depending on the antenna impedance. After the
first loop, the remaining ones are at quarter¬
In the past, transmission lines frequently have wavelength intervals.)
been classified as “tuned” or “untuned,” depend¬ Such a classification tends to be arbitrary, since
ing on whether or not the line had to be cut to a there is no well-defined value of SWR below which
certain length in order to have a substantially a line may be “untuned” and above which it must
resistive input impedance. As shown earlier, when be “tuned.” It is possible to couple power into anv
the SWR is high, the input impedance is resistive length of line, regardless of the standing-wave ratio,
(or mostly so) only at line lengths that are a if the principles already outlined are followed. If
multiple of a quarter wavelength, assuming the the SWR is high, special methods of reactance
load represented by the antenna is itself a substan¬ compensation may be required if the line length is
tially pure resistance. (If it is not, the resistive- unfavorable, as explained earlier, but it is not
input points correspond to the current and voltage necessary to cut the line to a particular length
loops, the first of which may occur at any distance (which may be an inconvenient length for installa¬
up to a quarter wavelength from the antenna, tion) in order to put power into it.

COUPLING THE LINE TO THE ANTENNA

Throughout the discussion of transmission-line “trap” circuits or multiple wires. Information on
principles in the first part of this chapter the these is given in later chapters. Such an antenna
operation of the line has been described in terms of can be assumed to be equivalent to a resonant
an abstract “load.” This load had the electrical half-wave dipole on each of the bands for which it
properties of resistance and, sometimes, reactance. is designed, and may be fed through a line, coaxial
It did not, however, have any physical attributes or otherwise, that has a Zo matching the antenna
that associated it with a particular electrical device. impedance, as described later in this chapter for
That is, it could be anything at all that exhibits simple dipoles.)
electrical resistance and/or reactance. The fact is Methods of coupling the line to the antenna
that so far as the line is concerned, it does not therefore divide, from a practical standpoint, into
matter what the load is, just so long as it will two classes. In the first, operation on several
accept power. amateur bands is the prime consideration and the
Many amateurs make the mistake of confusing standing-wave ratio is secondary. The SWR is
transmission lines with antennas, believing that normally rather large and the input impedance of
because two identical antennas have different kinds the line depends on the line length and the
of lines feeding them, or the same kind of line with operating frequency.
different methods of coupling to the antenna, the In the second class, a conscious attempt is
“antennas” are different. There may be practical made, when necessary, to transform the antenna
reasons why one system (including antenna, trans¬ impedance to a value that matches the characteris¬
mission line, and coupling method) may be pre¬ tic impedance of the line. When this is done the
ferred over another in a particular application. But line operates with a very low standing-wave ratio
to the transmission line, an antenna is just a load and its input impedance is essentially a pure
that terminates it, and the important thing is what resistance, regardless of the line length. A transmis¬
that load looks like to the line in terms of sion line can be considered to be “flat,” within
resistance and reactance. practical limits, if the SWR is not more than about
Any kind of transmission line can be used with 1.5 to 1.
any kind of antenna, if the proper measures are
taken to couple the two together. Losses
Frequency Range and SWR A principal reason for matching the antenna to
the line impedance is that a flat line operates with
Probably the principal factor that determines the least power loss. While it is always desirable to
the way a transmission line is operated is the reduce losses and thus increase efficiency, the
frequency range over which the antenna is to work. effect of standing waves in this connection can be
Very few types of antennaswill present essentially overemphasized. This is particularly true at the
the same load impedance to the line on harmonic¬ lower amateur frequencies, where the inherent loss
ally related frequencies. As a result, the builder in most types of lines is quite low even for runs
often is faced with choosing between (1) an that, in the average amateur installation, are rather
antenna system that will permit operating the long.
transmission line with a low standing-wave ratio, For example, 100 feet of 300-ohm receiving¬
but is confined to one operating frequency or a type Twin-Lead has a loss of only 0.18 dB at 3.5
narrow band of frequencies, and (2) a system that MHz, as shown by Table 3-1. Even with an SWR as
will permit operation in several harmonically relat¬ high as 10 to I the additional loss caused by
ed bands but with a large SWR on the line. (There standing waves is less than 0.7 dB, from Fig. 3-23.
are “multiband” systems which, in principle, make Since 1 dB represents the minimum detectable
one antenna act as though it were a half-wave change in signal strength, it does not matter from
dipole on each of several amateur bands, by using this standpoint whether the line is flat or not. But
Transmission Lines 105

at 144 MHz the loss in the same length of line

perfectly matched is 2.8 dB, and an SWR of 10 to
1 would mean an additional loss of 4 dB. At the
higher frequency, then, it is worthwhile to match
the antenna and line as closely as possible.

Power Limitations
Another reason for matching is that certain
types of lines, particularly those with solid dielec¬
tric, have definite voltage and current limitations.
At the lower frequencies this is a far more
compelling reason than power loss for at least
approximate matching. Where the voltage and
current must not exceed definite maximum values,
the amount of power that the line can handle is
inversely proportional to the standing-wave ratio.
If the safe rating on the 300-ohm line in the
example above is 500 watts when perfectly
matched, the line can handle only 50 watts with
equal safety when the SWR is 10 to 1. Thus,
despite the fact that the line losses are low enough
to make no appreciable difference in the signal
strength, the high SWR could be tolerated only
with low-power transmitters.

Line Radiation Fig. 3-46 - Center and end feed as used in simple
Aside from power considerations, there is a antenna systems.
more-or-less common belief that a flat line “does
not radiate” while one with a high SWR does matched 300-ohm line. When there are standing
radiate. This impression is quite unjustified. It is
waves, the safe power can be found by dividing
true that the radiation from a parallel-conductor
1200 by the SWR. In a center-fed half-wave
Une increases with the current in the line, and that
antenna, as in Fig. 3-46A, the SWR should not
the effective line current increases with the SWR. exceed about 5 to 1 (at the fundamental fre¬
However, the loss by radiation from a properly
quency), so receiving type 300-ohm Twin-Lead
balanced line is so small (and is, furthermore, would appear to be safe for power outputs up to
independent of the line length) that multiplying it
several times still does not bring it out of the 250 watts or so.
Since there is little point in using a mismatched
“negligible” classification. line to feed an antenna that is to operate on one
Whenever a line radiates it is because of faulty
amateur band only, the discussion to follow will be
installation (resulting in unbalance with parallel-
based on the assumption that the antenna is to be
conductor lines) or “antenna currents” on the line.
operated on its harmonics for multiband work.
Radiation from the latter cause can take place
from either resonant or nonresonant lines, parallel¬
conductor or coaxial. “Current” and “Voltage” Feed
Usual practice is to connect the transmission
UNMATCHED SYSTEMS line to the antenna at a point where either a
In many multiband systems or simple antennas current or voltage loop occurs. If the feed point is
where no attempt is made to match the antenna at a current loop the antenna is said to be current
impedance to the characteristic impedance of the fed; if at a voltage loop the antenna is voltage fed.
line, the customary practice is to connect the line These terms should not be confused with center
either to the center of the antenna (center feed) as feed and end feed, because they do not necessarily
indicated in Fig. 3-46A, or to one end (end feed) as have corresponding meanings. There is always a
shown in Fig. 3-46B. voltage loop at the end of a resonant antenna, no
Because the line operates at a rather high matter what the number of half wavelengths, so a
standing-wave ratio, the best type to use is the resonant end-fed antenna is always voltage fed.
open-wire line. Solid Twin-Lead of the 300-ohm This is illustrated at D and E in Fig. 3-47 for
receiving variety can also be used, but the power end-fed antennas a half wavelength long (antenna
limitations discussed in the preceding section fundamental frequency) and one wavelength long
should be kept in mind. Although the manufactur¬ (second harmonic). It would continue to be true
ers have placed no power rating on receiving-type for an end-fed antenna operated on any harmonic.
300-ohm line, it seems reasonable to make the However, Fig. 3-47F shows voltage feed at the
assumption, based on the conductor size, that a center of the antenna; in this case the antenna has
current of 2 A can readily be carried by a line a total length of two half wavelengths, each of
installed so that there is free air circulation about which is voltage fed. Voltage feed is determined
it. This corresponds to a power of 1200 watts in a not by the physical position of the transmission
 106
Chapter 3

Fig. 3-47 — Current and voltage

feed in antennas operated at the
fundamental frequency, 2 times
the fundamental, and 3 times the
fundamental. The current and
voltage distribution on the anten¬
na are identical with both
methods only at the fundamental
frequency.

VOLTAGE FEED

line on the antenna, but by the fact that a voltage Operation on Harmonics
loop occurs on the antenna at the feed point. Since In the usual case of an antenna operated on
voltage loops always occur at integral multiples of several bands, the point at which the transmission
a half wavelength from either end of a resonant line is attached is of course fixed. The antenna
antenna, feeding the antenna at any half-wave¬ length is usually such that it is resonant at some
length point constitutes voltage feed. frequency in the lowest frequency band to be used,
Typical cases of current feed are shown at A, B and the transmission Une is connected either to the
and C in Fig. 3-47. The feed point is at a current center or the end. The current and voltage distribu¬
loop, which always occurs at the midpoint of a tion along antennas fed at both points is shown in
half-wave section of the antenna. In order to feed Fig. 3-48. With end feed, A to F inclusive, there is
at a current loop the transmission line must be always a voltage loop at the feed point. Also, the
connected at a point that is an odd multiple of current distribution is such that in every case the
quarter wavelengths from either end of the reson¬ antenna operates as a true harmonic radiator of the
ant antenna. A center-fed antenna is also current type described in Chapter Two.
fed only when the antenna length is an odd With center feed, the feed point is always at a
multiple of half wavelengths. Thus the antenna in current loop on the fundamental frequency and all
Fig. 3-47B is both center fed and current fed since odd multiples of the fundamental. In these cases
it is three half wavelengths long. It would also be the current and voltage distribution are identical
center fed and current fed if it were five, seven, with the distribution on an end-fed antenna. This
etc., half wavelengths long. can be seen by comparing A and G, C and I, and E
To current feed a one-wavelength antenna, or and K, Fig. 3-48. (In I, the phase is reversed as
any resonant antenna having a length that is an compared with C, but this is merely for conve¬
even multiple of one-half wavelength, it is neces¬ nience in drawing; the actual phases of the currents
sary to shift the feed point from the center of the in each half-wave section reverse each half cycle so
antenna (where a voltage loop always occurs in it does not matter whether the current curve is
such a case) to the middle of one of the half-wave drawn above or below the line, so long as the
sections. This is indicated in Fig. 3-47C in the case relative phases are properly shown in the same
of a one-wavelength antenna; current feed can be antenna.) On odd multiples of the fundamental
used if the line is connected to the antenna at a frequency, therefore, the antenna operates as a
point 1/4 wavelength from either end. true harmonic antenna.
Transmission Lines 107

(A)

(B)

Fig. 3-48 — Current and voltage

distribution at the fundamental
frequency and various multiples,
with both end feed and center
feed. The distributions are the
same with both types of feed only
when the frequency is an odd
multiple of the fundamental.

(F)

On even multiples of the fundamental fre¬ case of the center-fed antenna can be explained
quency the feed point with center feed is always at with the aid of Fig. 3-49. It will be recalled from
a voltage loop. This is shown at H, J and L in Fig. Chapter Two that the direction of current flow
3-48. Comparing B and H, it can be seen that the reverses in each half wavelength of wire. Also, in
current distribution is different with center feed any transmission line the currents in the two wires
than with end feed. With center feed the currents always must be equal and flowing in opposite
in both half-wave sections of the antenna are in the directions at any point along the line. Starting
same phase, but with end feed the current in one from the end of the antenna, the current must be
half-wave section is in reverse phase to the current flowing in one direction throughout the first
in the other. This does not mean that one antenna half-wavelength section, whether this section is
is a better radiator than the other, but simply that entirely antenna or partly antenna and partly one
the two will have dif ferent directional characteris¬ wire of the transmission line. Thus, in A, Fig. 3-49,
tics. The center-fed arrangement is commonly the current flows in the same direction from P to
known as “two half-waves in phase,” while the Q, since this is all the same conductor. However,
end-fed system is a “one-wavelength antenna or one quarter wave is in the antenna and one in the
“second-harmonic” antenna. transmission line. The current in the other line
Similarly, the system at J has a different wire, starting from R, must flow in the opposite
current and voltage distribution than the system at direction in order to balance the current in the first
D, although both resonate at four times the wire, as shown by the arrow. And since the
fundamental frequency. A similar comparison can distance from R to .S' is 1/2 wavelength, the current
be made between F and L. The center-fed arrange¬ continues to flow in the same direction all the way
ment at J really consists of two one-wavelength to S. The currents in the two halves of the antenna
antennas, while the arrangement at L has two are therefore flowing in the same direction. Fur¬
2-wavelength antennas. These have different direc¬ thermore, the current is maximum 1/4 wavelength
tional characteristics than the 2-wavelength and from the ends of the antenna, as previously
4-wavelength antennas (D and F) that resonate at explained, and so both the currents are maximum
the same multiple, respectively, of the fundamental at the junction of the antenna and transmission
frequency. line. This makes the current distribution along the
The reason for this difference between odd and length of the antenna exactly the same as with an
even multiples of the fundamental frequency in the end fed antenna.
 108
Chapter 3

all even multiples of the fundamental frequency

there is a current minimum at the junction of the
line and antenna, with center feed, because there is
an integral number of half wavelengths in each side
of the antenna. The voltage is maximum at the
junction in such a case, and we have voltage feed.
Where the multiple of the fundamental is odd,
there is always a current maximum at the junction
of the transmission line and antenna, as demon¬
strated by A and C in Fig. 3-49. At these points the
voltage is minimum and we therefore have current
feed.

“Zepp” or End Feed

In the early days of short-wave communication
an antenna consisting of a half-wave dipole, end-
fed through a 1/4-wavelength transmission line,
was developed as a trailing antenna for Zeppelin
airships. In its utilization by amateurs, over the
years, it has become popularly known as the
“Zeppelin” or “Zepp” antenna. The term is now
applied to practically any resonant antenna fed at
the end by a two-wire transmission line.
The mechanism of end feed is perhaps some¬
what difficult to visualize, since only one of the
two wires of the transmission line is connected to
the antenna while the other is simply left free. The
difficulty lies in the natural tendency to think in
Fig. 3-49 — Showing how the type of feed changes terms of current flow in ordinary electrical circuits,
from current to voltage, with a center-fed antenna, where it is necessary to have a complete loop
on twice the fundamental frequency, and back to between both terminals of the power source before
current feed on three times the fundamental. The
same change occurs between all even and odd any current can flow at all. But as explained
frequency multiples. earlier, this limitation applies only to circuits in
which the electromagnetic fields reach the most
distant part of the circuit in a time interval that is
Fig. 3-49B shows the case where the overall negligible in comparison with the time of one
length of the antenna is one wavelength, making a cycle. When the circuit dimensions are comparable
half wave on each side. A half wavelength along the with the wavelength, no such complete loop is
transmission line also is shown. If we assume that necessary. The antenna itself is an example of an
the current is flowing downward in the line "open ” circuit in which large currents can flow.
conductor from Q to R, it must be flowing upward One way of looking at end feed is to consider
from S to T if the line currents are to balance. the entire length of wire, including both antenna
However, the distance from Q to P is 1/2 wave¬ and feeder, as a single unit. For example, suppose
length, and so the current in this section of the we have a wire one wavelength long, as in Fig.
antenna must flow in the opposite direction to the 3-50A, fed at a current loop by a source of rf
current flowing in the section from Q to R. The power. The current distribution will be as shown
current in section PQ is therefore flowing away by the curves, with the assumed directions indica¬
from Q. Also, the current in section TV must be ted by the arrows. If we now fold back the
flowing in the opposite direction to the current in 1/4-wavelength section to the left of the power
ST, and so is flowing toward T. The currents in the source, as shown at B, the overall current distribu¬
two half-wave sections of the antenna are therefore tion will be similar, but the currents in the two
flowing in the same direction. That is, they are in wires of the folded section will be flowing in
the same phase. opposite directions. The amplitudes of the currents
With the above in mind, the direction of at any point along the folded-back portion will be
current flow in a 1-1/2-wavelength antenna, Fig. equal in the two wires. The folded section,
3-49C, should be easy to follow. The center therefore, has become a 1/4-wavelength transmis¬
half-wave section QT corresponds to the half-wave sion line, since the fields from the equal and
antenna in A. The currents in the end sections, PQ opposite currents cancel. There is, however,
and TU simply flow in the opposite direction to nothing to prevent current from continuing to flow
the current in QT. Thus the currents are out of in the right-hand half-wavelength section, since
phase in alternate half-wave sections. there was current there before the left-hand section
The shift in voltage distribution between odd was folded.
and even multiplies of the fundamental frequency This picture, although showing how power can
can be demonstrated by a similar method, making flow from the transmission line to an antenna
allowance for the fact that the voltage is maximum through end feed, lacks completeness. It does not
where the current is minimum, and vice versa. On take into account the fact that the current /I in
Transmission Lines 109

the transmission line is greatly different from the SWR with Wire Antennas
current Iin the antenna. A more basic viewpoint is When a line is connected to a single-wire
the one already mentioned in Chapter Two: The antenna at a current loop the standing-wave ratio
current is caused by electromagnetic fields travel¬ can be estimated with good-enough accuracy with
ing along the wire and simply constitutes a the aid of the curve in Fig. 2-20. Although the
measurable manifestation of those fields; the cur¬ actual value of the radiation resistance, as mea¬
rent does not cause the fields. From this stand¬ sured at a current loop, will vary with the height of
point the transmission-line conductors merely serve the antenna above ground, the theoretical values
as “guides” for the fields so the electromagnetic given in Fig. 2-20 will at least serve to establish
energy will go where we want it to go. When the whether the SWR will be high or low.
energy reaches the end of the transmission line it With center feed the line will connect to the
meets another guide, in the form of the antenna, antenna at a current loop on the fundamental
and continues along it. However, the antenna is a frequency and all odd multiples, as shown by Fig.
different form of guide; it has a single conductor 3-48. At the fundamental frequency and usual
while the line has two; it has no provision for antenna heights, the antenna resistance should lie
preventing radiation while the line is designed for between 50 and 100 ohms, so with a line having a
that very purpose. This is simply another way of characteristic impedance of 450 ohms the SWR
saying that the impedance of the antenna differs will be Zo/R L = 450/50 = 9 to 1 as one limit, and
from that of the transmission line, so there will be 450/100 = 4.5 to 1 as the other. On the third
reflection when the energy traveling along the line harmonic the theoretical resistance as given by Fig.
arrives at the antenna. We are then back on familiar 2-20 is near 100 ohms, so the SWR should be
ground, in that we have a transmission line about 4.5 to 1. For 300-ohm line the SWR can be
terminated in an impedance different from its expected to be between 3 and 6 on the antenna
characteristic impedance. fundamental and about 3 to 1 on the third
harmonic.
The impedances to be expected at voltage loops
Feeder Unbalance are less readily determined. Theoretical values are
With end feed, the currents in the two line in the neighborhood of 5000 to 8000 ohms,
wires do not balance exactly and there is therefore depending on the antenna conductor size and the
some radiation from the line. The reason for this is number of half wavelengths along the wire. Such
that the current at the end of the free wire is zero experimental figures as are available indicate a
(neglecting a small charging current in the insulator lower order of resistance, with measurements and
at the end) while the current does not go to zero at estimates running from 1000 to 5000 ohms. In any
the junction of the “active” line wire and the event, there will be some difference between end
antenna. This is because not all the energy going feed and center feed, since the current distribution
into the antenna is reflected back from the far end, on the antenna is different in these two cases at
some being radiated; hence the incident and any given even multiple of the fundamental fre¬
reflected currents cannot completely cancel at a quency. Also, the higher the multiple the lower the
node. resistance at a voltage loop, so the SWR can be
In addition to this unavoidable line radiation a expected to decrease when an antenna is operated
further unbalance will occur if the antenna is not at a high multiple of its fundamental frequency.
exactly resonant at the operating frequency. If the Assuming 4000 ohms for a wire antenna two half
frequency is too high (antenna too long) the waves long, the SWR would be about 6 or 7 with a
current node does not occur at the junction of the 600-ohm line and around 12 with a 300-ohm line.
antenna and “live” feeder, but moves out on the However, considerable variation is to be expected.
antenna. When the frequency is too low the node
moves down the active feeder. Since the node on
the free feeder has to occur at the end, either case
is equivalent to shifting the position of the
standing wave along one feeder wire but not the
other. The further off resonance the antenna is
operating, the greater the unbalance and the
greater the line radiation. With center feed this
unbalance does not occur, because the system is
symmetrical with respect to the line.
To avoid line radiation it is always best to feed
the antenna at its center of symmetry. In the case
of simple antennas for operation in several bands,
this means that center feed should be used. End
feed is required only when the antenna is operated
on an even harmonic to obtain a desired directional
characteristic, and then only when it must be used
on more than one band. For single-band operation
it is always possible to feed an even-harmonic
antenna at a current loop in one of the half-wave Fig. 3-50 — Folded-antenna analogy of transmis¬
sections nearest the center. sion line for an end-fed antenna.
 110
Chapter 3

ANTENNA CURRENTS ON
TRANSMISSION LINES
In any discussion of transmission-line operation
it is always assumed that the two conductors carry
equal and opposite currents thoughout their
length. This is an ideal condition that may or may
not be realized in practice. In the average case the
chances are rather good that the currents will not
be balanced unless special precautions are taken.
Whether the line is matched or not has little to do
with the situation.
Consider the half-wave antenna shown in Fig.
3-51 and assume that it is somehow fed by a source
of power at its center, and that the instantaneous
direction of current flow is as indicated by the
arrows. In the neighborhood of the antenna is a
group of conductors disposed in various ways with
respect to the antenna itself. All of these conduc¬
tors are in the field of the antenna and are
therefore coupled to it. Consequently, when cur¬
rent flows in the antenna a voltage will be induced
Fig. 3-52 — The important length for resonance to
in each conductor. This causes a current flow antenna currents coupled from the antenna to the
determined by the induced voltage and the imped¬ line. In the center-fed system one side of the
ance of the conductor. antenna is part of the "parallel"-resonant system.
The degree of coupling depends on the position
of the conductor with respect to the antenna,
for the negligible spacing between them, the two
assuming that all the conductors in the figure are
conductors lie in the same position with respect to
the same length. The coupling between the antenna
the antenna. Therefore, identical voltages will be
and conductor IJ is greater than in any other case,
because IJ is close to and parallel with the antenna. induced in both, and the resulting currents will be
Ideally, the coupling between conductor GH and flowing in the same direction in both conductors.
the antenna is zero, because the voltage induced by It is only a short step to visualizing conductors EF
current flowing in the left-hand side of the antenna and KL as the two conductors of a section of
is exactly balanced by a voltage of opposite transmission line in the vicinity of the antenna.
polarity induced by the current flowing in the Because of coupling to the antenna, it is not only
right-hand side. This is because the two currents possible but certain that a voltage will be induced
are flowing in opposite directions with respect to in the two conductors of the transmission line in
GH. Complete cancellation of the induced voltages
parallel. The resulting current flow is in the same
can occur, of course, only if the currents in the direction in both conductors, whereas the true
transmission line currents are always flowing in
two halves of the antenna are symmetrically
opposite directions at each point along the line.
distributed with respect to the center of the
These “parallel” currents are of the same nature as
antenna, and also only if every point along GH is
the current in the antenna itself, and hence are
equidistant from any two points along the antenn;
called “antenna” currents on the line. They are
that are likewise equidistant from the center. This
responsible for most of the radiation that takes
cannot be true of any of the other conductors place from transmission lines.
shown, so a finite voltage will be induced in any
When there is an antenna current of appreciable
conductor in the vicinity of the antenna except
amplitude on the line it will be found that not only
one perpendicular to the antenna at its center.
are the line currents unbalanced but the apparent
SWR is different in each conductor, and that the
Transmission Line in the Antenna Field loops and nodes of current in one wire do not
Now consider the two conductors EF and KL, occur at corresponding points in the other wire.
which are parallel and very close together. Except Under these conditions it is impossible to measure
the true SWR.
It should be obvious from Fig. 3-51 that only in
the case of a center-fed antenna can the coupling
between the line and antenna be reduced to zero.
There is always some such coupling when the
antenna is end fed, so there is always the possibil¬
ity that antenna currents of appreciable amplitude
will exist on the line, contributing further to the
inherent line unbalance in the end-fed arrange¬
ment. But the center-fed system also will have
appreciable antenna-to-line coupling if the line is
Fig. 3-51 — Coupling between antenna and con¬ not brought off at right angles to the antenna for a
ductors in the antenna's field. distance of at least a half wavelength.
Transmission Lines
111

Antenna currents will be induced on lines of because the resonances are not extremely sharp.
any type of construction. If the line is coax, the Working close to resonance, although not exactly
antenna current flows only on the outside of the on it, will allow an appreciable “antenna” current
outer conductor; no current is induced inside the to flow even though it may not be as large as it
Une. However, an antenna current on the outside would if the Une were exactly resonant for it. For
of coax is just as effective in causing radiation as a the same reason the line length should be chosen to
similar current induced in the two wires of a fall in a range where there is a considerable
parallel-conductor line. distance between resonances. A length of 76 feet,
for example, would be definitely less susceptible to
resonance effects than a length of 96 feet.
Detuning the Line for Antenna Currents The lengths shown in Fig. 3-53 are subject to
The antenna current flowing on the line as a some modification in practice. Transmission lines
result of voltage induced from the antenna will be usually have bends, are at varying heights above
small if the overaU circuit, considering the line ground, etc., all of which will modify the resonant
simply as a single conductor, is not resonant at the frequency. It is advisable to check the system for
operating frequency. The frequency (or frequen¬ resonance at and near all operating frequencies
cies) at which the system is resonant depends on before assuming that the line is safely detuned for
the total length and whether the transmission line antenna currents. This can be done by temporarily
is grounded or not at the transmitter end. connecting the ends of the line together and
If the Une is connected to a coupling circuit coupling them through a small capacitance (not
that is not grounded, either directly or through a more than a few pF) to a resonance indicator such
capacitance of more than a few picofarads, it is as a grid-dip meter. Very short leads should be used
necessary to consider only the length of the between the meter and antenna. Fig. 3-54 shows
antenna and Une. In the end-fed arrangement, the method. Once the resonance points are known
shown at A in Fig. 3-52, the line length, L, should it is a simple matter to prune the feeders to get as
not be an integral multiple or close to such a far away as possible from resonance at any
multiple of a half wavelength. In the center-fed frequency to be used.
system, Fig. 3-52B, the length of the Une plus one Resonances in systems in which the coupling
side of the antenna should not be a multiple of a apparatus is grounded at the transmitter are not so
half wavelength. In this case the two halves of the easily predicted. The “ground” in such a case is
antenna are simply in parallel so far as resonance usually the metal chassis of the transmitter itself,
for the induced “antenna” current on the Une is not actual ground. In the average amateur station it
concerned, because the line conductors themselves is not possible to get a connection to real ground
act in parallel. When the antenna is to be used in without having a lead that is an appreciable
several bands, resonance of this type should be fraction of wavelength long. At the higher frequen¬
avoided at all frequencies to be used. Fig. 3-53 cies, and particularly in the vhf region, the distance
shows, as solid Unes along the length scale, the from the transmitter to ground may be one
lengths that avoid exact resonance on frequencies wavelength or more. Probably the best plan in such
from 3.5 to 29.7 MHz. These are based on the cases is to make the length Lin Fig. 3-52 equal to a
usual antenna-length formulas; the velocity factor multiple of a half wavelength. If the transmitter
of the line plays no part in establishing such has fairly large capacitance to ground, a system of
resonances since it applies only to true transmis¬ this length will be effectively detuned for the
sion-line currents. fundamental and all even harmonics when ground¬
Whenever possible, it is best to choose line ed to the transmitter at the coupling apparatus.
lengths, such as those indicated by the arrows, that However, the resonance frequencies will depend on
fall midway in the nonresonant range. This is the arrangement and constants of the coupling

Fig 3-53 — Lengths shown by solid lines along the horizontal axis avoid exact resonance at
frequencies in all amateur bands from 3.5 to 29.7 MHz, in systems where the coupling apparatus is not
grounded. Best operating lengths are at the centers of the wider ranges, as shown by the arrows. These
lengths correspond to Lin Fig. 3-52.
 112 Chapter 3

deal of freedom in that the antenna and line can be

selected independently. Its disadvantage is that it is
Line more complicated constructionally. Also, it some¬
times calls for a somewhat tedious routine of
measurement and adjustment before the desired
match is achieved.
GRID¬
DIP
METER
Operating Considerations
As pointed out earlier in this chapter, most
antenna systems show a marked change in resis¬
Fig. 3-54 - Using a grid-dip meter to check tance when going from the fundamental to multi¬
resonance of the antenna system for antenna ples of the fundamental frequency. For this reason
currents on the transmission line. The small capaci¬ it is usually possible to match the line impedance
tance may be a short length of wire connected to only on one frequency. A matched antenna system
the feed line, coupled lightly to the grid-dip- is consequently a one-band affair, in most cases. It
oscillator coil with a 1-turn loop.
can, however, usually be operated over a fan-
frequency range in a given band. The frequency
system even in such a case, and preferably should range over which the standing-wave ratio is low is
be checked by means of the grid-dip meter. If this determined by the impedance-vs.-frequency charac¬
test shows resonance at or near the operating teristic of the antenna. If the change in impedance
frequency, alternative grounds (to a heating radi¬ is small for a given change in frequency, the SWR
ator, for example) should be tried until a combina¬ will be low over a fairly wide band of frequencies.
tion is found that detunes the whole system. However, if the impedance change is rapid (a
It should be quite clear, from the mechanism sharply resonant or high-g antenna - see discus¬
that produces antenna currents on a transmission sion of Q later in this chapter) the SWR will also
line, that such currents are entirely independent of rise rapidly as the operating frequency is shifted to
the normal operation as a true transmission line. It one side or the other of the frequency for which
does not matter whether the line is perfectly the antenna is resonant and for which the line is
matched or is operated with a high standing-wave matched.
ratio. Nor does it matter what kind of line is used, Antenna Resonance
air-insulated or solid-dielectric, parallel-conductor
or coax. In every case, the antenna currents should A point that needs emphasis in connection with
be minimized by detuning the line if the line is to matching the antenna to the line is that, with the
fulfill only its primary purpose of transferring exception of a few special cases discussed later in
power to the antenna. this chapter, the impedance at the point where the
line is connected must be a pure resistance. This
Other Causes of Unbalance means that the antenna system must be resonant at
Unbalance in center-fed systems can arise even the frequency for which the line is to be matched.
when the line is brought away at right angles to the (Some types of long-wire antennas are exceptions,
antenna for a considerable distance. If both halves in that their input impedances are resistive over a
of the antenna are not symmetrically placed with wide band of frequencies. Such systems are essen¬
respect to nearby conductors (such as power and tially nonresonant.) The higher the Q of the
telephone wires, downspouting, etc.) the antenna antenna system, the more essential it is that exact
itself becomes unbalanced and the current distribu¬ resonance be established before an attempt is made
tion is different in the two halves. Because of this to match the line. This is particularly true of
unbalance a voltage will be induced in the line even close-spaced parasitic arrays. With simple dipole
if the Une is symmetrical with respect to the and harmonic antennas, the tuning is not so critical
antenna. and it is usually sufficient to cut the antenna to the

MATCHED LINES
Operating the transmission line at a low
standing-wave ratio requires that the line be ter¬
minated, at its output end, in a resistive load
matching the characteristic impedance of the line
as closely as possible. The problem can be ap¬
proached from two standpoints: (1) selecting a
transmission line having a characteristic impedance
that matches the antenna resistance at the point of
connection; or (2) transforming the antenna resis¬
tance to a value that matches the Zo of the line
selected.
The first approach is simple and direct, but its Fig. 3-55 — Half-wave dipole fed with 75-ohm
application is limited because the antenna imped¬ Twin-Lead, giving a close match between antenna
ance and line impedance are alike only in a few and line impedance. The leads in the "Y" from the
special cases. The second approach provides a good end of the line to the ends of the center insulator
should be as short as possible.
Transmission Lines
113

length given by the appropriate formula in Chapter Length in feet -

Two. The frequency should be selected to be at the
center of the range of frequencies (which may be Inner
Outer braid
the entire width of an amateur band) over which Conductor
the antenna is to be used.

DIRECT MATCHING — IS-ohm Coax

As discussed in Chapter Two, the impedance at

the center of a resonant half-wave antenna at
heights of the order of 1/4 wavelength and more is
resistive and is in the neighborhood of 70 ohms.
This is fairly well matched by transmitting-type Fig. 3-56 — Half-wave antenna fed with 75-ohm
Twin-Lead having a characteristic impedance of 75 coaxial cable. The outside of the outer conductor
ohms. It is possible, therefore, to operate with a of the line may be grounded for lightning protec¬
low SWR using the arrangement shown in Fig. tion.
3-55. No precautions are necessary beyond those
already described in connection with antenna-to- the antenna, a condition that is met fairly well at
line coupling. the lower amateur frequencies. However, it is not
This system is badly mismatched on even negligible in the vhf and uhf range, nor should it be
multiples of the fundamental frequency, since the ignored at 28 MHz. The current that flows on the
feed in such cases is at a high-impedance point. outside of the line because of this unbalance, it
However, it is reasonably well matched at odd should be noted, does not arise from the same type
multiples of the fundamental. For example, an of coupling as the “antenna” current previously
antenna resonant near the low-frequency end of discussed. The coupling pictured in Fig. 3-51 can
the 7-MHz band will operate with a low SWR over still occur, in addition. However, the remedy is the
the 21-MHz band (three times the fundamental). same in both cases - the system must be detuned
The same method may be used to feed a for currents on the outside of the line. This can be
harmonic antenna at any current loop along the done by choosing one of the recommended lengths
wire. For lengths up to three or four wavelengths in Fig. 3-53, or by an actual resonance check using
the SWR should not exceed 2 to 1 if the antenna is the method shown in Fig. 3-54.
1/4 or 1/2 wavelength above ground.
At the fundamental frequency the SWR should Balancing Devices
not exceed about 2 to 1 within a frequency range The unbalanced coupling described in the pre¬
±2% from the frequency of exact resonance. Such ceding paragraph can be nullified by the use of
a variation corresponds approximately to the entire devices that prevent the unwanted current from
width of the 7-MHz band, if the antenna is flowing on the outside of the coaxial line. This
resonant at the center of the band. A wire antenna may be done either by making the current cancel
is assumed. Antennas having a greater ratio of itself out or by choking it off. Devices of this type
diameter to length will have a lower change in SWR fall in a class of circuits usually termed baluns, a
with frequency. contraction for “balanced to unbalanced.” The
Coaxial Cable baluns described in an earlier section under “Coup¬
ling the Transmitter to the Line” perform the same
Instead of using Twin-Lead as just described,
function, but the techniques described here are
the center of a half-wave dipole may be fed generally more suitable for mechanical reasons in
through 75-ohm coaxial cable such as RG-ll/U, as
coupling the line to the antenna.
shown in Fig. 3-56. Cable having an impedance of
The voltages at the antenna terminals in Fig.
approximately 50 ohms, such as RG-8/Ú, also may
3-56 are equal in amplitude with respect to ground
be used, particularly in those cases where the
but opposite in phase. Both these voltages act to
antenna height is such as to lower the radiation
cause a current to flow on the outside of the coax,
resistance of the antenna, below 1/4 wavelength.
and if the currents produced by both voltages were
(See Chapter Two.) The principle is exactly the
equal, the resultant current on the outside of the
same as with Twin-Lead, and the same remarks as
line would be zero since the currents are out of
to SWR apply. However, there is a considerable
phase and would cancel each other. But since one
practical difference between the two types of line.
antenna terminal is directly connected to the cable
With the parallel-conductor line the system is
shield while the other is only weakly coupled to it,
symmetrical, but with coaxial line it is inherently
the voltage at the directly connected terminal
unbalanced. produces a much larger current, and so there is
Stated broadly, the unbalance with coaxial line
is caused by the fact that the outside of the outer relatively little cancellation.
conductor is not coupled to the antenna in the The two currents could be made equal in
same way as the inner conductor and the inside of amplitude by making a direct connection between
the outer conductor. The overall result is that the outside of the line and the antenna terminal
current will flow on the outside of the outer that is connected to the inner conductor, but if it
conductor in the simple arrangement shown in Fig. were done right at the antenna terminals the line
3-56. The unbalance is rather small if the line and antenna would be short-circuited. However, if
diameter is very small compared with the length of the connection is made through a conductor
 114
Chapter 3
parallel to the line and a quarter wavelength long, Xntenna
as shown in Fig. 3-57A, the second conductor and ¿ess than-Ji-
the outside of the line act as a quarter-wave
‘ insulator” for the normal voltage and current at ~l _ Matching
the antenna terminals. (This is because a quarter¬ Stub v
wave line short-circuited at the far end exhibits a
very high resistive impedance, as explained earlier -—Line
in this chapter.) On the other hand, any unbalance
current flowing on the outside of the line because
of the direct connection between it and the To Xmtn
antenna has a counterpart in an equal current
flowing on the second conductor, because the
latter is directly connected to the other antenna
terminal. Where the two conductors are joined Antenna
together at the bottom, the resultant of the two
currents is zero, since they are of opposite phase. Matching
Thus, no current flows on the remainder of the
transmission line.
(b) x r stub
-- Ceux
Note that the length of the extra conductor has
no particular bearing on its operation in balancing To Xmtr.
out the undesired current. The length is critical
only in respect to preventing the normal operation Fig. 3-58 — Combined matching stub and balun. A
of the antenna from being upset. — Basic arrangement; B — Balun arrangement
achieved by using a section of the outside of the
coax feed line as one conductor of a matching
stub.

length of wire. (A piece of coax may be used

instead, as in the balun in Fig. 3-5 7A.) The spacing
between the stub conductors can be two to three
inches. The stub of Fig. 3-58 is ordinarily much
shorter than a quarter wavelength, and the impe¬
dance match can be adjusted by adjusting the stub
length along with the antenna length. With simple
coax feed, even with a quarter-wave balun as in
Fig. 3-57, the match depends entirely on the actual
antenna impedance and the Zo of the cable; no
adjustment is possible.

Adjustment
Fig. 3-57 — Methods of balancing the termination
when a coaxial cable is connected to a balanced When a quarter-wave balun is used it is advis¬
antenna. able to resonate it before connecting the antenna.
This can be done without much difficulty if a
Combined Balun and Matching Stub grid-dip meter is available. In the system shown in
Fig. 3-57A, the section formed by the two parallel
In certain antenna systems the balun length can pieces of line should first be made slightly longer
be considerably shorter than a quarter wavelength; than the length given by the formula. The shorting
the balun is, in fact, used as part of the matching connection at the bottom may be installed per¬
system. This requires that the radiation resistance manently. With the grid-dip meter coupled to the
be fairly low as compared with the line Zo so that shorted end, check the frequency and cut off small
a match can be brought about by first shortening lengths of the shield braid (cutting both lines
the antenna to make it have a capacitive reactance, equally) at the open ends until the stub is resonant
and then using a shunt inductance across the at the desired frequency. In each case leave just
antenna terminals to resonate the antenna and enough inner conductor remaining to make a short
simultaneously raise the impedance to a value connection to the antenna. After resonance has
equal to the lineZo. (See later section on matching been established, solder the inner and outer con¬
stubs.) The balun is then made the proper length to ductors of the second piece of coax together and
exhibit the desired value of inductive reactance. complete the connections indicated in Fig. 3-5 7 A.
The basic matching method is shown at A in An alternative method is first to adjust the
Fig. 3-58, and the balun adaptation to coaxial feed antenna length to the desired frequency, with the
is shown at B. The matching stub in the latter case line and stub disconnected, then connect the balun
is a parallel-line section, one conductor of which is and recheck the frequency. Its length may then be
the outside of the coax between point X and the adjusted so that the overall system is again reso-
antenna; the other stub conductor is an equal nant at the desired frequency.
Transmission Lines 115

The diameter of the coaxial detuning sleeve in

B should be fairly large compared with the
diameter of the cable it surrounds. A diameter of
two inches or so is satisfactory with half-inch
cable. The sleeve should be symmetrically placed
with respect to the center of the antenna so that it
will be equally coupled to both sides. Otherwise a
current will be induced from the antenna to the
outside of the sleeve. This is particularly important
at vhf and uhf.
In both the balancing methods shown in Fig.
3-57 the quarter-wave section should be cut to be
resonant at exactly the same frequency as the
antenna itself. These sections tend to have a
beneficial effect on the impedance-frequency char¬
acteristic of the system, because their reactance
varies in the opposite direction to that of the
antenna. For instance, if the operating frequency is
slightly below resonance the antenna has capacitive
reactance, but the shorted quarter-wave sections or
stubs have inductive reactance. Thus the reactances
tend to cancel, which prevents the impedance from
changing rapidly and helps maintain a low stand¬
Fig. 3-59 - A balun which provides an impedance
ing-wave ratio on the line over a band of fre¬
step-up ratio of 4:1. quencies.

Impedance Step-up Balun

Construction
A coax-line balun may also be constructed to
In constructing a balun of the type shown in give an impedance step-up ratio of 4:1. This form
Fig. 3-57A, the additional conductor and the line of balun is shown in Fig. 3-59. If 75-ohm line is
should be maintained parallel by suitable spacers. used, as indicated, the balun will provide a match
It is convenient to use a piece of coax for the for a 300-ohm terminating impedance. The U-
second conductor; the inner conductor can simply shaped section of line must be an electrical half
be soldered to the outer conductor at both ends wave in length, taking the velocity factor of the
since it does not enter into the operation of the tine into account. In most installations using this
device. The two cables should be separated suffi¬ type of balun, it is customary to roll up the length
ciently so that the vinyl covering represents only a of tine represented by the U-shaped section into a
small proportion of the dielectric between them. coil of several inches in diameter. The coil turns
Since the principal dielectric is air, the length of may be bound together with electrical tape. Be¬
the quarter-wave section is based on a velocity cause of the bulk and weight of the balun, this
factor of 0.95, approximately. type is seldom used with wire-line antennas sus¬
pended by insulators at the antenna ends. More
Detuning Sleeves commonly it is used with multielement antennas,
The detuning sleeve shown in Fig. 3-57B also is where its weight may be supported by the boom of
essentially an air-insulated quarter-wave line, but of the antenna system.
the coaxial type, with the sleeve constituting the
outer conductor and the outside of the coax line
being the inner conductor. Because the impedance
at the open end is very high, the unbalanced
voltage on the coax line cannot cause much current
to flow on the outside of the sleeve. Thus the
sleeve acts like a choke coil in isolating the
remainder of the line from the antenna. (The same
viewpoint can be used in explaining the action of
the quarter-wave arrangement shown at A, but is
less easy to understand in the case of baluns less
than 1/4 wavelength long.)
A sleeve of this type may be resonated by
cutting a small longitudinal slot near the bottom,
just large enough to take a single-turn loop which
is, in turn, link-coupled to the grid-dip meter. If
the sleeve is a tittle long to start with, a bit at a
time can be cut off the top until the stub is Fig. 3-60 — An rf choke formed by coiling the feed
resonant. line at the point of connection to the antenna.
 116 Chapter 3

Coax-Line RF Choke solid-dielectric lines. Such a matching arrangement

As was discussed earlier in this section, the is popularly known as the “Q” matching system.
unbalanced coupling which results from connecting One application of this type of matching
coaxial line to a balanced antenna may be nullified section is in matching a half-wave antenna to a
by choking off the current from flowing on the 600-ohm line, as shown in Fig. 3-61. Assuming that
outside of the feed line. A direct approach to this the antenna has a resistive impedance in the
objective is shown in Fig. 3-60, where the line itself vicinity of 65 to 70 ohms, the required Zo of the
is formed into a coil at the antenna feed point. Ten matching section is approximately 200 ohms. A
turns of coax line coiled at a diameter of 6 inches section of this type can be constructed of parallel
has been found effective for the hf bands. The tubing, from the data in Fig. 3-30.
turns may be secured in a tight coil with electrical The 1/4-wave transformer may be adjusted to
tape. This approach offers the advantage of requir¬ resonance before being connected to the antenna
ing no pruning adjustments. The effectiveness of a by short-circuiting one end and coupling it induc¬
choke of this sort decreases at the higher fre¬ tively at that end to a grid-dip meter. The length of
quencies, however, because of the distributed the short-circuiting conductor lowers the fre¬
capacitance among the turns. quency slightly, but this can be compensated for
by adding half the length of the shorting bar to
QUARTER-WAVE TRANSFORMERS each conductor after resonating, measuring the
shorting-bar length between the centers of the
The impedance-transforming properties of a conductors.
quarter-wave transmission line can be used to good
advantage in matching the antenna impedance to Driven Beam Elements
the characteristic impedance of the line. As des¬
cribed earlier, the input impedance of a quarter¬ Another application for the quarter-wave
wave line terminated in a resistive impedance ZR is “linear transformer” is in matching the very low
antenna impedances encountered in close-spaced
directional arrays to a transmission line having a
characteristic impedance of 300 to 600 ohms. The
observed impedances at the antenna feed point in
Rearranging this equation gives such cases range from about 8 to 20 ohms. A
matching section having a Zo of 75 ohms is useful
Zo = VZrZS with such arrays. The impedance at its input
This means that any value of load impedance ZR terminals will vary from approximately 700 ohms
can be transformed into any desired value of with an 8-ohm load to 280 ohms with a 20-ohm
impedance ZS at the input terminals of a quarter¬ load.
wave line, provided the line can be constructed to Transmitting Twin-Lead is suitable for this
have a characteristic impedance Zo equal to the application; such a short length is required that the
square root of the product of the two impedances. loss in the matching section should not exceed
The factor that limits the range of impedances that about 0.6 dB even though the SWR in the
can be matched by this method is the range of matching section may be almost 10 to 1 in the
values for Zo that is physically realizable. The extreme case.
latter range is approximately 50 to 600 ohms.
Practically any type of line can be used for the DELTA MATCHING
matching section, including both air-insulated and
Among the properties of a coil-and-capacitor
resonant circuit is that of transforming impe¬
dances. If a resistive impedance, Z1 in Fig. 3-62, is
connected across the outer terminals AB of a
resonant LC circuit, the impedance Z2 as viewed
looking into another pair of terminals such as BC
will also be resistive, but will have a different value
depending on the mutual coupling between the
parts of the coil associated with each pair of
terminals. Z2 will be less than Z1 in the circuit
shown. Of course this relationship will be reversed
if Z1 is connected across terminals BC and Z2 is
viewed from terminals AB.
A resonant antenna has properties similar to
those of a tuned circuit. The impedance presented
between any two points symmetrically placed with
respect to the center of a half-wave antenna will
depend on the distance between the points. The
greater the separation, the higher the value of
Fig. 3-61 — Matching a half-wave antenna to a impedance, up to the limiting value that exists
600-ohm line through a quarter-wave linear trans¬ between the open ends of the antenna. This is also
former. This arrangement is popularly known as suggested in Fig. 3-62. The impedance ZA between
the "Q" matching system. terminals 1 and 2 is lower than the impedance Zb
Transmission Lines 117

between terminals 3 and 4. Both impedances,

however, are purely resistive if the antenna is
resonant.
This principle is utilized in the delta matching
system shown in Fig. 3-63. The center impedance
of a half-wave dipole is too low to be matched
directly by any practicable type of air-insulated
parallel-conductor line. However, it is possible to
find, between two points, a value of impedance
that can be matched to such a line when a
“fanned” section or delta is used to couple the line
and antenna. The antenna length, L, should be
based on the formula in Chapter Two, using the Fig. 3-62 — Impedance transformation with a
appropriate factor for the length/diameter ratio. resonant circuit, together with antenna analogy.
The ends of the delta or “Y” should be attached at
points equidistant from the center of the antenna. less rapidly with frequency changes away from
When so connected, the terminating impedance for resonance than a single-wire antenna. Therefore it
the line will be essentially purely resistive. is possible to operate over a wide band of
Based on experimental data for the case of a frequencies, while maintaining a low SWR on the
simple half-wave antenna coupled to a 600-ohm line, than with a simple dipole. This is partly
line, the total distance, A, between the ends of the explained by the fact that the two conductors in
delta should be 0.120x for frequencies below 30 parallel form a single conductor of greater effective
MHz, and 0.115X for frequencies above 30 MHz. diameter.
The length of the delta, distance B, should be For reasons described in Chapter Two, a folded
0.150X. These values are based on a wavelength in dipole will not accept power at twice the funda¬
air, and on the assumption that the center impe¬ mental frequency. However, the current distribu¬
dance of the antenna is approximately 70 ohms. tion is correct for harmonic operation on odd
The dimensions will require modification if the multiples of the fundamental. Because the radia¬
actual impedance is very much different. tion resistance is not greatly different for a
The delta match can be used for matching the three-half-wave antenna and a single half wave, a
driven element of a directive array to a transmis¬ folded dipole can be operated on its third har¬
sion line, but if the impedance of the element is monic with a low SWR in a 300-ohm line. A 7-MHz
low — as is frequently the case - the proper folded dipole, consequently, can be used for the
dimensions for A and B must be found by 21-MHz band as well.
experimentation.
The delta match is somewhat awkward to Length Adjustment of
adjust when the proper dimensions are unknown, Multi- and Unequal-Conductor Dipoles
because both the length and width of the delta
must be varied. An additional disadvantage is that Figs. 2-87 and 2-88 of Chapter Two show how
there is always some radiation from the delta. This a wide range of impedance step-up ratios is
is because the conductors are not close enough available by varying the size of the conductors
together to meet the requirement (for negligible and/or by using more than two. Because the
radiation) that the spacing should be very small in relatively large effective thickness of the antenna
comparison with the wavelength. reduces the rate of change of reactance with
frequency, the tuning becomes relatively broad. It
is a good idea, however, to check the resonant
FOLDED DIPOLES frequency with a grid-dip meter in making length
Basic information on the folded dipole antenna adjustments. The transmission line should be dis¬
appears in Chapter Two. The two-wire system in connected and the antenna terminals temporarily
Fig. 2-86 is an especially useful one because the short-circuited when this check is being made.
input impedance is so close to 300 ohms that it can
be fed directly with 300-ohm Twin-Lead or with
open line without any other matching arrange¬
ment, and the line will operate with a low
standing-wave ratio. The antenna itself can be built
like an open-wire line; that is, the two conductors
can be held apart by regular feeder spreaders. TV
“ladder” line is quite suitable. It is also possible to
use 300-ohm line for the antenna, in addition to
using it for the transmission line. Additional
construction information is contained in Chapter
Thirteen. Since the antenna section does not
operate as a transmission line, but simply as two
wires in parallel, the velocity factor of Twin-Lead
can be ignored in computing the antenna length.
The reactance of the folded-dipole antenna varies
118 Chapter 3

As shown by the charts of Figs. 2-87 and 2-88, reactance; as a consequence, if the antenna itself is
there are two special cases where the impedance exactly resonant at the operating frequency, the
ratio of the folded dipole is independent of the input impedance of the T will show inductive
spacing between conductors. These are for a ratio reactance as well as resistance. The reactance must
of 4:1 with the two-conductor dipole and a ratio be tuned out if a good match to the transmission
of 9:1 in the three-conductor case. In all other fine is to be secured. This can be done either by
cases, the impedance ratio can be varied by shortening the antenna to obtain a value of
adjustment of the spacing. The adjustment range is capacitive reactance that will reflect through the
quite limited when ratios near 4 and 9, respec¬ matching system to cancel the inductive reactance
tively, are used, but increases with the departure in at the input terminals, or by inserting a capacitance
either direction from these “fixed” values. This of the proper value in series at the input terminals
offers a means for final adjustment of the match to as shown in Fig. 3-65, upper drawing.
the transmission line when the antenna resistance is A theoretical analysis has shown that the part
known approximately but not exactly. of the impedance step-up arising from the spacing
If a suitable match cannot be obtained by and ratio of conductor diameters is approximately
adjustment of spacing, there is no alternative but the same as given for the folded dipole in Fig. 2-87.
to change the ratio of conductor diameters. The The actual impedance ratio is, however, consider¬
impedance ratio decreases with an increase in ably modified by the length A of the matching
spacing, and vice versa. Hence, if a match cannot section (Fig. 3-64). The trends can be stated as
be brought about by changing the spacing, such a follows:
change will at least indicate whether the ratio of 1) The input impedance increases as the dis¬
d2 Id, should be increased or decreased. tance A is made larger, but not indefinitely. There
is in general a distance A that will give a maximum
THE T AND GAMMA value of input impedance, after which further
increase in A will cause the impedance to decrease.
The “T” matching system shown in Fig. 3-64 2) The distance A at which the input impe¬
has a considerable resemblance to the folded dance reaches a maximum is smaller as d2/d,
dipole; in fact, if the distanced is extended to the (using the notation of Fig. 2-87) is made larger,
full length of the antenna the system becomes an and becomes smaller as the spacing between the
ordinary folded dipole. The T has considerable conductors is increased.
flexibility in impedance ratio and is more conveni¬ 3) The maximum impedance values occur in
ent, constructionally, than the folded dipole when the region where A is 40 to 60 percent of the
used with the driven element of a rotatable antenna length in the average case.
parasitic array. Since it is a symmetrical system it is 4) Higher values of input impedance can be
inherently balanced, and so is well suited to use realized when the antenna is shortened to cancel
with parallel-conductor transmission lines. If coaxi¬ the inductive reactance of the matching section.
al line is used, some form of balun, as described
earlier, should be installed. Alternatively, the gam¬ Simple Dipole Matching
ma form described below can be used with
unbalanced lines. For a dipole having an approximate impedance
The current flowing at the input terminals of of 70 ohms, the T matching-section dimensions for
the T consists of the normal antenna current matching a 600-ohm line are given by the following
formulas:
divided between the radiator and the T conductors
in a way that depends on their relative diameters 180.5
and the spacing between them, with a superim¬ A (feet) =
/(MHz)
posed transmission-line current flowing in each half
of the T and its associated section of the antenna. 114
B (inches) =
Each such T conductor and the associated antenna /(MHz)
conductor can be looked upon as a section of
transmission line shorted at the end. Since it is These formulas apply for wire antennas with the
shorter than 1/4 wavelength it has inductive matching section made of the same size wire. With
an antenna element of different impedance, or for
matching a line having a Zo other than 600 ohms,
the matching-section dimensions can be deter¬
mined experimentally.

The Gamma
The gamma arrangement shown in Fig. 3-66 is
an unbalanced version of the T, suitable for use
with coaxial lines. Except for the fact that the
matching section is connected between the center
and one side of the antenna, the remarks above
about the behavior of the T apply equally well.
The inherent reactance of the matching section can
Fig. 3-64 — The T matching system, applied to a be canceled either by shortening the antenna
half-wave antenna and 600-ohm line. appropriately or by using the resonant length and
Transmission Lines 119

dance transformation. (See bibliography at the end

of this chapter.) The procedure uses mathematical
equations and the Smith Chart (see earlier section
of this chapter), and consists of the following basic
Trans steps.
Line
1) Find the impedance step-up ratio for the
gamma rod and element diameters and spacing.
(See Fig. 2-87; use the rod diameter as d, and the
element diameter as d2.)
2) Determine the Zo of the “transmission line”
formed by the gamma rod and the element,
considering them as two parallel conductors. Use
the equation
Trans
2S
Line
Zo = 276 logi »- ohms

where the terms are the same as for Fig. 2-87.

Fig. 3-65 — Series capacitors for tuning out
residual reactance with the T and gamma matching 3) Assign (or assume) a length for the gamma
systems. A maximum capacitance of 150 pF in rod, expressed in electrical degrees. Call this angle
each capacitor should provide sufficient adjust¬ 0.
ment range, in the average case, for 14-MHz 4) Determine the increased impedance of the
operation. Proportionately smaller capacitance driven element over its center-point impedance,
values can be used on higher frequency bands. caused by its being fed off center. Use the equation
Receiving-type plate spacing will be satisfactory for
power levels up to a few hundred watts.
cos 20
installing a capacitor C, as shown in the lower where Z2 is the impedance at the tap point and Z1
drawing of Fig. 3-65. is the complex impedance at the center of the
The gamma match has been widely used for element.
matching coaxial cable to all-metal parasitic beams 5) Determine the “load” impedance at the
for a number of years. Because it is well suited to antenna end of the gamma “transmission line.”
“plumber’s delight” construction, where all the This is the resultant value of Step 1 above
metal parts are electrically and mechanically con¬ multiplied by the value for Z2, taken as R + jX
nected, it has become quite popular for amateur from Step 4. Normalize this impedance value to
arrays. the Zo of the gamma “transmission line” deter¬
Because of the many variable factors - driven- mined in Step 2. Plot this normalized impedance
element length, gamma rod length, rod diameter, on the Smith Chart.
spacing between rod and driven element, and value 6) Using the TOWARD GENERATOR wave¬
of series capacitor — there are a number of lengths scale of the Smith Chart, take the “trans¬
combinations which will provide the desired mission line” length (rod length) into account and
match. However, the task of finding a proper determine the normalized input impedance to this
combination can be a tedious one, as the settings line. This impedance represents the portion of the
are interrelated. A few “rules of thumb” have total impedance at the gamma feed point which
evolved which provide a starting point for the arises from the antenna alone.
various factors. For matching a multielement array 7) In shunt with the impedance from Step 6 is
made of aluminum tubing to 52-ohm line, the an inductive reactance caused by the short-circuit
length of the rod should be 0.04 to 0.05 X, its termination on the gamma “transmission line”
diameter 1/3 to 1/2 that of the driven element, and itself. Determine the normalized value of this
its spacing (center to center from the driven inductance either from the Smith Chart (taking 0 +
element), approximately 0.007 X. The capacitance /0 as the load and the rod length into account on
value should be approximately 7 pF per meter of the TOWARD GENERATOR wavelengths scale) or
wavelength, i.e., about 140 pF for 20-meter opera¬ from the equation
tion. The exact gamma dimensions and value for Xp = / tan 6 ohms
the capacitor will depend upon the radiation
resistance of the driven element, and whether or Also, Fig. 3-72 may be used.
not it is resonant. These starting-point dimensions
are for an array having a feed-point impedance of
about 25 ohms, with the driven element shortened
approximately 3% from resonance.

Calculating Gamma Dimensions

D. J. Healey, W3PG, has developed a method of
determining by calculations whether or not a Fig. 3-66 — The gamma match, as used with tubing
particular set of parameters for a gamma match elements. The transmission line may be either
will be suitable for obtaining the desired impe¬ 52-ohm or 75-ohm coax.
120 Chapter 3

As an example, assume a 20-meter Yagi beam is

to be matched to 50-ohm line. The driven element
is 1-1/2 inches in diameter, and the gamma rod is a
length of 1/2-inch tubing, spaced 6 inches from the
element (center to center). Initially, the rod length
is adjusted to 0.04 X, or 14.4° (33 inches). The
driven element has been shortened by 3% from its
resonant length.
Following Step 1, from Fig. 2-87, the impe¬
dance step-up ratio is 6.4.
From the equation of Step 2, the Zo of the
transmission line is 315 ohms.
From Step 3, 9 = 14.4°.
For Step 4, assume the antenna has a radiation
resistance of 25 ohms and a capacitive reactance
component of 25 ohms (about the reactance which
would result from the 3% shortening). The overall
impedance of the driven element is therefore 25 —
/25 ohms. Using this value for Z1 in the equation
of Step 4, Z2 is determined to be 26.6 /26.6.
In Step 5, the value obtained above for Z2,
26.6 - /26.6, is multiplied by the step-up ratio
Fig. 3-67 — Smith Chart calculation of gamma (Step 1), 6.4. The resultant impedance is 170 -
dimensions. See text.
/170 ohms. Normalized to the Zo of the gamma
“transmission line,” 315 ohms, this impedance is
8) Invert the line input impedance (from Step 0.54 - /0.54. This value is plotted on the Smith
6) to obtain the equivalent admittance, G + jB. Chart, shown at point A of Fig. 3-67.
This may be done by locating the point on the Taking the line length into account (0.04 X) as
Chart which is diametrically opposite that for the directed in Step 6, the normalized line input
plot of the impedance. (Remember that inductance impedance is found to be 0.44 - /0.29, as shown at
is considered to be a negative susceptance, and point B of Fig. 3-67.
capacitance a positive susceptance.) Similarly, in¬ From the equation in Step 7, Xp is found to be
vert the inductance value from Step 7. (This /0.257. This same value may be determined from
susceptance will simply be the reciprocal of the the Smith Chart, as shown at point C, or from Fig.
reactance.) 3-72 for a matching-section length of 14.4°.
9) Add the two parallel susceptance com¬ Point D is found on the Smith Chart as directed
ponents from Step 8, taking algebraic signs into in Step 8, and represents a normalized admittance
account. Plot the new admittance on the Smith of 1.6 +/1.07 mhos. The inductance from Step 7,
Chart, G(from Step 8) + jB (from this step). above, inverted to susceptance, is —/I/O.257 =
10) Invert the admittance of Step 9 to impe¬ -/3.89. (This same value may be read diametrically
dance by locating the diametrically opposite point opposite point C in Fig. 3-67, at point C'.)
on the Smith Chart. Convert the normalized Proceeding as indicated in Step 9, the admit¬
resistance and reactance components to ohms by tance components of the parallel combination are
multiplying each by the line Zo (from Step 2). This 1.6 +/1.07 - /3.89 = 1.6 —/2.82. This admittance
impedance is that which terminates the transmis¬ is plotted as shown at point E in Fig. 3-67.
sion line with no gamma capacitor. A capacitor Inverting the above admittance to its equivalent
having the reactance of the X component of the impedance, point F of Fig. 3-67, the normalized
impedance should be used to cancel the induc¬ value of 0.16 + /0.27 is read. Multiplying each
tance, leaving a purely resistive line termination. If value by 315 (from Step 2), the input impedance
the dimensions were properly chosen, this value to the gamma section is found to be 50.4 + /85
will be near the Zo of the coaxial feed line. ohms.
Thus, a series capacitor having a reactance of 85
ohms is required to cancel the inductance in the
gamma section (from the standard reactance equa¬
tion the required capacitance is 134 pF), and a
very good match is provided for 50-ohm line.

Adjustment
After installation of the antenna, the proper
constants for the T and gamma must be deter¬
mined experimentally. The use of the variable
series capacitors, as shown in Fig. 3-65, is recom¬
mended for ease of adjustment. With a trial
position of the tap or taps on the antenna, measure
the SWR on the transmission line and adjust C
(both capacitors simultaneously in the case of the
Transmission Lines 121

boom of the antenna array. The beta match is

electrically identical to the hairpin match, the
difference being in the mechanical construction of
the matching section. With the beta match, the
conductors of the matching section straddle the
boom, one conductor being located on either side,
and the electrically neutral point consists of a
sliding or adjustable shorting clamp placed around
the boom and the two matching-section conduc¬
tors.
The electrical operation of the hairpin match
has been treated extensively by Gooch, Gardner,
and Roberts (see bibliography at the end of this
chapter). The antenna is matched to the transmis¬
sion line by forming an equivalent parallel-resonant
circuit in which the antenna resistance appears in
series with the capacitance. The impedance of this
type parallel-resonant circuit varies almost inverse¬
ly with the series antenna resistance, and therefore
T) for minimum SWR. If it is not close to 1 to 1, can cause a very small antenna resistance to appear
try another tap position and repeat. It may be as a very large resistance at the terminals of the
necessary to try another size of conductor for the resonant circuit. The values of inductance and
matching section if satisfactory results cannot be capacitance are chosen to transform the antenna
secured. Changing the spacing will show which resistance to a resistance value equal to the
direction to go in this respect, just as in the case of characteristic impedance of the transmission line.
the folded dipole discussed in the preceding The capacitive portion of this circuit is pro¬
section. duced by slightly shortening the antenna driven
element. For a given frequency the impedance of a
THE OMEGA MATCH shortened half-wave element appears as the
The omega match is a slightly modified form of
the gamma match. In addition to the series - REFLECTOR
capacitor, a shunt capacitor is used to aid in
canceling a portion of the inductive reactance (A)
' ' -7 ,- FEO ELEMENT
introduced by the gamma section. This is shown in X y 'ï

Fig. 3-68. Cl is the usual series capacitor. The --DIRECTOR

addition of C2 makes it possible to use a shorter
gamma rod, or makes it easier to obtain the desired
match when the driven element is resonant. (The
effect of the shunt capacitor may be taken into
account when calculating gamma dimensions, as in
the foregoing section, during the performance of
Step 9.) During adjustment, C2 will serve primarily
to determine the resistive component of the load as
seen by the coax line, and Cl serves to cancel any
reactance.

THE HAIRPIN AND BETA MATCHES

The usual form of the hairpin match is shown
in Fig. 3-69. Basically, the hairpin is a form of an
¿-matching network. Because it is somewhat easier
to adjust for the desired terminating impedance
than the gamma match, it is preferred by many
amateurs. Its disadvantages, as compared to the
gamma, are that it must be fed with a balanced line
(a balun may be used with a coax feeder, as shown
in Fig. 3-69), and the driven element must be split
at the center. This latter requirement complicates
the mechanical mounting arrangement for the
element, by ruling out “plumber's delight” con¬ Fig. 3-70 — For the Yagi antenna shown at A, the
input impedance at its operating frequency is
struction. represented at B, if the driven element is shorter
As indicated in Fig. 3-69, the center point of than its resonant length. By adding an inductor, as
the hairpin is electrically neutral. As such, it may shown at C, a low value of RA is made to appear as
be grounded or connected to the remainder of the a higher impedance at terminals AB. At D, the
antenna structure. The hairpin itself is usually diagram of C is redrawn in the usual ¿-network
secured by attaching this neutral point to the configuration.
122 Chapter 3

spaced 1-1/2 inches, its characteristic impedance is

300 ohms (from Fig. 3-30 or the equation associa¬
ted with it). Normalizing the required 41-ohm
reactance to this impedance, 41/300 = 0.137.
Entering the chart of Fig. 3-72 with this value,
0.137, on the scale at the bottom, it may be seen
that the hairpin length should be 7.8 electrical
degrees, or 7.8/360 wavelength. For purposes of
these calculations, taking a 97.5% velocity factor
into account, the wavelength in inches is
1 1,500//m ¡( 2.If the antenna is to be used on 20
meters, the required hairpin length is x
36U 14
= 17.8 inches. The length of the hairpin affects
primarily the resistive component of the termina¬
ting impedance as seen by the feed line. Greater
resistances are obtained with longer hairpin sec¬
tions, and smaller resistances with shorter sections.
Reactance at the feed-point terminals is tuned out
by adjusting the length of the driven element, as
necessary. If a fixed-length hairpin section is in use,
a small range of adjustment may be made in the
effective value of the inductance by spreading or
squeezing together the conductors of the hairpin.
Spreading the conductors apart will have the same
effect as lengthening the hairpin, while placing
them closer together will effectively shorten it.

Fig. 3-71 — Reactance required to match various MATCHING STUBS

antenna resistances to common line or balun
impedances. As explained earlier in this chapter, a mismatch-

antenna resistance and a capacitance in series, as

indicated schematically in Fig. 3-70B. The induc¬
tive portion of the resonant circuit at C is a hairpin
of heavy wire or small tubing which is connected
across the driven-element center terminals. The
diagram of C is redrawn in D to show the circuit in
conventional ¿-network form. R A , the radiation
resistance, is a smaller value than R IN , the imped¬
ance of the feed line. (In ¿-network matching, the
higher of the two impedances to be matched is
connected to the shunt-arm side of the network,
and the lower impedance to the series-arm side.)
Instead of using a separate hairpin matching
section and a balun, as shown in Fig. 3-69, a
combined matching stub and balun may be used
with coax line. This is shown in Fig. 3-58B. The
principles of operation are discussed in a later
section of this chapter titled Stubs on Coaxial
Lines.
If the approximate radiation resistance of the
antenna system is known, Figs. 3-71 and 3-72 may
be used to gain an idea of the hairpin dimensions
necessary for the desired match. The curves of Fig.
3-71 were obtained from design equations for
¿-network matching. Fig. 3-72 is based on the
equation, %p = / tan 0, which gives the inductive
reactance as normalized to the Zo of the hairpin, Fig. 3-72 — Inductive reactance (normalized toZo
looking at it as a short-circuit-terminated length of of matching section), scale at bottom, versus
transmission line. For example, if an antenna¬ required matching-section length, scale at left. To
system impedance of 20 ohms is to be matched to determine the length in wavelengths, divide the
52-ohm line, Fig. 3-71 indicates that the inductive number of electrical degrees by 360. For open-wire
reactance required for the hairpin is 41 ohms. If line, a velocity factor of 97.5% should be taken
into account when determining the electrical
the hairpin is constructed of quarter-inch tubing length.
Transmission Lines 123

length long has an input impedance that is both

resistive and reactive. The equivalent circuit of the
line input impedance can be formed either of
resistance and reactance in series or resistance and
reactance in parallel. Depending on the line length,
the series-resistance component, Rs, can have any
value between the terminating resistance, ZR
(when the line has zero length) andZo 2/ZR (when
the line is exactly 1/4 wave long). The same thing
is true of Rp, the parallel-resistance component.
(RS and RP do not have the same values at the
same Une length, however, other than zero and 1/4
wavelength.) With either equivalent there is some
line length that will give a value of ÄS or Rp equal
to the characteristic impedance of the line. How¬ Fig. 3-73 — Use of open or closed stubs for
ever, there will always be reactance along with the canceling the parallel reactive component of input
resistance. But if provision is made for canceling or impedance.
“tuning out” this reactive part of the input
impedance, only the resistance will remain. Since
this resistance is equal to the Zo of the transmis¬ loop, as in Fig. 3-74A, ZR is less than Zo (in the
sion line, the section from the reactance-cancella¬ average case) and therefore an open stub is called
tion point back to the generator will be properly for, installed within the first quarter wavelength of
matched. line measured from the antenna. Voltage feed, as at
Tuning out the reactance in the equivalent B, corresponds to ZR greater than Zo and there¬
series circuit requires that a reactance of the same fore requires a closed stub.
value as Xs, but of opposite kind, be inserted in The Smith Chart may be used to determine the
series with the line. Tuning out the reactance in the length of the stub and its distance from the load
equivalent parallel circuit requires that a reactance (see later example). If the load is a pure resistance
of the same value as XP but of the opposite kind and the characteristic impedances of the line and
be connected across the line. In practice it is stub are identical, the lengths may be determined
convenient to use the parallel-equivalent circuit. by equations. For the closed stub when ZR is
The transmission line is simply connected to the greater than Zo, they are
load (which of course is usually a resonant
antenna) and then a reactance of the proper value tan A = x/SWR and cot B =
is connected across the line at the proper distance s/SWR
from the load. From this point back to the For the open stub when ZR is less than Zo
transmitter there are no standing waves on the line.
A convenient type of reactance to use is a cot A = x/SWR and tan B = 1.
section of transmission line less than one-quarter VSWR
wavelength long, either open-circuited or short- In these equations the lengths A and B are the
circuited, depending on whether capacitive reac¬ distance from the stub to the load and the length
tance or inductive reactance is called for. Reac¬ of the stub, respectively, as shown in Fig. 3-74.
tances formed from sections of transmission line These lengths are expressed in electrical degrees,
are called matching stubs, and are designated as equal to 360 times the lengths in wavelengths.
open or closed depending on whether the free end In using the Smith Chart or the above equations
is open- or short-circuited. The two types of it must be remembered that the wavelength along
matching stubs are shown in the sketches of Fig. the line is not the same as in free space. If an
3-73. open-wire line is used the velocity factor of 0.975
The distance from the load to the stub (dimen¬ will apply. When solid-dielectric line is used the
sion A in Fig. 3-73) and the length of the stub, B, free-space wavelength as given by the curves must
depend on the characteristic impedances of the line be multiplied by the appropriate velocity factor to
and stub and on the ratio of ZR to Zo. Since the obtain the actual length of A and B (See Table 3-1).
ratio of ZR to Zo is also the standing-wave ratio in Although the equations above do not apply
the absence of matching, the dimensions are a when the characteristic impedances of the line and
function of the standing-wave ratio. If the line and stub are not the same, this does not mean that the
stub have the same Zo, dimensions A and B are line cannot be matched under such conditions. The
dependent on the standing-wave ratio only. Con¬ stub can have any desired characteristic impedance
sequently, if the standing-wave ratio can be mea¬ if its length is chosen so that it has the proper value
sured before the stub is installed, the stub can be of reactance. Using the Smith Chart, the correct
properly located and its length determined even lengths can be determined without difficulty for
though the actual value of load impedance is not dissimilar types of line.
known. In using matching stubs it should be noted that
Typical applications of matching stubs are the length and location of the stub should be based
shown in Fig. 3-74, where open-wire line is being on the standing-wave ratio at the load. If the line is
used. From inspection of these drawings it will be long and has fairly high losses, measuring the SWR
recognized that when an antenna is fed at a current at the input end will not give the true value at the
124 Chapter 3

Fig. 3-74 — Application

of matching stubs to
common types of an¬
tennas.

load. This point was discussed earlier in this matching section.” The justification for this is that
chapter in the section on attenuation. a quarter-wave section of line is similar to a
resonant circuit, as described earlier in this chapter,
Reactive Loads and it is therefore possible to use it to transform
In this discussion of matching stubs it has been impedances by tapping at the appropriate point
assumed that the load is a pure resistance. This is along the line.
the most desirable condition, since the antenna Earlier equations give design data for matching
that represents the load preferably should be tuned sections, A being the distance from the antenna to
to resonance before any attempt is made to match the point at which the line is connected, and A + B
the line. Nevertheless, matching stubs can be used being the total length of the matching section. The
even when the load is considerably reactive. A curves apply only in the case where the characteris¬
reactive load simply means that the loops and tic impedances of the matching section and trans¬
nodes of the standing waves of voltage and current mission line are the same. Equations are available
along the line do not occur at integral multiples of for the case where the matching section has a
1/4 wavelength from the load. To use the equa¬ different Zo than the line, but are somewhat
tions above it is necessary to find a point along the complicated and will not be given here, since it is
line at which a current loop or node occurs. Then generally possible to make the line and matching
the first set of equations gives the stub length and section similar in construction.
distance toward the transmitter from a current
loop. The second set gives the stub length and Adjustment
distance toward the transmitter from a current In the experimental adjustment of any type of
node. matched line it is necessary to measure the
standing-wave ratio with fair accuracy in order to
Stubs on Coaxial Lines tell when the adjustments are being made in the
The principles outlined in the preceding section proper direction. In the case of matching stubs,
apply also to coaxial lines. The coaxial cases experience has shown that experimental adjust¬
corresponding to the open-wire cases shown in Fig. ment is unnecessary, from a practical standpoint, if
3-73 are given in Fig. 3-75. The equations given the SWR is first measured with the stub not
earlier may be used to determine the dimensions A connected to the transmission line, and the stub is
and B. In a practical installation the junction of the then installed according to the design data.
transmission line and stub would be a T connector.
A special case of the use of a coaxial matching
stub in which the stub is associated with the
transmission line in such a way as to form a balun
has been described earlier in this chapter (Fig.
3-58). The principles used are those just described.
The antenna is shortened to introduce just enough
reactance at its input terminals to permit the
matching stub to be connected at that point, rather
than at some other point along the transmission
line as in the general cases discussed here. To use
this method the antenna resistance must be lower
than the Zo of the main transmission line, since the
resistance is transformed to a higher value. In beam
antennas this will nearly always be the case.

Matching Sections
If the two antenna systems in Fig. 3-74 are
redrawn in somewhat different fashion, as shown
in Fig. 3-76, there results a system that differs in
no consequential way from the matching stubs
previously described, but in which the stub formed
by A and B together is called a “quarter-wave Fig. 3-75 — Open and closed stubs on coaxial lines.
Transmission Lines 125

DESIGNING STUB MATCHES the impedance; 1/16.8 ohms equals 0.060 mho. To
plot this point it is first normalized by multiplying
WITH THE SMITH CHART the conductance and susceptance values by the Zo
Fig. 3-75A shows the case of a line terminated of the line. Thus, (0.060 + /0 ) X 52 = 3.1 + /0.
in a load impedance less than the characteristic This admittance value is shown plotted at point B
impedance of the line, calling for an open (capaci¬ in Fig. 3-77. It may be seen that points A and Bare
tive) stub for impedance matching. As an example, diametrically opposite each other on the chart.
suppose that the antenna is a close-spaced array fed Actually, for the solution of this example, it
by a 52-ohm line, and that the standing-wave ratio wasn’t necessary to compute the values for either
has been determined to be 3.1:1. From this point A or point B as in the above paragraph, for
information, a constant-SWR circle may be drawn they were both determined from the known SWR
on the Smith Chart. Its radius is such that it value of 3.1. As may be seen in Fig. 3-77, the
intersects the lower portion of the resistance axis points are located on the constant-SWR circle
at the SWR value, 3.1, as shown in Fig. 3-77. which was already drawn, at the two places where
Since the stubs of Fig. 3-75 are connected in it intersects the resistance axis. The plotted value
parallel with the transmission line, determining the for point A, 0.32, is simply the reciprocal of the
design of the matching arrangement is simplified if value for point B, 3.1. However, an understanding
Smith Chart values are dealt with as admittances, of the relationships between impedance and admit¬
rather than impedances. (An admittance is simply tance is easier to gain with simple examples such as
the reciprocal of the associated impedance.) This this.
leaves less chance for errors in making calculations, In stub matching, the stub is to be connected at
by eliminating the need for making series-equiva¬ a point in the line where the conductive compo¬
lent to parallel-equivalent circuit conversions and nent equals the Zo of the line. Point B represents
back, or else for using complicated equations for the admittance of the load, which is the antenna.
determining the resultant value of two complex Various admittances will be encountered along the
impedances connected in parallel. line, when moving in a direction indicated by the
A complex impedance, Z, is equal to R + jX, as toward GENERATOR wavelengths scale, but all
described earlier in this chapter. The equivalent admittance plots must fall on the constant-SWR
admittance, Y,is equal to G - jB, where G is the circle. Moving clockwise around the SWR circle
conductance component and B the susceptance. from point B, it is seen that the line input
(Inductance is taken as negative susceptance, and conductance will be 1.0 (normalized Zo of the
capacitance as positive.) Conductance and suscep¬ line) at point C, 0.082 X toward the transmitter
tance values are plotted and handled on the Smith from the antenna. Thus, the stub should be
Chart in the same manner as are resistance and connected at this location on the line.
reactance. Because of the way in which the Smith The normalized admittance at point C, the
Chart is designed, the coordinates for an admit¬ point representing the location of the stub, is 1 —
tance will be located at a point which is diametri¬ /1.2 mhos, having an inductive susceptance compo¬
cally opposite the plot for its impedance counter¬ nent A capacitive susceptance having a normalized
part — on the same SWR circle, but on the value of+/1.2 mhos is required across the line at
opposite side of prime center. the point of stub connection, to cancel the
Assuming that the close-spaced array of the inductance. This capacitance is to be obtained
foregoing example has been resonated at the from the stub section itself; the problem now is to
operating frequency, it will present a purely determine how long the stub should be. This is
resistive termination for the load end of the done by first plotting the susceptance required for
52-ohm line. From earlier information of this cancellation, 0 + /1.2, on the Chart (point D in Fig.
chapter, it is known that the impedance of the 3-77). This point represents the input admittance
antenna equals Zo/SWR = 52/3.1 « 16.8 ohms. If as seen looking into the stub. The “load” or
this value were to be plotted as an impedance on termination for the stub section is found by
the Smith Chart, it would first be normalized moving in the TOWARD LOAD direction around
(16.8/52 = 0.32) and then plotted as 0.32 + /0. the chart, and will appear at the closest point on
Although not necessary for the solution of this the resistance/conductance axis, either at the top
example, this value is plotted at point A in Fig. or the bottom of the Chart. Moving counter¬
3-77. What is necessary is a plot of the admittance clockwise from point D, this is located at E, at the
for the antenna as a load. This is the reciprocal of top of the chart, 0.139 X away. From this we know

Fig. 3-76 — Application of match¬

ing sections to common antenna
types.
126 Chapter 3

The proper value of reactance may be deter¬

mined from Smith Chart information. In the
previous example, the required susceptance, nor¬
malized, was +/T.2 mhos. This is converted into
actual mhos by dividing by the line Zo; 1.2/52 =
0.023 mho, capacitive. The required capacitive
reactance is the reciprocal of this latter value,
1/0.023 = 43.5 ohms. If the frequency is 14.2
MHz, for instance, 43.5 ohms corresponds to a
capacitance of 258 pF. A 325-pF variable capacitor
connected across the line 0.082 wavelength from
the antenna terminals would provide ample adjust¬
ment range. The rms voltage across the capacitor is
E = >JP- Zo and for 500 watts, for example, would
be E = \/500 X 52 = 161 volts. The peak voltage is
1.41 times the rms value, or 227 volts.

FLEXIBLE SECTIONS FOR

ROTATABLE ARRAYS
When open-wire transmission line is used there
is likely to be trouble with shorting or grounding
of feeders in rotatable arrays unless some special
Fig. 3-77 — Smith Chart method of determining precautions are taken. Usually some form of
the dimensions for stub matching. insulated flexible line is connected between the
antenna and a stationary support at the top of the
tower or mast on which the antenna is mounted.
the required stub length. The “load” at the far end Such a flexible section can take several forms,
of the stub, from Fig. 3-75A, should be an open and it can be made to do double duty. Probably
circuit. This load, as represented on the Smith the most satisfactory system, for arrays that are
Chart, has a normalized admittance of 0 + ;0 mho, not designed to be fed with coaxial line, is to use a
which is equivalent to an open circuit. flexible section of coax with coaxial baluns at both
When the stub, having an input admittance of 0 ends. The outer conductor of the coax may be
+ /T.2 mhos, is connected in parallel with the line grounded to the tower or to the beam antenna
at a point 0.082 X from the load, where the line framework, wherever it is advantageous to do so.
input admittance is 1.0 - /T.2, the resultant Such a flexible section is shown in Fig. 3-78. If the
admittance is the sum of the individual admit¬ coaxial section is made any multiple of a half wave
tances. The conductance components are added in electrical length, the impedance of the array will
directly, as are the susceptance components. In this be repeated at the bottom of the flexible section.
case, 1.0 — /T.2 + /T.2 = 1.0 + /0 mho. Thus, the Another method is to use Twin-Lead for the
line from the point of stub connection to the flexible section. The 300-ohm tubular type
transmitter will be terminated in a load which designed for transmitting applications is recom¬
offers a perfect match. When determining the mended. Here, again, half-wave sections repeat the
physical line lengths for stub matching, it is antenna feed impedance at the bottom end. The
important to remember that the velocity factor for Twin-Lead section may also be made an odd
the type of line in use must be considered. multiple of a quarter wavelength, in which case it
will act as a Q section, giving an impedance
MATCHING WITH LUMPED CONSTANTS step-down between a 450-ohm line and an antenna
It was pointed out earlier that the purpose of a impedance of 200 ohms.
matching stub is to cancel the reactive component
of line impedance at the point of connection. In
other words, the stub is simply a reactance of the GROUND-PLANE ANTENNAS
proper kind and value shunted across the line. It The same principles discussed earlier also apply
does not matter what physical shape this reactance to an unsymmetrical system such as the grounded
takes. It can be a section of transmission line or a antenna or the ground-plane antenna. In the case
“lumped” inductance or capacitance, as desired. In of the quarter-wave ground-plane antenna a
the above example with the Smith Chart solution, straightforward design procedure for matching is
a capacitive reactance was required. A capacitor possible because the radiation resistance is essen¬
having the same value of reactance can be used just tially independent of the physical height of the
as well. There are cases where, from an installation system (provided the radiator is reasonably clear of
standpoint, it may be considerably more conven¬ other conductors in the vicinity) and there is no
ient to use a capacitor in place of a stub. This is ground-connection resistance to be included in the
particularly true when open-wire feeders are used. total resistance to be matched.
If a variable capacitor is used, it becomes possible The ground-plane antenna lends itself well to
to adjust the capacitance to the exact value direct connection to coaxial line, so this type of
required. Une is nearly always used. Several matching
Transmission Lines 127

Fig. 3'78 — Flexible

sections for rotatable
arrays. Coax may be
used, as at A. If the
coax section is any mul¬
tiple of a half wave¬
length, the antenna im¬
pedance will be re¬
peated at the bottom
end. Twin-Lead may be
used either as a Q sec¬
tion or as an impedance
repeater, as shown in B.

methods are available. If the antenna length can be attaching the line tap (2), the radiator tap (/) may
adjusted to resonance, the stub matching system be set for resonance at the operating frequency as
previously described is convenient. indicated by a grid-dip meter coupled to L. The
A second method of matching, particularly line tap (2) is then moved along the coil to find the
convenient for small antennas (28 MHz and higher point that gives minimum standing-wave ratio as
frequencies) mounted on top of a supporting mast indicated by an SWR indicator. To bring the SWR
or pole, requires shortening the antenna to the down to 1 to 1 it will usually then be necessary to
high-frequency side of resonance so that it shows a make a small readjustment of the radiator tap (7)
particular value of capacitive reactance at its base. and perhaps further “touch up” the line tap (2),
The antenna terminals are then shunted by an since the adjustments interact to some extent.
inductive reactance, which may have the physical This method is equivalent to tapping down on a
form either of a coil or a closed stub, to restore parallel-resonant circuit to match a low value of
resonance and simultaneously transform the radia¬ resistance to a higher value connected across the
tion resistance to the proper value for matching the whole circuit. The antenna impedance can be
transmission line. This concept is the same as for represented by a capacitance in parallel with a
the hairpin match, described in detail in an earlier resistance which is much higher than the actual
section. radiation resistance. The transformation of resis¬
tance comes about by utilizing the parallel equiva¬
Tapped-Coil Matching lent of the radiation resistance and capacitive
reactance in series, using the relationships given in
The matching arrangement shown in Fig. 3-79
Fig. 3-15.
is a more general form of the method just
mentioned, in that it does not require adjusting the
radiator height to an exact value. The radiator Matching by Length Adjustment
must be shortened so that the system will show Still another method of matching may be used
capacitive reactance, but any convenient amount when the antenna length is not fixed by other
of shortening can be used. This system is particu¬ considerations. As shown in Chapter Two under
larly useful on lower frequencies where it may not Special Antenna Types, the radiation resistance as
be possible to obtain a height approximating a measured at the base of a ground-plane antenna
quarter wavelength. increases with the antenna height, and it is possible
The antenna impedance is matched to the to choose a height such that the base radiation
characteristic impedance of the line by adjusting resistance will equal the Zo of the transmission line
the taps on L. As a preliminary adjustment, before to be used. The heights of most interest are a little
 128 Chapter 3

corresponding capacitance values for the frequency

in question can be determined from appropriate
charts or by equation. Variable capacitors of
sufficient range should be used.
In an analysis by Robert Stephens, W3MIR, the
impedance and reactance versus height data of
Chapter Two for ground-plane antennas has been
converted to conductance and susceptance infor¬
mation. For parallel-equivalent matching, an input
conductance of 1/52 mho is needed. For an
antenna length-to-diameter ratio of approximately
1000, there are two heights for which the conduc¬
tance is this value - one at 0.234 X and the other
at 0.255 X. At the shorter height, the susceptance
is positive (capacitive), 0.0178 mho, and at the
longer it is negative, 0.0126 mho. So far as
Fig. 3-79 — Matching to ground-plane antenna by radiation is concerned, one is as good as the other,
tapped coil. This requires that the antenna (but not and the choice becomes the one of the simpler
radials) be shorter than the resonant length. mechanical approach for the particular antenna. If
the antenna is made 0.234 wavelength high, its
over 100 degrees (0.28 wavelength), where the capacitive reactance may be canceled with a shunt
resistance is approximately 52 ohms, and about inductor having a reactance of 56 ohms at the
113 degrees (0.32 wavelength), where the resis¬ operating frequency, resulting in a 52-ohm termi¬
tance is 75 ohms, to match the two common types nation for the feed line. If the antenna Ls 0.255
of coaxial line. These heights are quite practicable wavelength, a 52-ohm match may be obtained with
for ground-plane antennas for 14 MHz and higher a shunt capacitor of 79 ohms. Similarly, a match
frequencies. The lengths (heights) in degrees as may be obtained for 75-ohm coax with a height of
given above do not require any correction for 0.23 wavelength and a shunt inductor of 56 ohms,
length/diameter ratio; i.e., they are free-space or with a height of 0.26 wavelength and a shunt
lengths. capacitor of 78 ohms. These reactance values may
Since the antenna is not resonant at these be obtained with lumped constants or with stubs,
lengths, its input impedance will be reactive as well as described earlier. As mentioned above, these
as resistive. The reactance must be tuned out in heights do not require any correction factor; they
order to make the line see a purely resistive load are free-space lengths.
equal to its characteristic impedance. This can be The adjustment of systems like these requires
done with a series capacitor of the proper value. only that the capacitance or inductance be varied
The approximate value of capacitive reactance until the lowest possible SWR is obtained. If the
required, for antennas of typical length/diameter lengths mentioned above are used, the SWR should
ratio, is about 100 ohms for the 52-ohm case and be close enough to 1 to 1 to make a Tine
about 200 ohms for the 75-ohm case. The adjustment of the length unnecessary.

BANDWIDTH AND ANTENNA Q

Although more properly a subject for discus¬ frequency the SWR will rise rapidly off resonance,
sion in connection with antenna fundamentals, the but if the rate of reactance change is small the shift
bandwidth of the antenna is considered here in SWR likewise will be small. Hence an antenna
because as a practical matter the change in antenna that has a relatively slow rate of reactance change
impedance with frequency is reflected as a change will cover a wider frequency band, for a given value
in the standing-wave ratio on the transmission line. of SWR at the band limits (such as 2 to 1 or 3 to
Thus when an antenna is matched to the line at 1), than one having a relatively rapid rate of
one frequency - usually in the center of the band reactance change.
of frequencies over which the antenna is to be used In the region around the resonant frequency of
— a shift in the operating frequency will be the antenna the impedance as measured at a
accompanied by a change in the SWR. This would current loop varies with frequency in essentially
not occur if the antenna impedance were purely the same way as the impedance of a series-resonant
resistive and constant regardless of frequency, but circuit using lumped constants. It is therefore
unfortunately no practical antennas are that “flat.” possible to define a quantity Q for the antenna in
In the frequency region around resonance the the same way as Q is defined in a series-resonant
resistance change is fairly small and, by itself, circuit. The Q of the antenna is a measure of the
would not affect the SWR enough to matter, antenna’s selectivity, just as the Q of an ordinary
practically. The principal cause of the change in circuit is a measure of its selectivity.
SWR is the change in the reactive component of The Q of the antenna can be found by
the antenna impedance when the frequency is measuring its input resistance and reactance at
varied. If the reactance changes rapidly with some frequency close to the resonant frequency (at
Transmission Lines 129

Fig. 3-80 — Bandwidth in terms of SWR limits, as a

function of antenna Q. The inset formula gives an
effective Q (QI determined by the fractional band
(A///o) and the actual antenna Q as defined in the
text.

exact resonance the antenna is purely resistive and

there is no reactive component). Then, for fre¬
quency changes of less than 5 percent from the
exact resonant frequency, the antenna Q is given
with sufficient accuracy by the following formula:

Q= í .X
R In

where X and R are the measured reactance and SWR LIMIT ALLOWED
resistance and nis the percentage difference,
expressed as a decimal, between the antenna’s
resonant frequency and the frequency at which X
and R were measured. For example, if the fre¬ reactance must be used in the matching section; in
quency used for the measurement differs from the other words, the Q of the matching section in such
resonant frequency by 2 percent, n = 0.02. cases tends to be higher than desirable. Simple
For an ordinary half-wave dipole over the range systems having direct matching, such as a dipole
of length/diameter ratios used in Fig. 2-7, Chapter fed with 75-ohm line or a folded dipole matched to
Two, the approximate Q values vary from about 14 the line, will have the greatest bandwidth, other
for curve A to about 8 for curve C. In parasitic things being equal, because no matching network is
arrays with close spacing between elements the required.
input Q may be well over 50, depending on the
spacing and tuning (see Chapter Four).
Bibliography
SWR vs. Q
Source material and more extended discussion
If the Q of the antenna is known, the variation of topics covered in this chapter can be found in
in SWR over the operating band can be determined the references given below and in the textbooks
from Fig. 3-80. It is assumed that the antenna is listed at the end of Chapter Two.
matched to the line at the center frequency of the
band. Conversely, if a limit is set on the SWR, the Armed Services Index of RF Transmission Lines
width of the band that can be covered can be and Fittings, Armed Services Electro-Standards
found from Fig. 3-80. As an example, suppose that Agency, Ft. Monmouth, N.J.
a dipole having a Q of 15 (more or less typical of a Belcher, “RF Matching Techniques, Design and
wire antenna) is to be used over the 3.5-4 MHz Example,” QST, October, 1972.
band and that it is matched with a I-to-1 SWR at Cholewski, “Some Amateur Applications of the
the band center. Then tsf/fo = 0.5/3.75 = 0.133 Smith Chart,” QST, January, 1960.
Geiser, “Resistive Impedance Matching with
and QI = 15 X 0.133 =2. The SWR that can be Quarter-Wave Lines,” QST, February, 1964.
expected at the band edges, 3.5 and 4 MHz, is Gooch, Gardner and Roberts, “The Hairpin
shown by the chart to be a bit over 4 to 1. If it Match,” QST, April, 1962.
should be decided arbitrarily that no more than a Grammer, “Simplified Design of Impedance-
2-to-l standing-wave ratio is allowable, QI is 0.75 Matching Networks,” QST, March, April, May,
and from the formula in Fig. 3-80 the total 1957.
bandwidth is found to be 187.5 kHz. Grammer, “Antennas and Feeders,” QST, October,
November, December, 1963.
Effect of Matching Network Hall, “Smith-Chart Calculations for the Radio
Amateur,” QST, January, February, 1966.
The measurement of resistance and reactance to Healey, “An Examination of the Gamma Match,”
determine Q should be made at the input terminals QST, April, 1969.
of the matching network, if one is required. The Kraus and Sturgeon, “The T-Matched Antenna,”
selectivity of the matching network has just as QST, September, 1940.
much effect on the bandwidth, in terms of SWR on Maxwell, “Another Look at Reflections,” QST,
the line, as the selectivity of the antenna itself. Apr., June, Aug., Oct. 1973, Apr., 1974.
Where the greatest possible bandwidth is wanted a McCoy, “The Ultimate Transmatch,” QST, July,
1970.
low-Ç matching network must be used. 1'his is not Stephens, “Admittance Matching the Ground-Plane
always controllable, particularly when the antenna Antenna to Coaxial Transmission Line,” Tech¬
resistance differs considerably from the Zo of the nical Correspondence, QST, April, 1973.
line to which it is to be matched. A large Turrin, "Application of Broad-Band Balun Trans¬
impedance ratio usually means that large values of formers,” QST, April, 1969.
Chapter 4

Multielement Directive Arrays

The gain and directivity that can be secured by current /, the field at P will be IE. With A, B and C
intentionally combining antenna elements into an operating, the field will be 3E, and with all four
array represent a worthwhile improvement both in operating with the same /, the field will be AE.
transmitting and receiving. Power gain in an an¬ Since the power received at P is proportional to the
tenna is the same as an equivalent increase in the square of the field strength, the relative power
transmitter power. But, unlike increasing the received at P is 1, 4, 9, and 16, depending on
power of one’s own transmitter, it works equally whether one, two, three or four dipoles are
well on signals received from the favored direction. operating.
In addition, the directivity reduces the strength of
signals coming from the directions not favored, and
so helps discriminate against a good deal of
interference.
One common method of securing gain and
directivity is to combine the radiation from a
group of half-wave dipoles in such a way as to
concentrate it in a desired direction. The way in
which such combinations affect the directivity has
been explained in Chapter Two. A few words of
rdditional explanation may help make it clear how
power gain is achieved. CM (B) (C)
In Fig. 4-1, imagine that the four circles, A, B,
C, and D represent four dipoles so far separated Fig. 4-2 — Parallel (A) and collinear (B) antenna
elements. The array shown at C combines both
from each other that the coupling between them is parallel and collinear elements.
negligible. The point P is supposed to be so far
away from the dipoles that the distance from P to
each one is exactly the same (obviously P would
have to be much farther away than it is shown in Now since all four dipoles are alike and there is
this drawing). Under these conditions the fields no coupling between them, the same power must
from all the dipoles will add up at P if all four are be put into each in order to cause the current Ito
fed rf currents in the same phase. flow. For two dipoles the relative power input is 2,
Let us say that a certain current, /, in dipole A for three dipoles it is 3, for four dipoles 4, and so
will produce a certain value of field strength, E, at on. The gain in each case is the relative received (or
the distant point P. The same current in any of the output) power divided by the relative input power.
other dipoles will produce the same field at P. Thus Thus we have:
if only dipoles A and B are operating, each with a
Relative Relative Gain
Output Input Power in
Dipoles Power Power Gain dB
A only 1 1 1 0
A and B 4 2 2 3
A, B
and C 9 3 3 4.8
A, B, C
and D 16 4 4 6

The power gain is directly proportional to the

number of elements used.
Fig. 4-1 — Fields from separate antennas combine It is well to have clearly in mind the conditions
at a distant point, P, to produce a field strength under which this relationship is true:
that exceeds the field produced by the same power 1) The fields from the separate antenna ele¬
in a single antenna. ments must be in phase at the receiving point.

130
Multielement Directive Arrays 131

2) The currents in all elements must be iden¬

tical.
3) The elements must be separated in such a
way that the current induced in one by another is
negligible, i.e., the radiation resistance of each
clement must be the same as it would have been
had the other elements not been there.
Very few antenna arrays meet all these con¬
ditions exactly. However, as a rough approxi¬
mation it may be said that the power gain of a
directive array consisting of dipole elements in
which optimum values of element spacing are used
is proportional to the number of elements. It is not
impossible, though, for an estimate based on this
rule to be in error by a factor of 2 or more.
Fig. 4-4 — Typical bidirectional (A) and unidirec¬
Definitions tional (B) directive patterns. These drawings also
The “element” in a multielement directive illustrate the application of the terms "major" and
array is usually a half-wave dipole. The length is "minor" to the pattern lobes.
not always an exact electrical half wavelength,
because in some types of arrays it is desirable that parallel, as at A in Fig. 4-2, or collinear (end-to-
the element show either inductive or capacitive end), Fig. 4-2B. Fig. 4-2C shows an array com¬
reactance. However, the departure in length from a bining both parallel and collinear elements. The
true half-wave is ordinarily small (not more than elements can be either horizontal or vertical,
5%, in the usual case) and so has no appreciable depending on whether horizontal or vertical polar¬
effect on the radiating properties of the element. ization is desired. There is seldom any reason for
Antenna elements in multielement arrays of the mixing polarization, so arrays are customarily
type considered in this chapter are always either constructed with all elements similarly polarized.
A driven element is one supplied power from
Diredion of
the transmitter, usually through a transmission
Max Radatm line. A parasitic element is one that obtains power
solely through coupling to another element in the
Axis Of _
(A) array because of its proximity to such an element.
A driven array is one in which all the elements
are driven elements. A parasitic array is one in
which one or more of the elements are parasitic
elements. At least one element in a parasitic array
has to be a driven element, since it is necessary to
introduce power into the array.
A broadside array is one in which the principal
direction of radiation is perpendicular to the axis
of the array and to the plane containing the
elements. An end-fire array is one in which the
principal direction of radiation coincides with the
(B) direction of the array axis. These definitions are
illustrated by Fig. 4-3.
A bidirectional array is one that radiates
equally well in either direction along the line of
maximum radiation. A bidirectional pattern is
shown in Fig. 4-4 at A. A unidirectional array is
one that has only one principal direction of
radiation, as illustrated by the pattern at B in Fig.
4-4.
Axis of Predion of The major lobes of the directive pattern are
Vier from oMj of •It monta ’m», rm those in which the radiation is maximum. Lobes of
(C) lesser radiation intensity are called minor lobes.
Axis of Diffdionof
The beamwidth of a directive antenna is the width,
Krrty Max. Rad. in degrees, of the major lobe between the two
directions at which the relative radiated power is
Sido View
equal to one half its value at the peak of the lobe.
At these “half-power points” the field intensity is
Fig. 4-3 — Representative broadside arrays are equal to 0.707 times its maximum value, or down
shown at A and B, the first with collinear elements, 3 dB from maximum. Fig. 4-5 is an example of a
the second with parallel elements. An end-fire
array is shown at C. Practical arrays may combine lobe having a beam width of 30 degrees.
both broadside and end-fire directivity, including Unless specified otherwise, the term “gain” as
both parallel and collinear elements. used in this chapter is the power gain over a
 132 Chapter 4

This is illustrated in Fig. 4-6. Assume that by

some means an identical voltage is applied to each
of the dipoles at the end marked A. Assume also
that the instantaneous polarity of the voltage is
such that the current is flowing away from the
point at which the voltage is applied. The arrows
show the assumed current directions. Then the
currents in elements 1 and 2 are completely in
phase, since they are flowing in the same direction
in space and arc caused by the same voltage.
However, the current in element 3 is flowing in the
opposite direction in space because the voltage is
Fig. 4-5 — The width of a beam is the angular applied to the opposite end of the element. The
distance between the directions at which the current in element 3 is therefore 180 degrees out
received or transmitted power is one half the of phase with the currents in elements 1 and 2.
maximum power.
The phasing of driven elements depends on the
direction of the element, the phase of the applied
half-wave dipole of the same orientation and height voltage, and the point at which the voltage is
as the array under discussion, and having the same applied. In the systems used by amateurs the
power input. Gain may either be measured experi¬ voltages applied to the elements are practically
mentally or determined by calculation. Experi¬ always exactly in or exactly out of phase with each
mental measurement is difficult and often subject other. Also, the axes of the elements are always in
to considerable error, since in addition to the the same direction, since parallel or collinear
normal errors in measurement (the accuracy of elements are invariably used. The currents in driven
simple rf measuring equipment is relatively poor, elements in such systems are therefore always
and even high-quality instruments suffer in accu¬ either exactly in or out of phase with the currents
racy compared with their low-frequency and de in other elements.
counterparts) the accuracy depends considerably It is possible to use phase differences of less
on conditions - the antenna site, including height, than 180 degrees in driven arrays - one important
terrain characteristics, and surroundings under case is where the voltage applied to one set of
which the measurements are made. Calculations arc elements differs by 90 degrees from the voltage
frequently based on the measured or theoretical applied to another set — but such systems have not
directive patterns (see Chapter Two) of the an¬ met with much application in amateur work. The
tenna. An approximate formula often used is reason probably is that making provision for
proper phasing is considerably more of a problem
„ 40,000 than in the case of simple 0- or 180-degree phasing.
, ’ e He Y
In parasitic arrays the phase of the currents in
where the parasitic elements depends on the spacing and
G = gain over an isotropic antenna. tuning, as described later.
0H = horizontal half-power beam width in
degrees. Ground Effects
pv = vertical half-power beam width in The effect of the ground is the same with a
degrees. directive antenna as it is with a simple dipole
This formula, strictly speaking, applies only to antenna. The reflection factors discussed in Chap¬
antennas having approximately equal and narrow E ter Two may therefore be applied to the vertical
and //-plane (see Chapter Two) beamwidths up pattern of an array, subject to the same modifica¬
to about 20 degrees - and no large minor lobes. tions mentioned in that chapter. In cases where the
The error may be considerable when the formula is array elements are not all at the same height, the
applied to simple directive antennas having rel¬ reflection factor for the mean height of the array
atively large beamwidths, The error is in the must be used. The mean height is the average of
direction of making the calculated gain larger than the heights measured from the ground to the
the actual gain. centers of the lowest and highest elements.
Front-to-back ratio means the ratio of the
power radiated in the favored direction to the A

power radiated in the opposite direction.

Phase
The term “phase” has the same meaning when
used in connection with the currents flowing in
antenna elements as it does in ordinary circuit
work. For example, two currents are in phase when
they reach their maximum values, flowing in the
same direction, at the same instant. The direction
of current flow depends on the way in which Fig. 4-6 — Illustrating phasing of currents in
power is applied to the element. antenna elements.
Multielement Directive Arrays 133

MUTUAL IMPEDANCE

Consider two half-wave dipoles that are fairly

close to each other. When power is applied to one
and current flows, a voltage will be induced in the
second by the electromagnetic field and current
will flow in it as well. The current in antenna No. 2
will in turn induce a voltage in antenna No. 1,
causing a current to flow in the latter. The total
current in No. 1 is then the sum (taking phase into
account) of the original current and the induced
current.
If the voltage applied to antenna No. 1 has not
changed, the fact that the amplitude of the current
flowing is different, with antenna No. 2 present,
than it would have been had No. 2 not been there
indicates that the presence of the second antenna
has changed the impedance of the first. This effect Fig. 4-7 — Radiation resistance measured at the
is called mutual coupling, and results in a mutual center of one element as a function of the spacing
impedance. The actual impedance of an antenna between two parallel half-wave self-resonant
element is the sum of its seif-impedance (the antenna elements.
impedance with no other antennas present) and its
mutual impedance with all other antennas in the raised depending on whether the induced current is
vicinity. mostly in or mostly out of phase with the original
The magnitude and nature of the mutual current.
impedance depends on the amplitude of the Except in the special cases when the induced
current induced in the first antenna by the second, current is exactly in or out of phase with the
and on the phase relationship between the original original current, the induced current causes the
and induced currents. The amplitude and phase of phase of the total current to shift with respect to
the induced current depend on the spacing be¬ the applied voltage. The mutual impedance, in
tween the antennas and whether or not the second other words, has both resistive and reactive com¬
antenna is tuned to resonance. ponents. Consequently, the presence of a second
antenna nearby may cause the impedance of an
Amplitude of Induced Current antenna to be reactive - that is, the antenna will
The induced current will be largest when the be detuned from resonance - even though its
two antennas are close together and are parallel. self-impedance is entirely resistive. The amount of
Under these conditions the voltage induced in the detuning depends on the magnitude and phase of
second antenna by the first, and in the first by the the induced current.
second, has its greatest value and causes the largest
current flow. The coupling decreases as the paral¬ Tuning Conditions
lel antennas are moved farther apart. A third factor that affects the impedance of
The coupling between collinear antennas is antenna No. 1 when No. 2 is present is the tuning
comparatively small, and so the mutual impedance of the latter. If No. 2 is not exactly resonant the
between such antennas is likewise small. It is not current that flows in it as a result of the induced
negligible, however. voltage will either lead or lag the phase it would
have had if the antenna were resonant. This causes
Phase Relationships an additional phase advance or delay that affects
the phase of the current induced back in No. 1.
When the separation between the two antennas Such a phase lag has an effect similar to change in
is an appreciable fraction of a wavelength a the spacing between self-resonant antennas. How¬
measurable period of time elapses before the field ever, a change in tuning is not exactly equivalent to
from antenna No. 1 reaches antenna No. 2 and a change in spacing because the two methods do
there is a similar time lapse before the field set up not have the same effect on the amplitude of the
by the current in No. 2 gets back to induce a
induced current.
current in No. 1. Hence the current induced in No.
1 by No. 2 will have a phase relationship with the Mutual Impedance and Gain
original current in No. 1 that depends on the
spacing between the two antennas. The mutual impedance between antennas is
The induced current can range all the way from important because it determines the amount of
being completely in phase with the original current current that will flow for a given amount of power
to being completely out of phase with it. In the supplied. It must be remembered that it is the
first case the total current is larger than the original current that determines the field strength from the
current and the antenna impedance is reduced. In antenna. Other things being equal, if the mutual
the second, the total current is smaller and the impedance between two antennas is such that the
impedance is increased. At intermediate phase currents are greater for the same total power than
relationships the impedance will be lowered or would be the case if the two antennas were not
 134 Chapter 4

about 0.7 wavelength. The maximum gain is

secured from a pair of such elements when the
spacing is in this region, because the current is
larger for the same power and the fields from the
two arrive in phase at a distant point placed on a
line perpendicular to the line joining the two
antennas (see Fig. 2-17, Chapter Two).
The broken curve in Fig. 4-7 representing two
antennas operated 180 degrees out of phase (end¬
fire), cannot be interpreted quite so simply. The
radiation resistance decreases with decreasing
Fig. 4-8 — Radiation resistance measured at the spacing in this case. However, the fields from the
center of one element as a function of the spacing two antennas add up in phase at a distant point in
between the ends of two collinear self-resonant the favored direction only when the spacing is
half-wave antenna elements operated in phase. one-half wavelength (in the range of spacings
considered). At smaller spacings the fields become
increasingly out of phase, so the total field is less
coupled, the power gain will be greater than in the than the simple sum of the two. The latter factor
case discussed at the beginning of this chapter. On decreases the gain at the same time that the
the other hand, if the mutual impedance is such as reduction in radiation resistance is increasing it. As
to reduce the current, the gain will be less than if shown later in this chapter, the gain goes through a
the antennas were not coupled. maximum when the spacing is in the region of 1/8
The calculation of mutual impedance between wavelength.
antennas is a difficult problem, but data are The curve for two collinear elements in phase,
available for several special cases. Two simple but Fig. 4-8, shows that the radiation resistance de¬
important ones are shown in Figs. 4-7 and 4-8. creases and goes through a broad minimum in the
These graphs do not show the mutual impedance region of 0.3- to 0.5-wavelength spacing between
but instead show a more useful quantity, the the adjacent ends of the antennas. Since the
radiation resistance measured at the center of an minimum is not significantly less than the radiation
antenna as it is affected by the spacing between resistance of an isolated antenna, the gain will not
two antennas. exceed the gain calculated on the basis of un¬
As shown by the solid curve in Fig. 4-7, the coupled antennas. That is, the best that two
radiation resistance at the center of either dipole, collinear elements will give, even with the optimum
when the two are self-resonant, parallel, and spacing, is a power gain of about 2 (3 dB). When
operated in phase, decreases rapidly as the spacing the separation between the ends is very small - the
between them is increased until the spacing is usual method of operation — the gain is reduced.

DRIVEN ARRAYS
Driven arrays may be either broadside or elements if the space occupied by the antenna is
end-fire, and may consist of collinear elements, not increased proportionally.
parallel elements, or a combination of both. The Generally speaking, the maximum gain in the
number of elements that it is practicable to use smallest linear dimensions will result when the
depends on the frequency and the space available antenna combines both broadside and end-fire
for the antenna. Fairly elaborate arrays, using as directivity and uses both parallel and collinear
many as 16 or even 32 elements, can be installed in elements. In this way the antenna is spread over a
a rather small space when the operating frequency greater volume of space, which has the same effect
is in the vhf range. At lower frequencies the as extending its length to much greater extent in
construction of antennas with a large number of one linear direction.
elements would be impracticable for most ama¬
teurs. Feeding Driven Arrays
It is characteristic of broadside arrays that the Not the least of the problems encountered in
power gain is proportional to the length of the constructing multielement driven arrays is that of
array but is substantially independent of the supplying the required amount of power to each
number of elements used, provided the optimum element and making sure that the currents in the
element spacing is not exceeded. This means, for elements are in the proper phase. The directive
example, that a 5-element array and a 6-element patterns given in this chapter are based on the
array will have the same gain, provided the assumption that each element carries the same
elements in both are spaced so that the overall current and that the phasing is exact. If the
array length is the same. Although this principle is element currents differ, or if the phasing is not
seldom used for the purpose of reducing the proper, the actual directive patterns will not be
number of elements, because of complications quite like those shown. Small departures will not
introduced in feeding power to each element in the greatly affect the gain, but may increase the beam
proper phase, it does illustrate the fact that there is width and introduce minor lobes — or emphasize
nothing to be gained by increasing the number of those that exist already.
Multielement Directive Arrays 135

If the directive properties of beam antennas are

to be fully realized, care must be used to prevent
antenna currents from flowing on transmission
lines (see Chapter Three) used as interconnections
between elements, as well as on the main
transmission line. If radiation takes place from
these lines, of if signals can be picked up on them,
the directive effects may be masked by such stray
radiation or pickup. Although this may not greatly
affect the gain either in transmission or reception,
received signals coming from undesired directions Fig. 4-9 — Gain of two collinear half-wave elements
will not be suppressed to the extent that is possible as a function of spacing between the adjacent ends.
with a well-designed system.
radiation to low vertical angles. For purposes of
COLLINEAR ARRAYS estimating the effect of ground reflection the
height is taken as the height of the center of the
Collinear arrays are always operated with the array. Applying the ground-reflection factor for
elements in phase. (If alternate elements in such an this height (using the reflection factors given in
array are out of phase, the system simply becomes Chapter Two for half-wave antennas) to the di¬
a harmonic-type antenna.) A collinear array is a rective pattern of the array will give the resultant
broadside radiator, the direction of maximum vertical pattern, taking into account ground re¬
radiation being at right angles to the line of the flection.
antenna. If a collinear array is mounted horizontally, the
Power Gain directive pattern in the vertical plane at right angles
Because of the nature of the mutual impedance to the array is the same as the vertical pattern of a
between collinear elements the radiation resistance simple half-wave antenna at the same height
is increased as shown in Fig. 4-8. For this reason (Chapter Two).
the power gain does not increase in direct pro¬
portion to the number of elements. The gain with Two-Element Array
two elements, as the spacing between them is The simplest and most popular collinear array is
varied, is shown by Fig. 4-9. Although the gain is one using two elements, as shown in Fig. 4-10. This
greatest when the end-to-end spacing is in the system is commonly known as “two half-waves in
region of 0.3 to 0.5 wavelength, the use of spacings phase,” and the manner in which the desired
of this order is inconvenient constructionally and current distribution is secured has been described
introduces problems in feeding the two elements. in Chapter Three. The directive pattern in a plane
As a result, collinear elements are almost always containing the wire axis is shown in Fig. 4-11.
operated with their ends quite close together - in Depending on the conductor size, height, and
wire antennas, usually with just a strain insulator similar factors, the impedance at the feed point can
between. be expected to be in the range from about 4000 to
With very small spacing between the ends of 6000 ohms, for wire antennas. If the elements are
adjacent elements the theoretical power gain of made of tubing having a low length/diameter ratio,
collinear arrays is approximately as follows: values as low as 1000 ohms are representative. The
2 collinear elements — 1.9 dB system can be fed through an open-wire tuned line
3 collinear elements - 3.2 dB with negligible loss for ordinary line lengths, or a
4 collinear elements - 4.3 dB matching section may be used if desired.

Three- and Four-Element Array

More than four elements arc rarely used.
When more than two collinear elements are
Directivity used it is necessary to connect “phasing” stubs
The directivity of a collinear array, in a plane between adjacent elements in order to bring the
containing the axis of the array, increases with its
length. Small secondary lobes appear in the pattern
when more than two elements are used, but the
amplitudes of these lobes are low enough so that
they are not important. In a plane at right angles to
the array the directive diagram is a circle, no
matter what the number of elements. Collinear
operation, therefore, affects only the directivity in
the plane containing the antenna. At right angles to
the wire the pattern is the same as that of the
half-wave elements of which it is composed. Fig. 4-10 — A two-element collinear array ("two
When a collinear array is mounted with the half waves in phase”). The transmission line shown
elements vertical the antenna radiates equally well would operate as a tuned line. A matching section
in all geographical directions. An array of such can be substituted and a nonresonant line used if
“stacked” collinear elements tends to confine the desired.
 136 Chapter 4

second and third elements makes the instantaneous

current direction correct in the third element. This
stub may be looked upon simply as the alternate
half-wave section of a long-wire antenna folded
back on itself to cancel its radiation. In Fig. 4-12A
the part to the right of the transmission line has a
total length of three half wavelengths, the center
half wave being folded back to form a quarter-wave
phase-reversing stub. No data are available on the
impedance at the feed point in this arrangement,
but various considerations indicate that it should
be over 1000 ohms.
An alternative method of feeding three
collinear elements is shown in Fig. 4-12B. In this
case power is applied at the center of the middle
element and phase-reversing stubs are used between
this element and both of the outer elements. The
impedance at the feed point in this case is
somewhat over 300 ohms and provides a close
match to 300-ohm line. The SWR will be less than
2 to 1 when 600-ohm line is used. Center feed of
Fig. 4-11 — Free-space directive diagram for a this type is somewhat preferable to the arrange¬
two-element collinear array. Field strength is ment in Fig. 4-12A because the system as a whole
shown on a relative basis. This is the horizontal is balanced. This assures more uniform power
pattern at low wave angles when the array is distribution among the elements. In A, the right¬
horizontal. hand element is likely to receive somewhat less
power than the other two because a portion of the
fed power is radiated by the middle element before
currents in all elements in phase. It will be recalled it can reach the one located at the extreme right.
from Chapter Two that in a long wire the direction A four-element array is shown in Fig. 4-12C.
of current flow reverses in each half-wave section. The system is symmeterical when fed between the
Consequently, collinear elements cannot simply be two center elements as shown. As in the three-
connected end to end; there must be some means element case, no data are available on the im¬
for making the current flow in the same direction pedance at the feed point. However, the SWR with
in all elements. In Fig. 4-12A the direction of a 600-ohm line should not be much over 2 to 1.
current flow is correct in the two left-hand Fig. 4-13 shows the directive pattern of a four-
elements because the transmission line is connected element array. The sharpness of the three-element
between them. The phasing stub between the pattern is intermediate between Figs. 4-11 and
4-13, with a small minor lobe at right angles to the
array axis.
Collinear arrays can be extended to more than
I— four elements. However, the simple two-element
I
collinear array is the type most used, for the reason
that it lends itself well to multiband operation.
More than two collinear elements are seldom used
because more gain can be obtained from other
(A) types of arrays.

Adjustment
In any of the collinear systems described the
lengths of the radiating elements in feet can be

Fig. 4-12 — Three- and four-element

collinear arrays. Alternative methods of
feeding a three-element array are shown
at A and B. These drawings also show
the current distribution on the antenna
elements and phasing stubs. A matched
transmission line can be substituted for
the tuned line by using a suitable match¬
ing section.
Multielement Directive Arrays 137

open, as described in Chapter Three. If more than

three or four elements are used it is best to add
elements two at a time (one at each end of the
array), resonating the system each time before a
new pair is added.
The Extended Double Zepp
An expedient that may be adopted to obtain
the higher gain that goes with wider spacing in a
simple system of two collinear elements is to make
the elements somewhat longer than 1/2 wave¬
length. As shown in Fig. 4-14, this increases the
spacing between the two in-phase half-wave sec¬
tions at the ends of the wires. The section in the
center carries a current of opposite phase, but if
this section is short the current will be small
because it represents only the outer ends of a
half-wave section. Because of the small current and
short length the radiation from the center is small.
The optimum length for each element is 0.64
wavelength. At greater lengths the system tends to
act as a long-wire antenna and the gain decreases.
Fig. 4-13 - Free-space directive diagram for a
four-element collinear array. Field strength is This system is known as the “extended double
shown on a relative basis. Zepp.” The gain over a half-wave dipole is approxi¬
mately 3 dB, as compared with slightly less than 2
found from the formula 468//(MHz). The lengths
of the phasing stubs can be found from the formu¬
las given in Chapter Three for the type of line used.
If the stub is open-wire line (500 to 600 ohms
impedance) it is satisfactory to use a velocity
factor of 0.975 in the formula for a quarter-wave
Une. On-the-ground adjustment is, in general, an
unnecessary refinement. If desired, however, the
following procedure may be used when the system
has more than two elements.
Disconnect all stubs and all elements except
those directly connected to the transmission line
(in the case of feed such as is shown in Fig. 4-12B
leave only the center element connected to the
line). Adjust the elements to resonance, prelerably
using the still-connected element or method de¬
scribed in Chapter Three. When the proper length
is determined, cut all other elements to the same
length. Make the phasing stubs slightly long and
use a shorting bar to adjust their length. Connect
the elements to the stubs and adjust the stubs to
resonance, as indicated by maximum current in the
shorting bars or by the positions of the standing
Fig. 4-15 — Free-space directive diagram for the
waves along the transmission line. When the whole extended double Zepp. This is also the horizontal
system is resonant the position of the first current directional pattern when the elements are horizon¬
or voltage maximum along the transmission line tal.
should be the same as when the line is shorted or
dB for two collinear dipoles. The directional
pattern in the plane containing the axis of the
antenna is shown in Fig. 4-15. As in the case of all
other collinear arrays, the free-space pattern in the
plane at right angles to the antenna elements is the
same as that of a half-wave antenna; Le., is circular.

BROADSIDE ARRAYS
WITH PARALLEL ELEMENTS

Fig. 4-14 — The extended double Zepp. This To obtain broadside directivity with parallel
system gives somewhat more gain than two half¬ elements the currents in the elements must all be in
wave collinear elements. phase. At a distant point lying on a line perpen-
 138
Chapter 4

The patterns are given in Chapter Two. When the

array elements are horizontal the vertical pattern is
the product of the broadside pattern for the
particular array used multiplied by the ground¬
reflection factors given in Chapter Two. For the
purpose of applying the ground-reflection factor
the height of the array is taken as the mean height
above ground. The horizontal directive pattern of a
horizontally polarized parallel-element broadside
array is the same as that of a simple dipole.

Two-Element Arrays
Fig. 4-16 - Power gain as a function of the spacing
between two parallel elements operated in phase The elements of a broadside array must be
(broadside). connected by transmission lines that supply power
in the proper phase to each element. Three
dicular to the axis of the array and also perpen¬ methods of interconnection for a two-element
dicular to the plane containing the elements, the array are given in Fig. 4-17. In A, the main
fields from all elements add up in phase. The transmission line is connected to the “phasing line”
situation is similar to that pictured in Fig. 4-1 in at its center. The two halves of the phasing line,
this chapter and in Fig. 2-17, Chapter Two. AB and AC are simply in parallel, as far as the main
Broadside arrays of this type theoretically can transmission line is concerned, so the currents in
have any number of elements. However, practical the phasing line flow in opposite directions, with
limitations of construction and available space respect to the junction A. This brings the currents
usually limit the number of broadside parallel in the array elements in phase. The phasing line can
elements to two, in the amateur bands below 30 be any convenient length in this case, so the line
MHz, when horizontal polarization is used. More spacing between the two elements can be any value
than four such elements seldom are used even at desired. Although no data are available on im¬
vhf. pedances, a rough estimate indicates that in most
practical cases the impedance will be well below
Power Gain 100 ohms at the point where the main transmission
line joins the phasing line, assuming a half-wave
The power gain of a parallel-element broadside phasing line having a Zo of about 600 ohms. If the
array depends on the spacing between elements as
well as on the number of elements. The way in
which the gain of a two-element array varies with
spacing is shown in Fig. 4-16. The greatest gain is
obtained when the spacing is in the vicintiy of 0.7
wavelength.
The theoretical gains of broadside arrays having
more than two elements are approximately as
follows:

No. of dB Gain dB Gain

Parallel withl/2-Wave with 3/4-Wave
Elements Spacing Spacing
3 5 7
4 6 8.5
5 7 10
6 8 11

The elements must, of course, all lie in the same

plane.

Directivity
The sharpness of the directive pattern depends
on spacing between elements and on the number of
elements. Larger element spacing will sharpen the
main lobe, for a given number of elements. The
two-element array has no minor lobes when the
spacing is 1/2 wavelength, but small minor lobes
appear at greater spacings. When three or more
elements are used the pattern always has minor
lobes.
The vertical directive pattern of such an array
when the elements are vertical is the same as that Fig. 4-1 7 — Two-element broadside arrays, showing
for a simple half-wave dipole at the same height. different methods of supplying power.
Multielement Directive Arrays 139

phasing line is not exactly 1/2 wavelength long the V if I? S if i? if if

impedance will be reactive as well as resistive.
In B, the main transmission line is connected at
the junction of the phasing line and one element.
In this case it is necessary to transpose the phasing
line somewhere along its length so that the
element currents will be in the proper phase. This
is shown by the arrows indicating relative direction
of current flow. The impedance at the feed point
will be resistive and of the order of a few thousand
ohms when the elements and phasing line all have
electrical lengths of 1/2 wavelength. With this type
of feed the spacing between the elements is Fig. 4-19 — Vertical pattern broadside to a
two-element in-phase array with horizontal ele¬
determined by the electrical length of the phasing ments. This pattern is for a mean height of 3/4
line; it must be an electrical half wave long to bring wavelength, i.e., lower element 1/2 wavelength
the element currents in proper phase. Open-wire high and upper element one wavelength high. At
lines are always used as phasing lines in this type of low wave angles, the horizontal pattern of such an
system because their electrical length is nearest to array is the same as for a half-wave dipole.
the length of an actual half wavelength in space. If
the velocity factor of the phasing line is much less resistive impedance seen by the transmission line is
than 1 the antenna elements will perforce be
6000/2 = 3000 ohms.
considerably less than a half wavelength apart, and The arrays shown in Fig. 4-17 may be installed
this will reduce the gain. either vertically or horizontally, depending on the
A third method of feeding is shown at C. This is type of polarization desired. The free-space di¬
the best of the three, insofar js symmetry is rective diagram given in Fig. 4-18 is also the
concerned. The spacing between the two elements horizontal pattern of the array at low wave angles
can be any desired value. However, when the when the elements are vertical. The vertical pattern
spacing is one-half wavelength the impedance at for a horizontally polarized two-element array for
the point where the main transmission line is a mean height of 3/4 wavelength is given in Fig.
connected is resistive and can be calculated with 4-19. The pattern for other heights may be found
the aid of Fig. 4-7. For example, Fig. 4-7 shows by multiplying the pattern of Fig. 4-18 by the
that the radiation resistance of each element is ground-reflection factor for the actual mean
approximately 60 ohms at half-wave spacing. If the height.
Zo of the phasing line is 600 ohms, the impedance
reflected at the transmission-line terminals will be
6002/60 = 6000 ohms, since with half-wave END-FIRE ARRAYS
spacing the phasing line is 1/4 wave long from the The term “end-fire” covers a number of differ¬
element to the junction. As the reflected resis¬
ent methods of operation, all having in common the
tances from both elements are in parallel, the
fact that the maximum radiation takes place along
the array axis, and that the array consists of a
number of parallel elements in one plane. End-fire
arrays can be either bidirectional or unidirectional.
In the bidirectional type commonly used by
amateurs there are only two elements, and these
are operated with currents 180 degrees out of
phase. Unidirectional end-fire driven arrays have
not had much amateur use because the element
phasing is neither 0 nor 180 degrees and tends to
be complicated from an adjustment standpoint.
(Instead, unidirectional antennas as used by ama¬
teurs are practically all based on the use of
parasitic elements as described later in this chap¬
ter.)
Two-Element Arrays
In the two-element array with equal currents
out of phase the gain varies with the spacing
between elements as shown in Fig. 4-20. The
maximum gain is in the neighborhood of 1/8-wave
spacing. Below .05-wave spacing the gain decreases
rapidly, since the system is approaching the spacing
Fig. 4-18 — Free-space directive diagram of a used for nonradiating transmission lines.
two-element broadside array for an element spac¬ The radiation resistance at the center of cither
ing of 1/2 wavelength. This drawing gives the
low-angle horizontal pattern of a vertically polar¬ element is very low at the spacings giving the
ized array. greatest gain, as shown by Fig. 4-7. The spacings
140
Chapter 4

conductor loss will mean that the theoretical gain

cannot be realized.
Three methods of feeding bidirectional end-fire
elements are shown in Fig. 4-21. In A, one section
of the phasing line is transposed to bring the ele¬
ment currents in proper phase. The method at B is
suitable for close-spaced (Le., 1/8-wave) arrays
because each half of the connecting wire is only
1/16 wave long and carries very little current.
Hence there is very little radiation from the wires
joining the ends of the elements to the trans¬
Fig. 4-20 — Gain of an end-fire array consisting of mission line even though the currents are in phase.
two elements fed 180 degrees out of phase, as a The center-fed arrangement shown at C is es¬
function of the spacing between elements. Maxi¬ pecially useful when the antenna is to be operated
mum radiation is in the plane of the elements and on two bands - for example, 14 and 28 MHz - the
at right angles to them at spacings up to 0.5 higher of which is the second harmonic of the
wavelength, but the direction changes at greater lower.
spacings. Because of the very low radiation resistance
when the spacing, S in Fig. 4-21, is 1/8 wave¬
most frequently used are 1/8 and 1/4 wavelength, length, the SWR on the transmission line is very
at which the resistances are about 8 and 32 ohms, high. No figures are available for the end-fed cases,
respectively. When the spacing is 1/8 wavelength it but it can be estimated to be 20 to 1 or higher.
is advisable to use good-sized conductors - prefer¬ With center feed using a 600-ohm line the SWR is
ably tubing - for the elements because with the over 30 to 1 ; if the transmission line has any
radiation resistance so low the heat losses in the considerable distance to run, it is advisable to
conductor can represent an appreciable portion of match it to the antenna by using a matching
the power supplied to the antenna. Excessive section of the type described in Chapter Three.
Such a matching section should be of open-wire

Fig. 4-22 — Free-space directive diagram of a

two-element end-fire array with 180-degree phas¬
ing, in the plane containing the two parallel
elements.

construction in view of the high SWR. The line

itself, of course, can be any type capable of
carrying the transmitter power. If the transmission
line does not have to run more than a wavelength
or two it may be of open-wire construction and
operated as a tuned line.
Fig. 4-21 - Parallel-element end-fire array with With 1/4-wave spacing the increased radiation
various methods of feed, showing current distribu¬ resistance will lower the SWR considerably. With
tion. Matching sections for making the transmis¬ center feed it will be about 10 to 1 (600 ohm line),
sion line nonresonant, as described in Chapter and should not exceed that figure with end feed.
Three, may be used in the first three cases. The In a close-spaced array fed through a tuned
distance S may be selected from Fig. 4-20. transmission line the element lengths are not
 141
Multielement Directive Arrays

critical; the only point to watch is to preserve the

symmetry of the system as a whole. When a
matching section is used, however, it is necessary
to adjust the system accurately to the particular
frequency to be used most. The low radiation
resistance makes the antenna a sharply tuned
affair, and so relatively small departures from the
design frequency will throw off the impedance
Another way of overcoming the high SWR on
the transmission line, and at the same time
reducing the resistance loss in the antenna ele¬
ments, is to use a folded-dipole (see Chapter Two)
arrangement as indicated at D in Fig. 4-21. In this
way the impedance at the element terminals is
stepped up, and then there is a further impedance
step-up in the section Q which is a quarter-wave
g-type matching transformer. A number of com¬
binations is listed below.
S, No. of Z o of 1/4- Zo of main
wave- conductors wave match- transmis-
Fig. 4-23 — Free-space directive diagram of a
length in dipole ing section sion line
two-element end-fire array with 180-degree phas¬
1/8 1 75 300 ing, in the plane at right angles to the plane
1/8 2 75 75 containing the elements.
1/8 3 300 600
1/8 4 300 300 one is an electrical quarter wavelength longer than
1/4 1 75 75 the other. The length L can be any convenient
1
14 2 300 300 value. Open quarter-wave matching sections are
1/4 3 600 600 shown, but half-wave shorted sections could be
used instead. The two transmission lines are con¬
In each case the SWR on the transmission line will nected in parallel at the transmitter coupling
be well below 2 to 1. Twin-Lead will be satisfac¬ circuit.
tory for the 75 and 300-ohm line. The velocity When the currents in the elements are neither in
factor of the line must be taken into account in phase nor 180 degrees out of phase the radiation
determining the physical length of the 1/4-wave resistances of the elements are not equal. This
matching transformer. In all the arrangements complicates the problem of feeding equal currents
listed above except those using plain dipoles as to the elements. If the currents are not equal one
elements the frequency characteristic of the or more minor lobes will appear in the pattern and
antenna will be broadened somewhat by the decrease the front-to-back ratio. The adjustment
folded-dipole construction. process is likely to be tedious and requires field¬
The free-space directive pattern in the plane strength measurements in order to get the best
containing the array is given in Fig. 4-22 and the performance.
corresponding pattern in the plane at right angles More than two elements can be used in uni¬
to the array plane is given in Fig. 4-23. Fig. 4-22 is directional end-fire array. The requirement for
also the horizontal directive pattern of the array at unidirectivity is that there must be a progressive
low wave angles when the elements are horizontal, phase shift in the element currents equal to the
while Fig. 4-23 is the horizontal pattern at low spacing, in electrical degrees, between the ele¬
wave angles when the elements are vertical. The ments, and the amplitudes of the currents in the
vertical pattern of a horizontally polarized array is various elements also must be properly related.
shown in Fig. 4-24.
Unidirectional End-Fire Arrays
Two parallel elements spaced 1/4 wavelength
apart and fed equal currents 90 degrees out of
phase will have a directional pattern, in the plane
at right angles to the plane of the array, as
represented in Fig. 4-25. The maximum radiation is
in the direction from the element in which the
current leads to the element in which the current
lags. In the opposite direction the fields from the
two elements cancel.
One way in which the 90-degree phase differ¬ Fig. 4-24 — Vertical pattern of a horizontally
ence can be obtained is shown in Fig. 4-26. Each polarized two-element end-fire array. Solid curve,
element must be matched to its transmission line, height 1/2 wavelength; broken curve, height 1
the two lines being of the same type except that wavelength.
 142 Chapter 4

only a few of the simpler types that are in common

use. The drawings that follow all show the ele¬
ments arranged for horizontal polarization, which
is customary on the frequencies below 30 MHz
where these arrays find their greatest application.
For vertical polarization the arrays should be
rotated 90 degrees so that the elements are vertical
— that is, “stood on end.”
Other methods of interconnecting elements
than those shown in the drawings may be used.
However, the methods shown are recommended
over others for two reasons: The antenna system is
symmetrical with respect to the feed point, thus
making the current distribution among elements as
uniform as possible; the lengths of phasing lines
(and antenna elements as well) are not critical so
long as the lengths of lines radiating from a
junction are all the same. With other feed methods
this may not be true, and it becomes necessary to
Fig. 4-25 — Representative pattern for a two- use the methods described in Chapter Three to
element end-fire array with 90-degree phasing, in ensure that elements and phasing lines are exactly
the plane perpendicular to the plane containing the resonant at the design frequency, if maximum
elements.

This requires “binomial” current distribution —

i.e., the ratios of the currents in the elements must
be proportional to the coefficients of the binomial
series. In the case of three elements, this requires
that the current in the center element be twice that
in the two outside elements, for 90-degree
(quarter-wave) spacing and element current
phasing. This antenna has an overall length of 1/2
wavelength.

COMBINATION DRIVEN ARRAYS TRANS LINS

Broadside, end-fire and collinear elements can Fig. 4-27 — A four-element array combining
readily be combined to increase gain and di¬ collinear broadside elements and parallel end-fire
rectivity, and this is in fact usually done when elements.
more than two elements are used in an array. Com¬
binations of this type give more gain, in a given
amount of space, than plain arrays of the types just performance is to be secured from the antenna.
described. This adjustment process can be rather difficult as
The combinations that can be worked out are well as tedious. If the feed arrangements shown in
almost endless, but in this section we shall describe the drawings are followed the lengths of wire
elements in feet can be found from 468// (MHz),
the element spacings from 984// (MHz) multiplied
by the fraction of wavelength desired, and the
phasing lines can simply be cut to fit, keeping all
lines the same length.

Gain of Combination Arrays

The accurate calculation of the power gain of a
multielement array requires a knowledge of the
mutai impedances between all elements. For ap¬
proximate purposes it is sufficient to assume that
each set (collinear, broadside, end-fire) will have
the gains as given earlier, and then simply add up
the gains for the combination. This neglects the
effects of cross-coupling between sets of elements.
However, the array configurations are such that the
mutual impedances from cross-coupling should be
relatively small, particularly when the spacings are
Fig. 4-26 — Unidirectional two-element end-fire 1/4 wavelength or more, so that the estimated
array and method of obtaining 90-degree phasing.
gain should be reasonably close to the actual gain.
Multielement Directive Arrays 143

Fig. 4-30 — Four-element broadside array ("lazy-

H") using collinear and parallel elements.

section at the point resulting in the lowest line

SWR. This point can be determined by trial.
With 1/4-wave spactng the SWR on a 600-ohm
Fig. 4-28 — Free-space directive diagram of the line is estimated to be in the vicinity of 3 or 4 to 1.
antenna shown in Fig. 4-27. in the plane of the This type of antenna can be operated on two
antenna elements. The pattern in the plane perpen¬ bands having a frequency ratio of 2 to 1, if a
dicular to the element plane is the same as Fig. resonant feed line is used. For example, if designed
4-23. for 28 MHz with 1/4-wave spacing between ele¬
ments it can be operated on 14 MHz as a simple
Four-Element End-Fire and Collinear Array end-fire array (Fig. 4-21Ç) having 1/8-wave spac¬
ing.
The array shown in Fig. 4-27 combines collin¬
ear in-phase elements with parallel out-of-phase Four-Element Broadside Array
elements to give both broadside and end-fire
directivity. It is popularly known as a “two-section The four-element array shown in Fig. 4-30 is
W8JK” or “two-section flat-top beam.” The ap¬ commonly known as the “lazy-H.” It consistsofa
proximate gain calculated as described above is 6.2 set of two collinear elements and a set of two
dB with 1/8-wave spacing and 5.7 dB with 1/4- parallel elements, all operated in phase to give
wave spacing. Directive patterns are given in Figs. broadside directivity. The gain and directivity will
depend on the spacing, as in the case of a simple
4-28 and 4-29.
The impedance between elements at the point parallel-element broadside array. The spacing may
where the phasing line is connected is of the order be chosen between the limits shown on the
of several thousand ohms. The SWR with an drawing, but spacings below 3/8 wavelength are
unmatched Une consequently is quite high, and this not worthwhrle because the gain is small. Estima¬
system should be constructed with open-wire line ted gains are as follows
(500 or 600 ohms) if the line is to be resonant. To
use a matched line a closed stub 3/16 wavelength 3/8-wave spacing - 4.4 dB
long can be connected at the transmission-line 1/2-wave spacing - 5.9 dB
junction shown in Fig. 4-27, and the transmission 5/8-wave spacing - 6.7 dB
line itself can then be tapped on this matching 3/4-wave spacing - 6.6 dB

Half-wave spacing is generally used. Directive pat¬

terns for this spacing are given in Figs. 4-31 and
4-32.
With half-wave spacing between parallel ele¬
ments the impedance at the junction of the phasing
line and transmission line is resistive and is in the
vicinity of 100 ohms. With larger or smaller spacing
the impedance at this junction will be reactive as
well as resistive. Matching stubs are recommended
in cases where a nonresonant line is to be used.
They may be calculated and adjusted as described
in Chapter Three.
The system shown in Fig. 4-30 may be used on
Fig. 4-29 — Vertical pattern of the four-element two bands having a 2-to-l frequency relationship.
antenna of Fig. 4-27 when mounted horizontally. It should be designed for the higher of the two
Solid curve, height 1/2 wavelength; broken curve,
height 1 wavelength. Fig. 4-28 gives the horizontal frequencies, using 3/4-wave spacing between paral¬
pattern. lel elements. It will then operate on the lower
144 Chapter 4

Fig. 4-31 - Free-space directive diagrams of the

four-element antenna shown in Fig. 4-30. The solid
curve is the horizontal directive pattern at low
wave angles when the antenna is mounted with the
elements horizontal. The broken curve is the
free-space vertical pattern of a horizontally po¬
larized array, broadside to the array. Actual pat¬
tern in the presence of ground may be found by
multiplying this pattern by the ground-reflection
factors given in Chapter Two.

C shows the effect of transposing the phasing

line. This transposition reverses the direction of
current flow in the lower pair of elements, as
compared with A, and thus changes the array from
a combination collinear and end-fire arrangement
frequency as a simple broadside array with 3/8- into a collinear-broadside array.
wave spacing. The drawing at D shows what happens when
An alternative method of feeding is shown in the transmission line is connected at the center of a
the small diagram in Fig. 4-30. In this case the section of phasing line. Viewed from the main
elements and the phasing line must be adjusted transmission line the two parts of the phasing line
exactly to an electical half wavelengh. The imped¬ are simply in parallel, so the half wavelength is
ance at the feed point will be resistive and of the measured from the antenna element along the
order of 2000 ohms. upper section of phasing line and thence along the
transmission line. The distance from the lower
Checking Phasing
elements is measured in the same way. Obviously
In the antenna diagrams earlier in this chapter the two sections of phasing line should be the same
the relative direction of current flow in the various length. If they are not, the current distribution
antenna elements and connecting lines was shown becomes quite complicated; the element currents
by arrows. In laying out any antenna system it is are neither in phase nor 180 degrees out of phase,
necessary to know that the phasing lines are and the elements at opposite ends of the lines do
properly connected; otherwise the antenna may not receive the same power. To change the element
have entirely different characteristics than antici¬ current phasing at D into the phasing at A, simply
pated. The phasing may be checked either on the transpose the wires in one section of the phasing
basis of current direction or polarity of voltages. line; this reverses the direction of current flow in
There are two rules to remember: the antenna elements connected to that section of
1) In every half-wave section of wire, starting phasing line.
from an open end, the current directions reverse.
In terms of voltage, the polarity reverses at each
half-wave point, starting from an open end.
2) Currents in transmission lines always must
flow in opposite directions in adjacent wires. In
terms of voltage, polarities always must be oppo¬
site.
Examples of the use of current direction and
voltage polarity are given at A and B, respectively,
in Fig. 4-33. The half-wave points in the system are
marked by the small circles. When current in one
section flows toward a circle, the current in the
next section must also flow toward it, and vice
versa. In the four-element antenna shown at A, the
current in the upper right-hand element cannot Fig. 4-32 — Vertical pattern of the four-element
flow toward the transmission line, because then the broadside antenna of Fig. 4-30, when mounted
current in the right-hand section of the phasing line with the elements horizontal and the lower set 1/2
would have to flow upward and thus would be wavelength above ground. "Stacked” arrays of this
flowing in the same direction as the current in the type give best results when the lowest elements are
at least 1/2 wavelength high. The gain is reduced
left-hand wire. The phasing line would simply act and the wave angle raised if the lowest elements are
like two wires in parallel in such a case. close to ground.
Multielement Directive Arrays 145

Fig. 4-33 — Methods of checking

the phase of currents in elements
and phasing lines.

PARASITIC ARRAYS
Multielement arrays containing parasitic ele¬ quently called a “Yagi” or “Yagi-Uda” antenna,
ments are called “parasitic” arrays even though at ¿■ter the inventors.
least one and sometimes more than one of the As explained earlier in this chapter in the
elements is driven. A parasitic element obtains its section on mutual impedance, the amplitude and
power through electromagnetic coupling with a phase of the current induced in an antenna element
driven element, as contrasted with receiving it by depend on its tuning and the spacing between it
direct connection to the power source. A parasitic and the driven element to which it is coupled. The
array with linear (dipole-type) elements is fre- fact that the relative phases of the currents in
driven and parasitic elements can be adjusted is
very advantageous. For example, the spacing and
tuning can be adjusted to approximate the condi¬
tions that exist when two driven elements 1/4
wavelength apart are operated with a phase difter-
ence of 90 degrees (which gives a unidirectional
pattern as shown in Fig. 4-25). However, complete
cancellation of radiation in the rear direction is not
possible when a parasitic element is used. This is
because it is usually not possible to make ampli¬
tude and phase both reach desired values simulta¬
neously. Nevertheless, a properly designed parasitic
array can be adjusted to have a large front-to-back
ratio.
The substantially unidirectional characteristic
and relatively simple electrical configuration of an
array using parasitic elements make it especially
useful for antenna systems that are to be rotated to
aim the beam in any desired direction.

Reflectors and Directors

Although there are special cases where a parasit¬
ic array will have a bidirectional (but usually not
symmetrical) pattern, in most applications the
pattern tends to be unidirectional. A parasitic
Fig. 4-34 — Antenna systems using a single element is called a director when it makes the
parasitic element. In A the parasitic element acts as radiation maximum along the perpendicular line
a director, in B as a reflector. The arrows show the from the driven to the parasitic element, as shown
direction in which maximum radiation takes place. at A in Fig. 4-34. When the maximum radiation is
 146
Chapter 4

in the opposite direction - that is, from the In only two cases are the gains shown in Fig.
parasitic element through the driven element as at 4-35 secured when the parasitic element is self-
B - the parasitic element is called a reflector. resonant. These occur at 0.1- and 0.25-wavelength
Whether the parasitic element operates as a spacing, with the parasitic element acting as direc¬
director or reflector is determined by the relative tor and reflector, respectively. For reflector opera¬
phases of the currents in the driven and parasitic tion, it is necessary to tune the parasitic element to
elements. At the element spacings commonly used a lower frequency than resonance to secure maxi¬
(1/4 wavelength or less) the current in the parasitic mum gain at all spacings less than 0.25 wavelength,
element will be in the right phase to make the while at greater spacings the reverse is true. The
element act as a reflector when its tuning is closer the spacing the greater the detuning re¬
adjusted to the low-frequency side of resonance quired. On the other hand, the director must be
(inductive reactance). The parasitic element will detuned toward a higher frequency (that is, its
act as a director when its tuning is adjusted to the length must be made less than the self-resonant
high-frequency side of resonance (capacitive reac¬ length) at spacings greater than 0.1 wavelength in
tance). The proper tuning is ordinarily accom¬ order to secure maximum gain. The amount of
plished by adjusting the lengths of the parasitic detuning necessary becomes greater as the spacing
elements, but the elements can be “loaded” at the is increased. At less than 0.1-wavelength spacing
center with lumped inductance or capacitance to the director must be tuned to a lower frequency
achieve the same purpose. If the parasitic element than resonance to secure the maximum gains
is self-resonant the element spacing determines indicated by the curve. (Generally these require¬
whether it will act as a reflector or director. ments for maximum gain are not followed in
practice for close-spaced directors or wide-spaced
THE TWO-ELEMENT BEAM reflectors; instead, forward gain is sacrificed for a
The maximum gain theoretically obtainable higher front-to-back ratio or greater bandwidth, as
with a single parasitic element, as a function of the discussed in a later section.)
spacing, is shown in Fig. 4-35 (from analysis by G.
H. Brown). The two curves show the greatest gain Input Impedance
to be expected when the element is tuned for The radiation resistance at the center of the
optimum performance either as a director or driven element varies as shown in Fig. 4-36 for the
reflector. This shift from director to reflector, with spacings and tuning conditions that give the gains
the corresponding shift in direction as shown in indicated by the curves of Fig. 4-35. These values,
Fig. 4-34, is accomplished simply by tuning the especially in the vicinity of 0.1-wavelength spacing,
parasitic element - usually, in practice, by chang¬ are quite low. The curves coincide at 0.1 wave¬
ing its length. length, both showing a value of 14 ohms.
With the parasitic element tuned to act as a The low radiation resistance at the spacings
director, maximum gain is secured when the giving highest gain tends to reduce the radiation
spacing is approximately 0.1 wavelength. When the efficiency. This is because, with a fixed loss
parasitic element is tuned to work as a reflector, resistance, more of the power supplied to the
the spacing that gives maximum gain is about 0.15 antenna is lost in heat and less is radiated, as the
wavelength, with a fairly broad peak. The director radiation resistance approaches the loss resistance
will give slightly more gain than the reflector, but in magnitude.
the difference is less than 1/2 dB. The loss resistance can be decreased by using
low-resistance conductors for the antenna ele¬
ments. This means, principally large-diameter con¬
ductors, usually tubing of aluminum, copper, or
copper-plated steel. Such conductors have
mechanical advantages as well, in that it is relative¬
ly easy to provide adjustable sliding sections for
changing length, while the fact that they can be
largely self-supporting makes them well adapted
for rotatable antenna construction. With half-inch
or larger tubing the loss resistance in any two-
element antenna should be small.
With low radiation resistance the standing
waves of both current and voltage on the antenna
reach considerably higher maximum values than is
the case with a simple dipole. For this reason losses
in insulators at the ends of the elements become
more serious. The use of tubing rather than wire
Fig. 4-35 — The maximum possible gain obtainable helps reduce the end voltage, and furthermore the
with a parasitic element over a half-wave antenna tubing does not require support at the ends, thus
alone, assuming that the parasitic element tuning is eliminating the insulators.
adjusted for greatest gain at each spacing. These The mutual impedance between two parallel
curves assume no ohmic losses in the elements. In antenna elements contains reactance as well as
practical antennas the gain is less, particularly at resistance, so that the presence of a director or
close spacings
reflector near the driven element affects not only
Multielement Directive Arrays 147

about 0.14-wavelength spacing. That is, at closer

spacings the parasitic element is principally a
director, while at greater spacings it is chiefly a
reflector. At 0.14 wavelength the radiation is the
same in both directions; in other words, the
antenna is bidirectional with a theoretical gain of
about 4 dB.
The front-to-back ratios that can be secured
with the parasitic element self-resonant are not
very great except in the case of extremely close
spacings. Spacings of the order of .05 wavelength
are not very practicable with outdoor construction
since it is difficult to make the elements sufficient¬
ly stable mechanically. Ordinary practice is to use
spacings of at least 0.1 wavelength and detune the
parasitic element for greatest attenuation in the
backward direction.
The radiation resistance increases rapidly for
spacings greater than 0.15 wavelength, while the
gain, with the parasitic element acting as a reflec¬
tor. decreases quite slowly. If front-to-back ratio is
not an important consideration, a spacing as great
as 0.25 wavelength can be used without much
Fig. 4-36 — Radiation resistance at center of driven
reduction in gain. At this spacing the radiation
element as a function of element spacing, when the resistance approaches that of a half-wave antenna
parasitic element is adjusted for the gains given in alone. Spacings of this order are particularly suited
Fig. 4-35. to antennas using wire elements, such as multi¬
element arrays consisting of combinations of col¬
linear and broadside elements.
the radiation resistance of the driven element but
also introduces a reactive component (assuming Front-to-Back Ratio
that the driven element length is such as to be The tuning conditions that give maximum gain
resonant if the parasitic element were not there). forward do not give maximum signal reduction, or
In other words, the parasitic element detunes the attenuation, to the rear. It is necessary to sacrifice
driven element. The degree of detuning depends on some gain to get the highest front-to-back ratio.
the spacing and tuning of the parasitic element, The reduction in backward response is brought
and also on the length/diameter ratios of the about by adjustment of the tuning or length of the
elements. parasitic element. With a reflector, the length must
With the parasitic element tuned for maximum be made slightly greater than that which gives
gain, the effect of the coupled reactance is to make maximum gain, at spacings up to 0.25 wavelength.
the driven element “look” more inductive with the The director must be shortened somewhat to
parasitic element tuned as a reflector than it does achieve the same end, with spacings of 0.1 wave-
when the parasitic element is tuned as a director.
That is, the driven element should be slightly
longer, if the parasitic element is a director, than
when the parasitic element is a reflector. These
remarks apply to spacings between about 0.1 and
0.25 wavelength, but are not necessarily true for
other spacings.

Self-Resonant Parasitic Elements

The special case of the self-resonant parasitic
element is of interest, since it gives a good idea of
the performance as a whole of two-element
systems, even though the results can be modified
by detuning the parasitic element. Fig. 4-37 shows
gain and radiation resistance as a function of the
element spacing for this case. Relative field
strength in the direction A of the small drawing is
indicated by Curve A; similarly for Curve B. The
front-to-back ratio at any spacing is the difference
between the values given by curves A and B at that
spacing. Whether the parasitic element is function¬ Fig. 4-37 — Theoretical gain of a two-element
ing principally as a director or reflector is deter¬ parasitic array over a half-wave dipole as a function
mined by whether Curve A or Curve B is on top; it of element spacing when the parasitic element is
can be seen that the principal function shifts at self-resonant.
 148 Chapter 4

length is approximately adjusted for optimum gain

as a director, for self-resonance, for optimum gain
as a reflector, and for optimum front-to-back ratio
as a reflector. Over this range of adjustment the
width of the main beam does not change signifi¬
cantly. These patterns are based on experimental
measurements by J. L. Gillson, W3GAU.

Bandwidth
The bandwidth of the antenna can be specified
in various ways, such as the width of the band over
which the gain is higher than some stated figure,
the band over which at least a given front-to-back
ratio is obtained, or the band over which the
standing-wave ratio on the transmission line can be
maintained below a chosen value. The latter is
probably the most useful, since the SWR not only
determines the percentage power loss in the trans¬
mission line but also affects the coupling between
the transmitter and the line.
The bandwidth from this latter standpoint
Fig. 4-38 — Experimentally determined horizon¬ depends on the Q of the antenna (see Chapter
tal directive patterns of horizontally polarized Three). The Q of close-spaced parasitic arrays is
two-element parasitic arrays at a height of 1-1/4 quite high, with the result that the frequency range
wavelengths. These patterns are for a wave angle of over which the SWR will stay below a specified
12 degrees. The curves represent the following maximum value is relatively narrow. The data in
conditions, approximately:
A — Parasitic element tuned for maximum gain Table 4-1, for a driven element and close-spaced
as a director. director, are from experimental measurements
B — Parasitic element self-resonant. made by J. P. Shanklin. The antenna with .075-
C — Parasitic element tuned for maximum gain wavelength spacing will, through a suitable match¬
as a reflector ing device, operate with an SWR of less than 3 to 1
D — Parasitic element tuned for maximum over a band having a width equal to about 3 per
front-to-back ratio as areflector. cent of the center frequency (this corresponds to
The spacing between elements is 0.1 wavelength. the width of the 14-MHz band for a 14-MHz
The patterns should not be compared for gain, antenna, for example) and maintain a front-to-back
since they are plotted on a relative basis to an
arbitrarily chosen maximum of 1.0. ratio of approximately 10 dB or better over this
band. At greater element spacings than those

length and more. The tuning condition, or element

length, which gives maximum attenuation to the
rear is considerably more critical than that for
maximum gain, so that a good front-to-back ratio
can be secured without sacrificing more than a
small part of the gain.
For the sake of good reception, general practice
is to adjust for maximum front-to-back ratio rather
than for maximum gain. Larger front-to-back ratios
can be secured with the parasitic element operated
as a director rather than as a reflector. With the
optimum director spacing of 0.1 wavelength, the
front-to-back ratio with the director tuning adjust¬
ed for maximum gain is only 5.5 dB (the back
radiation is equal to that from the antenna alone).
By proper director tuning, however, the ratio can
be increased to 17 dB; the gain in the desired
direction is in this case 4.5 dB, or 1 dB less than
the maximum obtainable.

Directional Patterns
The directional patterns obtained with two-
element arrays will vary considerably with the
tuning and spacing of the parasitic element. Typi¬ Fig. 4-39 — Vertical patterns of a horizontally
polarized two-element array under the conditions
cal patterns are shown in Figs. 4-38 and 4-39, for given in Fig. 4-38. These patterns are in the vertical
four cases where the parasitic element tuning or plane at right angles to the antenna elements.
Multielement Directive Arrays 149

indicated a maximum
gain of slightly more
TABLE 4-I than 7 dB (Uda and
Mushiake). A number
Feed Impedance and Front-to-Back Ratio of a Fed Dipole of experimental investi¬
with One Director gations has shown that
Fed Input Front-to-Back the optimum spacing
Element Dipole Director Resistance Ratio between the driven ele¬
Spacing Length Length at Band Center Q at Band Center ment and reflector is in
13.2 ohms 53.2 20.0 ( 26 dB) the region of 0.15 to
0.050 X. 0.509 X 0.484 X
0.075 0.504 0.476 24.4 29.4 8.3 (18 dB) 0.25 wavelength, with
0.100 0.504 0.469 28.1 20.0 4.3 (12.7 dB) 0.2 wavelength repre¬
senting probably the
best overall choice.
shown in the table the Q is smaller and the With 0.2-wavelength reflector spacing, Fig. 4-11
bandwidth consequently greater, but the front-to- shows the variation in gain with director length,
back ratio is smaller. This is to be expected from with the director also spaced 0.2 wavelength from
the trend shown by the curves of Fig. 4-37. The the driven element, and Fig. 4-42 shows the gain
gain is practically constant at about 5 dB for all variation with director spacing. (These curves are
spacings shown in Table 4-1. from work by Carl Greenblum.) It is obvious that
The same series of experimental measurements the director spacing is not especially critical, and
showed that with the parasitic element tuned as a that the overall length of the array (bloom length
reflector for maximum front-to-back ratio the in the case of a rotatable antenna) can be anywhere
optimum spacing was 0.2 wavelength. I he maxi¬ between 0.35 and 0.45 wavelength with no appre¬
mum front-to-back ratio was determined to be 16 ciable difference in gain.
dB. In both the director and reflector cases the Wide spacing of both elements is desirable not
front-to-back ratio decreased rather rapidly as the only because it results in high gain but also because
operating frequency was moved away from the adjustment of tuning or element length is less
frequency for which the system was tuned. With critical and the input resistance of the driven
the reflector at 0.2-wavelength spacing and tuned element is higher than with close spacing. The
for maximum front-to-back ratio the input resis¬ latter feature improves the efficiency of the anten¬
tance was found to be 72 ohms and the Q of the na and makes a greater bandwidth possible. How¬
antenna was 4.7. ever, a total antenna length, director to reflector,
The antenna elements used in these measure¬ of more than 0.3 wavelength at frequencies of the
ments had a length/diameter ratio of 330. A order of 14 MHz introduces considerable difficulty
smaller length/diameter ratio will decrease the rate from a constructional standpoint, so lengths of
of reactance change with length and hence decrease 0.25 to 0.3 wavelength are frequently used for this
the Q, while a larger ratio will increase the Q. The band, even though they are less than optimum.
use of fairly thick elements is desirable when In general, the gain of the antenna drops off
less rapidly when the reflector length is increased
maximum bandwidth is sought.
beyond the optimum value than it does for a
corresponding decrease below the optimum value.
THE THREE-ELEMENT BEAM The opposite is true of a director, as shown by Fig.
It is readily possible to use more than one 4-41. It is therefore advisable to en, if necessary,
parasitic element in conjunction with a single on the long side for a reflector and on the short
driven element. With two parasitic elements the side for a director. This also tends to make the
optimum gain and directivity result when one is antenna performance less dependent on the exact
used as a reflector and the second as a director. frequency at which it is operated, because an
Such an antenna is shown in Fig. 4-40.
As the number of parasitic elements is in¬
creased, the problem of determining the optimum
element spacings and lengths to meet given specifi¬
cations - i.e„ maximum gain, maximum front-to-
back ratio, maximum bandwidth, and so on -
becomes extremely tedious because of the large
number of variables. In general, it can be said that
when one of these quantities - gain, front-to-back
ratio, or bandwidth - is maximized the other two
cannot be. Also, if it is desired to design the
antenna to have a specific input impedance for
matching a transmission line, the other three
cannot be maximized.

Power Gain Fig. 4-40 - Antenna system using a driven element

A theoretical investigation of the 3-element and two parasitic elements, one as a reflector and
case (director, driven element and reflector) has one as a director.
 150
Chapter 4

resistance tends to be lower the closer the spacing

between the parasitic and driven elements. Values
of the order of 10 ohms are typical with a
3-element beam having 0.1-wavelength director
spacing, when the director length is adjusted for
maximum gain. This can be raised considerably —
to 50 ohms or more — by sufficient change in
director length at a sacrifice of gain. The minimum
value of resistance increases with increased director
spacing, and is of the order of 30 ohms at a spacing
of 0.25 wavelength.
As in the case of the two-element beam, tuning
and spacing of the parasitic elements affect the
reactance of the driven element; that is, a change in
the spacing or length of the parasitic elements will
tend to change the resonant frequency of the
driven element. It is generally found, however, that
the resonant length of the driven element with the
Fig. 4-41 — Gain of a 3-element Yagi over a dipole parasitic elements properly tuned does not differ
as a function of the director length for 0.2-wave- greatly from its resonant length with the parasitic
length spacing between driven element and director elements removed. This is because the two parasitic
and the same spacing between driven element and elements reflect opposite kinds of reactance into
reflector. These curves show how the element the driven element and hence tend to cancel each
thickness affects the optimum length; p is the other’s effects in this respect.
element radius expressed as a fraction of the Fig. 4-43 shows the results of experimental
wavelength. Ä/500 corresponds to an element
radius of approximately 1/2 inch at 50 MHz. measurements made by J. P. Shanklin of the input
Where the fractional-wavelength radius is smaller, resistance of 3-element arrays having an overall
as on the lower frequencies, the optimum director length (director to reflector) of 0.3 wavelength.
length will be somewhat greater. The curves give resistance contours as a function
of the spacing between the driven element and the
reflector and the length of the director. So long as
increase above the design frequency has the same
the reflector length was in the optimum region for
effect as increasing the length of both parasitic
good front-to-back ratio, as described in the next
elements, while a decrease in frequency has the
section, small changes in reflector length were
same effect as shortening both elements. By
found to have only a comparatively small effect on
making the director slightly short and the reflector
slightly long, there will be a greater spread between the input resistance. In using Fig. 4-43, it is to be
understood that the spacing between the director
the upper and lower frequencies at which the gain
starts to show a rapid decrease. and driven element is equal to the difference
between 0.3 wavelength and the selected driven-
element-to-reflector spacing, since the length of the
Input Impedance array was held constant.
The radiation resistance as measured at the The elements used in obtaining the data in Fig.
center of the driven element of a 3-element array 4-43 had a length/diameter ratio of 330.
can vary over a fairly wide range since it is a
function of the spacing and tuning of the parasitic Front-to-Back Ratio
elements. There are, however, certain fairly The element lengths and spacings are more
well-defined trends. (1) The resistance tends to critical when a high front-to-back ratio is the
reach a minimum at the parasitic-element tuning objective than when the antenna is designed for
condition that gives maximum gain, becoming maximum gain. Some gain must be sacrificed for
larger as the element is detuned in either direction the sake of a good front-to-back ratio, just as in the
- that is, made longer or shorter. (2) The case of the two-element array. In general, a high
front-to-back ratio requires fairly
close spacing between the director
and driven element, but consider¬
«a ably larger spacings are optimum
for the reflector.
* The front-to-back ratio will
* change more rapidly than the gain
S s
when the operating frequency
differs from that for which the
L15 0.20 0.25 0.30 antenna was adjusted.
SPACING FROM DIRECTOR TO FED ELEMENT IN A The front-to-back ratio tends
to decrease with increased spacing
Fig. 4-42 — Gain of 3-element Yagi versus director between the elements. However,
spacing, the reflector spacing being fixed at 0.2 with a director spacing of about
wavelength.
0.2 wavelength, it is possible to
Multielement Directive Arrays 151

In these measurements it was found that the

length of the reflector for optimum front-to-back
ratio did not vary over much of a range. In the
“low-g” region of Fig. 4-44 it was 0.51 wave¬
length, increasing to 0.525 wavelength in the
“high-g” region. The proper driven-element length
was found to be 0.49 wavelength (at the center of
the band) for all conditions.
Similar tests were made on antennas having
overall lengths of 0.2 and 0.4 wavelength. The
conclusion was (1) that the smaller length would
give high front-to-back ratios but with high Q
values and consequently small bandwidth; (2) with
0.4-wavelength overall the Q values were low
FED-TO-REFLECTOR SPACING, WAVELENGTHS enough for good bandwidth but the front-to-back
ratio was smaller.
Fig. 4-43 — Resonant resistance of fed dipole in a The Q values given by the chart can be used as
3-element parasitic antenna, overall length 0.3 described in Chapter Three to find the bandwidth
wavelength. over which the SWR will not exceed a specified
value.
secure a very good front-to-s/de ratio, which may
be a useful feature in some locations. As in the case
of front-to-back ratio, the reflector spacing has a
considerably lesser effect.
Bandwidth
The bandwidth with respect to input imped¬
ance, as evidenced by the change in standing-wave
ratio over a band of frequencies, will in general be
smaller the smaller the input resistance. This in
turn becomes smaller when the spacing between
elements is decreased. Hence close spacings are
usually associated with small bandwidths, especial¬
ly when the element lengths are adjusted for
maximum gain.
Fig. 4-44, also from data obtained in the
experimental measurements of J. P. Shanklin,
shows how the Q of a 3-element antenna having a
total length of 0.3 wavelength varies as a function
of spacing and director length. The data are for an
element length/diameter ratio of 330, but should
hold sufficiently well for ratios between 200 and
400. From the standpoint of impedance band¬ Fig. 4-45 — Measured radiation patterns of a
width, the upper right-hand region of the chart is horizontally polarized 3-element array having the
director 0.1 wavelength and the reflector 0.15
best since this region is associated with low values wavelength from the driven element. Element
of Q. tuning was adjusted for maximum gain. These
"horizontal" patterns are for the wave angles at
which the lowest lobe (Fig. 4-46) has its maximum.
The wave angles are 28 degrees at a height of 1/2
wavelength and 12 degrees at a height of 1
wavelength.

The low values of radiation resistance are

accompanied by a high degree of selectivity in the
antenna; that is, its impedance is constant over
only a small frequency range. These changes in
impedance make it troublesome to couple power
from the transmitter to the line. Such difficulties
can be reduced by using wider spacing - in
particular, using spacings of the order of 0.2
wavelength or more.
FED DIPOLE TO REFLECTOR SPACING, WAVELENGTHS
Directive Patterns
Fig. 4-44 — Q of input impedance of fed dipole in
a 3-element parasitic antenna, overall length 0.3 The directive patterns of Figs. 4-45 to 4-48
wavelength. inclusive are, like those of Figs. 4-38 and 4-39
152 Chapter 4

Fig. 4-46 — Vertical patterns of the antenna of Fig. Fig. 4-48 — Vertical patterns of the antenna of Fig.
4-45 in the vertical plane at right angles to the 4-47 in the vertical plane at right angles to the
direction of the antenna elements. direction of the antenna elements.

based on experimental measurements made by mination at 50 MHz, uses the following spacings:
W3GAU at vhf. They show that the beam is Driven element to reflector - 0.2 wavelength
somewhat sharper, as is to be expected, when the Driven element to first director — 0.2 wave¬
parasitic-element tuning is adjusted for maximum length
gain. Increasing the height of the antenna will of First director to second director - 0.25 wave¬
course lower the wave angle since the shape and length
amplitude of the vertical lobes are determined by Using a length/diameter ratio of about 100 for the
the ground-reflection factors given in Chapter Two elements, the element lengths for maximum gain
as well as by the free-space pattern of the antenna were found to be
itself. Reflector - 0.51 wavelength
Driven element — 0.47 wavelength
FOUR-ELEMENT ARRAYS First director - 0.45 wavelength
Parasitic arrays having a driven element and Second director — 0.44 wavelength
three parasitic elements - reflector and two The input resistance with the above spacings and
directors - are frequently used at the higher dimensions was of the order of 30 ohms and the
frequencies, 28 MHz and up. This type of antenna antenna gave useful gain over a total bandwidth
is shown in Fig. 4-49. equal to about 4 percent of the center frequency.
Close spacing is undesirable in a four-element
antenna because of the low radiation resistance. An LONG YAGIS
optimum design, based on an experimental deter¬
Parasitic arrays are not limited as to the number
of elements that can be used, although it is hardly
practical to use more than four at frequencies
below 30 MHz. However, on the vhf bands an array
that is long in terms of wavelength is often of
practicable physical size. Several independent in¬
vestigations of the properties of multielement Yagi
antennas have shown that in a general way the gain
of the antenna expressed as a power ratio is
proportional to the length of the array, provided
the number, lengths, and spacings of the elements
are properly chosen.
The results of one such study (by Carl Green-
blum) are shown in terms of the number of
elements in the antenna in Figs. 4-50 and 4-51. In
every case the antenna consists of a driven element,
one reflector, and a series of directors properly
spaced and tuned. Thus if the antenna is to have a
gain of 12 dB, Fig. 4-50 shows that 8 elements -
driven, reflector, and six directors — will be
required, and Fig. 4-51 shows that for such an
8-element antenna the array length required is 1.75
wavelength.
Table 4-II shows the optimum element spacings
Fig. 4-47 — Measured radiation patterns of a
horizontally polarized 3-element parasitic array determined from the Greenblum investigations.
having the director 0.1 wavelength from the driven There is a fair amount of latitude in the placement
element and the reflector 0.15 wavelength from of the elements along the length of the array,
the driven element. Element tuning was adjusted although the optimum tuning of the element will
for maximum front-to-back ratio. Heights and vary somewhat with the exact spacing chosen.
wave angles are same as in Fig. 4-45. Within the spacing ranges shown, the gain will not
Multielement Directive Arrays 153

TABLE 4 11
Optimum Element Spacings for Multielement Yagi Arrays

No.
Elements R-DE DE-Dt Dt -D 2 d2-d< Da -Ds Ds-Ds

2 0.15X-0.2X
2 0.07X-0.11X
3 0.16 —0.23 0.16 -0.19
4 0.18 -0.22 0.13 -0.17 0.14X-0.18X
5 0.18 -0.22 0.14 -0.17 0.15 -0.20 0.17X-0.23X
6 0.16 -0.20 0.14 -0.17 0.16 -0.25 0.22 -0.30 0.25X-0.32X
-0.20 0.14 -0.16 0.18 -0.25 0.25 -0.35 0.27 -0.32 0.27X-0.33X 0.30X-0.40X
8 0.16
-0.20 0.14 -0.16 0.18 -0.25 0.25 -0.35 0.27 -0.32 0.27 -0.33 0.35 -0.42
8 to N 0.16

DE — Driven Element; R — Reflector; D - Director; N - any number; director spacings beyond D* should be 0.35-0.42X.

vary more than 1 dB provided the director lengths In another study of long Yagi antennas at vhf,
are suitably adjusted. J. A. Kmosko, W2NLY, and H. G. Johnson,
The optimum director lengths are in general W6QKI, reached essentially the same general con¬
greater, the closer the particular director is to the clusions concerning the relationship between over¬
driven element, but the length does not uniformly all antenna length and power gain, although their
decrease with increasing distance from the driven gain figures differ from those of Greenblum. The
element. Fig 4-52 shows the experimentally deter- comparison is shown in Fig. 4-53. The Kmosko-
Johnson results are based on a somewhat different
element spacing and a construction in which thin
director elements are supported above the metal
boom rather than running through it. In their
optimum design the first director is spaced 0.1
wavelength from the driven element. The next two
directors are slightly over 0.1 wavelength apart, the
fourth director is approximately 0.2 wavelength
from the third, and succeeding directors are spaced
0.4 wavelength apart. The Kmosko-Johnson figures
are based on a simplified method of computing
gain from the beamwidth of the antenna pattern,
the beamwidths having been measured experimen¬
tally. The Greenblum data is from experimental
measurement of gain.
Experimental gain figures based on measure¬
ments made at 3.3-centimcter wavelength by H. W.
Fig. 4-49 — A four-element antenna system, using Ehrenspeck and H. Poehler, shown by a third curve
two directors and one reflector in conjunction with in Fig. 4-53, indicate lower gain for a given antenna
a driven element. length but confirm the gain-vs.-length trend. These
measurements were made over a large ground plane
mined lengths for various element diameters, based using elements of the order of one-quarter wave¬
on cylindrical elements supported by mounting length high. The general conclusions of this study
through a cylindrical metal boom two or three were (1) that the reflector spacing and tuning is
times the element diameter. The curves probably independent of the other antenna dimensions, the
would not be useful for other shapes. optimum fed-element to reflector spacing being in

Fig. 4-50 — Gain in dB over a

half-wave dipole vs. the number of
elements of the Yagi array, assuming
the array length is as given in Fig.
4-51.
 154 Chapter 4

Fig. 4-51 — Optimum length of

Yagi antenna as a function of
number of elements.

the neighborhood of 0.25 wave¬

length but not critical; (2) for a
given antenna length the gain is
practically independent of the
number of directors provided the
director-to-director spacing does
not exceed 0.4 wavelength; (3)
that the optimum director tuning
differs with different director
spacings but that for constant
spacings all directors can be simi¬
larly tuned; and (4) a slight im¬
provement in gain results from
using an extra director spaced
about 0.1 wavelength from the
driven element.
The agreement between these three sets of and gain. The increase in gain that can be realized
measurements is not as close as might be wished, is dependent on the spacing between the individual
which simply confirms the difficulty of deter¬ arrays. It is assumed, of course, that all the
mining optimum design where a multiplicity of individual arrays making up the stacked system are
elements is used, and of measuring gain with a identical, and that in the case of broadside stacking
degree of accuracy that will permit reconciliation the corresponding elements are parallel and lie in
of the results obtained by various observers. There planes perpendicular to the axis of the individual
is, however, agreement on the general principle arrays. In collinear stacking, it is assumed that the
that length is of greater importance than the corresponding elements are collinear and all ele¬
number of elements, within the limit of a maxi¬ ments of the individual arrays lie in the same plane.
mum element spacing of 0.4 wavelength. In both cases the driven elements must be fed in
It is an interesting fact that the feed-point phase.
impedance and bandwidth of long Yagis depends The decrease in beamwidth of the main radia¬
almost entirely on the two or three parasitic tion lobe that accompanies stacking is, in the
elements closest to the driven element, presumably general case, accompanied by a splitting off of one
because those farther from the driven element are or more sets of side lobes. These will have an
relatively loosely coupled to it. In this respect, amplitude depending on the shape of the directive
therefore, the information already given in connec¬ pattern of the unit array, the number of unit
tion with three-element arrays is quite applicable. arrays, and their spacing. An optimum spacing is
STACKED YAGIS one which gives as much gain as possible on the
condition that these side lobes do not exceed some
Parasitic arrays can be stacked either in broad¬ specified amplitude relative to the main lobe. Fig.
side or collinear fashion for additional directivity 4-54 shows the optimum spacings for three such
Multielement Directive Arrays 155

conditions (no side lobes, side lobes down 10 dB,

and side lobes down 20 dB) as a function of the
half-power beamwidth of the unit array, from
calculations by H. W. Kaspar, K2GAL. Maximum
gain occurs when the side lobes are approximately
10 dB down, as indicated in the figure.
A single 3-element array will have a half-power
beamwidth (free space) of approximately 75 de¬
grees, and from Fig. 4-54 it can be determined that
the optimum spacing for maximum gain will be
slightly over 3/4 wavelength. Measurements by
Greenblum have shown that the stacking gain that
can be realized with two such Yagi antennas is Fig. 4-53 — Gain of long Yagi antennas as a
approximately 3 dB, remaining practically constant function of overall length. The antenna consists of
at spacings from 3/4 to 2 wavelengths, but decreas¬ a driven element, a single reflector spaced approxi¬
ing rapidly at spacings less than 3/4 wavelength. mately one-fourth wavelength from the driven
With spacings less than about 1/2 wavelength element, and a series of directors spaced as
stacking does not give enough gain to make the described in the text. The three curves represent
construction of a stacked array worthwhile. the results of three independent studies.
If reduction of side-lobe amplitude is the
principal consideration rather than gain, smaller
spacings are optimum, as shown by the curves. A imum front-to-back ratio were given earlier in this
similar set of curves for four stacked unit arrays is chapter. The principal adjustment that actually
given in Fig. 4-55. needs to be made is to match the antenna to the
The spacings in Figs. 4-54 and 4-55 are mea¬ transmission line so the standing-wave ratio on the
sured between the array centers. When the unit latter is minimized.
arrays are stacked in a collinear arrangement, a Methods of Feed
spacing of less than 1/2 wavelength is physically
impossible with full-size elements, since at 1/2- The driven element in a parasitic array is a load
wavelength spacing the ends of the collinear for the transmission line in the same way that a
elements will be practically touching. driven element in any antenna system is such a
load. It differs from the load presented by a simple
dipole only in that the resistance may be quite low,
FEEDING AND ADJUSTMENT especially if close spacings are used between
The problems of matching and adjusting para¬ elements, and the rate of change of reactance as
sitic arrays for maximum performance are the same the operating frequency is moved away from the
in principle as with other antenna systems. Adjust¬ design frequency may be greater. With low input
ment of element lengths for optimum performance resistance, a fairly large impedance step-up is
usually necessitates measurements of relative field required for matching practicable lines, and the
strength. However, the experience of a great many amount of mismatch will increase more rapidly
amateurs who have followed the rather laborious than with a simple dipole when the applied
procedure (adusting each element a little at a time, frequency is varied from that at which the line is
and measuring the relative field after each such matched.
change) has accumulated a large amount of data on Practically any of the matching systems de¬
optimum lengths. Depending on the objective in tailed in Chapter Three are applicable. For exam¬
designing the antenna - i.e., max¬
imum gain, maximum front-to-
back ratio, etc. - it is possible to
predetermine the actual element
lengths for a given center fre¬
quency and thus avoid the neces¬
sity for such adjustments. Charts
giving proper element lengths for
3-element beams are shown in the
chapter on 14-, 21-, and 28-MHz
antennas for the maximum-gain
condition, and data for max-

Fig.4-54 — Optimum stacking

spacing for two-unit arrays. The
spacing for no side lobes, especial¬
ly for small beamwidths, may
result in almost no gain improve¬
ment with stacking.
 156 Chapter 4

varying amounts of reactance coupled

into the driven element, as well as the
fact that the radiation resistance at the
center of the driven element is often
very low, cause the impedance to change
rapidly when the applied frequency is
varied above or below the design fre¬
quency.
This impedance change can be made
less rapid by using fairly wide spacing
between elements, as already mentioned.
It is also beneficial to use elements
having a fairly large ratio of diameter to
length because, as explained in Chapter
20 40 60* 80* lOO* 120* 140*
160 iso* tWOi the impedance change with fre¬
HALF POWER BEAMWIDTH OF UNIT ARRAY
quency is reduced when the antenna
Fig. 4-55 — Optimum stacking spacing for four- conductor has a large diameter.
unit arrays. The use of a folded-dipole driven element is
beneficial in broadening the frequency characteris¬
tic of the antenna because of the smaller effective
pie, an open quarter-wave matching section can be length/diameter ratio that results from using two
used if the transmission line is to be 300- to or more conductors instead of one.
600-ohm parallel-conductor line. The quarter-wave
transformer (Q) method also can be used, with Adjusting Parasitic Arrays
75-ohm Twin-Lead for the transformer. This par¬
ticular value of Zo will match a driven element There are two separate processes in adjusting an
having a resistive impedance of 9 to 10 ohms to a array with parasitic elements. One is the determina¬
600-ohm open-wire line, and will result in only a tion of the optimum element lengths, depending
2-to-l mismatch if the driven-element impedance is on whether maximum gain or maximum front-to-
as low as 4 or 5 ohms or as high as 18 to 20 ohms. back ratio is desired. The other is matching the
The loss in a quarter-wave section of 75-ohm antenna to the transmission line. The second is
transmitting-type Twin-Lead, even though the usually dependent on the first, and the results
SWR in the matching transformer is rather high, observed on adjusting the element tuning may well
will not be of any real consequence. Alternatively, be meaningless unless the line is equally well
the matching transformer may be made of 50- or matched under all tuning conditions.
75-ohm coaxial cable. However, in such case some As stated earlier, the element tuning for maxi¬
method of line balancing (such as those described mum gain is not excessively critical, and the
in Chapter Three under baluns) should be used. dimensions given by the following formulas have
The delta match may also be used. been found to work well in practice for 3-element
The folded dipole used as the driven element antennas:
also furnishes a useful method of transforming a 475
low antenna resistance to a value suitable for Driven element: Length (ft) = yyYjj^)
matching a transmission line. Design details are
given in Chapter Three. Director: Length (ft) = y-^-
The choice of a matching system is affected by
constructional considerations, since parasitic arrays
Reflector: Length (ft) = y
are usually built to be rotated in operation. The
T match and gamma match are favorites with
many amateurs because they fit in well, construc¬ These are average lengths determined experimental¬
tionally, when the driven element is made of ly for elements having a length/diameter ratio of
tubing. Another matching system suitable for 200 to 400, and with element spacings from 0.1 to
continuously rotatable antennas uses two large 0.2 wavelength.
inductively coupled loops, one fixed and one Many amateurs have found that very satisfac¬
rotatable. Open-wire line is preferred for such tory results are secured simply by cutting the
coupling, however, so this method is seldom used elements to the lengths given by these formulas. It
these days. Because of the attendant problems with has been a rather common experience that, after a
installation of open-wire lines, most amateurs considerable amount of time has been spent in
prefer coaxial lines for carrying rf power up a trying all possible adjustments, the dimensions
tower or other support and use rotators with finally determined to be optimum are very close to
mechanical stops to avoid entangling the feed line those given by the formulas above or the charts
around the tower. contained in Chapter Nine (the difference between
the charts and the formulas above amount to only
Broadening the Response about one percent in most cases), and the actual
difference in gain is negligible. It appears safe to
It has already been pointed out that the tuning say, therefore, that in the average case there is
conditions giving maximum gain with parasitic probably little to be realized, in the way of
elements are not highly critical. However the increased gain, by spending much time in adjusting
Multielement Directive Arrays 157

element lengths. The front-to-back ratio can often

be improved, however, since it depends much more
on the exact element tuning. In general, the
reflector tuning is the more critical.
If the array is put up by formula, the only
adjustment that need be made is to match the
driven element to the transmission line. The
adjustment procedure for each type of matching
arrangement is described in Chapter Three.
Test Setup
The only practicable method of adjusting para¬
sitic element lengths for best performance is to FREQUENCY

measure the field strength from the antenna as Fig. 4-56 — Field-strength measurement setup. The
adjustments are made. Measurements on a relative folded dipole should be at least as high as the
basis are entirely satisfactory for the purpose of antenna under test and should be three or more
determining the operating conditions that result in wavelengths away. R should be a 300-ohm compo¬
the maximum output or greatest front-to-back sition resistor to provide a proper load for the line,
ratio. For this purpose the measuring equipment so that a line of any desired length can be used. If
does not need to be calibrated; the only require¬ the sensitivity is not high enough with this arrange¬
ment the alternative connections at the right will
ment is that it indicate whether the signal is result in increased meter readings. The taps are
stronger or weaker. adjusted for maximum reading, keeping the
If the help of a nearby amateur owning a transmission-line taps spaced equally on either side
receiver with an S meter can be enlisted, the of the coil center tap. The indicating meter, M may
S-meter indications can be used to indicate the be either a microammeter or 0—1 milliammeter.
relative field strength. A few precautions must be
taken if this method is to be reliable. The receiving
antenna must have the same polarization as the the indicator, to avoid stray pickup. This pickup
transmitting antenna under test (that is usually can be checked as described in the preceding
horizontal) and should be reasonably high above paragraph. If the distance between the two anten¬
its surroundings. The receiving system should be nas is such that greater sensitivity is needed a
checked for pickup on the transmission line to reflector may be placed 1 /4 wavelength behind the
make sure that the indications given by the receiver receiving dipole.
are caused entirely by energy picked up by the
receiving antenna itself. This can be checked by Adjustment Procedure
temporarily disconnecting the line from the anten¬
na (but leaving it in place) and observing the signal It is advisable first to set the element lengths to
strength on the S meter. If the reading is not those given by the formulas and then match the
several S units below the reading with the antenna driven element to the transmission line obtaining as
connected, the readings cannot be relied upon low an SWR as possible. In subsequent adjustments
when adjusting the transmitting antenna for maxi¬ a close watch should be kept on the SWR and the
mum gain. In checking the front-to-back ratio, the transmitter power input should be maintained at
stray pickup at the receiving installation must be exactly the same figure throughout. If the SWR
well below the smallest signal received via the changes enough to affect the coupling at the
antenna, if the adjustments are to mean anything transmitter when an adjustment is made, but not
at all. enough to raise the line loss significantly (see Fig.
Another method of checking field strength is to 3-23), readjust the coupling to bring the input back
use a field-strength indicator of the diode-detector to the same value. If the line loss increases more
type. The preferable method of using such an than a fraction of a decibel, rematch at the driven
indicator is to connect it to a dipole antenna element. If this is not done, the results may be
mounted some distance away and at a height at entirely misleading; it is absolutely necessary to
least equal to that of the transmitting antenna. maintain constant power input to the driven
There should be no obstructions between the two element if adjustment of directors or reflectors is
antennas, and both should have the same polariza¬ to give meaningful results.
tion. The receiving dipole need not be a half wave The experience of most amateurs in adjusting
long, although that length is desirable because it parasitic arrays indicates that there is not a great
will increase the ratio of energy picked up on the deal of preference in the order in which elements
antenna to energy picked up by stray means. To are tuned, but that there is slightly less interaction
prevent coupling effects the distance between the if the director is first adjusted to give maximum
two antennas should be at least three wavelengths. gain and the reflector is then adjusted to give either
At shorter distances the mutual impedance may be maximum gain or maximum front-to-back ratio,
large enough to cause the receiving antenna to tend whichever is desired. After the second parasitic
to become part of the transmitting system, which element has been adjusted, go back and check the
can lead to false results. A recommended type of tuning of the first to make sure that it has not been
indicating system is shown in Fig. 4-56. The thrown out of adjustment by the mutual coupling.
transmission line should drop vertically down to If there are three parasitic elements, the other two
 158 Chapter 4

selectivity, even though some gain is sacrificed in

so doing.
Adjustment by Reception
As an alternative to applying power to the array
and checking the field strength, it is possible to
adjust the array by measuring received signal
strength. It is impracticable to do this on distant
signals because of fading. The most reliable method
is to erect a temporary antenna of the type
recommended for field-strength measurements (Fig.
4-56) and excite it from a low-power oscillator.
The same precautions with respect to distance
between the two antennas apply.
Fig. 4-57 — The basic quad antenna, with driven In this method, as in the one Where the
loop and reflector loop. The loops are electrically transmitting antenna is excited, it is necessary to
one wavelength in circumference (1/4 wavelength minimize line radiation and pickup if the results
on a side). Both configurations shown give horizon¬
tal polarization; for vertical polarization, the driven are to be reliable. The same tests may be applied.
element should be fed at one of the side corners in However, it is less easy to keep the SWR under
the arrangement at the left, or at the center of a control. In the receiving case the SWR on the
vertical side in the "square" quad at the right. transmission line depends on the load presented by
the receiver to the line. Under most conditions the
SWR will be reasonably constant over an amateur
should be checked each time an appreciable change band, although its value may not be known.
is made in one. The actual lengths should not be However, the energy transfer from the antenna to
very far from those given by the formulas when the the line depends on the mismatch between the
optimum settings are finally determined. As al¬ driven element and the line. There is no convenient
ready pointed out, the reflector length may be way to check this in the receiving case. About all
somewhat greater when adjusted to give maximum that can be done is to apply power to the array
front-to-back ratio. after a set of tuning conditions has been reached,
Radiation from the transmission line must be and then rematch at the driven element if neces¬
eliminated, or at least reduced to a very low value sary. After rematching, the measurement will have
compared with the radiation from the antenna to be repeated. Thus double checking is necessary
itself, if errors are to be avoided. Conditions are if the results are to be comparable with those
usually favorable to low line radiation in horizon¬ obtained by the field-strength method.
tally polarized rotatable parasitic arrays because
the line is usually symmetrical with respect to the THE QUAD ANTENNA
antenna and is brought away perpendicular to it, at In this chapter it has been assumed that the
least for a half wavelength or so. Nevertheless the various antenna arrays have been assemblies of
line radiation can be appreciable unless the line is linear half-wave (or approximately half-wave) di¬
detuned as described in Chapter Three. With pole elements. However, other element forms may
coaxial line some method of line balancing at the be used according to the same basic principles. For
antenna always should be incorporated to avoid example, loops of various types may be combined
“skewing” the beam pattern or lowering the into directive arrays. A popular type of parasitic
front-to-back ratio. array using loops is the quad antenna, in which
After arriving at the optimum adjustments at loops having a perimeter of one wavelength are
the frequency for which the antenna was designed, used in much the same way as dipole elements in
the performance should be checked over a frequen¬ the Yagi antenna.
cy range either side of the design frequency to The quad antenna was originally designed by
observe the sharpness of response. If the field Moore, W9LZX, in the late 1940s. Since its
strength falls off rapidly with frequency, it may be inception there has been extensive controversy
desirable to shorten the director a bit to increase whether the quad is a better performer than a Yagi.
the gain at frequencies above resonance and length¬ This argument continues, but over the years several
en the reflector slightly to increase it at frequencies facts have become apparent. For example, Lindsay
below resonance. Do not confuse the change in (W7ZQ, ex-W0HTH) has made many comparisons
SWR with the change in antenna gain. The antenna between quads and Yagis. His data, Fig. 4-58,
itself may give good gain over a considerable shows that the quad has a gain of approximately 2
frequency range, but the SWR may vary between decibels over a Yagi for the same array length.
wide limits in this range. To check the antenna Another argument that has existed is that for a
behavior, keep the power input to the transmission given array height, the quad has a lower angle of
line constant and rematch the driven element to radiation than a Yagi. Even among authorities
the line, as suggested above, whenever the line there is disagreement on this point. However, as
losses increase appreciably. If such rematching is Fig. 4-59 shows, the //-plane pattern of a quad is
found necessary over the band of frequencies to be slightly greater than a Yagi at the half-power
used, it may be advisable to retune the system to points. This means that the quad covers a wider
give a higher input resistance and thus decrease the area in the vertical plane.
Multielement Directive Arrays 159

The full-wave loop has been discussed in Chap¬ about 4.5 percent greater than the wavelength in
ter Two. Two such loops, one as a driven element free space, as compared to the approximately 2
and one as a reflector, are shown in Fig. 4-57. This percent increase in the formula above for the
is the original version of the quad; in subsequent driven element.
development, loops tuned as directors have been In any case, on-the-ground adjustment is re¬
added in front of the driven element. The square quired if optimum results are to be secured,
loops may be mounted either with the corners especially with respect to front-to-back ratio. The
lying on horizontal and vertical lines, as shown at method of adjustment parallels that outlined pre¬
the left, or with two sides horizontal and two viously for the Yagi antenna.
vertical (right). The feed points shown for these Element spacings of the order of 0.14 to 0.2
two cases will result in horizontal polarization, wavelength are generally used, the smaller spacings
which is commonly used. being employed in antennas having more than two
The parasitic element is tuned in much the elements, where the structural support for ele¬
same way as the parasitic element in a Yagi ments with larger spacings tends to become dif¬
antenna. That is, the parasitic loop is tuned to a ficult. The feed-point impedances of antennas
lower frequency than the operating frequency having element spacings of this order have been
when the element is to act as a reflector, and to a found to be in the 40- to 60-ohm range, so the
higher frequency when it acts as a director. Fig. driven element can be fed through coaxial cable
4-57 shows the parasitic element with an adjustable with only a small mismatch. For spacings of the
tuning stub, a convenient method of tuning since order of 0.25 wavelength (physically feasible for
the resonant frequency can be changed simply by two elements, or for several elements at 28 MHz)
changing the position of the shorting bar on the the impedance more closely approximates the
stub. In practice, it has been found that the length impedance of a driven loop alone (see Chapter
around the loop should be approximately 3 per¬ Two) — that is, 80 or 90 ohms.
cent greater than the self-resonant length if the The feed methods described in Chapter Three
element is a reflector, and about 3 percent shorter can be used, just as in the case of the Yagi.
than the self-resonant length if the parasitic ele¬
ment is a director. Approximate formulas for the Directive Patterns and Gain
loop lengths in feet are The small gain of a one-wavelength loop over a
half-wave dipole (see Chapter Two) carries over
Driven.element
. 1005 = . into arrays of loops. That is, if a quad parasitic
f (MHz) array and a Yagi having the same overall length
(boom length) are compared the quad will have
1030 approximately 2 dB greater gain than the Yagi, as
Reflector =
/(MHz) mentioned earlier. This assumes that both antennas
have the optimum number of elements for the
975 antenna length; the number of elements is not
Director = ,,.. u necessarily the same in both when the antennas are
/ (MHz)
long. Fig. 4-58 shows the results of comparative
for quad antennas intended for operation below 30 experimental measurements on the two antenna
MHz. At vhf, where the ratio of loop circumfer¬ types at 440 MHz, made by Lindsay (from QST,
ence to conductor diameter is usually relatively May, 1968). The curves show directivity vs. array
small, the circumference must be increased in length, which may be read as feet instead of
comparison to the wavelength. For example, a centimeters if the frequency is taken as 14 MHz
one-wavelength loop constructed of quarter-inch instead of 440 MHz. Gain over a half-wave dipole
tubing for 144 MHz should have a circumference (assuming negligible ohmic loss) can be found by

Fig. 4-58 — Comparative directivity

(and gain) of the Yagi and quad as a
function of overall array length. Al¬
though measured with circular loops,
the performance with the square
loops used in the quad is comparable.
The measurements were made on
model antennas at 440 MHz (W7ZQ)
but also apply at lower frequencies.
 160 Chapter 4

5 ELEMENT YAGI
30 CM BOOM 55 CM BOOM
(A) (B)

Fig. 5-59 Measured patterns of 4-element quad and 5-element Yagi antennas, showing
approximately equivalent beamwidths. Measurements made on model antennas at 440 MHz by W7ZQ.

subtracting 2.14 dB. The gains were calculated the same for the quad and Yagi when the overall
from the measured patterns shown in Fig. 4-59. length of the latter is about twice (more closely,
The experimentally measured pattern in Fig. 1.8 times) the length of the quad, indicating that
4-59 shows that the beamwidths are approximately under such conditions the gains are about equal.

THE LOG-PERIODIC DIPOLE ARRAY

The log-periodic dipole array (LPDA) consists frequency, there is a smooth transition along the
of a system of driven elements, but not all array of the elements which comprise the active
elements in the system are active on a single region. The following information was provided by
frequency of operation. Depending upon its design Peter Rhodes, K4EWG.
parameters, the LPDA can be operated over a range A good LPDA may be designed for any band,
of frequencies having a ratio of 2:1 or higher, and hf to uhf, and can be built to meet the amateur’s
over this range its electrical characteristics - gain, requirements at nominal cost: high forward gain,
feed-point impedance, front-to-back ratio, etc. — good front-to-back ratio, low VSWR, and a boom
will remain more or less constant. This is not true length equivalent to a full sized three-element Yagi.
of any of the types of antennas discussed earlier in The LPDA exhibits a relatively low SWR (usually
this chapter, for either the gain factor or the not greater than 2 to 1) over a wide band of
front-to-back ratio, or both, deteriorate rapidly as frequencies. A well-designed LPDA can yield a
the frequency of operation departs from the design 1.3-to-l SWR over a 1.8-to-l frequency range with
frequency of the array. And because the antenna a typical directivity of 9.5 dB. (Directivity is the
designs discussed earlier are based upon resonant ratio of maximum radiation intensity in the for¬
elements, off-resonance operation introduces reac¬ ward direction to the average radiation intensity
tance which causes the SWR in the feeder system from the array. Assuming no resistive losses in the
to increase. antenna system, 9.5 dB directivity equates to 9.5
As may be seen in Fig. 4-60, the log-periodic dB gain over an isotropic radiator or approximately
array consists of several dipole elements which are 7.4 dB gain over a half-wave dipole.
each of different lengths and different relative
spacings. A distributive type of feeder system is
used to excite the individual elements. The element Basic Theory
lengths and relative spacings, beginning from the The LPDA is frequency independent in that the
feed point for the array, are seen to increase electrical properties such as the mean resistance
smoothly in dimension, being greater for each level, /?o, characteristic impedance of the feed line,
element than for the previous element in the array. Zo, and driving-point admittance, To, vary period¬
It is this feature upon which the design of the ically with the logarithm of the frequency. As the
LPDA is based, and which permits changes in frequency f, is shifted to another frequency f2
frequency to be made without greatly affecting the within the passband of the antenna, the relation¬
electrical operation. With changes in operating ship is f2 = /t, where
Multielement Directive Arrays 161

r = a design parameter, a constant; r<1.0. Also

(Eq.1)
n= 1, 2, 3,. . . n
= lowest frequency
n = highest frequency
f

The design parameter tis a geometric constant

near 1.0 which is used to determine the element
lengths, /, and element spacings, d, as shown in Fig. Fig. 4-60 — Schematic diagram of log-periodic
4-60. That is. dipole array, with some of the design parameters
indicated. Design factors are:
4 = rl,
In _ dn n ~ 1
= rh
41 - 1 dn - 2,n - 1

n - i
_ dn,
'n = T4n _ 1) (Eq. 2)
241-1

where l
n - shortest element length, and h ^^n .
nn =-y , where

dis = rd, 2
I= el. length
dj4 = rdt3 h = el. half length
d = element spacing
T = design constant
a - relative spacing constant
dn - l,n = Tdn - 2,n - 1 (Eq. 3) S= feeder spacing
Zo = char, impedance of antenna feeder
where d¡ 3 = spacing between elements 2 and 3.
Each element is driven with a phase shift of feeder and act as parasitic directors. The inductive
180° by switching or alternating element connec¬ current components in longer elements 4 and 5 are
tions, as shown in Fig. 4-60. The dipoles near the dominant and they act like parasitic reflectors.
input, being nearly out of phase and close together, Elements 6, 7 and 8 receive most of their power
nearly cancel each others’ radiation. As the ele¬ from the feeder and act like driven elements. The
ment spacing, d, expands there comes a point along amplitudes of the currents in the remaining ele¬
the array where the phase delay in the transmission ments are small and they may be ignored as
line combined with the 180’ switch gives a total of primary contributors to the radiation field. Hence,
360°. This puts the radiated fields from the two we have a generalized Yagi array with seven
dipoles in phase in a direction toward the apex. elements comprising the active region. It should be
Hence, a lobe coming off the apex results. noted that this active region is for a specific set of
This phase relationship exists in a set of dipoles design parameters (r = 0.93, a= 0.175). The
known as the “active region.” If we assume that an number of elements making up the active region
LPDA is designed for a given frequency range, then
that design must include an active region of dipoles
for the highest and lowest design frequency. It has
a bandwidth which we shall call (bandwidth of
the active region).
Assume for the moment that we have a
12-element LPDA. Currents flowing in the ele¬
ments are both real and imaginary, the real
current flowing in the resistive component of
the impedance of a particular dipole, and the
imaginary flowing in the reactive component.
Assume that the operating frequency is such that
element number 6 is near to being half-wave
resonant. The imaginary parts of the currents in
shorter elements 7 to 12 are capacitive, while those
in longer elements 1 to 6 are inductive. The
capacitive current components in shorter elements
9 and 10 exceed the conductive components;
hence, these elements receive little power from the Fig. 4-61.
162 Chapter 4

term antenna feeder is defined as that line which

connects each adjacent element. The feed line is
that line between antenna and transmitter. The
characteristic impedance of the antenna feeder, Zo,
must be determined so that the feed-line imped¬
ance and type of balun can be determined. The
antenna-feeder impedance Zo depends on the mean
radiation resistance level Ro (required input imped¬
ance of the active region elements - see Fig. 4-63)
and average characteristic impedance of a dipole,
Za. (Za is a function of element radius aand the
resonant element half length, where h = X/4. See
Fig. 4-64.) The relationship is as follows:
Fig. 4-62.
Zo =
8 a Za + (Eq.
y \8 a Z^) 9)

will vary with rand a. Adding additional elements where Zo = characteristic impedance of feeder
on either side of the active region cannot signifi¬
cantly modify the circuit or field properties of the Ro = mean radiation resistance level or
array. required input impedance of the
This active region determines the basic design active region.
parameters for the array, and sets the bandwidth Za = average characteristic impedance of a
for the structure, Äs. That is, for a design¬ dipole
frequency coverage of bandwidth B, there exists an
associated bandwidth of the active region such that = 120 (/n — - 2.55) (Eq. 10)
a

Äs = BX B^ (Eq. 4) h= el. half length

f a= radius of el.
where B = operating bandwidth ~— (Eq. 5) a = mean spacing factor = ^=- (Eq. 11)

/i = lowest freq., MHz

fn = highest freq., MHz From Fig. 4-63 we can see that Ro decreases
with increasing rand increasing Also theVSWR
B^ varies with rand “ as shown in Fig. 4-61. with respect to Ro has a minimum value of about
Element lengths which fall outside B^ play an 1.1 to 1 at a optimum, and a value of 1.8 to 1 at a
insignificant role in the operation of the array. The = .05. These SWR values are acceptable when using
gain of an LPDA is determined by the design standard RG-8/U 52-ohm and RG-ll/U 72-ohm
parameter r and the relative element spacing coax for the feed line. However, a one-to-one
constant a. There exists an optimum value for a, VSWR match can be obtained at the transmitter
uOpt> f°r eac h T in the range 0.8 < r < 1.0, for end using a coax-to-coax Transmatch. A Trans¬
which the gain is maximum; however, the increase match will enable the transmitter low-pass filter to
in gain achieved by using a opt and rnear 1.0 (i.e., see a 52-ohm load on each frequency within the
r = 0.98) is only 3 dB above isotropic (3 dBi) when array passband. The Transmatch also eliminates
compared with the minimum a(omin = -05) and r possible harmonic radiation caused by the fre¬
= 0.9, shown in Fig. 4-62. quency-independent nature of the array.
An increase in rmeans more elements and
optimum ameans a long boom. A high-gain (8.5
dBz) LPDA can be designed in the hf region with r
= 0.9 and a= .05. The relationship of r, a, and “ is
as follows:
a= (1/4) (1 - t) cot “ (Eq. 6)

where « = 1/2 the apex angle

r= design constant
a= relative spacing constant

aiso«^^-^ (Eq. 7)
L, n - 1

c
Opt = 0.258t — .066 (Eq. 8)

The method of feeding the antenna is rather

simple. As shown in Fig. 4-60, a balanced feeder is
required for each element, and all adjacent ele¬
ments are fed with a 180° phase shift by alter¬
nating element connections. In this section the
Multielement Directive Arrays 163

Once the value of Zo has been determined for

each band within the array passband, the balun and
feed line may be chosen. That is, if Zo = 100 ohms,
a good choice for the balun would be 1 to 1
balanced to unbalanced, and 72-ohm coax feed
line. If Zo = 220 ohms, choose a 4 to 1 balun, and
52-ohm coax feed line, and so on. The balun may
be omitted if the array is to be fed with an
open-wire feed line.
The terminating impedance, Zt, may be omit¬
ted. However, if it is used, it should have a length
no longer than X max /8. The terminating imped¬
ance tends to increase the front-to-back ratio for
the lowest frequency used. For hf-band operation a
6-inch shorting jumper wire may be used for Zt. THICKNESS RATIO k/a.
When Zt is simply a short-circuit jumper the Fig. 4-64.
longest element behaves as a passive reflector. It
also might be noted that one could increase the
front-to-back ratio on the lowest frequency by needs. If the array is too large, increase by 5°
moving the passive reflector (No. 1 element) a and repeat steps 2 through 6.
distance of 0.15 to 0.25 X behind element No. 2, as 7) Determine the terminating stub ZL (Note:
would be done in the case of an ordinary Yagi For hf arrays short out the longest element with a
parasitic reflector. This of course would necessitate 6-inch jumper. For vhf and uhf arrays use:
lengthening the boom. The front-to-back ratio
increases somewhat as the frequency increases. 2t = Xnux/8’
This is because more of the shorter inside elements
form the active region, and the longer elements 8) Once the final values of r and oare found,
become additional reflectors. the characteristic impedance of the feeder Zo must
be determined so the type of balun and feed line
Design Procedure can be found. Use Eq. 9. Determine Ro from Fig.
4-63, Za from Fig. 4-64 and o' from Eq. 11. Note:
A systematic step-by-step design procedure of Values for h/a, Za, and Zo must be determined for
the LPDA follows. This procedure may be used for each amateur band within the array passband.
designing any LPDA for any desired bandwidth. Choose the element half-length hnearest h = X/4,
1) Decide on an operating bandwidth B be¬ at the center frequency of each amateur band.
tween ft , lowest frequency and f n,highest fre¬ Once Zo is found for each band, choose whatever
quency, using Eq. 5. combination of balun and feed line will give the
2) Choose r and a to give a desired gain (Fig. lowest SWR on each band.
4-62). 9) Solve for the remaining element lengths from
0.8 < r < 0.98 Eq. 2.
.05 o ^opt 10) Determine the element spacing di i from

The value of a opt may be determined from Eq. 8. du = 1/2 (/,-/,) cot <x (Eq. 14)
3) Determine the apex half-angle «
. _ 4q and the remaining element-to-element spacings
cot “ = I-
1 — T from Eq. 3.
4) Determine the bandwidth of the active group
B^ from Fig. 4-61.
5) Determine the structure (array) bandwidth
ßs from Eq. 4.
6) Determine the boom length, L, number of
elements, N, and longest element length, li.

(Eq. 12)

(Eq. 13)

I- 492
Fig. 4-65 - Measured radiation pattern for the
lowest frequency band (14 MHz) of a 12-element
13-30 MHz log-periodic dipole array. For its design
where X max = longest free-space wavelength = parameters, r = 0.9 and o= .05. The measured
984//!. Examine L, N and h and determine front-to-back ratio is 14.4 dB at 14 MHz, and
whether or not the array size is acceptable for your increases to 21 dB at 28 MHz.
 164 Chapter 4

This completes the design. Construction infor¬ Ehrenspek and Poehler, “A New Method of Ob¬
mation for an array designed by this procedure is taining Maximum Gain from Yagi Antennas,”
contained in Chapter Nine. The measured radiation I.R.E. Transactions on Antennas and Propaga¬
pattern for a 12-element LPDA is shown in Fig. tion, October, 1959.
Gillson, “Parasitic-Array Patterns,” QST, March,
4-65. 1949.
This section has dealt with the basic LPDA Greenblum, “Notes on the Development of Yagi
system. However, there are several high-gain array Arrays,” QST, Part I, August, 1956; Part II,
possibilities using this type of antenna as a basis. September, 1956.
Tilting the elements toward the apex will increase Hall and Myers, “Phased Verticals in a 40-Meter
the gain 3 to 5 dB. Adding parasitic directors and a Beam-Switching Array,” QST, August, 1972.
reflector will increase both gain and front-to-back Isbell, “Log Periodic Dipole Arrays,” IRE Trans¬
ratio for a specific frequency within the passband. actions on Antennas and Propagation, Vol.
AP-8, No. 3, May, 1960, pp. 260-267.
The LPDA-Yagi combination is very simple. Use
Kasper, “Optimum Stacking Spacings in Antenna
the LPDA design procedures within the set of Arrays,” QST, April, 1958.
driven elements, and place parasitic elements at King, Mack, and Sandler, Arrays of Cylindrical
normal Yagi spacings from the LPDA end ele¬ Dipoles, pp. 244-269, Cambridge Univ. Press,
ments. Use standard Yagi design procedures for the London, 1968.
parasitic elements. An example of a single-band Kmosko and Johnson, "Long Long Yagis,” QST,
high-gain LPDA-Yagi would be a two- or three- January, 1956.
element LPDA for 21.0 to 21.45 MHz with the Kraus, “Directional Antennas with Closely-Spaced
addition of 2 or 3 parasitic directors and one Elements,” QST, January, 1938.
Ladner and Stoner, Short- Wave Wireless Communi¬
parasitic reflector. The combinations are endless. cation, John Wiley & Sons, Inc., New York,
N.Y.
Laport, Radio Antenna Engineering, McGraw-Hill
Book Co., New York, N.Y.
Lawson, “Simple Arrays of Vertical Antenna
Bibliography Elements,” QST, May, 1971.
Lindsay, “Quads and Yagis,” QST, May, 1968.
Source material and more extended discussion Rhodes, “The Log-Periodic Dipole Array,” QST,
of topics covered in this chapter can be found in Nov., 1973.
the references given below. Romander, “The Extended Double Zepp Anten¬
Brown, “Directional Antennas,” Proc. I.R.E., Jan¬ na,” QST, June, 1938.
uary, 1937 Rumsey, Frequency Independent Antennas, pp.
Carrel, “The Design of Log-Periodic Dipole Anten¬ 71-78, Academic Press, N.Y., 1966.
nas,” 1961 IRE International Convention Southworth, “Certain Factors Affecting the Gain
Record, Part 1, Antennas and Propagation, pp. of Directive Antennas,” Proc. I.R.E., Septem¬
61-75; also Ph.D. thesis, University of Ill., ber, 1930.
Urbana, Ill., 1961. Terman, Radio Engineering, McGraw-Hill Book
Carter, “Circuit Relations in Radiating Systems Co., New York, N.Y.
and Applications to Antenna Problems,” Proc. Uda and Mushiake, Yagi-Uda Antenna, Sasaki
I. R.E., J une, 1932. Publishing Co., Sendai, Japan.
 Chapter 5

Long-Wire Antennas

The power gain and directive characteristics that there is a diversity effect with a long-wire
of the harmonic wires (which are “long” in terms antenna because it is spread out over a large
of wavelength) described in Chapter Two make distance, rather than being concentrated in a small
them useful for long-distance transmission and space; this may raise the average level of received
reception on the higher frequencies. In addition, energy for ionospheric propagation. Another factor
long wires can be combined to form antennas of is that long-wire antennas have directive patterns
various shapes that will increase the gain and that are sharp in both the horizontal and vertical
directivity over a single wire. The term “long planes, and tend to concentrate the radiation at the
wire,” as used in this chapter, means any such low vertical angles that are most useful at the
configuration, not just a straight-wire antenna. higher frequencies. This is not true of some types
of multielement arrays.
Long Wires vs. Multielement Arrays
General Characteristics of
In general, the gain obtained with long-wire Long-Wire Antennas
antennas is not as great, when the space available
for the antenna is limited, as can be secured from Whether the long-wire antenna is a single wire
the multielement arrays in Chapter Four. To offset running in one direction or is formed into a V,
this the long-wire antenna has advantages of its rhombic, or some other configuration, there are
own. The construction of long-wire antennas is certain general principles that apply and some
simple both electrically and mèchanically, and performance features that are common to all types.
there are no especially critical dimensions or The first of these is that the power gain of a
adjustments. The long-wire antenna will work well long-wire antenna as compared with a half-wave
and give satisfactory gain and directivity over a dipole is not considerable until the antenna is
2-to-l frequency range; in addition, it will accept really “long” - long, that is, when the lengths are
power and radiate it well on any frequency for measured in wavelengths, rather than in a specific
which its overall length is not less than about a half number of feet. The reason for this is that the
wavelength. Since a wire is not “long,” even at 28 fields radiated by elementary lengths of wire along
MHz, unless its length is at least equal to a half the antenna do not combine, at a distance, in as
wavelength on 3.5 MHz, any long-wire antenna can simple a fashion as the fields from half-wave
be used on all amateur bands that are useful for dipoles used as described in Chapter Four. There is
long-distance communication. no point in space, for example, where the distant
As between two directive antennas having the fields from all points along the wire are exactly in
same theoretical gain, one a multielement array phase (as they are, in the optimum direction, in the
and the other a long-wire antenna, many amateurs case of two or more collinear or broadside dipoles
have found that the long-wire antenna seems more when fed with in-phase currents). Consequently,
effective in reception. One possible explanation is the field strength at a distance always is less than

Fig. 5-1 — Theoretical gain of

a long-wire antenna over a
dipole as a function of wire
length. The angle, with respect
to the wire, at which the
radiation intensity is maxi¬
mum also is shown.

165
166 Chapter 5

gain in the favored direction. For amateur work,

particularly with directive antennas that cannot be
rotated, the minor lobes of a long-wire antenna
have some advantages. In most directions the
antenna will be as good as a half-wave dipole, and
in addition will give high gain in the most favored
direction; thus a long-wire antenna (depending on
the design) frequently is a good all-around radiator
in addition to being a good directive antenna.
In the discussion of directive patterns of long-
wire antennas in this chapter, it should be kept in
mind that the radiation patterns of resonant long
wires are based on the assumption that each
half-wave section of wire carries, a current of the
same amplitude. As pointed out in Chapter Two,
this is not exactly true, since energy is radiated as
it travels along the wire. For this reason it is to be
anticipated that, although the theoretical pattern is
bidirectional and identical in both directions,
actually the radiation (and reception) will be best
in one direction. This effect becomes more marked
as the antenna is made longer.
Fig. 5-2 - Angles at which radiation from long
wires is maximum (solid curves) and zero (broken
curves). The major lobe. No. 1, has the power gains Wave Angles
given by Fig. 5-1. Secondary lobes have smaller
amplitude, but the maxima may exceed the radia¬ The wave angle at which maximum radiation
tion intensity from a half-wave dipole. takes place from a long-wire antenna depends
largely on the same factors that operate in the case
would be obtained if the same length of wire were of simple dipoles and multielement antennas. That
cut up into properly phased and separately driven is, the directive pattern in the presence of ground is
dipoles. As the wire is made longer the fields found by multiplying the free-space vertical-plane
combine to form an increasingly intense main lobe, pattern of the antenna by the ground-reflection
but this lobe does not develop appreciably until factors for the particular antenna height used.
the wire is several wavelengths long. This is These factors are discussed in Chapter Two.
indicated by the curve showing gain in Fig. 5-1. As mentioned a few paragraphs ago, the free-
The longer the antenna, the sharper the lobe space radiation pattern of a long-wire antenna has a
becomes, and since it is really a cone of radiation major lobe that forms a cone around the wire. The
about the wire in free space, it becomes sharper in angle at which maximum radiation takes place
all planes. Also, the greater the length, the smaller becomes smaller, with respect to the wire, as the
the angle with the wire at which the maximum wire length is increased, also shown in Fig. 5-1. For
radiation occurs. this reason a long-wire antenna is primarily a
low-angle radiator when installed horizontally
Directivity above the ground. Its performance in this respect is
improved by selecting a height that also tends to
Because many points along a long wire are concentrate the radiation at low wave angles.
carrying currents in different phase (usually with Antenna systems formed from ordinary hori¬
different current amplitude as well) the field zontal dipoles that are not stacked have, in most
pattern at a distance becomes more complex as the cases, a rather broad vertical pattern; the wave
wire is made longer. This complexity is manifested angle at which the radiation is maximum therefore
in a series of minor lobes, the number of which depends chiefly on the antenna height. However,
increases with the wire length. The intensity of with a long-wire antenna the wave angle at which
radiation from the minor lobes is frequently as the major lobe is maximum can never be higher
great as, and sometimes greater than, the radiation than the angle at which the first null occurs (see
from a half-wave dipole. The energy radiated in the Fig. 5-2), even if the antenna height is very low.
minor lobes is not available to improve the gain in (The efficiency may be less at very small heights,
the major lobe, which is another reason why a partly because of increased losses in the ground
long-wire antenna has to be long to give appre¬ and partly because the pattern is affected in such a
ciable gain. way as to put a greater porportion of the total
Driven and parasitic arrays of the simple types power into the minor lobe.) The result is that when
described in Chapter Four do not have minor lobes radiation at wave angles below 15 or 20 degrees is
of any great consequence. For that reason they under consideration, a long-wire antenna is less
frequently seem to have better directivity than sensitive to height than are the multielement arrays
long-wire antennas, because their response in direc¬ or a simple dipole. To assure good results, however,
tions other than that at which the antenna is aimed the antenna should have a height equivalent to at
is well down. This will be so even if a multielement least a half wavelength at 14 MHz - that is, a
array and a long-wire antenna have the same actual minimum height of about 30 feet. Greater heights
Long-Wire Antennas 167

will give a worthwhile improvement at wave angles range including several amateur bands. The antenna
below 10 degrees. characteristics change very slowly with length,
With an antenna of fixed physical length and except when the wires are short (i.e., around one
height, both length and height increase, in terms of wavelength), and there is no need to try to estab¬
wavelength, as the frequency is increased. The lish exact resonance at a particular frequency.
overall effect is that both the antenna and the The formula for harmonic wires given in Chap¬
ground reflections tend to keep the system opera¬ ter Two is quite satisfactory for determining the
ting at high effectiveness throughout the frequency lengths of any of the antenna systems to be
range. At low frequencies the wave angle is raised, described. For convenience, the formula is re¬
but high wave angles are useful at 3.5 and 7 MHz. peated here in the following form:
At high frequencies the converse is true. Good 984 (N - 0.025)
all-around performance usually results on all bands Length (feet) =
Freq. (MHz)
when the antenna is designed to be optimum in the
14-MHz band. where IV is the number of full waves on the
antenna. In cases where exact resonance is desired
Calculating Length for some reason (for obtaining a resistive load for
In this chapter lengths are always discussed in a transmission line at a particular frequency, for
terms of wavelengths. There can obviously be example) it is best established by trimming the
nothing very critical about wire lengths in an wire length until measurement of the resonant
antenna system that will work over a frequency frequency shows it to be correct.

LONG SINGLE WIRES

In Fig. 5-1 the solid curve shows that the gain exists. Only one of the two lobes is considered in
in decibels of a long wire increases almost linearly this drawing, and its lower half is cut off by the
with the length of the antenna. The gain does not ground. The maximum intensity of radiation in the
become appreciable until the antenna is about four remaining half occurs through the broken-line
wavelengths long, where it is equivalent to doub¬ semicircle; that is, the angle B (between the wire
ling the transmitter power (3 dB). The actual gain direction and the line marked wave direction) is
over a half-wave dipole when the antenna is at a the angle given by Fig. 5-1 for the particular
practical height above ground will depend on the antenna length used.
way in which the radiation resistances of the In the practical case, there will be some wave
long-wire antenna and the comparison dipole are angle (A) that is optimum for the frequency and
affected by the height. The exact way in which the the distance between the transmitter and receiver.
radiation resistance of a long wire varies with Then for that wave angle the wire direction and the
height depends on its length. In general, the optimum geographical direction of transmission are
resistance does not fluctuate as much, in terms of related by the angle C. If the wave angle is very
percentage, as does the resistance of a half-wave low, B and C will be practically equal. But as the
antenna. This is particularly true at heights from wave angle becomes higher the angle C becomes
one-half wavelength up. smaller; in other words, the best direction of
The nulls bounding the lobes in the directive transmission and the direction of the wire more
pattern of a long wire are fairly sharp and are nearly coincide. They coincide exactly when Cis
frequently somewhat obscured, in practice, by zero; that is, when the wave angle is the same as
irregularities in the pattern. The locations of nulls the angle given by Fig. 5-1.
and maxima for antennas up to eight wavelengths The maximum radiation from the antenna can
long are shown in Fig. 5-2. be aligned with a particular geographical direction

Orientation Mwe Direction

The broken curve of Fig. 5-1

shows the angle with the wire at
which the radiation intensity is To distant
maximum. As shown in Chapter \ point
Two, there are two main lobes to ground)
the directive patterns of long-wire
antennas; each makes the same
angle with respect to the wire.
The solid pattern, considered in
free space, is the hollow cone Fig. 5-3 — Showing how the cone
formed by rotating the wire on its of radiation from a long wire results
axis. in maximum radiation at different
When the antenna is mounted wave angles when the angle (C)
horizontally above the ground, between the wire and the distant
the situation depicted in Fig. 5 3 point assumes different values.
168 Chapter 5

make the lobe and wave angle coincide. This is

shown in Fig. 5-4, for the case of a one-wavelength
antenna tilted so that the maximum radiation from
one lobe is horizontal to the left, and from the
other is horizontal to the right (zero wave angle).
The solid pattern can be visualized by imagining
the plane diagram rotating about the antenna as an
axis.
Since the antenna is neither vertical nor hori¬
zontal in this case, the radiation is part horizon¬
tally polarized and part vertically polarized. Com¬
puting the effect of the ground becomes com¬
plicated, because the horizontal and vertical com¬
ponents must be handled separately. In general, the
Fig. 5-4 - Alignment of lobes for horizontal directive pattern at any given wave angle becomes
transmission by tilting a long wire in the vertical unsymmetrical when the antenna is tilted. For
plane.
small amounts of tilt (less than the amount that
directs the lobe angle horizontally) and for low
wave angles the effect is to shift the optimum
at a given wave angle by means of the following direction closer to the line of the antenna. This is
formula: true in the direction in which the antenna slopes
cos B downward. In the opposite direction the low-angle
cos C =- radiation is reduced.
cos A
In most amateur work the chief requirement is that
the wave angle should be as low as possible, Feeding Long Wires
particularly at 14 MHz and above. In such case it is
usually satisfactory to make angle C the same as is It has been pointed out in Chapter Three that a
given by Fig 5-1. harmonic antenna can be fed only at the end or at
It should be borne in mind that only the a current loop. Since a current loop changes to a
maximum point of the lobe is represented in Fig. node when the antenna is operated at any even
5-3. Radiation at higher or lower wave angles in multiple of the frequency for which it is designed,
any given direction will be proportional to the way a long-wire antenna will operate as a true long wire
in which the actual pattern shows the field strength on all bands only when it is fed at the end.
to vary as compared with the maximum point of A common method of feeding is to use a
the lobe. resonant open-wire line, as described in Chapter
Three. This system will work on all bands down to
Tilted Wires the one, if any, at which the antenna is only a half
wave long. Any convenient line length can be used
Fig. 5-3 shows that when the wave angle is if the transmitter is matched to the line input
equal to the angle which the maximum intensity of impedance by the methods described in Chapter
the lobe makes with the wire, the best transmitting Three.
or receiving direction is that of the wire itself. If Two arrangements for using nonresonant lines
the wave angle is less than the lobe angle, the best are given in Fig. 5-5. The one at A is useful for one
direction can be made to coincide with the band only since the matching section must be a
direction of the wire by tilting the wire enough to quarter wave long, approximately, unless a dif¬
ferent matching section is used for each band. In B,
the Q-section impedance should be adjusted to
match the antenna to the line as described in
Chapter Three, using the value of radiation resis¬
tance given in Fig. 2-20. This method is best suited
to working with a 600-ohm transmission line.
Although it will work as designed only on one
band, the antenna can be used on other bands by
treating the line and matching transformer as a

(B)
resonant line. In such case, as mentioned earlier,
the antenna will not radiate as a true long wire on
even multiples of the frequency for which the
matching system is designed.
i Transformer The end-fed arrangement, although the most
convenient when tuned feeders are used, suffers
the disadvantage that there is likely to be a
soon Hon res onant hne considerable antenna current on the line, as des¬
cribed in Chapter Three. In addition, the antenna
reactance changes rapidly with frequency for the
Fig. 5-5 — Methods of feeding long single-wire reasons outlined in Chapter Two (Figs. 2-9 and
antennas.
2-10). Consequently, when the wire is several
 169
Long-Wire Antennas

wavelengths long a relatively small change in between those at which the antenna is exactly
frequency - a fraction of the width of a band - resonant. This leads to a considerable amount of
may require major changes in the adjustment of radiation from the line. The unbalance can be
the transmitter-to-line coupling apparatus. Also, overcome by using two wires in one of the
the line becomes unbalanced at all frequencies arrangements described in succeeding sections.

COMBINATIONS OF RESONANT LONG WIRES

The directivity and gain of long wires may be eliminated, so that the pattern becomes essentially
increased by using two wires so placed in relation bidirectional.
to each other as to make the fields from both The directional pattern of an antenna of this
combine to produce the greatest possible field type is sharper in both the horizontal and vertical
strength at a distant point. The principle is similar planes than the patterns of the individual wires
to that used in forming the multielement arrays composing it. Maximum radiation in both planes is
described in Chapter Four. However, the maximum along the line bisecting the V. There are minor
radiation from a long wire occurs at an angle of less lobes in both the horizontal and vertical patterns
than 90 degrees with respect to the wire, so but if the legs are long in terms of wavelength the
different physical relationships must be used. amplitude of the minor lobes is small. When the
antenna is mounted horizontally above the ground
Parallel Wires (the usual method) the wave angle at which the
radiation from the major lobe is maximum is
One possible method of using two (or more) determined by the height, but cannot exceed the
long wires is to place them in parallel, with a angle values shown in Fig. 5-1 for the leg length
spacing of 1/2 wavelength or so, and feed the two used. Only the minor lobes give high-angle radia¬
in phase. In the direction of the wires the fields tion.
will add up in phase. However, since the wave angle The gain and directivity of a V depend on the
is greatest in the direction of the wire, as shown by length of the legs. An approximate idea of the gain
Fig. 5-3, this method will result in rather high-angle for the V antenna may be obtained by adding 3 dB
radiation unless the wires are several wavelengths to the gain value from Fig. 5-1 for the corres¬
long. The wave angle can be lowered, for a given ponding leg length. The actual gain will be modi¬
antenna length, by tilting the wires as described fied by the mutual impedance between the sides of
earlier. With a parallel arrangement of this sort the the V, and will be somewhat higher than indicated
gain should be about 3 dB over a single wire of the by the values determined as above, especially at the
same length, at spacings in the vicinity of 1/2 longer leg lengths. With 8-wavelength legs, the gain
wavelength. is approximately 4 dB greater than that indicated
for a single wire in Fig. 5-1.
THE V ANTENNA
Lobe Alignment
Instead of using two long wires parallel to each
other they may be placed in the form of a It is possible to align the lobes from the
horizontal V, with the angle at the apex of the V individual wires with a particular wave angle by the
equal to twice the angle given by Fig. 5-1 for the method described in connection with Fig. 5-3. At
particular length of wire used. The currents in the very low wave angles the change in the apex angle
two wires should be out of phase. Under these is extremely small; for example, if the desired wave
conditions the plane directive patterns of the angle is 5 degrees the apex angles of twice the value
individual wires combine as is indicated in Fig. 5-6. given in Fig. 5-1 will not be reduced more than a
Along a line in the plane of the antenna and degree or so, even at the longest leg lengths which
bisecting the V the fields from the individual wires might be used.
reinforce each other at a distant point. The other When the legs are long, alignment does not
pair of lobes in the plane pattern is more or less necessarily mean that the greatest signal strength

Fig. 5-6 — Two long

wires and their respec¬
tive patterns are shown
at the left. If these two
wires are combined to
form a V whose apex *
angle is twice that of
the major lobes of the
wires and the wires are
excited out of phase,
the radiation along the
bisector of the V adds
and the radiation in the
other directions tends
to cancel.
 170
Chapter 5

will be secured at the wave angle for which the a way as to increase the gain of the antenna. For
apex angle is chosen. It must be remembered that example, the optimum apex angle in the case of
the polarization of the radiated field is the same as 1-Ä legs is 90 degrees.
that of a plane containing the wire. As illustrated
by the diagram of Fig. 5-3, at any wave angle other Multiband Design
than zero the plane containing the wire and passing
through the desired wave angle is not horizontal. In When a V antenna is used over a range of
the limiting case where the wave angle and the frequencies - such as 14 to 28 MHz - its
angle of maximum radiation from the wire are the characteristics over the frequency range will not
same the plane is vertical, and the radiation at that change greatly if the legs are sufficiently long at
wave angle is vertically polarized. At in-between the lowest frequency. The apex angle, at zero wave
angles the polarization consists of both horizontal angle, for a 5-wavelength V (each leg approxi¬
and vertical components. mately 350 feet long at 14 MHz) is 44 degrees. At
When two wires are combined into a V the 21 MHz, where the legs are 7.5 wavelengths long,
polarization planes have opposite slopes. In the the optimum angle is 36 degrees, and at 28 MHz
plane bisecting the V, this makes the horizontally where the leg length is 10 wavelengths it is 32
polarized components of the two fields add to¬ degrees. Such an antenna will operate well on all
gether numerically, but the vertically polarized three frequencies if the apex angle is about 35
components are out of phase and cancel com¬ degrees. From Fig. 5-7, a 35-degree apex angle with
pletely. As the wave angle is increased the hori¬ a 5-wavelength V will align the lobes at a wave
zontally polarized components become smaller, so angle of something over 15 degrees, but this is not
the intensity of horizontally polarized radiation too high when it is kept in mind that the maximum
decreases. On the other hand, the vertically polar¬ radiation actually will be at a lower angle. At 28
ized components become more intense but always MHz the apex angle is a little large, but the chief
cancel each other. The overall result is that effect will be a small reduction in gain and a slight
although alignment for a given wave angle will broadening of the horizontal pattern, together with
increase the useful radiation at that angle, the wave a tendency to reduce the wave angle at which the
angle at which maximum radiation occurs (in the radiation is maximum. The same antenna can be
direction of the line bisecting the V) is always used at 3.5 and 7 MHz, and on these bands the fact
below the wave angle for which the wires are that the wave angle is raised is of less consequence,
aligned. As shown by Fig. 5-7, the difference since high wave angles are useful. The gain will be
between the apex angles required for optimum small, however, because the legs are not long at
alignment of the lobes at wave angles of zero and ♦hese frequencies.
15 degrees is rather small, even when the legs are
many wavelengths long. Other V Combinations
For long-distance transmission and reception The gain can be increased about 3 dB by
the lowest possible wave angle usually is the best. stacking two Vs one above the other, a half
Consequently, it is good practice to choose an apex wavelength apart, and feeding them in phase with
angle between the limits represented by the two each other. This will result in a lowered angle of
curves in Fig. 5-7. The actual wave angle at which radiation. The bottom V should be at least a
the radiation is maximum will depend on the shape quarter wavelength above the ground, and prefer¬
of the vertical pattern and the height of the ably a half wavelength.
antenna above ground. Two V antennas can be broadsided to form a
When the leg length is small, there is some W, giving an additional 3-dB gain. However, two
advantage in reducing the apex angle of the V transmission lines are required and this, plus the
because this changes the mutual impedance in such fact that five poles are needed to support the

Fig. 5-7 — Apex angle

of V antenna for align¬
ment of main lobe at
different wave angles,
as a function of leg
length in wavelengths.
Long-Wire Antennas 171

system, renders it normally impractical for the

amateur.
The V antenna can be made unidirectional by
using another V placed an odd multiple of a
quarter wavelength in back of the first and exciting
the two with a phase difference of 90 degrees. The
system will be unidirectional in the direction of the
antenna with the lagging current. However, the V
reflector is not normally employed by amateurs at
low frequencies because it restricts the use to one
band and requires a fairly elaborate supporting
Fig. 5-8 — The resonant rhombic or diamond¬
structure. Stacked Vs with driven reflectors could, shaped antenna. All legs are the same length, and
however, be built for the 200- to 500-MHz region opposite angles of the diamond are equal.
without much difficulty. The overall gain for such
an antenna (two stacked Vs, each with a V
reflector) is about 9 dB greater than the gains given desired to align the lobes from individual wires
in Fig. 5-1. with the wave angle, the curves of Fig. 5-7 may be
used, again using the length of one leg in taking the
Feeding the V data from the curves. The diamond-shaped antenna
The V antenna is most conveniently fed by also can be operated as a nonresonant antenna, as
tuned feeders, since they permit multiband opera¬ described later in this chapter, and much of the
tion. Although the length of the wires in a V beam discussion in that section applies to the resonant
is not at all critical, it is important that both wires rhombic as well.
be of the same electrical length. If it is desired to The direction of maximum radiation with a
use a nonresonant line, probably the most appro¬ resonant rhombic is given by the arrows in Fig. 5-8,
priate matching system is that using a stub or i.e., the antenna is bidirectional. There are minor
quarter-wave matching section. The adjustment is lobes in other directions, their number and inten¬
as described in Chapter Three. sity depending on the leg length. When used at
frequencies below the vhf region the rhombic
antenna is always mounted with the plane contain¬
ing the wires horizontal. The polarization in this
THE RESONANT RHOMBIC ANTENNA plane, and also in the perpendicular plane that
The diamond-shaped or rhombic antenna bisects the rhombic, is horizontal. At 144 MHz and
shown in Fig. 5-8 can be looked upon as two above, the dimensions are such that the antenna
acute-angle Vs placed end-to-end. This arrangement can be mounted with the plane containing the
has two advantages over the simple V that have wires vertical if vertical polarization is desired.
caused it to be favored by amateurs. For the same When the rhombic antenna is to be used on
total wire length it gives somewhat greater gain several amateur bands it is advisable to choose the
than the V; a rhombic 4 wavelengths on a leg, for apex angle, A, on the basis of the leg length in
example, has a gain of better than 1 dB over a V wavelengths at 14 MHz. This point is covered in
antenna with 8 wavelengths on a leg. And the more detail in connection both with the V and
directional pattern of the rhombic is less affected with the nonresonant rhombic. Although the gain
by frequency than the V when the antenna is used on higher frequency bands will not be quite as
over a wide frequency range. This is because a favorable as if the antenna had been designed for
change in frequency causes the major lobe from the higher frequencies, the system will radiate well
one leg to shift in one direction while the lobe at the low angles that are necessary at such
from the opposite leg shifts the other way. This frequencies. At frequencies below the design fre¬
tends to make the optimum direction stay the quency the greater apex angle of the rhombic (as
same over a considerable frequency range. The compared with a V of the same total length) is
disadvantage of the rhombic as compared with the more favorable to good radiation than in the case
V is that one additional support is required. of the V.
The same factors that govern the design of the The resonant rhombic antenna can be fed in the
V antenna apply in the case of the resonant same way as the V. Resonant feeders are necessary
rhombic. The angle A in the drawing is the same as if the antenna is to be used in several amateur
that for a V having a leg length equal to L. If it is bands.

NONRESONANT LONG-WIRE ANTENNAS

All the antenna systems previously considered In Fig. 5-9 let us suppose that the wire is
have been based on resonant operation, that is, parallel with the ground (horizontal) and is termi¬
with standing waves of current and voltage along nated by a load Z equal to its characteristic
the wire. Although most antenna designs are based impedance, Zo. The load Z can represent a receiver
on using resonant wires, resonance is by no means matched to the line. The resistor R is also equal to
a necessary condition for the wire to radiate and the Zo of the wire. A wave coming from the
intercept electromagnetic waves. direction X will strike the wire first at its far end
172
Chapter 5

1/2 wavelength long; the higher the multiple, the

smaller the response.

Directional Characteristics
The explanation above considered the phase
but not the relative amplitudes of the individual
currents reaching the load. When the appropriate
correction is made, the angle with the wire at
which radiation or response is maximum is given
by the curve of Fig. 5-10. The response drops off
gradually on either side of the maximum point,
resulting in lobes in the directive pattern much like
those for harmonic antennas, except that the
system is substantially unidirectional. Typical pat¬
terns are shown in Fig. 5-11. When the antenna
length is 3/2 wavelength or greater there are also
Fig. 5-9 — Nonresonant long-wire antenna. angles at which secondary maxima occur; these
secondary maxima (minor lobes) have their peaks
approximately at angles for which the length AB,
and sweep across the wire at an angle until it Fig. 5-9, is less than AC by any odd multiple of
one-half wavelength. When AB is shorter than AC
reaches the end at which Z is connected. In doing
so it will induce voltages in the antenna and by an even multiple of a half wavelength, the
induced currents cancel each other completely at
currents will flow as a result. The current flowing
Z, and in such cases there is a null for waves
toward Z is the useful output of the antenna, while
the current flowing toward R will be absorbed inarriving in the direction perpendicular to BC.
R. The same thing is true of a wave coming from The antenna of Fig. 5-9 responds to horizon¬
tally polarized signals when mounted horizontally.
the direction X'. In such an antenna there are no
standing waves, because all received power is If the wire lies in a plane that is vertical with
absorbed at either end. respect to the earth it responds to vertically
polarized signals. By reciprocity, the characteristics
The greatest possible power will be delivered to
for transmitting are the same as for receiving. For
the load Z when the individual currents induced as
the wave sweeps across the wire all combine average conductor diameters and heights above
properly on reaching the load. The currents will ground, 20 or 30 feet, the Zo of the antenna is of
the order of 500 to 600 ohms.
reach Z in optimum phase when the time required
for a current to flow from the far end of the It is apparent that an antenna operating in this
way has much the same characteristics as a
antenna to Z is exactly one-half cycle longer than
the time taken by the wave to sweep over the transmission line. When it is properly terminated at
both ends there are traveling waves, but no
antenna. Since a half cycle is equivalent to a half
wavelength in space, the length of the antenna standing waves, on the wire. Consequently the
current is substantially the same all along the wire.
must be one-half wavelength greater than the
distance traversed by the wave from the instant it
strikes the far end of the antenna to the instant
that it reaches the near end. This is shown by the
small drawing, where AC represents the antenna,
BC is a line perpendicular to the wave direction,
and AB is the distance traveled by the wave in
sweeping past AC. AB must be one-half wavelength
shorter than AC. Similarly, AB‘ must be the same
length as AB for a wave arriving from X'.
A wave arriving at the antenna from the
opposite direction Y(or T'), will similarly result in
the largest possible current at the far end. How¬
ever, the far end is terminated in R, which is equal
to Z, so all the power delivered to R by a wave
arriving from Ywill be absorbed in R. The current
traveling to Z will produce a signal in Z in
proportion to its amplitude. If the antenna length
is such that all the individual currents arrive at Zin
such phase as to add up to zero, there will be no
current through Z. At other lengths the resultant
current may reach appreciable values. The lengths LENGTH IN WAVELENGTHS
that give zero amplitude are those which are odd
multiples of 1/4 wavelength, beginning at 3/4 Fig. 5-10 — Angle with respect to wire axis at
wavelength. The response from the Ydirection is which the radiation from a nonresonant long-wire
greatest when the antenna is any even multiple of antenna is maximum.
Long-Wire Antennas 173

Actually, it decreases in the direction in which the

current is flowing because of energy loss by
radiation as well as by ohmic loss in the wire and
the ground. The antenna can be looked upon as a
transmission line terminated in its characteristic
impedance, but having such wide spacing between
conductors (the second conductor in this case is
(A)
the image of the antenna in the ground) that
radiation losses are by no means inconsequential.
A wire terminated in its characteristic imped¬
ance will work on any frequency, but its direc¬
tional characteristics change with frequency as
shown by Fig. 5-10. To give any appreciable gain
over a dipole the wire must be at least a few AnizMA Axis
wavelengths long. The angle at which maximum
response is secured can be in any plane that
contains the wire axis, so in free space the major
lobe will be a cone. In the presence of ground, the
discussion given in connection with Fig. 5-3
applies, with the modification that the angles of
best radiation or response are those given in Fig.
5-10, rather than by Figs. 5-1 or 5-2. As compari¬
son of the curves will show, the difference in
optimum angle between resonant and nonresonant
wires is quite small.

THE NONRESONANT RHOMBIC ANTENNA

The highest development of the long-wire
antenna is the nonresonant rhombic, shown sche¬ Fig. 5-11 — Typical radiation patterns (cross
matically in Fig. 5-12. It consists of four conduc¬ section of solid figure) for terminated long wires.
(A) length two wavelengths, (B) four wavelengths,
tors joined to form a diamond, or rhombus. All both for an idealized case in which there is no
sides of the antenna have the same length and the decrease of current along the wire. In practice, the
opposite corner angles are equal. The antenna can pattern is somewhat distorted by wire attenuation.
be considered as being made up of two V antennas
placed end to end and terminated by a noninduc- Fig. 5-13 shows the tilt angle as a function of
tive resistor to produce a unidirectional pattern. the antenna leg length. The curve marked “0°” is
The terminating resistor is connected between the for a wave angle of zero degrees; that is, maximum
far ends of the two sides, and is made approxi¬ radiation in the plane of the antenna. The other
mately equal to the characteristic impedance of the curves show the proper tilt angles to use when
antenna as a unit. The rhombic may be constructed aligning the major lobe with a desired wave angle.
either horizontally or vertically, but practically For a wave angle of 5 degrees the difference in tilt
always is horizontal at frequencies below 54 MHz, angle is less than one degree for the range of
since the pole height required is considerably less.
lengths shown. Just as in the case of the resonant V
Also, horizontal polarization is equally, if not and resonant rhombic, alignment of the wave angle
more, satisfactory at these frequencies. and lobes always results in still greater radiation at
Tire basic principle of combining lobes of
a lower wave angle, and for the same reason, but
maximum radiation from the four individual wires
also results in the greatest possible radiation at the
constituting the rhombus or diamond is the same
in either the nonresonant type shown in Fig. 5-12, desired wave angle.
The broken curve marked “Optimum Length”
or the resonant type described earlier in this
shows the leg length at which maximum gain is
chapter. The included angles should differ slightly
because of the differences between resonant and
nonresonant wires, as just described, but the
differences are almost negligible.

Tilt Angle
It is a matter of custom, in dealing with the
nonresonant or terminated rhombic, to talk about
the “tilt angle” (0 in Fig. 5-12), rather than the
angle of maximum radiation with respect to an
individual wire. The tilt angle is simply 90 degrees
minus the angle of maximum radiation. In the case
of a rhombic antenna designed for zero wave angle
the tilt angle is 90 degrees minus the values given in
Fig. 5-12 - The nonresonant rhombic antenna.
Fig. 5-10.
 174
Chapter 5
1 5|-

__—
—
_ — —

-•
OPT MUM LENG
z

✓ /
? / y

« WAVE
JANGLE
TILT

4$

IS 2 25 3 15 4 45 5
LEG LENGTH IN WAVELENGTHS

Fig, 5-13 — Rhombic-antenna design chart. For any given leg length, the curves show the proper tilt
angle to give maximum radiation at the selected wave angle. The broken curve marked "Optimum
Length shows the leg length that gives the maximum possible output at the selected wave angle The
optimum length as given by the curves should be multiplied by 0.74 to obtain the leg length for which
the wave angle and main lobe are aligned (see text, "Alignment of Lobes").

secured at a chosen wave angle. Increasing the leg For example, suppose the antenna is to be used at
length beyond the optimum will result in lessened frequencies up to and including the 28 MHz band,
gain, and for that reason the curves do not extend and that the leg length is to be 6 wavelengths on
beyond the optimum length. Note that the opti¬ the latter frequency. For zero wave angle the
mum length becomes greater as the desired wave optimum tilt angle is 68 degrees, and the calculated
angle is smaller. Leg lengths over 6X are not free-space directive pattern in the vertical plane
recommended because the directive pattern bisecting the antenna is shown in the right-hand
becomes so sharp that the antenna performance is drawing of Fig. 5-14. At 14 MHz this same antenna
highly variable with small changes in the angle, has a leg length of three wavelengths, which calls
both horizontal and vertical, at which an incoming for a tilt angle of 58.5 degrees for maximum
wave reaches the antenna. Since these angles vary radiation at zero wave angle. The calculated pat¬
to some extent in ionospheric propagation, it does terns for tilt angles of 58.5 and 68 degrees are
not pay to attempt to use too great a degree of shown in the left-hand drawing in Fig. 5-14, and it
directivity. is seen that if the optimum tilt for 28-MHz
operation is used the gain will be reduced and the
wave angle raised at 14 MHz. In an attempt at a
Multiband Design compromise, we might select a wave angle of 15
degrees, rather than zero, for 14 MHz since, as
When a rhombic antenna is to be used over a shown by Fig. 5-13, the tilt angle is larger and thus
considerable frequency range it is worth paying more nearly coincides with the tilt angle for zero
some attention to the effect of the tilt angle on the wave angle on 28 MHz. From the chart, the tilt
gain and directive pattern at various frequencies. angle for 3 wavelengths on a leg and a 15-degree
Long-Wire Antennas 175
determined. (Also see later discussion on lobe
alignment.)

Power Gain
The theoretical power gain of a nonresonant
rhombic antenna over a dipole, both in free space,
is given by the curve of Fig. 5-16. This curve is for
zero wave angle and includes an allowance of 3 dB
for power dissipated in the terminating resistor.
The actual gain of an antenna mounted horizon¬
tally above the ground, as compared with a dipole
at the same height, can be expected to vary a bit
Fig. 5-14 — Showing the effect of tilt angle on the either way from the figures given by the curve. The
free-space vertical pattern of a nonresonant rhom¬ power lost in the terminating resistor is probably
bic antenna having a leg length of 3 wavelengths at less than 3 dB in the average installation, since
one frequency and 6 wavelengths at twice the
frequency. These patterns apply only in the direc¬ more than half of the power input is radiated
tion of the antenna axis. before the end of the antenna is reached. However,
there is also more power loss in the wire and in the
wave angle is 61.5 degrees. The patterns with this ground under the antenna than in the case of a
tilt angle axe shown in Fig. 5-14 for both the 14- simple dipole, so the 3 dB figure is probably a
and 28-MHz cases. The effect at 28 MHz is to representative estimate of overall loss.
decrease the gain at zero wave angle by more than
6 dB and to split the radiation in the vertical plane
into two lobes, one of which is at a wave angle too
high to be useful at this frequency.
Inasmuch as the gain increases with the leg
length in wavelengths, it is probably better to favor
the lower frequency in choosing the tilt angle. In
the present example, the best compromise proba¬
bly would be to split the difference between the
optimum tilt angle for the 15-degree wave angle at
14 MHz and that for zero wave angle at 28 MHz;
that is, use a tilt angle of about 64 degrees. Design
dimensions for such an antenna are given in Fig.
5-15.
The patterns of Fig. 5-14 are in the vertical
plane through the center of the antenna only. In
vertical planes making an angle with the antenna
axis, the patterns may differ considerably. The
effect of a tilt angle that is smaller than the
optimum is to broaden the horizontal pattern, so
at 28 MHz the antenna in the example would be Fig. 5-16 - Theoretical gain of a nonresonant
less directive in the horizontal plane than would be rhombic antenna over a half-wave dipole in free
the case if it were designed for optimum perfor¬ space. This curve includes an allowance of 3 dB for
mance at that frequency. It should also be noted loss in the terminating resistor.
that the patterns given in Fig. 5-14 are free-space
patterns and must be multiplied by the ground¬
reflection factors for the actual antenna height Termination
used, if the actual vertical patterns are to be Although there is no marked difference in the
gain obtainable with resonant and nonresonant
rhombics of comparable design, the nonresonant
antenna has the advantage that over a wide
frequency range it presents an essentially resistive
and constant load to the transmitter coupling
apparatus. In addition, nonresonant operation
makes the antenna substantially unidirectional,
while the unterminated or resonant rhombic is
always bidirectional, although not symmetrically
so. In a sense, it can be considered that the power
dissipated in the terminating resistor is simply
power that would have been radiated in the other
direction nad the resistor not been there, so the
Fig. 5-15 — Rhombic antenna dimensions for a fact that some of the power (about one-third) is
compromise design between 20- and 10-meter
requirements, as discussed in the text. The leg used up in heating the resistor does not mean an
length is 6 X on 10 meters, 3 X on 20. actual loss in the desired direction.
 176
Chapter 5

The characteristic impedance of an ordinary for adjustment rather than at the top of the pole.
rhombic antenna, looking into the input end, is of Resistance wire may be used for this line, so that a
the order of 700 to 800 ohms when properly portion of the power will be dissipated before it
terminated in a resistance at the far end. The reaches the resistive termination, thus permitting
terminating resistance required to bring about the the use of lower wattage lumped resistors. The line
matching condition usually is slightly higher than length is not critical, since it operates without
the input impedance because of the loss of energy standing waves and hence is nonresonant.
through radiation by the time the far end is
reached. The correct value usually will be found to Multiwire Rhombics
be of the order of 800 ohms, and should be
determined experimentally if the flattest possible The input impedance of a rhombic antenna
antenna is desired. However, for average work a constructed as in Fig. 5-12 is not quite constant as
noninductive resistance of 800 ohms can be used the frequency is varied. This is because the varying
with the assurance that the operation will not be separation between the wires causes the character¬
far from optimum. istic impedance of the antenna to vary along its
The terminating resistor must be practically a length. The variation in Zo can be minimized by a
pure resistance at the operating frequencies; that is, conductor arrangement that increases the capaci¬
its inductance and capacitance should be negligible. tance per unit length in proportion to the separa¬
Ordinary wire-wound resistors are not suitable tion between the wires.
because they have far too much inductance and The method of accomplishing this is shown in
distributed capacitance. Small carbon resistors have Fig. 5-17. Three conductors are used, joined
satisfactory electrical characteristics but will not together at the ends but with increasing separation
dissipate more than a few watts and so cannot be as the junction between legs is approached. As used
used, except when the transmitter power does not in commercial installations having legs several
exceed 10 or 20 watts or when the antenna is to be wavelengths long, the spacing between wires at the
used for reception only. The special resistors center is 3 to 4 feet. Since all three wires should
designed either for use as “dummy” antennas or have the same length, the top and bottom wires
for terminating rhombic antennas should be used will be slightly farther from the support than the
in other cases. To allow a factor of safety, the total middle wire. Using three wires in this way reduces
rated power dissipation of the resistor or resistors the Zo of the antenna to approximately 600 ohms,
should be equal to half the power output of the thus providing a better match for a practicable
transmitter. open-wire line, in addition to smoothing out the
To reduce the effects of stray capacitance it is impedance variations over the frequency range.
desirable to use several units, say three, in series A similar effect, although not quite so good, is
even when one alone will safely dissipate the obtained by using two wires instead of three. It has
power. The two outer units should be identical and been found that, with the 3-wire system, the gain
each should have one fourth to one third the total of the antenna is slightly greater (of the order of 1
resistance, with the center unit making up the dB) than when only a single conductor is used.
difference. The units should be installed in a
weatherproof housing at the end of the antenna to Front-to-Back Ratio
protect them and to permit mounting without It is theoretically possible to obtain an infinite
mechanical strain. The connecting leads should be front-to-back ratio with a terminated rhombic
short so that little extraneous inductance is intro¬ antenna, and in practice very large values can
duced. actually be secured. However, when the antenna is
Alternatively, the terminating resistance may be terminated in its characteristic impedance the
placed at the end of an 800-ohm line connected to infinite front-to-back ratio can be secured only at
the end of the antenna. This will permit placing the frequencies for which the leg length is an odd
resistors and their housing at a point convenient multiple of a quarter wavelength, as described in
the section on nonresonant long wires. The front-
to-back ratio is smallest at frequencies for which
the leg length is a multiple of a half wavelength.
When the leg length is not an odd multiple of a
quarter wave at the frequency under consideration,
the front-to-back ratio can be made very high by
slightly decreasing the value of terminating resis¬
tance. This permits a small reflection from the far
end of the antenna which cancels out, at the input
end, the residual response. With large antennas the
front-to-back ratio may be made very large over
the whole frequency range by experimental adjust¬
ment of the terminating resistance. Modification of
the terminating resistance can result in a splitting
Fig. 5-17 — Three-wire rhombic antenna. Use of of the back null into two nulls, one on either side
multiple wires improves the impedance characteris¬ of a small lobe in the back direction. Changes in
tic of a nonresonant rhombic and increases the gain the value of terminating resistance thus permit
somewhat.
“steering” the back null over a small horizontal
Long-Wire Antennas 177

range so that signals coming from a particular spot

not exactly to the rear of the antenna may be
minimized.

Ground Effects
Reflections from the ground play exactly the
same part in determining the vertical directive
pattern of a horizontal rhombic antenna that they
play with other horizontal antennas. Consequently,
if a low wave angle is desired it is necessary to
make the height great enough to bring the reflec¬
tion factor into the higher range of values given by
the charts in Chapter Two.

Alignment of Lobes, Wave Angle, and

Ground Reflections
When maximum antenna response is desired at
a particular wave angle (or maximum radiation is
desired at that angle) the major lobe of the antenna
cannot only be aligned with the wave angle as Fig. 5-18 - Antenna height to be used for securing
previously described but also with a maximum in maximum radiation at a desired wave angle. This
the ground-reflection factor. When this is done it is curve applies to any type of horizontal antenna.
no longer possible to consider the antenna height
independently of other aspects of rhombic design.
The wave angle, leg length, and height become radiation at 20 degrees. The reduction in gain is
mutually dependent. about 1.5 dB.
This method of design is of particular value
Methods of Feed
when the antenna is built to be used over fixed
transmission distances for which the optimum If the broad frequency characteristic of the
wave angle is known. It has had wide application in rhombic antenna is to be fully utilized the feeder
commercial work with nonresonant rhombic anten¬ system used with it must be similarly broad. This
nas, but seems less desirable for amateur use where, practically dictates the use of a transmission line of
for the long-distance work for which rhombic the same characteristic impedance as that shown at
antennas are built, the lowest wave angle that can the antenna input terminals, or approximately 700
be obtained is the most desirable. Alignment of all to 800 ohms. Data for the construction of such
three factors is limited in application because it lines will be found in Chapter Three.
leads to impracticable heights and leg lengths for While the usual matching stub can be used to
small wave angles. Consequently, when a fairly provide an impedance transformation to more
broad range of low wave angles is the objective it is satisfactory line impedances, this limits the opera¬
more satisfactory to design for a low wave angle tion of the antenna to a comparatively narrow
and simply make the antenna as high as possible. range of frequencies centering about that for which
Fig. 5-18 shows the lowest height at which the stub is adjusted. On the whole, the best plan is
ground reflections make the radiation maximum at to connect a 600-ohm line directly to the antenna
a desired wave angle. It can be used in conjunction and accept the small mismatch which results. The
with Fig. 5-13 for complete alignment of the operation of the antenna will not be adversely
antenna. For example, if the desired wave angle is affected, and since the standing-wave ratio is quite
20 degrees, Fig. 5-18 shows that the height must be low (1.33 to 1) the additional loss over the
0.75 wavelength. From Fig. 5-13, the optimum leg perfectly matched condition will be inappreciable
length is 4.2 wavelengths and the tilt angle is just even for rather long lines.
under 70 degrees. A rhombic antenna so designed Bibliography
will have the maximum possible output that can be
obtained at a wave angle of 20 degrees; no other Source material and more extended discussion
set of dimensions will be as good. However, it will of topics covered in this chapter can be found in
have still greater output at some angle lower than the references given below.
20 degrees, for the reasons given earlier. When it is Bruce, “Developments in Short-Wave Directive
desired to make the maximum output of the Antennas,” Proc. I.R.E., August, 1931.
antenna occur at the 20-degree wave angle, it may Bruce, Beck and Lowry, “Horizontal Rhombic
Antennas,” Proc. I.R.E., January, 193S.
be accomplished by using the same height and tilt Carter, Hansell and Lindenblad, “Development of
angle, but with the leg length reduced by 26 Directive Transmitting Antennas by R.C.A.
percent. Thus for such alignment the leg length Communications,” Proc. I.R.E., October, 1931.
should be 4.2 X 0.74 = 3.1 wavelengths. However, Harper, Rhombic Antenna Design, D. Van Nos¬
the output at the 20-degree wave angle will be trand Co., Inc., New York, N. Y.
smaller than with 4.2-wavelength legs, despite the Laport, “Design Data for Horizontal Rhombic
fact that the smaller antenna has its maximum Antennas,” RCAReview, March, 19S2.
 Chapter 6

Multiband Antennas

For operation in a number of bands such as much as one-quarter wavelength, and still be a very
those between 3.5 and 30 MHz it would be efficient radiator at the lowest frequency.
impracticable, for most amateurs, to put up a In addition, methods have been devised for
separate antenna for each band. Nor is it necessary; making a single antenna structure operate on a
a dipole, cut for the lowest frequency band to be number of bands while still offering a good match
used, readily can be operated on higher frequencies to a transmission line, usually of the còaxial type.
if one is willing to accept the fact that such It should be understood, however, that a “multi¬
harmonic-type operation leads to a change in the band antenna” is not necessarily one that will
directional pattern of the antenna (see Chapter match a given line on all bands on which it is
Two), and if one is willing to use “tuned” feeders. intended to be used. Even a relatively short whip
A center-fed single-wire antenna can be made to type of antenna can be operated as a multiband
accept power and radiate it with high efficiency on antenna with suitable loading, which may be in the
any frequency higher than its fundamental reso¬ form of a coil at its base on those frequencies
nant frequency and, with some reduction in where loading is needed, or which may be incor¬
efficiency and bandwidth, on frequencies as low as porated in the tuned feeders which run from the
one half the fundamental. transmitter to the base of the antenna.
In fact, it is not necessary for an antenna to be This chapter describes a number of systems that
a full half-wavelength long at the lowest frequency. can be used on two or more bands. Beam antennas
It has been determined that an antenna can be are not included, these being treated separately in
considerably shorter than a half wavelength, as later chapters.

DIRECTLY FED ANTENNAS

The simplest multiband antenna is a random
length of No. 12 or No. 14 wire. Power can be fed
to the wire on practically any frequency by one or
the other of the methods shown in Fig. 6-1. If the
wire is made either 67 or 135 feet long, it can
also be fed through a tuned circuit as in Fig. 6-2. It
is advantageous to use an SWR bridge or other
indicator in the coax line at the point marked X.
If a 28- or 50-MHz rotary beam has been
installed, in many cases it will be possible to use
the beam’s feed line as an antenna on the lower
frequencies. Connecting the two wires of the
feeder together at the station end will give a
random-length wire that can be conveniently
coupled to the transmitter as in Fig. 6-1. The
rotary system at the far end will serve only to “end
load” the wire and will not have much other effect.
One disadvantage of all such directly fed
systems is that part of the antenna is practically
Fig. 6-1 — (A) Random-length wire driven directly within the station, and there is a good chance that
from the pi-network output of a transmitter. (B> L
network for use in cases where sufficient loading trouble with rf feedback will be encountered. The
cannot be obtained with (A). C1 should have about rf within the station can often be minimized by
the same plate spacing as the final tank capacitor; a choosing a length of wire so that a current loop
maximum capacitance of 100 pF is sufficient if L1 occurs at or near the transmitter. This means using
is 20 to 25 pH. A suitable coil would consist of 30 a wire length of a quarter wavelength (65 feet at 80
turns of No. 12, 2-1/2 inches in diameter, 6 turns meters, 33 feet at 40 meters), or an odd multiple
per inch. Bare wire should be used so the tap can of a quarter wavelength (3/4 wavelength is 195 feet
be placed as required for loading transmitter. at 80 meters, 100 feet at 40 meters). Obviously

178
Multiband Antennas 179

portion becomes fairly critical if radiation from the

line is to be held to a minimum. Such an antenna
system for multiband operation is the “end-fed” or
“Zepp-fed” antenna shown in Fig. 6-3. 1 he anten¬
na length is made a half wavelength at the lowest
operating frequency. The feeder length can be
anything that is convenient, but feeder lengths that
XMTR are multiples of a quarter wavelength generally'give
X trouble with parallel currents and radiation from
the feeder portion of the system. The feeder can be
an open-wire Une of No. 14 solid copper wire
spaced 4 or 6 inches with ceramic or plastic
spacers. Open-wire TV line (not the type with a
Fig. 6-2 — If the antenna length is 135 feet, a solid web of dielectric) is a convenient type to use.
parallel-tuned coupling circuit can be used on each This type of line is available in approximately 300-
amateur band from 3.5 through 30 MHz. C1 and 450-ohm characteristic impedance.
should duplicate the final-tank tuning capacitor If one has room for only a 67-foot flat top and
and L1 should have the same dimensions as the yet wants to operate in the 3.5-MHz band, the two
final-tank inductor on the band being used. If the feeder wires can be tied together at the transmitter
wire is 67 feet long, series tuning can be used on end and the entire system treated as a random¬
3.5 MHz as shown at the left; parallel tuning will length wire fed directly, as in Fig. 6-1.
be required on 7 MHz and higher frequency bands. The simplest insurance against parallel currents
C2 and L2 will in general duplicate the final tank
tuning capacitor and inductor, the same as with which could cause feed-line radiation is to use a
parallel tuning. The L network shown in Fig. 6-1B feeder length that is not a multiple of a quarter
is also suitable for these antenna lengths. wavelength, e.g., a 40-foot length would be suitable
for all amateur frequencies between 3.5 and 30
this can be done for only one band in the case of
even harmonically related bands, since the wire
length that gives a current loop at the transmitter
will give a voltage loop at two (or four) times that
frequency.
When operating with a random-wire antenna, as
in Figs. 6-1 and 6-2, it is wise to try different types
of grounds on the various bands, to see which will
give the best results. In many cases it will be
satisfactory to return to the transmitter chassis for
the ground, or directly to a convenient water pipe.
If neither of these works well (or the water pipe is Fig. 6-4 — A center-fed antenna system for
not available), a length of No. 12 or No. 14 wire multiband use.
(approximately 1/4 wavelength long) can often be
used to good advantage. Connect the wire at the MHz. A Transmatch (described in a later section of
point in the circuit that is shown grounded, and this chapter) can be used to provide multiband
run it out and down the side of the house, or coverage with an end-fed antenna with any length
support it a few feet above the ground if the of open-wire feed line, as shown in Fig. 6-3.
station is on the first floor or in the basement. It
should not be connected to actual ground at any CENTER-FED ANTENNAS
point.
The simplest and most flexible (and also most
END-FED ANTENNAS inexpensive) all-band antennas are those using
open-wire parallel-conductor feeders to the center
When a straight-wire antenna is fed at one end of the antenna, as in Fig. 6-4. Because each half of
by a two-wire line, the length of the antenna the flat top is the same length, the feeder currents
will be balanced at all frequencies unless, of course,
unbalance is introduced by one half of the antenna
being closer to ground (or a grounded object) than
the other. For best results and to maintain feed¬
current balance, the feeder should run away at
right angles to the antenna, preferably for at least a
quarter wavelength.
Center feed is not only more desirable than end
feed because of inherently better balance, but
generally also results in a lower standing-wave ratio
on the transmission line, provided a parallel¬
conductor line having a characteristic impedance of
450 to 600 ohms is used. TV-type open-wire Une is
Fig. 6-3 — An end fed antenna for multiband use. satisfactory for all but possibly high-power installa-
180 Chapter 6

tions (over 500 watts), where heavier wire and that an amateur is likely to encounter. The unit
wider spacing is desirable to handle the larger can be used with either open-wire feeders, balanced
currents and voltages. lines, coaxial lines, or even an end-fed single wire.
The length of the antenna is not critical nor is Frequency range of the unit is from 3 to 30 MHz,
the length of the line. As mentioned earlier, the accomplished without the use of bandswitching.
length of the antenna can be considerably less than Basically, the circuit is designed for use with
one-half wavelength and still be very effective. If unbalanced lines, such as coax. For balanced Unes,
the overall length is at least one-quarter wavelength a 1:4 (unbalanced-to-balanced) balun, Fig. 6-8, is
at the lowest frequency, a quite usable system will connected to the output of the Transmatch.
result. The only difficulty that may exist with this The chassis used for the Transmatch is made of
type of system is the matter of coupling the a 16 X 25-inch sheet of aluminum. When bent to
antenna-system load to the transmitter. Most form a U, the completed chassis measures 16x13
modern transmitters are designed to work into a X 6 inches. When mounting the variable capacitors,
50- to 70-ohm coaxial load. With this type of the roller inductor and the balun, allow at least
antenna system a coupling network such as a 1/2-inch clearance to the chassis and adjoining
Transmatch is required. components. The capacitors should be mounted on
standoff insulators. The balun can be mounted on
FEED-LINE RADIATION a cone insulator or piece of Plexiglas.
The preceding sections have pointed out means Amidon type T-200-2 cores are used in making
of reducing or eliminating feed-line radiation. the balun. The balun requires three ferrite cores
However, it should be emphasized that any radia¬ stacked for 2-kW or two cores for 1-kW power
tion from a transmission line is not “lost” energy levels. Each core should be covered with two layers
and is not necessarily harmful. Whether or not of 3M No. 27 glass-cloth insulating tape. Next, the
feed-line radiation is important depends entirely on cores are stacked and covered with another layer of
the antenna system being used. For example, the tape. The winding consists of 13 bifilar turns of
feed-line radiation is not desirable when a directive No. 14 or No. 16 Teflon-covered wire. Approxi¬
array is being used. Such radiation can distort the mately 20 feet of wire (two 10-foot lengths) are
desired pattern of such an array, producing required.
responses in unwanted directions. In other words, An SWR indicator is necessary for adjusting the
Transmatch during operation. Construction of such
one wants radiation only from the directive array.
On the other hand, in the case of a multiband indicators is described in Chapter Seventeen. In the
Transmatch pictured (Fig. 6-6), an indicator is
dipole where general coverage is desired, if the feed
built into the enclosure.
line happens to radiate, such energy could actually
have a desirable effect. Antenna purists may For coax-to-coax feeder matching, the antenna
dispute such a premise, but from a practical feed line should be connected to J2, Fig. 6-7. Cl
standpoint where one is not concerned with a and C2 should be set at maximum capacitance and
directive pattern, much time and labor can be power applied to the transmitter. The SWR indica¬
saved by ignoring possible transmission-line radia¬ tor should be switched to read reflected power.
tion. Then adjust L3 until there is a drop in the
reflected reading. Cl and C2 should then be reset,
FEEDER-TO-TRANSMITTER along with L3, until a perfect match is obtained. It
MATCHING NETWORKS will be found that with many antenna systems,
several different matching combinations can be
The Transmatch shown in Figs. 6-5, 6-6, and obtained. Always use the matching setting that
6-7 is designed to handle practically any mismatch uses the most capacitance from Cl and C2, as

Fig. 6-5 — The universal Trans¬

match shown here will couple a
transmitter to almost any antenna
system. The meter is part of an
optional built in SWR indicator.
Suitable indicators for inclusion,
if the builder desires, are given in
Chapter Seventeen.
Multiband Antennas 181

Fig. 6-6 — Interior view of the

Transmatch. Both C1 and C2
must be mounted on insulated
standoffs and insulated shaft
couplers used between the capaci¬
tors and the panel knobs. Like¬
wise, T1 should be installed on an
insulated mounting. An isolantite
cone is used in the unit shown
(the balun could be mounted on a
piece of Plexiglas). Feedthrough
isolantite insulators, mounted
through the rear deck, are used
for the antenna connectors. The
components at the left are those
of the SWR indicator.

maximum C provides the best harmonic attenua¬ handle them. However, any antenna system can be
tion. modified, quite simply, to put the system within
End-fed wires should be connected to J3. Use the matching limits of the Matchbox.
the same adjustment procedures for setting up the When one installs a multiband antenna system
Transmatch as outlined above. For balanced feed¬ of any type, a complex load may exist at the input
ers, the feed line should be connected to J4 and J5, end of the feed line. Depending on the frequency
and a jumper must be connected between J3 and in use and the line length, this load can be very
J4 (see Fig. 6-7 at C). high or very low impedance, or anywhere in
A slight modification will permit this Trans¬ between. If the impedance happens to be outside
match to be used on the 160-meter band. Fixed the range of the Matchbox, it will be impossible to
capacitors, 100 pF each (Centralab type obtain a match. Changing the feed-line length will
85OS-1OON), can be installed across each of the change the line input impedance, and thus may
stator sections of Cl, providing sufficient C to tune bring it into the range of the Matchbox. The goal is
to 1.8 MHz. But, the fixed capacitors must be to obtain a length that will put the antenna system
removed when using the Transmatch on the hf within the matching range of the unit. If, for
bands. example, one finds that a match is possible on 80
meters but not on 40, add or remove some feed
Commercial Matching Networks line and recheck the matching on both bands. It is
The Transmatch shown in Fig. 6-7 will match possible with experimentation to find a feed-line
length that will provide a match on most bands.
any antenna system. However, some commercial
types have limited matching ranges and amateurs If it is inconvenient to alter the length of the
encounter problems in matching multiband sys¬ feeder, a match may still be obtained in most cases
tems. As an example, one such unit, which is very
popular in amateur use, is the Johnson Matchbox.
The Matchbox uses a very efficient circuit but “if- COAX TO
ANTENNA
Ji
has limited matching ranges. The 250-watt version
will match loads from approximately 12 ohms to
2500 ohms, and the kilowatt unit will handle loads
from 25 to 1200 ohms, 80 through 10 meters. TRANS.
Many antenna systems present loads which fall
outside these ranges and the Matchbox will not
TO CZ

Fig. 6-7 — Circuit diagram of the Transmatch. (B)

C1 — Dual-section or air variable, 200 pF per
section (E. F. Johnson 154-507 or Millen
16250). TO C2
C2 — Air variable 350 pF, (E. F. Johnson 154-10
or Millen 16520A). (C)
J1, J2 — Coax chassis connector, type SO-239. JUMPER
BALANCED
J3, J4, J5 — Isolantite feedthrough insulators. FEEDERS
L1 — Roller inductor, 28 pH (E. F. Johnson
229-203).
T1 — Balun transformer; see text and Fig. 6-8.
 182 Chapter 6

or so of line usually will help cool things off in

such cases.
A newer version of the off-center-fed antenna
(miscalled “Windom”) uses 300-ohin TV Twin-
Lead instead of a single-wire line. This system is
shown in Fig. 6-9B. The claim has been made that
the 300-ohm line is matched by the antenna
impedance at the connection point both on the
Fig. 6-8 — Details of the balun bifilar windings for antenna’s fundamental frequency and on harmon¬
the Transmatch. The drawing shows the connectors ics, but there is little theoretical justification for
required. In the actual balun, the turns should be this. The system is particularly susceptible to
close spaced on the inside of the core and spread parallel line currents because of the unsymmetrical
evenly on the outside. feeder connection, and probably in many cases the
line acts more like a single-wire feeder than a
parallel-conductor one. The parallel currents on the
by connecting a lumped-value reactance directly line can be choked off by using balun coils (see
across the line-input terminals, at the point where
they are connected to the matching network. Quite
often the line-input impedance will contain a
high-value reactive component, either inductive or
capacitive, along with the resistive component.
Connecting a reactance of the opposite type and
proper value will cancel the high reactance, and
may thereby put the resultant impedance within
the matching range of the network. The type and
value of the reactance for each amateur band must
be determined experimentally. Since pure reac¬
tances consume no power, efficiency of the
antenna system will be affected only by the losses
in the reactive components. Therefore, high-Ç coils
and low-loss capacitors, such as transmitting mica
or ceramic types, are preferred. Quite high rf
voltages may be developed across these added
components, so they should be selected and
connected with this possibility in mind.

OFF-CENTER FEED-THE
"WINDOM"ANTENNA
A multiband antenna that enjoyed considerable
popularity in the 1930s is the “off-center feed” or
“Windom,” named after the amateur who wrote a
comprehensive article about it. Shown in Fig.
6-9A, it consists of a half-wavelength antenna on
the lowest-frequency band to be used, with a
single-wire feeder connected off center as shown.
The antenna will operate satisfactorily on the
even-harmonic frequencies, and thus a single
antenna can be made to serve on the 80-, 40-, 20-
Fig. 6-9 — Two versions of the off-center-fed
and 10-meter bands. The single-wire feeder shows antenna. (A) Single-wire feed. The single-wire
an impedance of approximately 600 ohms to feeder can be connected directly to the "hot"
ground, and since the return circuit for the feed output terminal of a pi network in the transmitter.
system is through the earth, a good ground Alternatively, the link-coupled circuit shown may
connection is important to the effective operation be used with a separate ground connection as
of the antenna. Also, the system works best when indicated; this type of coupling helps reduce
installed over ground having high conductivity. troubles from rf currents on the station equipment.
The matching circuit described earlier in this
Theoretically the single-wire feeder can be any chapter also can be used, with unbalanced output.
convenient length, since its characteristic imped¬ Both circuits should be adjusted as described for
ance is matched by the antenna impedance at the the tapped matching circuit. (B) Two-wire feed,
point where the feeder is connected. However, this using 300-ohm TV line. The balun coils may be
type of feeder is susceptible to parallel-type cur¬ omitted and the 300-ohm line connected directly
rents just as much as the two-conductor type (see to the output terminals of a pi network in the
Chapter Three), and some feeder lengths will lead transmitter, but this is not recommended because
to “rf in the shack” troubles, especially when the it leads to rf troubles of the type described in the
feeder goes directly to a pi network in the text. The matching circuit shown earlier in this
chapter may be substituted for the balun coils if
transmitter. Adding or subtracting 1/8 wavelength desired.
Multiband Antennas 183

Fig. 6-10 — Multiband an¬

tenna using paralleled di
poles all connected to a
common low-impedance
transmission line. The half¬
wave dimensions may be
either for the centers of the
various bands or selected to
fit favorite frequencies in
each band. Length of half
wave in feet is 468/fre-
quency in MHz.

Chapter Three) as shown in Fig. 6-9B. The same dipoles. Thus, two lengths, as shown in the sketch,
balun will transform the impedance to 75 ohms, in serve to make dipoles for four bands. Ribbon with
cases where the line actually shows a resistive input copper-clad steel conductors (Amphenol type
impedance of 300 ohms. 14-022) should be used because all of the weight,
With either of the off-center-fed systems the including that of the feed line, must be supported
feeder should be brought away from the antenna at by the uppermost wire.
right-angles for at least a quarter wavelength before Two pieces of ribbon are first cut to a length
any bends are made. Any necessary bends should suitable for the two halves of the longest dipole.
be made gradually. Then one of the conductors in each piece is cut to
proper length for the next band higher in fre¬
MULTIPLE-DIPOLE ANTENNAS quency. The excess wire and insulation is stripped
away. A second pair of lengths is prepared in the
The antenna system shown in Fig. 6-10 consists
same manner, except that the lengths are appropri¬
of a group of center-fed dipoles all connected in
ate for the next two higher frequency bands.
parallel at the point where the transmission line
A piece of thick polystyrene sheet drilled with
joins them. One such dipole is used for each band
holes for anchoring each wire serves as the central
on which it is desired to work, and as many as four
insulator. The shorter pair of dipoles is suspended
have been used, as indicated in the sketch. It is not
the width of the ribbon below the longer pair by
generally necessary to provide a separate dipole for
clamps also made of poly sheet. Intermediate
the 21-MHz band since a 7-MHz dipole works
spacers are made by sawing slots in pieces of poly
satisfactorily as a third-harmonic antenna on this
sheet so that they will fit the ribbon snugly.
band. Dimensions, as determined by use of a grid-dip
Although there is some interaction between the
oscillator, are shown in Table 6-1. The system
dipoles it has been found in practice that the ones
showed an impedance of close to 70 ohms on all
that are not resonant at the frequency actually
bands, and the SWR on a.75-ohm line was low and
applied to the antenna have only a small effect on
nearly constant. However, 50-ohm-impedance
the feed-point impedance of the “active” dipole.
cable can be used with only a slight difference in
This impedance is therefore approximately that of
SWR.
a single dipole, or in the neighborhood of 60-70
ohms, and the system can be fed through a 50- or
75-ohm line with a satisfactorily low standing-wave TABLE 6-1
ratio on the line.
Since the antenna system is balanced, it is Twin-Lead Parallel-Dipole Antenna Dimensions
desirable to use a balanced transmission line to Frequency Length Each Half
feed it. The most desirable type of line is 75-ohm (MHz) Meters Feet In.
solid-dielectric Twin-Lead. The transmitting variety 7.1 9.95 32 8
of line should be used, since the 75-ohm receiving¬ 14.1 4.60 15 1
type line has rather high loss, even when matched. 21.2 3.44 11 3
However, either 52-ohm or 75-ohm coaxial line can 28.2 2.34 7 8
be used; coax line introduces some unbalance, but
this is not intolerable on the lower frequencies.
TRAP DIPOLES
The separation between the dipoles for the
various frequencies does not seem to be especially By using tuned circuits of appropriate design
critical, so far as experience indicates. One set of strategically placed in a dipole, the antenna can be
wires can be suspended from the next larger set, made to show what is essentially fundamental
using insulating spreaders (of the type used for resonance at a number of different frequencies.
feeder spreaders) to give a separation of a few The general principle is illustrated by Fig. 6-12.
inches. The two inner lengths of wire, X, together form a
An interesting method of construction used simple dipole resonant at the highest band desired,
successfully by ON4UF is shown in Fig. 6-11. The say 14 MHz. The tuned circuits Ll-Cl are also
antenna has four dipoles (for 7, 14, 21 and 28 resonant at this frequency, and when connected as
MHz) constructed from 300-ohm ribbon transmis¬ shown offer a very high impedance to rf current of
sion line. A single length of ribbon makes two that frequency which may be flowing in the
 184 Chapter 6

Fig. 6-11 — Sketch showing

how the Twin-Lead multiple¬
dipole antenna system is
assembled. The excess wire
and insulation are stripped
away.

section X-X Effectively, therefore, these two tuned having a characteristic impedance of the same
circuits act as insulators for the inner dipole, and order as the feed-point impedance of a simple
the outer sections beyond Ll-Cl are inactive. dipole can be connected at the center of the
However, on the next lower frequency band, antenna and will be satisfactorily matched on all
say 7 MHz, Ll-Cl shows an inductive reactance three bands, and so will operate at a low SWR on
and is the electrical equivalent of a coil. If the two all three. A line of 75-ohm impedance is satisfac¬
sections marked Yare now added and their length tory; coax may be used, but Twin-Lead will
adjusted so that, together with the loading coils maintain better balance in the system since the
represented by the inductive reactance of Ll-Cl, antenna itself is symmetrical.
the system out to the ends of the Y sections is
resonant at 7 MHz. This part of the antenna is Trap Losses
equivalent to a loaded dipole on 7 MHz and will Since the tuned circuits have some inherent
exhibit about the same impedance at the feed losses the efficiency of this system depends on the
point as a simple dipole for that band. The tuned 0s of the tuned circuits. Low-loss (high-0) coils
circuit L2-C2 is resonant at 7 MHz and acts as a should be used, and the capacitor losses likewise
high impedance for this frequency, so the 7-MHz should be kept as low as possible. With tuned
dipole is in turn insulated, for all practical pur¬ circuits that are good in this respect — comparable
poses, from the remaining outer parts of the with the low-loss components used in transmitter
antenna. tank circuits, for example — the reduction in
Carrying the same reasoning one step farther, efficiency as compared with the efficiency of a
L2-C2 shows inductive reactance on the next lower simple dipole is small, but tuned circuits of low Q
frequency band, 3.5 MHz, and is equivalent to a can lose an appreciable portion of the power
coil on that band. The length of the added supplied to the antenna.
sections, Z-Z, is adjusted so that, together with the
two sets of equivalent loading coils now indicated Dimensions
in C, the whole system is resonant as a loaded The lengths of the added antenna sections, Y
dipole on 3.5 MHz. A single transmission line and Z in the example, must in general be deter-

(c)-mm-mm-- o_mm_mm_

FEED
POINT

(d) -l(-1(-OO-)|-
FEED 1 "
POINT

Fig. 6-12 - Development of the "trap" dipole for

operation on fundamental-type resonance in sev¬
eral bands.
Multiband Antennas 185
mined experimentally. The length required for
resonance in a given band depends on the length/
diameter ratio of the antenna conductor and on
the L/C ratio of the trap acting as a loading coil.
The effective reactance of an LC circuit on half the
frequency to which it is resonant is equal to 2/3
the reactance of the inductance at the resonant
frequency. For example, if LI-Cl resonates at 14
MHz and LI has a reactance of 300 ohms at 14
MHz, the inductive reactance of the circuit at 7 Fig. 6-13 — Five-band (3.5 through 28 MHz) trap
MHz will be equal to 2/3 X 300 = 200 ohms. The dipole for operation with 75-ohm feeder at low
added antenna section, Y, would have to be cut SWR (W3DZZ). The balanced (parallel-conductor)
line indicated is desirable, but 75-ohm coax can be
to the proper length to resonate at 7 MHz with this substituted with some sacrifice of symmetry in the
amount of loading. Since any reasonable L/C ratio system. Dimensions given are for resonance (lowest
can be used in the trap without affecting its SWR) at 3750, 7200, 14,150, and 29,500 kHz.
performance materially at its resonant frequency, Resonance is very broad on the 21-MHz band, with
the L/C ratio can be varied to control the added SWR less than 2:1 throughout the band.
antenna length required. The added section will be
shorter with high-£ trap circuits and longer with
high-C traps. another type of trap devised by Lattin (see
bibliography at the end of this chapter), the coil is
Higher Frequencies supported inside an aluminum tube, and the trap
capacitor is obtained in the form of capacitance
On bands higher than that for which the inner
between the coil and the outer tube. This type of
dipole is resonant all traps in the system show
trap is inherently weatherproof.
capacitive reactance. Thus at such frequencies the
A simpler type of trap, easily assembled from
antenna has the equivalent circuit shown at D in
Fig. 6-12. The capacitive reactances have the effect readily available components, is shown in Fig.
6-14. A small transmitting-type ceramic capacitor
of raising the resonant frequency of the system as
compared with a simple dipole of the same overall is used, together with a length of commercially
available coil material, these being supported by an
length.
This effect is greatest near the resonant fre¬ ordinary antenna strain insulator. The circuit con¬
quency of the inner dipole X-X and becomes less stants and antenna dimensions differ slightly from
marked as the frequency is increased, since the those of Fig. 6-13, in order to bring the antenna
capacitive reactance decreases with increasing fre¬ resonance points closer to the centers of the
quency. The system therefore can be used on various phone bands. Construction data are given
higher frequency bands as a harmonic-type in Fig. 6-15. If a 10-turn length of inductor is used,
antenna, but obtaining resonance with low imped¬ a half turn from each end may be used to slip
ance will require careful balancing of the trap L/C through the anchor holes in the insulator to act as
ratios and the lengths of the various antenna leads.
The components used in these traps are suffi¬
sections. ciently weatherproof in themselves so that no
Five-Band Antenna additional treatment for this purpose has been

One such system has been worked out by

W3DZZ for the five amateur bands from 3.5 to 30
MHz. Dimensions are given in Fig. 6-13. Only one
set of traps is used, resonant at 7 MHz to isolate
the inner (7-MHz) dipole from the outer sections,
which cause the overall system to be resonant in
the 3.5-Mhz band. On 14, 21 and 28 MHz the
antenna works on the capacitive-reactance princi¬
ple just outlined. Using 75-ohm Twin-Lead as a
feeder, the SWR with this antenna was under 2 to
1 throughout the three high-frequency bands, and
the SWR was comparable with that obtained with
similarly fed simple dipoles on 7 and 3.5 MHz.

Trap Construction
Traps frequently are built with coaxial alumi¬
num tubes (usually with polystyrene tubing
between them for insulation) for the capacitor, and
with the coil either self-supporting or wound on a Fig. 6-14 — Easily constructed trap for wire
form of larger diameter than the tubular capacitor. antennas (W2CYK). The ceramic insulator is 4-1/4
The coil is then mounted coaxially with the inches long (Birnbach 668). The clamps are small
capacitor to form a unit assembly that can be service connectors available from electrical supply
supported at each end by the antenna wires. In and hardware stores (Burndy KS90 Servits).
186 Chapter 6

and the second, the ground plane. These antennas

are described in detail in Chapter Two.
The efficiency of any ground-mounted vertical
depends a great deal on earth losses. As pointed
out in Chapter Two, these losses can be reduced or
eliminated with an adequate radial system. Consid¬
erable experimentation has been conducted on this
subject by Sevick, and several important results
were obtained. It was determined that a radial
system consisting of 40 to 50 radials, two-tenths
wavelength long, would reduce the earth losses to
about 2 ohms when a quarter-wave radiator was
Fig. 6-15 — Layout of multiband antenna using being used. These radials should be on the earth’s
traps constructed as shown in Fig. 6-14. The surface, or if buried, placed not more than an inch
capacitors are 100 pF each, transmitting type, or so below ground. Otherwise, the rf current
5000-volt de rating (Centralab 850SL-100N). Coils would have to travel through the lossy earth before
are 9 turns No. 12, 2-1/2 inch diameter, 6 turns per reaching the radials. In a multiband vertical system,
inch (B&W 3029) with end turns spread as neces¬ the radials should be 0.2 wavelength long for the
sary to resonate the traps to 7200 kHz. These lowest band, i.e., 55 feet long for 80-meter
traps, with the wire dimensions shown, resonate operation. Any wire size may be used for the
the antenna at approximately the following fre¬
quencies on each band: 3900, 7250, 14,100, radials. The radials should fan out in a circle,
21,500 and 29,900 kHz (based on measurements radiating from the base of the antenna. A metal
by W9YJH). plate, such a piece of sheet copper, can be used at
the center connection.
The other common type of vertical is the
ground-plane antenna. Normally, this antenna is
found to be necessary. However, if it is desired to mounted above ground with the radials fanning out
protect them from the accumulation of snow or ice from the base of the antenna. The vertical portion
a plastic cover can be made by cutting two disks of of the antenna is usually an electrical quarter
polystyrene slightly larger in diameter than the wavelength, as is each of the radials. In this type of
coil, drilling at the center to pass the antenna antenna, the system of radials acts somewhat like
wires, and cementing a plastic cylinder on the an rf choke, to prevent rf currents from flowing in
edges of the disks. The cylinder can be made by the supporting structure, so the number of radials
wrapping two turns or so of .02-inch poly or is not as important a factor as it is with a
Lucite sheet around the disks, if no suitable ground-mounted vertical system. From a practical
ready-made tubing is available. standpoint, the customary number of radials is
four or five. In a multiband configuration, quarter¬
wave radials are required for each band of opera¬
Four-Band Trap Dipole tion with the ground-plane antenna. This is not so
In case there is not enough room available for with the ground-mounted antenna, where the
erecting the 100-odd-foot length required for the ground plane is relied upon to provide an image of
five-band antennas just described, Fig. 6-16 shows the radiating section. In the latter case, as long as
a four-band dipole operating on the same principle the ground-screen radials are approximately 0.2
that requires only half the linear space. The trap wavelength long at the lowest frequency, this
construction is the same as shown in Fig. 6-14. length will be more than adequate for the higher
With the dimensions given in Fig. 6-16 the reso¬ bands.
nance points are 7200, 14,100, 21,150 and 28,400
kHz. The capacitors are 27-pF transmitting-type SHORT VERTICAL ANTENNAS
ceramic (Centralab type 857). The inductors are 9
A short vertical antenna can be operated on
turns of No. 12, 2-1/2 inches in diameter, 6 turns
several bands by loading it at the base, the general
per inch (B&W 3029), adjusted so that the trap
arrangement being similar to Figs. 6-1 and 6-2.
resonates at 14,100 kHz before installation in the
That is, for multiband work the vertical can be
antenna.
handled by the same methods that are used for
random-length wires.
Vertical Antennas
However, a vertical antenna should not be
There are two basic types of vertical antennas longer than about 3/4 wavelength at the highest
and either type can be used in multiband configu¬ frequency to be used, if low-angle radiation is
rations. The first is the ground-mounted vertical wanted. If the antenna is to be used on 28 MHz

Fig. 6-16 — Sketch showing

dimensions of a trap dipole cover¬
ing the 40- through 10-meter
bands (K2GU).
Multiband Antennas 187

ing LI and Cl. The extra wire will of course be a

part of the antenna, and since it may have to run
VMTICAL
-ANTtHHA (U>>
through unfavorable surroundings it is best to
TO APS 2t ft) avoid its use if at all possible.
This system is best adjusted with the help of an
SWR indicator. Connect the coax line across a few
turns of LI and take a trial position of the shorting
tap until the SWR reaches its lowest value. Then
vary the line tap similarly; this should bring the
SWR down to a low value. Small adjustments of
both taps then should reduce the SWR to close to
1 to 1. If not, try adding Cl and go through the
same procedure, varying Cl each time a tap
position is changed.
Fig. 6-17 — Multiband vertical antenna system
using base loading for resonating on 10 to 80 Trap Verticals
meters. L1 should be wound with bare wire so it
can be tapped at every turn, using No. 12 wire. A The trap principle described in Fig. 6-12 for
convenient size is 2-1/2 inches in diameter, 6 turns center-fed dipoles also can be used for vertical
per inch (such as B&W 3029). Number of turns antennas. There are two principal differences: Only
required depends on antenna and ground lead one half of the dipole is used, the ground connec¬
length, more turns being required as the antenna tion taking the place of the missing half, and the
and ground lead are made shorter. For a 25-foot feed-point impedance is one half the feed-point
antenna and a ground lead of the order of 5 feet, impedance of a dipole. Thus it is in the vicinity of
L1 should have about 30 turns. The use of C1 is 30 ohms (plus the ground-connection resistance),
explained in the text. The smallest capacitance that
will permit matching the coax cable should be so 52-ohm cable should be used since it is the
used; a maximum capacitance of 100 to 150 pF commonly available type that comes closest to
will be sufficient in any case. matching.
As in the case of any vertical antenna, a good
ground is essential, and the ground lead should be
and lower frequencies, therefore, it should not be short. Some amateurs have reported successfully
more than approximately 25 feet high, and the using a ground plane dimensioned for the lowest
shortest possible ground lead should be used. If the frequency to be used; for example, if the lowest
base of the antenna is well above actual ground, frequency is 7 MHz, the ground-plane radials can
the ground lead should run to the nearest water or be approximately 34 feet long.
heating pipe. Traps of the type shown in Fig. 6-14 can be
Another method of feeding is shown in Fig. used in a vertical wire suspended from a support,
6-17. LI is a loading coil of adjustable inductance but a trap similar in construction to the loading
so the antenna can be tuned to resonate on the coil of Fig. 10-14 should be used in tubing verticals.
desired band. It is tapped for adjustment of tuning,
and a second tap permits using the coil as a
Combining Vertical and Horizontal Conductors
transformer for matching a coax line to the
transmitter. Capacitor Cl is not strictly necessary, The performance of vertical antennas such as
but may be helpful on the lower frequencies, 3.5 just described depends a great deal on the ground
and 7 MHz, if the antenna is quite short. In that connection. You have no way of knowing whether
case C1 makes it possible to tune to resonance with or not you have a “good” ground, in the rf sense.
a coil of reasonable dimensions at LI. Cl may also If you can eliminate the ground connection as a
be useful on other bands as well, if the system part of the antenna system, it simplifies things. Fig.
cannot be matched to the feed line with a coil 6-18 shows how it can be done. Instead of a
alone. ground, the system is completed by a wire —
The coil and capacitor preferably should be preferably, but not necessarily, horizontal - of the
installed at the base of the antenna, but if this same length as the antenna. This makes a balanced
cannot be done a wire can be run from the antenna system somewhat like the center-fed dipole.
base to the nearest convenient location for mount-

Fig. 6-18 — Vertical

and horizontal con¬
ductors combined.
This system can be
used on all bands
from 3.5 to 28 MHz
with good results.
 188 Chapter 6

in the line if you try to use coax, or even 300-ohm

Twin-Lead.
This system can be tuned up by using an SWR
indicator in the coax line between the transmitter
and a Transmatch.

HARMONIC RADIATION FROM

MULTIBAND ANTENNAS
Since a multiband antenna is intentionally
designed for operation on a number of different
frequencies, any harmonics or spurious frequencies
that happen to coincide with one of the antenna’s
resonant frequencies will be radiated with very
little, if any, attenuation. Particular care should be
exercised, therefore, to prevent such harmonics
from reaching the antenna.
Multiband antennas using tuned feeders have a
certain inherent amount of built-in protection
against such radiation since it is nearly always
necessary to use a tuned coupling circuit between
Fig. 6-19 — One method of mounting the vertical the transmitter and the feeder. This adds consider¬
section on a rooftop. The mounting base dimen¬ able selectivity to the system and helps to dis¬
sions can be adjusted to fit the pitch of the roof. criminate against all frequencies other than the
The 1 X 1 pieces should fit snugly around the desired one.
bottom of the bottle to keep it from shifting Multiple dipoles and trap antennas do not have
position. this feature, since the objective in design is to make
the antenna show as nearly as possible the same
resistive impedance in all the amateur bands the
It is desirable that the length of each conductor antenna is intended to cover. It is advisable to
be of the order of 30 feet, as shown in the drawing, conduct tests with other amateur stations to
if the 3.5-MHz band is to be used. At 7 MHz, this determine whether harmonics of the transmitting
length doesn’t really represent a compromise, since frequency can be heard at a distance of, say, a mile
it is almost a half wavelength on that band. or so. If they can, more selectivity should be added
Because the shape of the antenna differs from that to the system since a harmonic that is heard
of a regular half-wave dipole, the radiation charac¬ locally, even if weak, may be quite strong at a
teristics will be different, but the efficiency will be distance because of propagation vagaries. A Trans¬
high on 7 MHz and higher frequencies. Although match circuit of the type described earlier in this
the radiating part is only about a quarter wave¬ chapter will add enough selectivity to take care of
length at 3.5 MHz the efficiency on this band, too, harmonics of the strength generated by trans¬
will be higher than it would be with a grounded mitters of good design and construction.
system. If one is not interested in 3.5 MHz and
can’t use the dimensions shown, the lengths can be
reduced. Fifteen feet in both the vertical and
horizontal conductors will not do too badly on 7 Bibliography
MHz and will not be greatly handicapped, as
compared with a half-wave dipole, on 14 MHz and Source material and more extended discussion
higher. of topics covered in this chapter can be found in
The vertical part can be mounted in a number the references given below.
of ways. However, if it can be put on the roof of Bell, “Trap Collinear Antenna,” QST, Aug., 1963.
your house, the extra height will be worthwhile. Berg, “Multiband Operation with Parallelled Di¬
Fig. 6-19 suggests a simple base mounting using a poles,” QST, July, 1956.
Buchanan, “The Multimatch Antenna System,”
soft-drink bottle as an insulator. Get one with a QST, March, 1955.
neck diameter that will fit into the tubing used for Greenberg, “Simple Trap Construction for the
the vertical part of the antenna. To help prevent Multiband Antenna,” QST, October, 1956.
possible breakage, put a piece of some elastic Lattin, “Multiband Antennas Using Decoupling
material such as rubber sheet around the bottle Stubs,” QST, Dec., 1960.
where the tubing rests on it. Lattin, “Antenna Traps of Spiral Delay Line,”
The wire conductor doesn’t actually have to be QST, November, 1972.
horizontal. It can be at practically any angle that Richard, “Parallel Dipoles of 300-Ohm Ribbon,”
will let you pull it off in a straight line to a point QST, March, 1957.
Sevick, “The W2FMI Ground-Mounted Short Ver¬
where it can be secured. Use an insulator at this tical,” QST, March, 1973.
point, of course. Shafer, “Four-Band Dipole with Traps,” QST,
TV ladder line should be used for the feeder in October, 1958.
this system. On most bands the standing-wave ratio Wrigley, “Impedance Characteristics of Harmonic
will be high, and you will lose a good deal of power Antennas, QST, February, 1954.
 Chapter 7

Antennas for 1.8 MHz

Any of the antennas or arrays commonly used still true that to cover the greatest possible
on the higher frequencies are suitable for the distance, the waves must enter the ionosphere at
1.8-MHz band. However, practical considerations low angles.
with regard to height and size usually limit the Polarization
selection to a few basic types. These are the dipole, It was mentioned in Chapter One that a ground
vertical wire, end-fed wire, loop, and various wave must be vertically polarized, so that the
combinations of these four. Further compromises radiation from an antenna which is to produce a
are often necessary since even these antennas are good ground wave likewise must be vertically
still quite large. As the size and height decrease, so polarized. This dictates the use of an antenna
does the radiating effectiveness, and particular care system of which the radiating part is mostly
should be taken to reduce undesired losses to a vertical. Horizontal polarization will produce prac¬
minimum. The most significant losses result from tically no ground wave, and it is to be expected
induced ground currents, conductor resistance, that such radiation will be ineffective for daytime
losses in matching networks and loading coils, and communication. This is because absorption in the
absorption of rf energy by surrounding objects. ionosphere in the daytime is so high at these
The type of antenna installation finally selected is frequencies that the reflected wave is too weak to
often dictated by those losses most easily elimina¬ be useful. At night a horizontal antenna will give
ted. For example, vertical antennas are usually better results than it will during the day. Ionos¬
considered the most desirable ones to use on 1.8 pheric conditions permit the reflected wave to re¬
MHz but if a suitable ground system is not feasible turn to earth with less attenuation.
the ground losses will be very high. In such a case, Some confusion over the term ground wave
an ordinary dipole may give superior performance exists, since there are a number of propagation
even though the angle of maximum radiation is modes that go by this name. Here, only the type
further from optimum than that of a vertical. that travels over and near a conducting surface will
Some experimentation is often necessary to find be considered. If the surface is flat and has a very
the best system, and the purpose of this chapter is high conductivity, the attenuation of the wave
to aid the reader in selecting the best one for his follows a simple inverse-distance law. That is, every
particular station. time the distance is doubled, the field strength
drops by 6 dB. This law also holds for spherical
PROPAGATION ON 1.8 MHZ surfaces for some distance, and then the field
While important, propagation characteristics on strength drops quite rapidly. For the earth, the
1.8 MHz are secondary to system losses since, as break point is approximately 100 miles (160 km).
pointed out previously, the latter may offset any The conductivity of the surface is an important
attempt to optimize for the angle of radiation. factor in ground-wave propagation. For example,
Generally speaking, the 1.8-MHz band has similar sea water can almost be considered a perfect con¬
properties to those of the broadcast band (550 to ductor for this purpose, at frequencies well up into
1600 kHz) but with greater significance of the sky the hf range. However, there may be as much as an
wave. In this respect, it is not unlike the higher additional 20 dB of attenuation for a 1- to 10-mile
amateur frequencies such as 3.5 MHz, and most path over poor-conducting earth, compared to an
nighttime contacts over distances of a few hundred equivalent path over the sea. The conductivity of
miles on 160 m will be by sky-wave propagation. sea water is roughly 400 times as great as good¬
During the daytime, absorption of the sky wave in conducting land (agricultural regions) and 4000
the D region is almost complete but reliable times better than poor land (cities and industrial
communication is still possible by means of the areas).
ground wave. After sundown, the propagation depends upon
With respect to sky-wave transmission, 160- both the ground wave and sky wave. At the limit
meter waves entering the ionosphere, even verti¬ of the ground-wave region, the two may have equal
cally, are reflected back to earth, so that there is field strengths and may either aid or cancel each
no such phenomenon as skip distance on these other. The result is severe and rapid fading in this
frequencies. However, as at higher frequencies, it is zone. While of less importance in amateur applica-

189
190
Chapter 7

Fig. 7-1 - Drawing showing how earth currents affect the losses in a shortened-vertical antenna
system. In A, the current through the combination of Ce and Re may be appreciable if Ce is much
greater than Cw, the capacitance of the vertical to the ground wires. This ratio can be improved (up to
a point) by using more radials. By raising the entire antenna system off the ground. Ce (whichconsists
of the senes combination of Ce, andCe 2 ) is decreased while Cw stays the same. The radial system
shown at B is sometimes called a counterpoise.

tions, this effect limits the useful nighttime range of the parameters that results in low coil Q is
of broadcast stations. Antenna designs have been radiation. But the latter is exactly the desired
developed over the years which minimize sky-wave result in an antenna system. The radiation from a
radiation and maximize the ground wave. For coil increases as its length-to-diameter ratio in¬
broadcast work, a vertical antenna of 0.528-wave- creases. In some instances, the entire antenna may
length height is optimum over a good ground consist of a single coil. A helically wound vertical is
system. However, caution should be exercised in an example of this type. In any of the loading coils
applying this philosophy to amateur installations that are part of the radiating system, the conductor
since effective antenna systems, even for DX work, diameter should be as large as possible, and very
are possible with relatively high angles of radiation. close spacing between turns should be avoided.
The effect of the earth on antenna loss can best
be seen by examination of Fig. 7-1A. If a vertical
radiator that is short compared with a wavelength
REDUCING LOSSES IN is placed over a ground plane, the antenna current
THE ANTENNA SYSTEM will consist of two components. Part of the current
flows through Cw, which is the capacitance of the
As the length of an antenna becomes small vertical to the radial wires, and part flows through
compared with the wavelength being used, the Ce, the capacitance of the vertical to the earth. For
radiation resistance, Ra, drops to a very low value, a small number of radials, Ce will be much greater
as discussed in Chapter Two. The various losses can than Cw, and most of the current will flow through
be represented by a resistance, R^, in series with the circuit consisting of Ce and Re (the earth
Ra. R^ may be larger than Ra in practical cases. resistance). Power will be dissipated in Re which
Therefore, in an antenna system with high losses, will not contribute to the radiation. The solution
most of the applied power is dissipated in the loss to the problem is to increase the number of radials.
resistance and very little is radiated in Ra. Since This will increase Cw, but, of more importance,
Ra is mostly dependent upon antenna construc¬ will reduce Re by providing more return paths.
tion, efforts to reduce the loss resistance will Theory and experiments have shown that the ideal
normally not affect the radiation resistance. Effi¬ radial system with a 0.528-A vertical consists of
ciency can be improved significantly by keeping approximately 120 radials, each a half-wave long.
the loss resistance as low as possible. If fewer radials are used (12), little is to be gained
The simplest losses to reduce are the conductor by running them out so far. The converse is also
losses. Since electrically short antennas, such as true. If space restricts the length of the radials,
dipoles and end-fed wires, exhibit a large series increasing the number much over 12 will have little
capacitive reactance, a loading coil is commonly effect for an antenna of this height. Since the cur¬
used to tune out the reactance. If not part of the rent is greatest near the base of the antenna, a
radiating system, the coil should have as high a Q ground screen will also help if only a few radials
as possible. Incorporating the loading coil into the are used.
radiating system not only simplifies loading-coil Another method to reduce the ground currents
construction, but may actually increase the effi¬ is shown in Fig. 7-IB. By raising the antenna and
ciency by redistributing the current in the antenna. ground plane off the earth, Cw stays the same in
Such loading coils are designed for low loss, rather value but Ce is considerably reduced (such a
than high Q. The reason for this is the fact that one system is sometimes called a counterpoise). This
Antennas for 1.8 MHz 191

decreased influence of the earth is also the reason the “downleads” also acting as a feed point. If the
why as few as three radials are sufficient for hf and verticals are closely spaced (in comparison with a
vhf ground-plane antennas that are several feet wavelength), the entire system can be considered
above the earth. to be one vertical with N times the current of one
The simple lumped antenna-capacitance analy¬ of the downleads taken alone. The result is that the
sis is a good approximation to actual operation if radiation resistance is N2 x Ra, where Ra is the
the vertical is electrically short, but analysis radiation resistance of a single vertical. This is the
becomes more complicated for greater antenna same principle as acquiring an impedance step-up
lengths. For instance, the maximum ground loss in a multiconductor folded dipole. If the ground
for the 0.528-wave broadcast vertical mentioned losses are also considered, the effective loss resis¬
earlier occurs at a point 0.35 wavelength away tance (Æg) would also be transformed by the same
from the base. amount. However, since the current distribution in
Location of the antenna is perhaps more critical the ground is usually improved by using this
with regard to receiving applications than transmit¬ method, the ratio of RalR^ is also improved. The
ting ones. Sources of strong local noise, such as TV disadvantages of the system are increased complex¬
sets and power lines, can cause considerable diffi¬ ity and difficulty of adjustment. While little if any
culty on 1.8 MHz. However, the proximity of use of this principle has been applied to amateur
rf-absorbing objects such as steel buildings may cut systems for 1.8 MHz, it offers some interesting
down on transmitting efficiency also. Since most possibilities where a good ground system is imprac¬
installations are tailored to the space available, tical. The construction approach shown at B of
little can be done about the problem except to see Fig. 7-3 may be used for the erection of an
that the other losses are kept to a minimum. experimental antenna of this type.
Horizontal Antennas
ANTENNA TYPES
In cases where a good ground system is not
A popular misconception is that antennas for practical and when most of the operation will rely
160 meters have to be much larger, higher, and on sky-wave propagation, horizontal antennas can
more elaborate than those for the higher bands. be used (see Fig. 7-4). The relative simplicity of
When one considers that even the gigantic antennas
used for vlf work have radiating efficiencies of
approximately one percent, it is not surprising that
many contacts on 160 meters can be made with
little more than a piece of wire a few feet off the
ground, or even from mobile installations. While it
is true that a larger and more sophisticated system
may perform better than a smaller one, the point
here is that space restrictions should not discourage
the use of the band.

Verticals
One of the most useful antennas for 160 meters
is a vertical radiator over a ground plane. A typical
installation is shown in Fig. 7-2. Some form of
loading should be used since economics would not
justify a full-sized quarter-wave vertical. One of the
disadvantages of the vertical is the necessity of a
good ground system. Some improvement has been
noted by using a combination of radials and
ground rods where full-length radials were imprac¬
tical. The exact configuration will vary from one
installation to another, and the optimum place¬
ment of the ground rods will have to be deter¬
Fig. 7-2 — Physical layout of a typical vertical
mined by experimentation. antenna suitable for 1.8-MHz operation. Without
For verticals less than an electrical quarter wave the top-loading structure or capacitive hat, the
in height, the input reactance without loading will radiation resistance would be approximately 1 ohm
be capacitive. A simple series loading coil should be for a 30-foot vertical, and 3.3 ohms for a 50-foot
used to tune this reactance out and the coil may be height. The loading inductance for the 30-foot
the only matching network necessary. vertical is approximately 400 pH. Once the anten¬
Normally, matching to the feed line or trans¬ na is approximately tuned to resonance with the
mitter can be accomplished with simple networks base loading coil, a suitable tap near the low end of
or a Transmatch. However, a rather unique method the coil can be found which will give the best
match for the transmitter. The radiation resistance
is used with certain vlf antennas which also can be increased by the use of a top-loading
improves the radiating efficiency as well. The structure consisting of the guy wires (broken up
technique is called multiple tuning, and is illus¬ near the top by insulators) which are connected by
trated in Fig. 7-3A. A series of verticals is fed a horizontal wire, as shown. The radial system
through a common flat-top structure with one of consists of wires buried a few inches underground.
 192 Chapter 7

Fig. 7-3 — Possible configurations for a multiple-tuned vertical antenna for 1.8 MHz. Used extensively
in vlf systems, little experimentation has been performed with it by amateurs. The principle is similar
to that of the folded dipole where an impedance transformation occurs from a lower to higher value
simplifying matching. The ratio is equal to N2 where N is the number of elements. In the system
shown at B, the step-up ratio would be 16, since the total number of elements is four. The exact
values of the loading inductors should be found experimentally, being such that the current in each leg
is the same.
construction of an end-fed wire antenna makes it connected to ground, part of the return circuit
an attractive one for portable operation or where may be through the power line. This increases the
supporting structures are without much height. chances of interference from applicances such as
An is the case with electrically short verticals, TV sets and fluorescent lamps on the same circuit.
the input impedance of horizontal end-fed anten¬ Also, since there are usually connections on the
nas less than a quarter wave in length can be power service that are not soldered, rectification
considered to be a resistance in series with a may take place. The result is mixing of local
capacitive reactance. Matching networks for the broadcast stations with products on 160 meters.
end-fed wire are identical to those used for Filtering will not eliminate the problem because
verticals. the products are in the same band with the desired
Balanced center-fed antennas are also useful, signals. Problems of this type are usually less severe
even though they may be electrically short for 160 as the electrical length of the doublet approaches a
meters and at heights typical of those used at the half wave.
higher bands. For example, an 80-meter doublet
fed with open-wire line may also be used on 160 Combinations of Vertical and
meters with the appropriate matching network at Horizontal Antennas
the transmitter. Care should be taken to preserve The L and T antennas are the most common
the balanced configuration of the doublet in examples where combinations of horizontal and
matching to this type. If one side of the feed line is vertical radiators can be used to advantage. Various
types are shown in Fig. 7-5. Here, the philosophy is
usually to run the vertical portion up as high as
possible with the horizontal part merely acting as a
top-loading structure. Such a system can be con¬
sidered to be equivalent to a vertical, and perfor¬
mance should be improved by the use of a ground
system. Running the horizontal portion out to
great distances may or may not improve the
performance, unless the height is also increased.
A dipole fed with coaxial cable for a higher
frequency band can be used as a T antenna by
tying the feed-line conductors together at the
transmitter. This will also work with dipoles fed
with open-wire line; however, they may work just
Fig. 7-4 — Two matching networks suitable for use as well (or better) by using them in the most
with random-length horizontal (or vertical) wire conventional manner discussed earlier. The inver-
antennas. If the electrical length is less than 1/4 ted-V antenna has also given good results on 160
wave, the input impedance will be equivalent to a
resistance in series with a capacitive reactance and meters. While the center of the antenna should be
the circuit at A should be used. For lengths in the as high as possible, the total angle of the V should
vicinity of 1/2 wave, the input impedance will be not be less than 90 degrees at the apex. This will be
fairly high and may have reactances which are determined by the height of the apex and how high
either inductive or capacitive. For this case, the the ends of the antenna are located above the
parallel-tuned circuit in B should be used. ground. For angles less than 90 degrees, the
Antennas for 1.8 MHz 193

Fig. 7-5 — L, T and inverted-V antennas. The type

of matching network suitable for the L antenna
will depend upon the length Land is the same for a
straight horizontal antenna (see Fig. 7-4). By tying
the feed-line conductors together, an hf-band dipole
can be used as a T antenna for 160 meters. The
exact form that the matching network will take
depends on the lengths of both the horizontal and
vertical portions. Considering only the length of
one leg should be sufficient for the majority of
cases, however, and the equivalent L- antenna
network can be used. The arrangement at C shows
two different methods of feeding an inverted V. In
either case, the apex angle, 0, should be greater
than 90 degrees.

In order to obtain the sharp bidirectional

pattern of a small loop, the overall length of the
radiation efficiency drops very rapidly. The conductor must not exceed .08 wavelength. The
drawings in Fig. 7-4 show some of the configura¬ loop of Fig. 7-7 has a conductor length of 20
tions of the various antenna types discussed. feet. At 1.810 MHz, 20 feet is .036 wavelength.
With this style of loop, .036 wavelength is the
maximum practical dimension if one is to tune the
A RECEIVING LOOP FOR 160 METERS
element to resonance. This limitation results from
Small shielded loop antennas can be used to the distributed capacitance between the shield and
improve reception under certain conditions, inner conductor of the loop. RG-59/U was used for
especially at the lower amateur frequencies. The the loop element in this example. The capacitance
foregoing is particularly true when high levels of per foot for this cable is 21 pF, resulting in a total
man-made noise are prevalent, when the second- distributed capacitance of 420 pF. An additional
harmonic energy from a nearby broadcast station 100 pF was needed to resonate the loop at 1.810
falls in the 160-meter band, or when interference MHz. Therefore, the approximate inductance of
exists from some other amateur station in the the loop is 15 pH. The effect of the capacitance
immediate area. A properly constructed and tuned becomes less pronounced at the higher end of the
small loop will exhibit approximately 30 dB of hf spectrum, provided the same percentage of a
front-to-side response, the maximum response wavelength is used in computing the conductor
being at right angles to the plane of the loop. length. The ratio between the distributed capa¬
Therefore, noise and interference can be reduced citance and the lumped capacitance used at the
significantly or completely nulled out, by rotating feed point becomes greater at resonance. These
the loop so that it is sideways to the interference¬ facts should be contemplated when scaling the
causing source. Generally speaking, small shielded loop to those bands above 160 meters.
loops are far less responsive to man-made noise The radiation resistance of small loops is
than are the larger antennas used for transmitting extremely low, thereby rendering them quite in¬
and receiving. But, a trade-off in performance must efficient for transmitting applications. However,
be accepted when using the loop, for the strength they can be used for that purpose if one is willing
of received signals will be 10 or 15 dB less than to accept a sacrifice in signal level, and if the
when using a full-size resonant antenna. This impedance of the system is matched satisfactorily.
condition is not a handicap on 160 or 80 meters, A discussion of this subject appeared in QST for
provided the station receiver has normal sensitivity March, 1968 (McCoy, “The Army Loop in Ham
and overall gain. Because a front-to-side ratio of 30 Communication”).
dB may be expected, a shielded loop can be used There will not be a major difference in the
to eliminate a variety of receiving problems if made construction requirements of the loop if coaxial
rotatable, as shown in Fig. 7-6. cables other than RG-58/U are used. The line
 194 Chapter 7

impedance is not significant with respect to the

loop element. However, various types of coaxial
line exhibit different amounts of capacitance per
foot, thereby requiring more or less capacitance
across the feed point to establish resonance.
Shielded loops are not affected noticeably by
nearby objects, and therefore they can be installed
indoors or out after being tuned to resonance.
Moving them from one place to another does not
significantly affect the tuning.
In the model shown here it can be seen that a
supporting structure was fashioned from bamboo
poles. The X frame is held together at the center
by means of two U bolts. The loop element is
taped to the cross arms to form a square. It is
likely that one could use metal cross arms without
degrading the antenna performance. Alternatively,
wood can be used for the supporting frame.
A Minibox was used at the feed point of the
loop to contain the resonating variable capacitor.
In this model a 50- to 400-pF compression trimmer Fig. 7-7 — Schematic diagram of the loop antenna.
is used to establish resonance. It is necessary to The dimensions are not critical provided overall
weatherproof the box for outdoor installations. length of the loop element does not exceed
The shield braid of the loop coax is removed approximately .04 wavelength. Small loops which
for a length of one inch directly opposite the feed are one half or less the size of this one will prove
point. The exposed areas should be treated with a useful where limited space is a consideration.
sealing compound once this is done.
antenna-selector switch. Reception of European
In operation this receiving loop has been very DX stations with the loop has been possible from
effective in nulling out second-harmonic energy New England at times when other antennas were
from local broadcast stations. During DX and totally ineffective because of noise.
contest operation on 160 meters it helped prevent It was also discovered that the effects of
receiver overloading from nearby 160-meter approaching storms (with attendant atmospheric
stations that share the band. The marked reduction noise) could be nullified considerably by rotating
in response to noise has made the loop a valuable the loop away from the storm front. It should be
station accessory when receiving weak signals. It is said that the loop does not exhibit meaningful
not used all of the time, but is available when directivity when receiving sky-wave signals. The
needed by connecting it to the receiver through an directivity characteristics relate primarily to
ground-wave signals. This characteristic is discussed
further in Chapter Sixteen. This is a bonus feature
in disguise, for when nulling out local noise or
interference, one is still able to copy sky-wave
signals from all compass points!
For receiving applications it is not necessary to
match the feed line to the loop, though doing so
may enhance the performance somewhat. If no
attempt is made to secure an SWR of 1, the builder
can use 50- or 75-ohm coax for a feeder, and no
difference in performance will be observed. The Q
of this loop is sufficiently low to allow the
operator to peak it foi resonance at 1900 kHz and
use it across the entire 160-meter band. The
degradation in performance at 1800 and 2000 kHz
will be so slight that it will be difficult to discern.
Bibliography
Source material and more extended discussion
of topics covered in this Chapter can be found in
the references given below.
Brown, “The Phase and Magnitude of Earth Cur¬
rents Near Radio Transmitting Antennas/’/Vo-
ceedings of the I.R.E. Feb., 1935.
Brown, l ewis, and Epstein, “Ground Systems as a
Factor in Antenna Efficiency,“ Proceedings of
the I.R.E. June, 1937.
Fig. 7-6 — The 160-meter shielded loop. Bamboo “Some Notes on Ground Systems for 160 Meters,”
cross arms are used to support the antenna. QST, April, 1965.
 Chapter 8

Antennas for 3.5 and 7 MHz

Multiband antennas constructed as described in the half-wave antenna fed with untuned line is one
Chapter Six obviously will be useful on 3.5 and 7 of the most popular systems on the 3.5- and 7-MHz
MHz, and, in fact, the end-fed and center-fed bands. If the antenna is a single-wire affair, its
antennas shown in Chapter Six are quite widely impedance is in the vicinity of 60 ohms. The most
used for 3.5- and 7-MHz operation. The center-fed logical way to feed the antenna is with 72-ohm
system is better because it is inherently balanced Twin-Lead or 50- or 72-ohm coaxial line. The
on both bands and there is less chance for feeder heavy-duty Twin-Lead and the coaxial line present
radiation and rf feedback troubles, but either support problems, but these can be overcome by
system will give a good account of itself. On these using a small auxiliary pole to take the weight of
frequencies the height of the antenna is not too the line. The line should come away from the
important, and anything over 35 feet will work antenna at right angles, and it can be of any length.
well for average operation. This chapter is A “folded dipole” shows an impedance of 300
concerned principally with antennas designed for ohms, and so it can be fed directly with any length
use on one band only. of 300-ohm TV line. The line should come away
from the antenna at as close to a right angie as
possible. The folded dipole can be made of
HALF-WAVELENGTH ANTENNAS ordinary wire spaced by light-weight wooden or
An untuned or “flat” feed line is a logical plastic spacers, 4 or 6 inches long, or a piece of
choice on any band, because the losses are low, but 300-ohm TV line can be used for the folded dipole.
it generally limits the use of the antenna to one A folded dipole can be fed with a 600-ohm
band. Where only single-band operation is wanted, open-wire line with only a 2-to-l SWR, but a

Fig. 8-1 — Half-wavelength

antennas for single-band opera¬
tion. The multiwire types shown
in B, C and D offer a better match
to the feeder over a somewhat
wider range of frequencies but
otherwise the performances are
identical. The feeder should run
away from the antenna at a right
angle for as great a distance as
possible. In the coupling circuits
shown, tuned circuits should re¬
sonate to the operating fre¬
quency. In the series-tuned cir¬
cuits of A, B , and C, high L and
low C are recommended, and in D
the inductance and capacitance Three No -12 or No ¡4
should be similar to the output¬ wires, spaced 4 to
6 inches with wood
amplifier tank, with the feeders Na ¡4 spaced orp lastic spacers
5 inches or No ¡2
tapped across at least 1/2 the coil. soo-Ohm spaced b inches
The tapped-coil matching circuit f Twin-Lead.-
or the Transmatch, both shown in any length.
Chapter Six, can be substituted in
each case.
(c)

195
 196 Chapter 8

nearly perfect match can be otained with 600-ohm

open line and a three-wire dipole.
The three types of half-wavelength antennas
just discussed are shown in Fig. 8-1. One advantage
of the two- and three-wire antennas over the single
wire is that they offer a better match over a band.
This is particularly important if full coverage of the
3.5-MHz band is contemplated.
While there arc many other methods of
matching lines to half-wavelength antennas, the
three mentioned are the most practical ones. It is
possible, for example, to use a quarter-wavelength Fig. 8-3 — The inverted-V dipole. The length and
transformer of 150-ohm Twin-Lead to match a apex angle should be adjusted as described in the
text.
single-wire half-wavelength antenna to 300-ohm
feed line. But if 300-ohm feed line is to be used, a
folded dipole offers an excellent match without
the necessity for a matching section. Inverted-V Dipole
The formula shown above each antenna in Fig.
The halves of a dipole may be sloped to form
8-1 can be used to compute the length at any
an inverted V, as shown in Fig. 8-3. This has the
frequency, or the length can be obtained directly
advantages of requiring only a single high support
from the charts in Fig. 8-2.
and less horizontal space. K7GCO and others have
also reported that the dipole in this form is more
effective than a horizontal antenna, especially for
frequencies of 7 MHz and lower.
Sloping of the wires results in a decrease in the
resonant frequency and a decrease in feed-point
impedance and bandwidth as the angle between the
two wires is decreased. Thus, for the same
frequency, the length of the dipole must be
decreased somewhat. The angle at the apex is not
critical, although it should probably be made no
smaller than 90 degrees. Because of the lower
impedance, a 50-ohm line should be used, and the
usual procedure is to adjust the angle for lowest
SWR while keeping the dipole resonant by
adjustment of length. Bandwidth may be increased
by using multiconductor elements, such as the cage
configuration.

VERTICAL ANTENNAS
For 3.5-MHz work, the vertical can be a quarter
wavelength long (if one can get the height), or it
can be something less than this and “top-loaded.”
The bottom of the antenna has only to clear the
ground by inches. Probably the cheapest construc¬
tion of a quarter-wavelength vertical involves
running copper or aluminum wire alongside a
wooden mast. A metal tower can also be used as a
radiator. If the tower is grounded, the antenna can
be “shunt-fed,” as shown in B of Fig. 8-4. The
“gamma” matching system described in Chapter
Three may also be used. A good ground system is
helpful in feeding a quarter-wavelength vertical
antenna, and the ground can be either a convenient
water-pipe system or a number of radial wires
extending out from the base of the antenna for
about a quarter wavelength.
«s
The Ground Plane
The size of a ground-plane antenna makes it a
little impractical for 3.5-MHz work, but one can be
Fig. 8-2 — The above charts can be used to used at 7 MHz to good advantage, particularly for
determine the length of a half-wave antenna of DX work. This type of antenna can be placed
wire. higher above ground than an ordinary vertical
Antennas for 3.5 and 7 MHz 197

Fig. 8-4 — Vertical antennas are effective for 3.5-

or 7-MHz work. The quarter-wavelength antenna
shown at A is fed directly with 50-ohm coaxial
line, and the resulting standing-wave ratio is usually
less than 1.5 to 1, depending on the ground
resistance. If a grounded antenna is used as at B,
the antenna can be shunt-fed with either 50- or
75-ohm coaxial line. The tap for best match and
the value of C will have to be found by experi¬
ment; the line running up the side of the antenna
should be spaced from 6 to 12 inches from the
antenna. The length (height) of the antenna can be
computed from the formula, or it can be obtained
from Fig. 8-2 by using just one half the length
indicated in the chart. For example, at 3.6 MHz,
the length is 13072 = 65'.

without decreasing the low-angle radiation. The directional pattern. The following design for
vertical member can be a length of self-supporting 40-meter phased verticals is contributed by Gary
tubing at the top of a short mast, and the radials Elliott, KH6HCM/W7UXP. An 80-meter version
can be lengths of wire used also to support the can be constructed by proper scaling. There are
mast. The radials do not have to be exactly practical ways that verticals for 40 meters can be
horizontal, as shown in Fig. 8-5. combined, end-fire and broadside. In the broadside
The ground-plane antenna can be fed directly configuration, the two verticals are fed in phase,
with 50-ohm cable, although the resulting SWR on producing a figure-eight pattern that is broadside
the line will not be as low as it will if the antenna is to the plane of the verticals. In an end-fire
designed with a stub matching section, as described
in Chapter Three. However, the additional loss
caused by an SWR as high as 2 to 1 will be
inappreciable even in cable runs of several hundred
feet when the frequency is as low as 7 MHz.

PHASED VERTICALS
Two or more vertical antennas spaced a half
wavelength apart can be operated as a single
antenna system to obtain additional gain and a
Fig. 8-6 - Pattern for two 1/4-A verticals spaced
one-half wavelength apart fed 180 degrees out of
phase. The arrow represents the axis of the
elements.

arrangement, the two verticals are fed out of phase,

and a figure-eight pattern is obtained that is in line
with the two antennas, Fig. 8-6. However, an
end-fire pair of verticals can be fed 90 degrees out
of phase and spaced a quarter wavelength apart,
and the resulting pattern will be unidirectional.
The direction of maximum radiation is in line with
the two verticals, and in the direction of the
vertical receiving the lagging excitation; see Fig.
8-7.
Construction
Physically, each vertical is constructed of
telescoping aluminum tubing that starts off at
Fig. 8-5 — A ground-plane antenna is effective for 1-1/2-inch dia and tapers down to 1/4-inch dia at
DX work on 7 MHz. Although its base can be any the top. The length of each vertical is 32 feet. Each
height above ground, losses in the ground vertical is supported on two standoff insulators set
underneath will be reduced by keeping the bottom
of the antenna and the ground plane as high above on a 2 by 4, 6 feet long and strapped to a fence.
ground as possible. Feeding the antenna directly An alternative method of mounting would be a 2
with 50-ohm coaxial cable will result in a low by 4 about 8 feet long and set about 2 feet in the
standing-wave ratio. The length of the vertical ground.
radiator can be computed from the formula, or it Originally each vertical element was 32 feet, 6
can be obtained from Fig. 8-2 by using just one inches long, 2Mlf (MHz). After one vertical was
half the length indicated in the chart. The radial mounted on the 2 X 4 it was raised into position
wires are 2.5% longer, For example, at 7.1 MHz,
the radiator is 65’ 11'72 « 33'; the radials are and the resonant frequency was checked with an
1.025 X 33 = 33' 10". antenna noise bridge. It was found that the vertical
 198 Chapter 8

Three in the section on electrical length.)

Obviously, 22 feet, 10 inches of coax cannot be
used, as the verticals are spaced 34.6 feet apart.
This is overcome and a 90-degree lag is still
obtained by using a 3/4-wavelength section of
transmission line between the two verticals, Fig.
8-9A. The SWR is less than 1.25 to 1 across the
entire band, using 52-ohm coax and no matching
network.

Fig. 8-7 - Pattern for two 1/4-Ä verticals spaced PHASED ARRAYS
1/4 wavelength apart and fed 90 degrees out of Phased arrays with horizontal elements can be
phase. The arrow represents the axis of the used to advantage at 7 MHz, if they can be placed
elements, with the element on the right being the
one of lagging phase. at least 40 feet above ground. Any of the usual
combinations will be effective. If a bidirectional
characteristic is desired, the W8JK type of array,
resonated too low in frequency, about 6.9 MHz. shown at A in Fig. 8-10, is a good one. If a
This was to be expected as the fundamental unidirectional characteristic is required, two
equation for the quarter-wave vertical, 234/y, is elements can be mounted about 20 feet apart and
only reasonably correct for very small-diameter provision included for tuning one of the elements
tubing or antenna wire. When larger diameter
tubing (1-1/4 inch and larger) is used, the physical
length will be shorter than this, as described in
Chapter Two. Using the antenna noise bridge, an
inch at a time was cut off the top until the
resonant frequency was 7100 kHz. This resulted in
6 inches being cut off, thus making the vertical
exactly 32 feet long.
The ground system is very important in the
operation of a vertical. The two usual methods of
obtaining a ground system with verticals are shown
in Fig. 8-8.

Feed System TO
TRASS
In order to obtain the unidirectional pattern
shown in Fig. 8-7, the two verticals must be Fig. 8-9 — Two methods of feeding the phased
separated by a quarter wavelength, and one vertical verticals.
must be fed 90 degrees behind the other. Two
suggested feed methods are shown in Fig. 8-9. An as either a director or reflector, as shown in Fig.
electrical section of line cannot be used by itself to 8-10B. The parasitic element is tuned at the end of
connect the two verticals together to obtain the its feed line with a series- or parallel-tuned circuit
90-degree lag because of the velocity factor of (whichever would normally be required to couple
RÖ-8/U. The length of an electrical wavelength of power into the line), and the proper tuning
transmission line is based on the calculation: condition can be found by using the system for
246 X 0.66 _ 22’ jq' receiving and listening to distant stations along the
line of maximum radiation of the antenna. Tuning
the feeder to the parasitic element will peak up the
(Further information concerning velocity factor signal.
and transmission lines can be found in Chapter
40-METER LOOP
An effective but simple 40-meter antenna that
has a theoretical gain of approximately 2 dB over a
dipole is a full-wave, closed loop. A full-wavelength
closed loop need not be square. It can be
trapezoidal, rectangular, circular, or some distorted
configuration in between those shapes. For best
results, however, the builder should attempt to
make the loop as square as possible. The more
rectangular the shape the greater the cancellation
Fig. 8-8 — An 8- to 10-ft. ground rod may provide of energy in the system, and the less effective it
a satisfactory ground system in marshy or beach will be. The effect is similar to that of a dipole, its
areas, but in most locations a system of radial wires effectiveness becoming impaired as the ends of the
will be necessary. dipole are brought closer and closer together. The
Antennas for 3.5 and 7 MHz 199

Fig. 8-10 — Directional antennas for 7 MHz. To

realize any advantage from these antennas, they
should be at least 40 feet high. The system at A is
bidirectional, and that at B is unidirectional in a
direction depending upon the tuning conditions of
the parasitic element. The length of the elements in
either antenna should be exactly the same, but any
length from 60 to 150 feet can be used. If the
length of the antenna at A is between 60 and 80
feet, the antenna will be bidirectional along the
same line on both 7 and 14 MHz. The system at 8
can be made to work on 7 and 14 MHz in the same
way, by keeping the length between 60 and 80
feet.

practical limit can be seen in the “inverted-V” Optimum directivity occurs at right angles to
antenna, where a 90-degree apex angle between the the plane of the loop, or in more simple terms,
legs is the minimum value ordinarily used. Angles broadside from the loop. Therefore, one should try
that are less than 90 degrees cause serious to hang the system from available supports which
cancellation of the rf energy. will enable the antenna to radiate the maximum
The loop can be fed in the center of one of the amount in some favored direction.
vertical sides if vertical polarization is desired. For Just how the wire is erected will depend on
horizontal polarization it is necessary to feed either what is available in one’s yard. Trees are always
of the horizontal sides at the center. handy for supporting antennas, and in many
instances the house is high
enough to be included in
the lineup of solid objects
from which to hang a ra¬
diator. If only one sup¬
porting structure is
available it should be a
simple matter to put up an
A frame or pipe mast to
use as a second support.
(Also, tower owners see
Fig. 8-11 inset.)
The overall length of
the wire used in a loop is
determined in feet from
the formula 1005//(MHz).
Hence, for operation at
7125 kHz the overall wire
length will be 141 feet.
The matching transformer,
an electrical quarter wave¬
length of 75-ohm coax
cable, can be computed by
dividing 246 by the opera¬
ting frequency in MHz,
then multiplying that num¬
ber by the velocity factor
of the cable being used.
Thus, for operation at
7125 kHz, 246/7.125 MHz
= 34.53 feet. If coax with
solid polyethylene in¬
sulation is used a velocity
factor of 0.66 must be
Fig. 8-11 — Details of the full-wave loop. The dimensions given are for employed. Foam-
operation at the low end of 40 meters (7050 kHz). The height above polyethylene coax has a
ground was 7 feet in this instance, though improved performance should velocity factor of 0.80.
result if the builder can install the loop higher above ground without Assuming RG-59/U is used,
sacrificing length on the vertical sides. The inset illustrates how a single the length of the matching
supporting structure can be used to hold the loop in a diamond-shaped
configuration. Feeding the diamond at the lower tip provides radiation in transformer becomes 34.53
the horizontal plane. Feeding the system at either side will result in vertical (feet) X 0.66 = 22.79 feet
polarization of the radiated signal. or 22 feet, 9-1/2 inches.
 200 Chapter 8

This same loop antenna

may be used on the
twenty- and fifteen-meter
bands, although its pattern
will be somewhat different
than on its fundamental
frequency. Also, a slight
mismatch will occur, but
this can be overcome by a
simple matching network.
When the loop is mounted
in a vertical plane, it tends
to favor low-angle signals. RELAY
SWITCH
If a high-angle system is BOX

desired, say for 80 meters,

the full-wave loop can be
mounted in a horizontal 7 EACH
plane, thirty or more feet I FEED LINE
above ground. This * 36' OF RG-
B OR RG-38
arrangement will direct COAX

most of the energy vir¬

tually straight up, pro¬ 60* TYP.

viding optimum sky-wave TO XMTR

coverage on a short-haul
basis.

40-METER "SLOPER" SYSTEM

One of the more popular antennas for 3.5 and 7

MHz is the sloping dipole. David Pietraszewski,
K1THQ, has made an extensive study of sloping
dipoles at different heights with reflectors at the
3-GHz frequency range. From his experiments, he
developed the novel 40-meter antenna system
described here. With several sloping dipoles
supported by a single mast and a switching TOP VIEW
network, an antenna with directional characteris¬
tics and forward gain can be simply constructed. Fig. 8-12 — Five sloping dipoles suspended from
This 40-meter system uses several “slopers” equally one support. Directivity and forward gain can be
spaced around a common center support. Each obtained from this simple array. Top view shows
how the elements should be spaced around the
dipole is cut to a half wavelength and fed at the support.
center with 52-ohm coax. The length of each feed
line is 36 feet. This length is just over 3/8 X, which
provides a useful quality. All of the feed lines go to from the tower, suspended by rope, and makes an
a common point on the support (tower) where the angle of 60 degrees with the ground. In Fig. 8-13,
switching takes place. At 7 MHz, the 36-foot the switch box is shown containing all the
length of coax looks inductive to the antenna when necessary relays required to select the proper feed
the end at the switching box is open circuited. This line for the desired direction. One feed line is
has the effect of adding inductance at the center of selected at a time and opens the feed lines of those
the sloping dipole element, which electrically remaining. In this way the array is electrically
lengthens the element. The 36-foot length of feed rotated. These relays are controlled from inside the
line serves to increase the length of the element shack with an appropriate power supply and rotary
about 5%. This makes any unused element appear switch. For safety reasons and simplicity, 12-volt
to be a reflector. de relays are used. The control line consists of a
The array is simple and effective. By selecting five conductor cable, one wire used as a common
one of the slopers through a relay box located at connection; the others go to the four relays. By
the tower, the system becomes a parasitic array using diodes in series with the relays and a
which can be electrically rotated. All but one dual-polarity power supply, the number of control
element of the array become reflectors, while one wires can be reduced, as shown in Fig. 8-15B.
element is driven. Measurements indicate that this sloper array
The basic physical layout is shown in Fig. 8-12. provides up to 20 dB front-to-back ratio and
The height of the support point should be about forward gain of about 4 dB. If one direction is the
60 feet, but can be less and still give reasonable only concern, the switching system can be
results. The upper portion of the sloper is five feet eliminated and the reflectors should be cut 5
Antennas for 3.5 and 7 MHz 201

Fig. 8-13 — Inside view of relay box. Four relays

provide control over five antennas. See text. The Fig. 8-14- The basic materials required for the
relays pictured here are Potter and Brumfield type sloper system. Control box at left and relay box at
MR11D. right.

percent longer than the resonant frequency. The materials required for the sloper system are shown
one feature which is worth noting is the good in Fig. 8-14.
front-toback ratio. By arranging the system
properly, a null can be placed in an unwanted
direction, thus making it an effective receiving Bibliography
antenna. In the tests conducted with this antenna, Source material and more extended discussion
the number of reflectors used were as few as one of topics covered in this chapter can be found in
and as many as five. The optimum combination the references given below.
appeared to occur with four reflectors and one Elliott, “Phased Verticals for 40,” QST, April,
driven element. No tests were conducted with 1972.
more than five reflectors. This same array can be Hubbell, “Feeding Grounded Towers as Radia¬
scaled to 80 meters for similar results. The basic tors,” QST, June, 1960.

Fig. 8-15 - Schematic diagram for sloper control

system. All relays are 12-volt de, dpdt with 8-A
contact ratings. In A, the basic layout, excluding
control cable and antennas. Note that the braid of
the coax is also open-circuited when not in use.
Each relay is bypassed with .001-pF capacitors.
The power supply is a low-current type. In 8,
(B) diodes are used to reduce the number of control
wires when using de relays. See text.
 Chapter 9

Antennas for 14, 21, and 28 MHz

The antenna systems discussed in Chapter Six described in Chapter Eight, it is ordinarily useful
can, of course, be used on 14, 21 and 28 MHz with only on the band for which it is designed. Suitable
good results. The half-wave antenna for 3.5 MHz, lengths for wire antennas are given in Fig. 9-1, and
fed with tuned feeders, becomes a multiwavelength Eig. 3-53 should be referred to if feeder resonances
antenna at these higher frequencies, and the are to be avoided. Vertical antennas (and the
directional characteristics become a little more ground plane in particular) can be used at these
apparent than at the lower frequencies. Similarly, a frequencies and will give good low-angle radiation.
7-MHz half-wave antenna using a tuned feed line However, with a vertical receiving antenna, man¬
likewise can be used on the harmonically related made noise pickup is likely to be greater than with
higher frequency bands. a horizontal antenna.
The directional pattern of a half-wavelength
Half-Wave Dipoles horizontal antenna becomes apparent at these
The half-wave dipole fed with a matched frequencies, and it is not unwise to provide two
transmission line is often used on the 14-, 21-, and half-wave horizontal antennas for these bands at
28-MHz bands. Like its low-frequency counterpart right angles to each other, with a suitable switching
ï arrangement to permit using one antenna or the
other, depending upon the direction of the desired
signal. Better still is to use a single half-wave
antenna that can be rotated at least 135°.

A BEAM ANTENNA FOR 15 METERS

Half-wave antennas, as previously discussed in
LENCm

this chapter, are usually simple to construct from

wire and provide reasonable performance. When

FREQUENCY (MHzl

Fig. 9-1 — Charts for determining the length of a

half-wave wire antenna at 14, 21 and 28 MHz
FROMNCY (MHz) based on 468/f (MHz).

202
Antennas for 14, 21, and 28 MHz 203

NO. « CLAMPS

Fig. 9-3 — Gamma-matching-section dimensions for

the 15-meter 4-element array (also see Fig. 9-7).

rejection characteristics. Overall dimensions are

given in Fig. 9-2. Construction is straightforward
using commonly available tubing material which
normally is 12 feet long. The center of each
element is made from a 12-foot length of 6061-T6
aluminum alloy which is 1-1/8 inches in diameter.
The overall length of the element is determined by
the distance the telescoping section is extended
beyond the end of the center piece. Each of these
Fig. 9-2 — Overall dimensions for the 4-element telescoping sections is one inch OD and six feet
15-meter array. long to provide the proper fit. Table I in Chapter
Fourteen provides a guide for determining proper
the amateur is interested in constructing an array sizes for element material diameters. When tele¬
which can be rotated, aluminum tubing is used for scoping sections are needed, the difference between
the elements. The mechanical problems encoun¬ joining pieces (in terms of diameters) should be
tered are usually not much greater for several about .009 inch. An additional section of 7/8-inch
elements than with single-element rotatable antenna OD material is used at the tips of the reflector
systems. element to meet the dimensions specified. The two
This four-element Yagi antenna provides appre¬ 7/8-inch pieces extend about nine inches beyond
ciable power gain and exhibits significant back- the one-inch diameter stock.
204 Chapter 9

Fig- 9'5 Element lengths for 3-element Yagis. These lengths will hold closely for tubing elements
supported at or near the center. The radiation resistance (D) is useful information in planning for a
matching system, but it is subject to variation with height above ground and must be considered an
approximation. The driven-element length (C) may require modification for tuning out reactance if a
gamma- or hairpin-match feed system is used.
A 0.2D-0.2R beam cut for 28.6 MHz would have a director length of 452/28 6= 158= 15 feet 10
inches a reflector length of 490/28.6 = 17.1 = 17 feet 1 inch, and a driven-element length of
470.5/28.6 - 16.45 = 16 feet 5 inches.
Each element is held in place with two U bolts Pittman, Nevada. While gain measurements are
which clamp it to a six-inch long piece of alum¬ impossible without a test facility, the estimated
inum stock. These pieces of angle material are then power gain of this system is on the order of 9 dB.
fastened to the boom with automotive muffler The front-to-back ratio is typically 20 or 25 dB.
clamps. The size of the muffler clamp depends on
the size of the boom. For this model, a two-inch
diameter boom size should be satisfactory for all AN INTERLACED YAGI
but the roughest climate conditions. A three-inch FOR 20 AND 15 METERS
diameter boom does have advantages, however, as
explained in Chapter Fourteen. Many times it is desirable to install more than
Matching a feed line to the driven element can one antenna on top of a single tower or mast.
be accomplished by using the dimensions given in Stacking antennas, one above the other, creates a
Fig. 9-3. Final adjustment of the gamma system large stress on the mast and the rotor. With large
should be made after the antenna is mounted in arrays, it is desirable to reduce the weight and wind
place atop the mast by placing a power meter (or loading characteristics in every possible way to
SWR indicator) in series with the feed line at the lower damage possibilities from ice, wind, and
input connector and adjusting the capacitor along other undesirable weather creations. One simple
with the tap point for minimum reflected power as solution to the problem is to mount two complete
described in Chapter Three. antennas on one boom.
The mechanical dimensions for the Yagi des¬ Installing elements for two different antennas
cribed here were developed by Wilson Electronics, on one boom has been popular for many years.

Fig. 9-6 — Suggested proportions for

one side of tapered Yagi elements.
The other side is identical, of course,
and the center section of the element
can be a single piece twice as long as
the length shown here for the first
(largest diameter) section. Ap¬
propriate overall element lengths may
be determined from the graphs of
Fig. 9-5. See Table 14-1 for aluminum
tubing details.
Antennas for 14, 21, and 28 MHz 205

Fig g.7 _ Constructional details of a gamma-matching section for 52-ohm coax line. The
reactance-compensating capacitor is in tubular form. It is made by dividing the gamma rod or bar into
two telescoping sections separated by a length of polystyrene tubing, which serves as the dielectric.

Most commercially manufactured triband antennas same as specified with the two antennas described
use this technique. The question which develops, earlier in this chapter.
however, is whether or not interaction between Fig. 9-5 gives suitable dimensions for a three-
elements for different bands causes detrimental element antenna. Hardware sizes should approxi¬
effects. It is generally accepted that interaction, if mate the values given in Fig. 9-6. A gamma¬
any, is very minimal between bands which are not matching system is easy to construct and adjust
harmonically related. The example shown here is a correctly. The dimensions shown in Fig. 9-7 are
Wilson Electronics Model DB-54 designed to opera¬ typical of the requirements for a working system.
te on both 15 and 20 meters. Two driven elements Should the builder desire to use a hairpin match
are required and each is fed independently with arrangement, Chapter Three should be consulted.
separate transmission lines. The boom is 40 feet
long and is three inches OD. Smaller boom sizes are A THREE-BAND QUAD
not recommended. ANTENNA SYSTEM
Constructional details of this system are similar
to those given for the 15-meter antenna described Quads have been popular with amateurs during
earlier in this chapter except the element center the past few decades because of their light weight,
sections for 20 meters begin with 1-1/4 inch relatively small turning radius, and their unique
material and telescope down in size. The 15-meter ability to provide good DX performance when
elements are identical to the ones described earlier. mounted close to the earth. A two-element three-
All of the critical dimensions are given in Fig. 9-4. band quad, for instance, with the elements
A long boom needs to have additional support
given to it if appreciable sag or droop is to be
avoided. The truss can be made of any suitable
steel wire and should be connected to points about
10 feet in each direction from the boom-to-mast
plate. Turnbuckles should be used at the mast to
create suitable tension for the wires. Each of the
interlaced arrays can be treated as separate an¬
tennas for the purposes of tune-up. Since there is
little (if any) interaction between elements, the
15-meter section could be removed from the boom
if only a 20-meter monoband system is needed.

THE THREE-ELEMENT MONOBAND YAGI

Perhaps the most popular type of antenna used
on 20, 15 and 10 meters is the three-element array.
Three elements offer the best compromise between
gain, size, weight, wind loading, and front-to-back
ratio. Constructional techniques should be the Fig. 9-8 - The three-band quad antenna.
 206 Chapter 9

TABLE 9-1

Three-Band Quad Loop Dimensions

Driven First Second Third
Band Reflector Element Director Director Director
20
Meters (A172' 8" (B) 71'3" <C) 69' 6”
15
Meters (D) 48' 6’4" (E) 47' 7%" (F) 46' 5" (G) 46' 5"
10
Meters (HI 36' 2’A" (I) 35' 6" (J) 34' 7" (K) 34' 7" (L) 34' 7"
Letters indicate loops identified in Fig. 9-9.

mounted only 35 feet above the ground, will give are necessary if the antenna is to survive a winter
good performance in situations where a triband season. Another stumbling block for would-be
Yagi will not. Fig. 9-8 shows a large quad antenna quad builders is the installation of a three dimen¬
which can be used as a basis for design for either sional system (assuming a Yagi has only two
smaller or larger arrays. important dimensions) on top of a tower -
Five sets of element spreaders are used to especially if the tower needs guy wires for support.
support the three-element 20-meter, four-element With proper planning, however, many of these
15-meter, and five-element 10-meter wire-loop obstacles can be overcome. Le., a tram system may
system. The spacing between elements has been be used (see Eichman, QST, March, 1974).
chosen to provide optimum performance con¬ One question which comes up quite often is
sistent with boom length and mechanical con¬ whether to mount the loops in a diamond or a
struction. Each of the parasitic loops is closed square configuration. In other words, should one
(ends soldered together) and requires no tuning. spreader be horizontal to the earth, or should the
All of the loop sizes are listed in Table 9-1 and are wire be horizontal to the ground (spreaders
designed for a center frequency of 14.1, 21.1, and mounted in the fashion of an X)? From the
28.3 MHz. Since quad antennas are rather broad¬ electrical point of view, it is probably a trade-off.
tuning devices excellent performance is achieved in While the square configuration has its lowest point
both cw and ssb band segments of each band (with higher, above ground than a diamond version
the possible exception of the very high end of 10 (which may lower the angle of radiation slightly),
meters). Changing the dimensions to favor a the top is also lower than that of a diamond shaped
frequency 200 kHz higher in each band to create a array. Some authorities indicate that separation of
“phone” antenna is not necessary. the current points in the diamond system gives
The most obvious problem related to quad slightly more gain than is possible with a square
antennas is the ability to build a structurally sound layout. It should be pointed out, however, that there
system. If high winds or heavy ice are a normal never has been any substantial proof in favor of
part of the environment, then special precautions one or-the other, electrically.

io- meter
DIRECTOR
15-METER
SECOND DIRECTOR
 207
Antennas for 14, 21, and 28 MHz
Another consideration enters into the selection
of a design for a quad. The support itself, if guyed,
•-FIBER-GLASS will require a diamond quad to be mounted a short
ARM
Fig. 9-10 - Details of distance higher on the mast or tower than an
one of two assemblies equivalent square array if the guy wires are not to
for a spreader frame. interfere with rotation.
--ANGLE ALUMINUM
The two assemblies are The quad array shown in Fig. 9-8 uses fiber-glass
jointed to form an X spreaders available from Kirk Electronics, East
with a muffler clamp Haddam, Connecticut. Bamboo is a suitable sub¬
mounted at the po¬
sition shown. stitute (if economy is of great importance). How¬
MOUNTING HOLES
FOR MUFFLER
ever, the additional weight of the bamboo spread¬
CLAMP ers over fiber glass is an important consideration. A
typical 12-foot bamboo pole weighs about two
pounds; the fiber-glass type weighs less than a
pound. By multiplying the difference times eight for
a two-element array, 12 times for a three-element
antenna, and so on, it quickly becomes apparent

•-FIBER-GLASS
ARM

From the mechanical point of view there is no

question which version is better. The diamond
quad, with the associated horizontal and vertical
spreader arms, is capable of holding an ice load
much better than a system where no vertical
support exists to hold the wire loops upright.
Stated differently, the vertical poles of a diamond
array, if sufficiently strong, will hold the rest of
the system erect. When water droplets are accu¬
mulating and forming into ice, it is very reassuring
to see water running down the wires to a corner
and dripping off, rather than just sitting there on
the wires and freezing. The wires of the loop (or
several loops, in the case of a multiband antenna)
help support the horizontal spreaders under a load
of ice. A square quad will droop severely under
heavy ice conditions because there is nothing to Fig. 9-12 — An alternative method of assembling
hold it up straight. the wire of a quad loop to the spreader arm.

that fiber glass is worth the investment if weight is

an important factor. Properly treated, bamboo has
a useful life of three or four years,while fiber-glass
life is probably ten times that amount.
Spreader supports (sometimes called spiders)
are available from many different manufacturers. If
the builder is keeping the cost at a minimum, he
should consider building his own. The expense is
about half that of a commercially manufactured
equivalent and, according to some authorities, the
homemade arm supports described below are less
likely to rotate on the boom as a result of wind
pressure.
A three-foot long section of one-inch-per-side
steel angle stock is used to interconnect the pairs
of spreader arms. The steel is drilled at the center
to accept a muffler clamp of sufficient size to
clamp the assembly to the boom. The fiber glass is
attached to the steel angle stock with automotive
hose clamps, two per pole. Each quad-loop spread¬
er frame consists of two assemblies of the type
shown in Fig. 9-10.
Connecting the wires to the fiber glass can be
done in a number of different ways. Holes can be
 208 Chapter 9

possible. First, three individual sections of coaxial

cable may be used. Quarter-wave transformers of
75-ohm line are recommended for this service.
Second, a relay box may be installed at the center
of the boom. A three-wire control system may be
used to apply power to the proper relay for the
purpose of changing bands. The circuit diagram of
a typical configuration is presented in Fig. 9-14,
and its installation is shown in Fig. 9-15.
The quarter-wave transformers mentioned
above are necessary to provide a match between
the wire loop and a 50-ohm transmission line. It is
simply a section of 75-ohm coax cable placed in
series between the 50-ohm line and the antenna
feed point, as shown in Fig. 9-16. A pair of PL-259
connectors and a barrel interconnector may be
used to splice the cables together. The connectors
and the barrel should be wrapped well with plastic
tape and then sprayed with acrylic for protection
against the weather.
Fig. 9-13 - Assembly details of the fed element of Every effort must be placed upon proper
a quad loop. construction if freedom from mechanical problems
is to be expected. Hardware must be secure or
drilled at the proper places on the spreader arms vibrations created by the wind may cause unjoining
and the wires run through them. A separate wrap of assemblies. Solder joints should be clamped in
wire should be included at the entry/exit point to place to keep them from flexing, which might
prevent the loop from slipping. Details are pre¬ fracture a connection point.
sented in Fig. 9-11. Some amateurs have exper¬
ienced cracking of the fiber glass, which might be a
result of drilling holes through the material. How¬
ever, this seems to be the exception rather than the
rule. The model described here has no holes in the
spreader arms; the wires are attached to each arm
with a few layers of plastic electrical tape and then
wrapped approximately 20 times in a crisscross
fashion with 1/8-inch diameter nylon string, as
shown in Fig. 9-12. The wire loops are left open at
the bottom of each driven element where the
coaxial cable is attached. See Fig. 9-13. All of the
parasitic elements are continuous loops of wire; the
solder joint is at the base of the diamond.
A triband system requires that each driven
element be fed separately. Two methods are
Fig. 9-15 — The relay box is mounted on the boom
near the center. Each of the spreader-arm fiber-glass
poles is attached to steel angle stock with hose
clamps.

THE LOG-PERIODIC DIPOLE ARRAY

The antenna system shown in Figs. 9-17 and
9-18 was originally described in QST for Novem¬
ber, 1973. Additional information on the design of
a log-periodic dipole array (LPDA) is given in
Chapter Four.
The characteristics of the triband antenna are:
Frequency range, 13-30 MHz
Half-power beamwidth, 43° (14 MHz)
Fig. 9-14 — Suitable circuit for relay switching of
bands for the three-band quad. A 3-wire control Operating bandwidth, B = 30/13 = 2.3
cable is required. K1, K2 - Any type of relay Design parameter r = 0.9
suitable for rf switching, coaxial type not required Relative element spacing constant a= .05
(Potter and Brumfeld MR11A acceptable;although Apex half-angle « = 25°, cot = = 2.0325
this type has double-pole contacts, mechanical Bandwidth of active group, B^ = 1.4
arrangements of most single-pole relays make Bandwidth of structure, ßs = 3.22
them unacceptable for switching of rf ). Boom length, L= 26.5 ft
Antennas for 14, 21, and 28 MHz 209 i

PL-259 COAXIAL
CONNECTORS DOUBI E FEMALE
BARREL CONNECTOR

RG-8A/U Fig. 9-17 — The log-periodic dipole array (K4EWG).

(ANY LENGTH)
TO XMTR OR
RELAY BOX
The mechanical assembly uses materials readily
Fig. 9-16 — Showing installation of quarter-wave available from most local hardware stores or
75-ohm transformer section. aluminum supply houses. The materials needed are
given in Table 9-III. In the construction diagram,
Fig. 9-18, the materials are referenced by their
Longest element = 38 ft (a tabulation of element respective material list number. The photograph
lengths and spacings is given in Table 9-11) shows the overall construction picture, and the
Total weight, 116 pounds drawings show the details. Table 9-1V gives the
Wind-load area, 10.7 sq. ft required tubing lengths to construct the elements.
Required input impedance (mean resistance), Ko =
67 ohms, Zt = 6-inch jumper No. 18 wire
Average characteristic dipole impedance: Zaj4 Bibliography
MHz = 450 ohms; Za 2i MHz = 420 ohms; Source material and more extended discussion
^a28 MHz = 360 ohms of topics covered in this chapter can be found in
Mean spacing factor o' = .0527 the references given below.
Impedance of the feeder: Zoj4 mHz = 95 ohms; Bergren, “The Multielement Quad,” QST, May,
Zo 21 MHz = 97 ohms; Zo 28 MHz = 103 1963.
ohms Reynolds, "Simple Gamma Match Construction,”
Using a toroid balun at the input terminals and a QST, July, 1957.
72-ohm coax feeder the SWR is 1.4 to 1 Rhodes, "The Log-Periodic Dipole Array,” QST,
(maximum). November, 1973.
 210
Chapter 9

TABLE 9-II - ARRAY DIMENSIONS, FEET

El. No. ln h dn _\ n (spacing) nearest
resonant
1 38.0 19 0
2 34.2 17.1 3.862 = d 12 14 MHz
3 30.78 15.39 3.475 = d 23
4 27.7 13.85 3.13
5 24.93 12.465 2.815
6 22.44 11.22 2.533 • 21 MHz
7 20.195 1 0.098 2.28
8 18.175 9.088 2.05
9 16.357 8.179 1.85 • 28 MHz
10 14.72 7.36 1.663
11 13.25 6.625 1.496
12 11.924 5.962 1.347 = d11< i2

TABLE 9-III - MATERIALS LIST

7. 1-1/4" top rail of chain-link fence 26.5 lineal feet
Material Description Quantity
8. 1:1 toroid balun 1 ea.
1. Aluminum tubing — .047" wall thickness
1" - 12'or 6'lengths 126 lineal feet 9. 6 — 32 X 1" stainless-steel screws
7/8" — 12' lengths 96 lineal feet 24 ea.
6— 32 stainless-steel nuts 48 ea.
7/8" — 6' or 12' lengths 66 lineal feet No. 6 solder lugs
3/4" — 8' lengths 16 lineal feet 24 ea.

2. Stainless-steel hose clamps - 2" max. 48 ea. 10. No. 12 copper feeder wire 60 lineal feet

3. Stainless-steel hose clamps — 1-1/4" max. 26 ea. 11.

A. 12" X 8" X 1/4" aluminum plate 1 ea.
4. TV-type U-bolts 14 ea. B. 6" X 4" X 1/4" alum, plate 1 ea.

5. U-bolts, galv. type 12.

A. 3/4" galv. pipe 3 lineal feet
5/16" X 1-1/2" 4 ea. B. 1" galv. pipe — mast 5 lineal feet
1/4" X 1" 2 ea.
13. Galv. guy wire 50 lineal feet
6. 1" ID polyethelene water-service pipe —
160 psi test, approx. 1-1/4" OD 20 lineal feet 14. 1/4" X 2” turnbuckles 4 ea.
A. 1-1/4" X 1-1/4" X 1/8" aluminum angle — 15. 1/4" X 1-1/2" eye bolts 2 ea.
6' lengths 30 lineal feet
B. 1" X 1/4" aluminum bar - 6' lengths 12 lineal feet 16. TV guy clamps and eye bolts 2 ea.

TABLE 9-IV - ELEMENT MATERIAL REQUIREMENTS

El. No. 1 ” tubing 7/8" tubing 3/4" tubing 1-1/4" angle 1" bar

Lth. Qty. Lth. Qty. Lth. Qty. Lth. Lth.

1 6' 2 6' 2 8' 2 3' r

2 6' 2 12’ 2 3' r
3 6' 2 12' 2 3' r
4 6' 2 8.5' 2 3' r
5 6' 2 7' 2 3' r
6 6' 2 6' 2 3' 1'
7 6' 2 5' 2 2' r
8 6' 2 3.5' 2 2' r
9 6' 2 2.5' 2 2' r
10 3' 2 5' 2 2' r
11 3' 2 4' 2 2' r
12 3' 2 4' 2 2' r
 Chapter 10

HF Antennas for Restricted Space

It is not always practical to erect full-size pass without being observed, and wdl provide good
antennas for the hf bands. Those who live in performance if matched to the transmitter.
apartment buildings may be restricted to the use of A primary consideration with any antenna
miniscule radiators because of house rules, or system, makeshift or permanent, is safety to those
simply because the required space for full-size who might come in contact with the system.
antennas does not exist. Other amateurs may desire Always keep the antenna well out of reach. The
small antennas for aesthetic reasons, perhaps to QRP operator who uses only one or two watts of
prevent the neighbors in their residential areas power is not likely to create a hazardous situation
from becoming annoyed at the sight of a high with his antenna, but should observe the rules of
tower and beam antenna of full dimensions. There safety just the same.
are many reasons why some amateurs prefer to use Another invisible antenna can be realized by
physically shortened antennas, and this chapter erecting a flagpole and using it as a vertical
offers information of various schemes that can be antenna. Alternatively, the halyard can be the kind
employed to realize that goal. of plastic clothesline which contains a wire core,
It is important to consider that few compro¬ the halyard, thereby, serving as the antenna. The
mise antennas are capable of delivering the perfor¬ latter is especially useful when wooden or fiber¬
mance one can expect from the full-size variety. glass flagpoles are used. Hollow, nonmetallic flag¬
But the patient and skillful operator can often do poles lend themselves nicely to containing an
as well as some fellows who are equipped with high internal wire or length of copper tubing which can
power and full-size antennas. The former may not be used as an antenna.
be able to “bore a hole” in the band as often, and Another possibility which should not be over¬
with the commanding dispatch enjoyed by his looked is that of using the TV antenna and its feed
well-equipped brothers, but DX can be worked line as a vertical antenna. The antenna and feeder
successfully when band conditions are suitable. should be insulated from the supporting mast, and
standoff insulators should be used to keep the
"Invisible”Antennas 300-ohm Twin-Lead from touching the house or
other objects. The entire antenna system can be
Situations arise in nearly every amateur’s life tuned to hf-band resonance by means of a Trans¬
which call for discreet antenna installations. That match and treated as a random-length wire. In the
is, rather than arouse the ire of some neighboring interest of safety, the supporting mast or tower
nonamateur, it might save a lot of explanation and should be grounded. A high-quality lightning arres¬
discussion merely to put up a temporary radiating tor should be used between the feed line and the
system which does not resemble the classic wire grounded mast. Vhf operators can modify many
antenna. A 120-foot length of No. 28 enamel wire, TV antennas to work as beam antennas on 6 or 2
strung between the window of a motel unit and meters by cutting the elements to the correct
some supporting object at the far end, will usually length and adding a matching device, such as a
quarter-wave universal stub.
Fig. 10-1 — Dipole antenna length¬ Some amateurs have reported good results
ened electrically with off-center when using the downspout and gutter system of a
loading coils. For a fixed dimension wooden house as a radiator in hf-band work. Still
A, greater efficiency will be realized others have used plastic clothesline (with a steel
with greater distance B, but as Bis
increased, L must be larger in value
to maintain resonance. If the two
coils are placed at the ends of the
antenna, in theory they must be
infinite in size to maintain resonance.
Capacitive loading of the ends, either
through proximity of the antenna to
other objects or through the addition
of capacitance hats, will reduce the
required value of the coils.

211
212 Chapter 10

Fig. 10-2 — Chart for deter¬

mining approximate induc¬
tance values for off-center-
loaded dipoles. At the inter¬
section of the appropriate
curve from the body of the
chart for dimension A and the
proper value for the coil posi¬
tion from the horizontal scale
at the bottom of the chart,
read the required inductive
reactance for resonance from
the scale at the left. See Fig.
10-1. Dimension A is ex¬
pressed as percent length of
the shortened antenna with
respect to the length of a
half-wave dipole of the same
conductor material. Dimen¬
sion B is expressed as the
percentage of coil distance
from the feed point to the end
of the antenna. For example, a
shortened antenna which is
50% or half the size of a
half-wave dipole (one-quarter
wavelength overall) with load¬
ing coils positioned midway
between the feed point and
each end (50% out) would
require coils having an induc¬
tive reactance of approximate¬
core) between the clothes poles in the back yard, ly 950 ohms at the operating
placing an insulator at each point where the line is frequency for antenna reso¬
supported by a pole. nance.
Enterprising amateurs should be able to con-
trive many schemes for installing invisible anten¬ match in the attic and bring coaxial cable down
nas. The possibilities are unlimited. But always - from the antenna system to the operating position.
Safety First!
It is seldom practical to route the radiating portion
Indoor Systems of a wire antenna through the walls to some lower
level in the house. Electrical wiring and water pipes
Perhaps the simplest indoor antenna one can
will have an adverse effect on the efficiency of the
use is the attic-installed random-length wire. For antenna.
single-band use it is convenient to install a Trans¬
Physically shortened dipoles are practical and
should be of interest to the indoor-antenna user.
When there is insuffient area to mount a full-size
dipole, one can install a loading inductor in each
leg of the doublet and tune the system to
resonance by adjusting the number of coil turns.
Fig. 10-1 is a drawing of such an antenna. The
longer the overall length, dimension A, and the
farther the loading coils are positioned from the
center of the antenna, dimension B, the greater the
efficiency of the antenna. However, the greater
is distance B (for a fixed overall antenna size), the
larger the inductors must be to maintain resonance.
Approximate inductance values for single-band
resonance may be determined with the aid of Fig.
10-2, but the final values will depend upon the
proximity of surrounding objects in individual
installations and must be determined experimental¬
ly. The use of high-Q low-loss coils is suggested. A
grid-dip meter, Macromatcher, or SWR indicator is
recommended for use during adjustment of the
system.
Fig. 10-3 — Various configurations for small indoor If the attic area is large enough to accommo¬
antennas. A discussion of installation and tuning date an almost full-size dipole, simply erect as
methods is contained in the text. much of the antenna as possible in a straight line,
HF Antennas for Restricted Space 213
tors. The information gathered and presented here
was provided by Sevick, W2FMI. (See the bibliog¬
raphy for reference to his QST articles on the
subject of shortened antennas.)
A short vertical antenna, properly designed and
installed, approaches the efficiency of a full-size
resonant quarter-wave antenna. Even a six-foot
vertical on 40 meters can produce an exceptional
signal. Theory tells us that this should be possible,
but the practical achievement of such a. result
requires an understanding of the problems of
ground losses, loading, and impedance matching,
treated in the theory chapters of this book.
The key to success with shortened vertical
antennas lies in the efficiency of the ground system
with which the antenna is used. A system of 60
wire radials is recommended for best results,
though the builder may want to reduce the number
at some expense to performance, The radials can
be tensioned and pinned at the far ends to permit
on-the-ground installation, which will enable the
amateur to mow the lawn without the wires
Fig. 10-4 - W2FMI adjusts the 6-foot, 40-meter becoming entangled in the mower blades. Alter¬
vertical. natively, the wires can be buried in the ground
where they will not be visible. There is nothing
critical about the wire size for the radials. No. 28,
then bend the ends of the dipole up, down, or
22, or even 15-gauge, will provide the same results.
sideways from the main portion of the system. It is
The radials should be at least 0.2 wavelength long
recommended that the installer attempt to main¬
tain symmetry in the system by bending the ends (27 feet or greater.)
of the antenna in equal amounts. Some ideas for A top hat is formed as illustrated in Fig. 10-5.
indoor installations can be gained from Fig. 10-3. The diameter is 7 feet, and a continuous length of
Some amateurs living in wooden-frame dwell¬ wire is connected to the spokes around the outer
ings have enjoyed reasonable success when loading circumference of the wheel. A loading coil con¬
sisting of 14 turns of B&W 3029 Miniductor stock
the metal screen of a large window. Here again, a
(2-1/2 inch dia, 6 TPI, No. 12 wire) is installed six
Transmatch is almost mandatory in tuning the
inches below the top hat (see Fig. 10-4). This
window screen to resonance. The ground system
antenna exhibits a feed-point impedance of 3.5
can be the cold-water pipes, the third wire
ohms at 7.21 MHz. For operation above or below
(ground) of the electrical system, or both.
this frequency, the number of coil turns must be
decreased or increased, respectively. Matching is
Outdoor Antennas
it is possible to reduce the physical size of an
antenna by 50% or more and still obtain good
results. Use of an outdoor off-center-loaded dipole,
as described in the previous section, will permit a
city-size lot to accommodate a doublet antenna on
the lower frequency hf bands, 40, 80, or even 160
meters. Short vertical antennas can also be made
quite effective by using lumped inductance to
obtain resonance, and by using a capacitance hat to
increase the feed-point impedance of the system.
As is the case with full-size vertical quarter-wave
antennas, the ground-radial network should be as
effective as possible. Ground-mounted vertical radi¬
ators should be used in combination with several
buried radials. Above-ground vertical antennas
should be worked against at least four quarter¬
wavelength radials.

A SIX-FOOT-HIGH 40-METER
Fig. 10-5 — Construction details for the top hat.
VERTICAL ANTENNA For a diameter of 7 feet, half-inch aluminum
tubing is used. The hose clamp is of stainless steel
Figs. 10-4 through 10-7 give details for building and available at Sears. The rest of the hardware is
short, effective vertical quarter-wavelength radia- aluminum.
 214 Chapter 10

accomplished by increasing the feed-point imped¬

ance to 14 ohms through addition of a 4:1
transformer, then matching 14 ohms to 50 ohms
(feeder impedance) by means of a pi network.
Bandwidth for this antenna is approximately 100
kHz (range of frequency where SWR is less than
2:1).
More than two hundred contacts with the
six-foot antenna have strongly indicated the effi¬
ciency and capability of a short vertical. Invariably
at distances greater than 500 or 600 miles, the
short vertical yielded excellent signals. Similar
antennas can be scaled and constructed for bands
other than 40 meters. The 7-ft-dia top hat was
tried on an 80-m vertical, with an antenna height Fig. 10-6 — Standing-wave ratio of the six-foot
of 22 ft. The loading coil had 24 turns and was vertical using a 7-foot top hat and 14 turns of
placed two feet below the top hat. On-the-air loading 6 inches below the top hat.
results duplicated those on 40 meters. The band¬
width was 65 kHz.
Short verticals such as these have the ability to bile exhaust pipe. Most muffler shops can supply
radiate and receive almost as well as a full-size the materials as well as bend the pipe to specifica¬
quarter-wave antenna. The differences are practi¬ tions. Table 10-1 lists the dimensions required for
cally negligible. Trade-offs are in lowered input operation from 160 to 2 meters, inclusive. The
impedances and bandwidths. However, with a good following technique illustrates how a 40-meter
image plane and a proper design, these trade-offs model can be assembled (from English, W6WYQ
can be entirely acceptable. QST for Dec., 1971).
In forming the ring to these dimensions, four
10-foot lengths of tubing are used. A 10-degree
THE DDRR ANTENNA bend is made at 9-inch intervals in three of the
Another physically small but effective antenna lengths. The fourth length is similarly treated
is the DDRR (directional discontinuity ring radia¬ except for the last 18 inches which is bent at right
tor), described in Electronics, January, 1963. An angles to form the upright leg of the ring. One end
in-depth mathematical analysis of this low-profile of each section is flared so that the sections can be
antenna was given by Dome in QST for July, 1972. coupled together by slipping the end of one into
Fig. 10-8 shows details for constructing à DDRR the flare of its mate.
antenna. The required flares are easily made at the
In this example the radiating element of the muffler shop with the aid of the forming tools.
antenna is made from two-inch diameter automo- Another task which can best be completed at the
shop is to weld a flange onto the end of the upright
leg. This flange is to facilitate attaching the leg to
the mounting plate which provides a chassis for the
tuning mechanism and the coaxial-feed coupler.
After bending and flaring is complete, the ring is
assembled and minor adjustments made to bring it
into round and to the proper dimensions. This can
best be done by drawing a circle on the floor With
chalk and fitting the ring inside the circle. The
circle must be slightly larger than the center-to-
center diameter so that the reference line can be
seen easily. For example, with two-inch tubing the
diameter of the reference circle must be 9 feet, 2
inches. When a satisfactory fit is obtained between
the tubing ring and the chalk ring, drill a 1/4-inch
hole through each of the joints to accept a 1 /4-inch
bolt. These bolts will clamp the sections together.
Also, they can be used to attach the insulators
which support the ring at a fixed height above the
Fig. 10-7 — Base of the vertical antenna showing ground plane.
the 60 radials. The aluminum disk is 15 inches in Insulators for the antenna are made from
diameter and 1/4-inch thick. Sixty tapped holes for
1/4-20 aluminum hex head bolts form the outer 11-inch lengths of 2-inch PVC pipe inserted into a
ring and 20 form the inner ring. The insulator is standard cap of the same material. The PVC caps
polystyrene material (phenolic or Plexiglas suit¬ are first drilled through the center to accept the
able) with a one-inch diameter. Also shown is the 1/4-inch bolt previously installed at the joints. The
impedance bridge used for measuring input resis¬ caps are then slipped onto the bolts and nuts are
tance. installed and tightened to secure the caps in place.
HF Antennas for Restricted Space 215

TABLE 10-1
Dimensions for 1/4-Wavelength DDRR Elements
Band (Meters) 160 80 40 20 15 10 6 2
Feed Point (FP)* 12" 6" 6" 2" 1.5" 3" 1" 1/2'
Gap (G) 16" 7" 5" 3" 2.5" 2" 1.5" 1"
Capacitor. pF (C) 150 100 70 35 15 15 10 5
Spacing (Height) (SP) 48" 24" 11" 6" 4 3/4 3" 1 1/2" 1"
Tubing Diameter (D) 5" 4" 2" 1" 3/4" 3/4" 1/2" 1/4'
Ring Diameter (RD) 36' 18' 9' 4.5' 3’4" 2'4" 16 1/4" 6"
•See Fig. 10-8 for explanation of designations.

The 11-inch length of pipe, when inserted into the volts de the shaft of this motor turns at about 1
cap and pressed firmly until it touches bottom, rpm, which is ideal for DDRR tuning. The gears
results in a total insulator length of 12 inches. Four used were surplus items. If you cannot obtain
insulators are required, one at each of the joints gears, string and pulley drive will do almost as well,
and one near the open end of the ring for support. or you can mount both the motor and the
It is wise to locate this insulator as far back from capacitor in line and use direct coupling. Of course,
the end of the ring as possible because of the if you operate on a fixed frequency, or within a
increasing high rf voltage that develops as the end 40- to 50-kHz segment of the band, you can
of the ring is approached.f As a final measure, the dispense with the motor entirely and simply tune
bottom ends of the insulators are sealed to prevent the capacitor manually. In any case, the tuning
moisture from forming on the inside surfaces. unit must be protected from the weather. A plastic
Standard PVC caps may be used here, but plastic refrigerator box may be used to house the tuning
caps from 15-ounce aerosol cans fit well. capacitor and its drive motor.
A mounting plate is required to provide good
mechanical and electrical connections for the Electrical Connections and the Ground Plane
grounded leg of the radiator, the coaxial feed-line The connection between the open end of the
connection, and the tuning mechanism. If you are ring and the tuning capacitor is made with No. 12
using aluminum tubing, you should use an alumi¬ wire or larger. On the end of the base plate
num plate, and for steel tubing, a steel plate to opposite the tuning unit, and directly under the
lessen corrosion from the contacting of dissimilar ring about 8 inches from the grounded post, install
metals. Dimensions for the plate are shown in Fig. a bracket for a coaxial connector. The connector
10-10. The important consideration here is that should be oriented so that the feed line will lead
good, solid mechanical and electrical connections away from the ring at close to 90 degrees. Install a
are made between the ground side at the coaxial
connector, the ring base, and the tuning capacitor.
In the installation shown in Fig. 10-9 the 9-foot
ring resonated easily with approximately 20 pF of
capacitance between the high end of the ring and
the base plate or ground. Any variable capacitor
which will tune the system to resonance and which
will not arc under full power should be satisfac¬
tory. Remember, the rf voltage at the high imped¬
ance end of this antenna can reach 20 to 30 kV
with high power, so if you are using the maximum
legal limit, you would do well to consider using a
vacuum variable capacitor. To provide for full band
coverage, a 35-pF variable capacitor was coupled to
a reversible, slow-speed motor which enabled the
antenna to be tuned remotely from the operating
position in the antenna pictured. An indicated Fig. 10-8- RD = 0.078X128° ); SP = 0.11 D (2.5°);
SWR of 1.1 to 1 was achieved easily over the entire FP = 0.25SP (see Note 1); C = (see Note 2); D =
(see Note 3); G = (see Table 10-1). Notes: (1)
40-meter band. The motor used was a surplus item Actual dimension must be found experimentally.
made by Globe Industries of Dayton, Ohio. At 20 (2) Value to resonate the antenna to the operating
t Because of the danger of rf burns in the event freq. (3) D ranges upward from 1/2". The larger D
of accidental contact with the antenna, precautions is, the higher the efficiency is. Use largest practical
should be taken to prevent random access to the size, e.g., 1/2" for 10 meters, 5” or 6" for 80 or
completed installation. 160 meters.
216 Chapter 10

installation shown a near-perfect match was ob¬

tained with the feed line connected to the ring
about 12 inches from the grounded post. During
testing, when the antenna was set up on a concrete
surface without the ground plane, a match was
found when the feed line was connected nearly 7
feet from the post!
As shown in the photo, the compactness of the
antenna is readily apparent. The ground plane is
made up of three 12-foot lengths of chicken wire,
each 4 feet wide, which are bonded along the edges
at about 6-inch intervals. In this installation the
antenna, with the ground plane, could be disman¬
tled in about 30 minutes. If portability is not
important, it is best to bond all of the joints in the
tubing so that good electrical continuity is assured.
After all construction is completed, the antenna
should be given a coat of primer paint to minimize
rust. If it suits you, there is no reason why a final
Fig. 10-9 — The chicken-wire ground plane is coat of enamel could not be applied.
evident in the background. The base plate can be
seen at lower right. Note the relative positions of
the 52-ohm coaxial feed at the left end of the Tuning Procedures
plate, the flange on the foot of the post, and the
tuning unit at the right hand end of the plate. Once the mechanical construction is completed,
the antenna should be erected in its intended
operating location. Coupling to the station may be
accomplished with either 52- or 72-ohm coaxial
clamp on the ring directly above the coaxial cable. Tune and load the transmitter as with any
connector. Connect a lead of No. 12 or larger wire antenna. While observing an SWR meter in the line,
from the coaxial connector to the clamp. This wire operate the tuning motor. Indication of resonance
must have a certain amount of flexibility to is the noticeable decrease in indicated reflected
accommodate the movement necessary when power. At this point, note the loading of the
adjusting the match. The matching point must be transmitter; it will probably increase markedly as
found by experimentation. It will be affected by antenna resonance is approached. Retune the
the nature and quality of the ground plane over transmitter and move the feed-point tap on the
which the antenna is operating. The antenna will antenna for a further reduction in indicated reflec¬
function over earth ground, but a ground-plane ted power. There is interaction between the move¬
surface of chicken wire (laid under the antenna and ment at the feed tap and the resonance point;
bonded to the base plate) will provide a constant therefore, it will be necessary to operate the tuning
ground reference and improved performance. In a motor each time the tap is adjusted until the
roof-top location sheet-metal roofing should pro¬ lowest SWR is achieved. Don’t settle for anything
vide an excellent ground plane. A poor ground less than 1.1 to 1. With a good ground and proper
usually results in a matching point for the feed line tuning and matching, this ratio can be achieved and
far out along the circumference of the circle. In the maintained over the entire band. Once the proper

Fig. 10-10 — Drawing

of the base plate which
can be made from
either steel or alumi¬
num, as described in
the text. The lower
right portion of the
plate may be used for
the mounting of the
tuning capacitor (and
motor, if used).
HF Antennas for Restricted Space 217

feed point has been located, the only adjustment

necessary when changing frequency is retuning the
antenna to resonance by means of the motor. If
the antenna is to be fixed tuned, provide an
insulated shaft extension of 18 inches or so to the
tuning-capacitor shaft for manual adjustment. This
not only provides insulation from the high rf
voltage but also minimizes body-capacitance
effects during the tuning process.

SHORTENED YAGI BEAM ANTENNA

WITH LOADING COILS

At some sacrifice in bandwidth it is practical to

shrink the element dimensions of Yagi antennas.
Resonance can be established through the use of
loading inductors in the elements, or by using
inductors in combination with capacitance hats.
Though not as effective in terms of gain, the short
Yagi beam offers the advantage of effective height
above ground when compared to a wire dipole
hung from a convenient near-earth support. When
tower-mounted, the two-band Yagi of Figs. 10-11
through 10-15 requires only 22.5 feet maximum
turning diameter for its 16-foot elements and Fig. 10-11 — This short two-band Yagi can be
boom. The antenna consists of interlaced elements turned by a light-duty rotator.
for 15 and 20 meters. A driven element and
reflector are used for 15-meter operation. The
20-meter section is comprised of a driven element
and a director. Both driven elements are gamma

RES. FREQ.
14.050 MHz

RES. FREQ RES. FREQ RES. FREQ.

30.5 20.0 MHz 14.7 MHz 21.050 MHz
TURNS

16.5 C
Fig. 10-12 — Construc¬ TURNS t
23
TURNS
tional details for the 20- 20-METER

and 15-meter beam. The DRIVEN

ELEMENT 15-METER 20-METER
coils for each side of the REFLECTOR DIRECTOR
15-METER

element are identical. The 20-METER COAXIAL

CABLE
DRIVEN

gamma capacitors are each 6AMMA CAPACITOR

BOX
ELEMENT

140-pF variable units man¬

ufactured by E.F. Johnson
Co. The capacitors are
insulated from ground MAST 15-METER

within the container. Since PLATE

4'
GAMMA CAPACITOR
BOX
the antenna is one-half size
for each band, the tuning is
somewhat critical. The
builder is encouraged to

ï
follow carefully the dimen¬
sions given here.

15'
218 Chapter 10

The loading coils are wound on 1-1/8-inch

diameter Plexiglas rod. The rod slips into the
element tubing and is held in place with compres¬
sion clamps. Be sure to slit the end of the
aluminum where the compression clamps are
placed. See Fig. 10-14. The model shown in the
photographs has coils made of surplus Teflon-
-insulated miniature audio coaxial cable with the
shield braid and inner conductor shotted together.
A suitable substitute would be No. 14 enameled
copper wire wound to the same dimensions as
those given in Fig. 10-12.
All of the elements are secured to the boom
with common TV U-bolt hardware. Plated bolts are
desirable to prevent rust from forming. A 1/4-inch
thick boom-to-mast plate is constructed from a few
pieces of sheet aluminum cut into 10-ineh square
sheets and held together with No. 8 hardware.
Fig. 10-13 — The gamma assembly is held in place Several cookie tins could be used if sheet alumi¬
by means of a small U bolt. The capacitors are num is not available. A plate from a large electrical
mounted on etched circuit board. box might even be used as a boom-to-mast bracket.
Since it is galvanized, it is quite resistant to harsh
matched. A low-cost TV antenna rotator has weather.
sufficient torque to handle this light-weight array. A boom strut (sometimes called a truss) is
(From QST for September, 1973.) recommended because the weight of the elements
A misconception among amateurs is that any is sufficient to cause the boom to sag a bit. A
element short of full size is no good in an antenna
system. Reducing the size of an antenna by 50
percent does lower the efficiency by a decibel or
i!4"aluminum
two, but the gain capability of a parasitic array TUBING
outweighs this small loss in efficiency. Mounting
No.B SELF TAPPING
the antenna above the interference-generating SCREW
neighborhood can greatly reduce susceptibility to
HOSE CLAMP
man-made noise and certainly aids in the reduction
of rf heating to trees, telephone poles, and build¬ l'/ô" PLEXIGLAS
ings. Placing the antenna above these energy¬ ROD
absorbing objects is very desirable.
The dual-band beam has four elements, the
longest of which is 16 feet. All of the elements and
the boom are made from 1-1/4-inch diameter
aluminum tubing available at most hardware stores.
A complete parts list is given in Table 10-11.
Element sections and boom pieces are joined
together by slotting a 10-inch length of 1-1/4-inch Fig. 10-14 — Construction details
tubing with a nibbling tool and compressing it for a for the loading coils.
snug fit inside the element and boom tubing.
Coupling details are shown in Fig. 10-12.

1/8-inch diameter nylon line Is plenty strong. A

U-bolt clamp is placed on the mast several feet
above the antenna and provides the attachment
point for the center of the truss line. To reduce the
possibility of water accumulating in the element
tubing and subsequently freezing (rupture may be
the end result), crutch caps are placed over the
element ends. Rubber tips suitable for keeping
steel-tubing furniture from scratching hardwood
floors would serve the same purpose.
A heavy-duty steel mast should be used, such as
a one-inch-diameter galvanized water pipe. Steel
TV mast is also acceptable. Any conventional TV
type antenna rotator should hold up under load
conditions presented by this antenna. Nevertheless,
Fig. 10-15 — The boom-to-mast plate. certain precautions should be taken to assure
HF Antennas for Restricted Space 219

continued trouble-free service. For instance, when¬

ever possible, mount the rotator inside the tower TABLE 10-11
and extend the mast through the tower top sleeve. Complete Parts List for the Short Beam
This procedure relieves the rotator from having to
handle lateral pressures during windy weather QTY MATERIAL
conditions. A thrust bearing is desirable to reduce 9 Eight-foot lengths of aluminum tubing,
downward forces on the rotator bearing. 1-1/4" dia
The monoband nature of the beam requires the 11 U bolts
2 Variable capacitors, 140 pF (E.F.
use of two coaxial feed lines. The coaxial cable is Johnson)
attached to the 15-meter element (at the front of 4' Plexiglas cast rod, 1-1/8" dia
the beam) at the gamma-capacitor box. The other 16 Stainless steel hose clamps, 1-1/2" dia
end of the cable is connected to a surplus 28-V de 1 Aluminum plate, eight-inches square
single-pole coaxial switch. The cable for the 20- 10' Aluminum solid rod, 1/4"dia
meter element is connected in a similar fashion. 2 Refrigerator boxes, 4X4X4 inches
The switch allows the use of a single feed line from 25' Nylon rope, 1/8" dia
the shack to a point just below the antenna where 16 No. 8 sheet metal screws
16 No. 8 solder lugs
the switch is mounted. It is a simple matter to 8 Plastic (or rubber) end caps, 1-1/4” dia
provide voltage to the switch for operation on one
of the two bands. At the price of coaxial cable
today, a double run of feed line represents a
2:1 about 50 kHz in either direction from the
substantial investment and should be avoided if
center frequency. Although tests were not conduc¬
possible. ted at more than 150 watts input to the transmit¬
An etched circuit board was mounted inside an
aluminum Minibox to provide support and insula¬ ter there is no reason why the system would not
tion for each of the gamma tuning capacitors. operate correctly with a kilowatt of power sup¬
Plastic refrigerator boxes available from most plied to it.
After many months of testing this antenna,
department stores would serve just as well. The
several characteristics were noted. During this
capacitor housing is mounted to the boom by
period the antenna withstood several wind and ice
means of U bolts.
The builder is encouraged to follow the dimen¬ storms. Performance is what can be expected from
sions given in Fig. 10-12 as a starting point for the a two-element Yagi. The front-to-back ratio on 20
position of the gamma rods and shorting bar. meters is a bit less than 10 dB. On 15 meters the
Placing the antenna near the top of the tower and front-to-back ratio is considerably better - on the
then tilting it to allow the capacitors to be reached order of 15 dB. Gain measurements were not
makes it possible to adjust the capacitors for made.
minimum SWR as indicated by an SWR meter (or
A YAGI ANTENNA WITH HELICALLY
power meter) connected in the feed line at the
relay. If the SWR cannot be reduced below some WOUND ELEMENTS
nominal figure of approximately 1.1:1, a slight
repositioning of the gamma short might be re¬ Another practical approach in building short¬
quired. The dimensions given are for operation at ened Yagi antennas is the use of helically wound
14.050 and 21.050 MHz. The SWR climbs above elements. Bamboo poles or fiber-glass quad anten-

Fig. 10-16 — The short beam

with helically wound elements
for 40 meters is shown here
mounted on top of a 40-foot
tower. A nylon-rope cross
strut was not used with this
installation and a slight
amount of boom sag is
noticeable.
 220 Chapter 10

TOP VIEW
2 ELEMENT 40M HELICAL BEAM

DRIVEN ELEMENT %" ALUMINUM

PLATE

9—
THOSE WOODEN ' THOSE
/CLAMP DOWELS /CLAMP

|V////l///ll/f///»^^ llllllllIllllllllllllllllllHIIIIIIIIUti’ lllllllllllllllllllllllliii

^ELEMENT-TO- THOSE,
ALUMINUM / CLAMP
U-BOLT BOOM FIXTURE
WELDING ROD TAPERED FIBERGLASS
PLASTIC
ELEMENT AND WINDING LINK
REFRIGERATOR
(40 TURNS PER FOOT) BOX

_BOOM 2 -Bxl^" PIECES

OF ALUMINUM TUBING

15 8' MAST AND U-BOLTS

B00M-T0-MAST
FIXTURE

Fig. 10-17 — Overall dimensions for the 40-meter short beam. The boom consists of two pieces of
standard 1 -1 /4 inch dia Do-It-Yourself aluminum tubing.

na spreaders are utilized as forms for the spirally boxes are mounted at the center of each element
wound elements. The 40-meter beam illustrated in to protect the loading inductors and balun trans¬
Figs. 10-16 — 10-18, inch, is only 0.28 percent of former from the effects of weather. Two coats of
full size. The elements measure 18 feet, tip to tip, exterior spar varnish should be applied to the
and the boom is 16 feet in length. The feed-point helically wound elements after they are adjusted to
impedance is approximately 12 ohms, thereby resonance. This will keep the turns in place and
permitting the use of a 4:1 broad-band balun offer protection against moisture. Details for a
transformer to match the antenna to a 50-ohm 40-meter version of this style antenna are given
coaxial feed line. here, but the same approach can be used in
This antenna can be built for any 50-kHz fabricating short beams for the other hf bands.
segment of the 40-meter band and will operate Performance with the test model was excellent.
with an SWR of less than 2.5:1 across that range. The antenna was mounted 36 feet above ground
An SWR of 1 can be obtained at the center of the (rotatable) on a steel tower. Many European
50-kHz range to which the beam is adjusted, and a stations were worked nightly on 40-meter cw.
gradual rise in SWR will occur as the frequency of While using 100 watts rf output power with this
operation is changed toward the plus or minus antenna, reports from Europe ranged between RST
25-kHz points from center frequency. 559 and RST 589. Similar good results were
Ten-inch-long stubs of aluminum welding rod obtained when working South American stations
or No. 8 aluminum clothesline wire are used at the and U.S. amateurs on the West Coast. All tests
tips of each element to help lower the Q (in the were conducted from Newington, Conn.
interest of increased bandwidth). The stubs are
useful in trimming the elements to resonance after
the beam is elevated to its final height above Construction Details
ground. Coarse adjustment of the elements is The construction of the 40-meter beam is very
effected by means of tapped inductors located at simple and requires no special tools or hardware.
the center of each element. Plastic refrigerator Two fiber-glass 15-meter quad arm spreaders are
HF Antennas for Restricted Space 221
mounted on an aluminum plate with U bolts, as Tuning
shown in Fig. 10-17. A wooden dowel is inserted The parasitic element was adjusted to be about
approximately six inches into the end of each four percent lower in frequency than the driven
fiber-glass arm to prevent the U bolts from element. A grid-dip oscillator was coupled to the
crushing the poles. The aluminum mounting plate center loading coil and the stiff-wire element tips
is equipped with U-bolt hardware for attachment were trimmed (a quarter of an inch at a time) until
to the 1-1/4-inch diameter boom. resonance was indicated at 6.61 MHz. For phone¬
A plastic refrigerator box is mounted on each band use, the ends could be snipped for 6.91 MHz.
element support plate and is used to house a Adjusting the driven element is simple. Place an
Miniductor coil. No. 14 copper wire is used for the SWR meter or power meter at the input connector
elments. The wire is wound directly onto the and cut the end wires (or add some if necessary) to
fiber-glass poles at a density of 40 turns per foot obtain the best match between the line and the
(not turns per inch) for a total of 360 evenly antenna.
spaced turns. The wire is attached at each end
with an automotive hose clamp of the proper size
to fit the fiber-glass spreader. Since the fiber-glass
is tapered, care must be taken to keep the turns SHORT HELICALLY WOUND VERTICAL
from sliding in the direction of the end tips. ANTENNAS
Several pieces of plastic electrical tape were
wrapped around the pole and wire at intervals of The concept of size reduction can be applied to
about every foot. All of the element half sections vertical antennas as well as to Yagi beams. One has
are identical in terms of wire and pitch. Coil the option of using lumped L and C to achieve
dimensions and type are given in Figs. 10-17 and resonance in a shortened system, or the antenna
10-18. can be helically wound to provide a linear distri¬
The driven-element matching system consists of bution of the required inductance, as shown in Fig.
a 4:1 balun transformer and a tightly coupled link 10-19. No capacitance other than the amount
to the main-element Miniductor. Complete details existing between the radiator and ground is used in
are given in Fig. 10-18. establishing resonance at the operating frequency.
Mounted at the end of each element held in Shortened quarter-wavelength vertical antennas can
place by the hose clamp is a short section of stiff be made by forming a helix on a long cylindrical
wire material used for final tuning of the system. form of reasonable dielectric constant. The diam¬
Since the overall antenna is very small in relation eter of the helix must be very small in terms of
to a full-sized array, the SWR points of 2:1 are wavelength in order to prevent the antenna from
rather close to each other. The antenna shown in radiating in the axial mode. Acceptable form
the photograph provides an SWR of less than 2:1 diameters for hf-band operation are from one inch
within about 30 kHz either side of resonance. This (2.5 cm) to 10 inches (25 cm) when considering
particular antenna was tuned for 7.040 MHz and the practical aspects of antenna construction.
can be used throughout the cw portion of the Insulating poles of fiber glass, PVC tubing, treated
band. Tuning the antenna for phone-band opera¬ bamboo or wood, or phenolic are suitable for use
tion should not be difficult and the procedure in building helically wound radiators. If wood or
outlined below should be suitable. bamboo is used the builder should treat the

10 TURNS NO. 14 1:4 BALUN

BIFILAR WINDING
ON TOROIDAL
CORE

NO. 14
PLASTIC PLASTIC REFRIGERATOR
COVERED
BOX
1" MOSE i Bö W 3025 T'HOSE
CLAMP I9T z CL AMP

" ALUMINUM
TAPERED FIBERGLASS PLATE

ELEMENT ANO WINDING I WOODEN 1” WOODEN

( 40 TURNS PfR FOOT* U-BOLT DOW E L DOWEL
NO.14) BOOM
ANO U-BOLT

FRONT VIEW OF DRIVEN ELEMENT

2 ELEMENT 40 M HELICAL BEAM

Fig. 10-18 — Schematic diagram of the balun

assembly mounted inside the plastic utility box.
The core is a single T-200-2 Amidon. The 12-turn
link is wound directly over the 19-turn Miniductor.
 222 Chapter 10

half wavelength long, wound on the insulating

form with a linear pitch (equal spacing between
turns) will come close to yielding a resonant
quarter wavelength. Therefore, an antenna for use
on 160 meters would require approximately 260
feet of wire, spirally wound on the support. No
specific rule exists concerning the size or type of
wire one should use in making a helix. It is
reasonable to assume that the larger wire sizes are
preferable in the interest of minimizing PR losses
in the system. For power levels up to 1000 watts it
is wise to use a wire size of No. 16 or greater.
Aluminum clothesline wire is suitable for use in
systems where the spacing between turns is greater
than one wire diameter. Antennas requiring close¬
spaced turns can be made from enameled magnet
wire or No. 14 vinyl-jacketed, single-conductor
house wiring stock.
A short rod or metal disk should be made for
the top or high-impedance end of the vertical. This
is a necessary part of the installation to assure
reduction in antenna Q. This broadens the band¬
width of the system and helps prevent extremely
Fig. 10—19 — Details on how to build and hook up a
helically wound short vertical antenna. high amounts of rf voltage from appearing at the
far end of the radiator. (Some helical antennas
have acted like Telsa coils when used with high-
power transmitters, and have actually caught fire at
material with at least two coats of exterior spar the high-impedance end when a stub or disk was
varnish prior to winding the antenna element. The not used.) Since the ^-lowering device exhibits
completed structure should be given two more some additional capacitance in the system, it must
coats of varnish, regardless of the material used for be in place before the antenna is tuned.
the coil form. Application of the varnish will
weatherproof the antenna and prevent the coil
turns from changing position. Tuning and Matching
No strict rule has been established concerning Once the element is wound it should be
how short a helically wound vertical can be before mounted where it will be used, with the ground
a significant drop in performance is experienced. system installed. The feed end of the radiator can
As a general recommendation, one should use the be connected temporarily to the ground system.
greatest amount of length consistent with available Use a grid-dip meter and check the antenna for
space. A guideline might be to maintain an element resonance by coupling the dipper to the last few
length of .05 wavelength or more for antennas turns near the ground end of the radiator. Add or
which are electrically a quarter wavelength long.
remove turns until the vertical is resonant at the
Thus, use 13 feet (4 meters) or more of stock for desired operating frequency.
an 80-meter antenna, 7 feet (2 meters) for 40
It is impossible to predict the absolute value of
meters, and so on.
feed impedance for a helically wound vertical. The
A quarter-wavelength helically wound vertical
value will depend upon the length and diameter of
can be used in the same manner as a full-size
vertical. That is to say, it can be worked against an the element, the ground system used with the
above-ground wire-radial system (four or more antenna, and the size of the disk or stub atop the
radials), or it can be used in a ground-mounted radiator. Generally speaking, the radiation resis¬
manner where the radials are buried or lying on the tance will be very low - approximately 3 to 10
ground. Some operators have reported good results ohms. An L network of the kind shown in Fig.
when using antennas of this kind which employed 10-20 can be used to increase the impedance to 50
four helically wound radials which are cut for ohms. Constants are given for 40-meter operation
resonance at or slightly lower than the operating at 7.0 MHz. The (loaded Q) of the network
frequency. The latter technique should capture the inductors is low to provide reasonable bandwidth,
attention of those persons who must utilize indoor consistent with the bandwidth of the antenna.
antennas. Network values for other operating bands and
frequencies can be determined by using the re¬
Winding Information actance values listed belo w.The design center for the
network is based on a radiation resistance of 5
There is no hard-and-fast formula for deter¬ ohms. If the exact feed impedance is known the
mining the amount of wire needed to establish following equations can be used to determine
resonance in a helical antenna. Experience has precise component values for the matching net¬
indicated that a section of wire approximately one work:
HF Antennas for Restricted Space 223

*C1 =

Fig. 10-20 — An £ network suitable for matching

the low feed impedance of helically wound
where Xçi =Capacitive reactance of Cl verticals to a 50-ohm coaxial cable. The loaded Q
of this network is 3.
^C2 = Capacitive reactance of C2
^Ll = Inductive reactance of LI
Q = Loaded Q of network
comparable to that of many full-size quarter¬
= Radiation resistance of antenna wavelength vertical antennas. The major design
trade-off is in usable bandwidth. All shortened
Example: Find the network constants for a antennas of this variety are narrow-band devices.
helical antenna whose feed impedance is 5 ohms at At 7 MHz, in the example illustrated here, the
7 MHz, 0=3: bandwidth between the 2:1 SWR points will be on
the order of 50 kHz, half that amount on 80
Xci - 3x5 = 15 and meters, and twice that amount on 20 meters.
Therefore, the antenna should be adjusted for
operation in the center of the frequency spread of
XC2 = 50 / 5 = 50^/o.Hl = 50x 0.333 interest.
y 50 - 5

= 16.666 and Bibliography

Source material and more extended discussion
*L1 = 15 + = 15 + 15 = 30
\ 10.00 / of topics covered in this chapter can be found in
the references given below.
Therefore, Cl = 1500 pF, C2 = 1350 pF, and LI =
0.7 mH. The capacitors can be made from parallel Dome, “Increased Radiating Efficiency for Short
or series combinations of transmitting micas. LI Antennas,” QST, September, 1934.
can be a few turns of large Miniductor stock. At rf Dome, “A Study of the DDRR Antenna,” QST,
power levels of 100 W or less, large compression July, 1972.
trimmers can be used at Cl and C2 because the English, “A 40-Meter DDRR Antenna,” QST,
December, 1971.
maximum rms voltage at 100 W (across 50 ohms) King and Harrison, “The Impedance of Short,
will be 50. At, say, 800 W there will be approxi¬ Long, and Capacitively Loaded Antennas with a
mately 220 volts rms developed across 50 ohms. Critical Discussion of the Antenna Problem,”
This suggests the use of small transmitting variables Journal of Applied Physics, Vol. 15, February,
at Cl and C2, possibly paralleled with fixed values 1944.
of capacitance to constitute the required amount Myers and DeMaw, “The HW-40 Micro Beam,”
of capacitance for the network. By making some QST, February, 1974.
part of the network variable it will be possible to Myers and Greene, “A Bite Size Beam,” QST,
adjust the circuit for an SWR of 1 without September, 1973.
knowing precisely what the antenna feed im¬ Sevick, “The Ground-Image Vertical Antenna,”
QST, July, 1971.
pedance is.
Sevick, “The W2FMI Ground-Mounted Short
Fig. 10-19 illustrates the practical form a Vertical,” QST, March, 1973.
typical helically wound ground-plane vertical might Sevick, “The Constant-Impedance Trap Vertical”
take. Performance from this type antenna is QST, March, 1974.
Chapter 11

VHF and UHF Antenna Systems

While the basic principles of antenna design are length of the same type of line at 28 MHz. Thus it
essentially the same for all communication fre¬ may be more effective to use a high-gain array at
quencies, certain factors peculiar to vhf and uhf relatively low height, rather than a simpler array at
work call for changes in amateur antenna tech¬ great height above ground, particularly if the
nique for the frequencies above 50 megahertz. antenna location is not shielded by buildings or
Here the physical size of multielement arrays is heavy foliage in the immediate vicinity.
reduced to the point where an antenna system Height above ground is helpful, especially in
having some gain over a simple dipole is possible in cases where added height increases the distance to
almost every location, and the more progressive the visible horizon appreciably, but great height is
stations may employ arrays having much higher by no means so all-important as it was once
gain than is possible on lower frequencies. thought to be. Outstanding results have been
The importance of high-gain antennas in vhf obtained, particularly on 50 MHz, with relatively
work cannot be overemphasized. The reliable low antennas, and many 144-MHz stations are
working range of a station operating on 144 MHz, working out successfully with arrays not more than
for instance, may be only 30 miles or so when a 25 to 40 feet above ground.
simple dipole antenna is used, yet this same fellow The effectiveness of a vhf antenna system can
may increase his working radius to 100 miles or be increased markedly by stacking half-wave ele¬
more by the installation of a high-gain array. The ments one above the other and feeding them in
directive system introduces other advantages also. phase. Such stacking helps to lower the radiation
By restricting the field covered at any one position angle, an important factor in extending vhf cover¬
the beam antenna helps to reduce pickup of age, without changing the beamwidth in azimuth.
man-made noises, and it may be instrumental in Several examples of stacked arrays are shown in
correcting interference to broadcast and television the following pages.
reception, by permitting communication in direc¬ The physical size of a vhf array is an important
tions not coinciding with nearby antennas used on factor in its performance. In receiving, the larger
receivers for these services. A good antenna system the area presented to an incoming signal the greater
often makes the difference between routine operat¬ the strength of the signal at the receiver input
ing arid outstanding success in the vhf field, and it terminals, other factors being equal. Thus an array
is safe to say that by no other means can so large a for 432 MHz must be the same size as one for 144
return be obtained from a small investment as MHz, if an equal signal is to be received on both
results from the erection of a high-gain antenna bands. The array for the higher frequency will
system. require three times as many elements as the one for
the lower band, if similar element configurations
Design Considerations are used in both.
Antenna systems for the vhf range are usually
called upon to work over a wider frequency range
Polarization
than those used on lower bands; thus, antenna
bandwidth becomes an important consideration in Experience has shown that there is usually no
the design of a vhf array. It may be necessary, in marked difference in effective working radius with
some instances, to include this characteristic at the either horizontal or vertical polarization, though
expense of other attributes which might be con¬ there are indications that horizontal may give
sidered desirable, such as highest gain or front-to- somewhat higher signal levels over irregular terrain.
back ratio. The signal-to-noise ratio with horizontal systems is
A properly matched line is of utmost impor¬ likely to be better, in regions where man-made
tance in the proper functioning of the vhf antenna noise is a serious problem. Horizontal arrays also
system, because even with perfect matching the may have some mechanical advantages. It is gener¬
loss in a given line is essentially proportional to ally easier to build and rotate horizontal systems,
frequency (see Chapter Three). At 144 MHz, for particularly on the lower vhf bands. Simple 2-, 3-
example, the loss in a perfectly matched line is or 4-element arrays have proven very effective in
approximately five times what it is in the same 50-MHz work, and their use has reached the point

224
VHF and UHF Antenna Systems 225
of standardization on horizontal systems for that
band.
The picture is somewhat different on 144 MHz
and higher bands. Vertical arrays are more easily
constructed for these frequencies. Hundreds of
mobile stations on 144 MHz, nearly all using
vertical whip antennas, usually enjoy somewhat
wider coverage when the fixed stations also use
vertical antennas, though the loss from cross-
polarization may not be important in hilly terrain.
Where the 144-MHz band is used for emergency
communication, a logical antenna setup consists of FREQUENCY IN MHz
some sort of stacked (but nondirectional) collinear
vertical array for the control station, and vertical
whips for the portables and mobiles. Television and
fm reception, both sensitive to interference from
vhf stations, use horizontal antennas, and it can be
shown that interference is more troublesome when
the amateur stations also use horizontal systems.
Horizontal polarization is gaining ground in
amateur vhf work, however, and it appears that in
most areas its advantages outweigh the adverse
factors. Except for emergency net operation, much
of the normal 144-MHz operation is done with
horizontal antennas today. Areas on both coasts
still make use of vertical, however, and anyone
starting in on the vhf bands should determine
which polarization is in use in his locality before
investing heavily in antenna installations. There is
considerable polarization shift over mountainous
or irregular terrain, but generally speaking, best
results will be obtained when the same polarization
is used at both ends of a path.

ELEMENT LENGTHS AND SPACINGS

The resonant length of an ungrounded antenna
or antenna element is somewhat shorter than a half
wavelength for the frequency at which it is to be
used, as explained in Chapter Two. In dealing with
vhf antennas it is convenient to measure the length
in inches. The following formula gives the resonant
length of a half-wave element:
r .v , u(inches)
Length x 5904 X/l =-
Freq. (MHz)
where the factor K is dependent on the thickness
of the antenna conductor and the frequency at
which it is used. This factor is plotted in Fig. 2-4,
Chapter Two, and applies to cylindrical conductors
having uniform diameter throughout their length.
The length of a free-space half wavelength,
together with lengths as modified by the factor K
when it has the values 0.98, 0.96, and 0.94, are
shown graphically for the 50-, 144-, 220-, and FREQUENCY IN MHz
420-MHz bands in Fig. 11-1. Element spacings are
based on free-space lengths, which can readily be Fig. 11-1 — Frequency vs. length in inches for the
converted from the half-wavelength values shown 50-, 144-, 220-and 420-MHz bands, for a free-
in the charts. space half wavelength and resonant antenna lengths
The factor K as given in Fig. 2-4 is based on for various element length/thickness ratios (see
theoretical considerations which necessarily do not Chapter Two). Lengths for values of K other than
those given can be found by linear interpolation,
provide for different methods of mounting or e.g., the curve for K = 0.97 lies midway between
support for the element. The exact resonant length the curves for K = 0.96 and 0.98, etc. To find the
depends to some extent on constructional features free-space wavelength used in calculating spacings
of this nature. In average cases, the following multiply the length given for A/2 by 2.
226 Chapter 11

arrays in Chapter Four. The optimum lengths for

the parasitic reflector or director depend on the
element spacings. The spacings are not highly
critical, however, and the bandwidth is usually
greater when the spacings of elements near the
driven element are fairly large. Spacings of 0.2
wavelength (the half-wavelength figure given in Fig.
11-1 multiplied by 0.4) are customarily used in vhf
antennas. With these spacings the reflector will be
Fig. 11-2 — A simple method of providing for approximately 5% longer than the driven element
adjustment of element lengths. The insert is made and the first director will be about 5% shorter than
of the same size tubing as the element, but is the driven element. If additional directors are used
slotted and compressed to permit insertion into the they should be progressively shorter than the
ends of the elements.
driven element, as illustrated by the practical
arrays shown later in this chapter.
formula has been found to work out well in
practice: When the lengths of elements in an array are
given in terms of a decimal part of a wavelength,
T .uz-(inches)
Length u s = ---
5540- the wavelength used as a reference is usually the
Freq. (MHz) free-space wavelength. The free-space wavelength is
equal to the value given by the top curve in each
This corresponds closely to the curves for K = 0.94 graph in Fig. 11-1 multiplied by 2. When a “half
in Fig. 11-1.
wave” is referred to in connection with antennas,
Tapered elements, in which successively smaller the resonant length usually is meant.
sizes of tubing are used either for light construc¬
tion or in a collapsible element for portable use,
tend to exhibit K factors associated with the
smallest diameter of the taper. The exact resonant
length also depends on how the element is mount¬
ed; an element that is supported at its center
through a boom of appreciably larger diameter Unter msulllor
than that of the element, for example, will have a
slightly different resonant length than one that is 1-Melon
insulated from its support.
Fig. 11-4 — An exam¬
Driven Elements ple of the T match. 300 -oten fin«

The length of a driven element, whether the Arms should be about

element is a simple half-wave dipole or is the fed 25 inches long for
element in a parasitic array, is not especially 50-MHz arrays.
critical, since slight mistuning can easily be com¬
pensated for in the adjustment of the matching Adjustment of Element Length
system used between the element and the transmis¬
sion line. However, it is generally desirable that the When an antenna design given in this chapter is
element length be close to resonance at the median to be duplicated it is necessary to do nothing more
operating frequency. The graphs of Fig. 11-1 are than to cut the elements to the lengths given. If the
sufficiently accurate for this purpose. antenna is to be centered on a frequency in a
different part of the band, it will be sufficient to
Parasitic Elements scale all the lengths in proportion to the change in
Optimum spacings for the elements of a Yagi wavelength - that is, in inverse proportion to the
array are discussed in the section on parasitic ratio of the desired center frequency to that for
which the antenna was originally designed.
For experimental work it is desirable to have
some means for continuous adjustment of element
length. Usually the required adjustment range will
be not greater than 10% of the half wavelength,
and in the case of tubing elements a device suitable
for the purpose is shown in Fig. 11-2. It consists of
a short length of tubing, usually of the same stock
as the element, slotted lengthwise and compressed
Fig. 11-3 — Two ver¬ to make a tight fit in the element end.
sions of the "J" an¬
tenna, used in mobile
applications or in ver¬ PHASING AND MATCHING SECTIONS
tical arrays where par¬ Transmission-line lengths for such applications
asitic elements are ro¬ as phasing lines and Q sections can be determined
tated around a fixed from the formula
radiator.

Length (inches) = p5904 ,^ ,

Freq. (MHz)
VHF and UHF Antenna Systems 227

where V is the velocity factor of the type of line

used. For open-wire lines separated by insulating
spacers, V is approximately 0.975. Parallel¬
conductor lines of self-supporting tubing have a
velocity factor close to unity. The velocity factors
of various types of solid-dielectric lines, both
parallel-conductor and coaxial, are given in Chapter
Three.
Fig. 11-5 — Schematic version of the gamma
PRACTICAL ANTENNA DESIGNS match. Values for C and D are given in the text.
The element lengths and spacings in the anten¬
nas described in this chapter have been worked out wave vertical radiator fed by a quarter-wave stub
by experiment to meet practical operating require¬ matching section, as shown at A, Fig. 11-3. The
ments. Since these requirements often are conflict¬ spacing between the two sides of the matching
ing, in terms of antenna design, compromises are section should be two inches or less, and the point
necessary; for example, it is usually necessary to of attachment of the line will depend on the
sacrifice a small amount of gain for the sake of impedance of the line used. The feeder should be
reasonable bandwidth and a good front-to-back moved along the matching section until the point is
ratio. In general, this means that the element found that gives the best operation. The bottom of
lengths and to a lesser extent the spacings (the the matching section may be grounded for light¬
latter are less critical than the element lengths) ning protection.
may not always conform to the values that A variation of the “J” for use with coaxial-line
investigation has shown to be optimum for maxi¬ feed is shown at B in Fig. 11-3.
mum gain. The “J” is also useful in mobile applications,
It can be emphasized, however, that a design though a simple quarter-wave whip will usually
that has been carefully worked out will, if accur¬ suffice.
ately duplicated, give results identical with those
obtained with the original antenna. If the builder The Delta or “Y” Match
has other objectives than the designer originally Probably the simplest arrangement for feeding a
had, they may be achieved by individual adjust¬ dipole or parasitic array is the delta, or “Y” match.
ment; careful adjustment of lengths and spacings
for a desired result, whether it be maximum gain,
high front-to-back ratio, maximum bandwidth, or
whatever, is a process that will have a great deal of
appeal for the experimentally inclined amateur.

TRANSMISSION LINES AND

MATCHING METHODS
As mentioned at the beginning of this chapter,
it is important that the standing-wave ratio on the
transmission line be kept as low as possible.
Otherwise line losses may become prohibitive,
particularly when solid-dielectric lines are used.
Lines normally employed include open-wire lines
of 300 to 600 ohms impedance, usually spaced 1/4
to 2 inches, polyethylene-insulated flexible par¬
allel-conductor lines, available in 72-, 150- and
300-ohm impedances, and coaxial lines of 50 to 90
ohms.
Occasionally two coaxial lines are used side by
side, with the inner conductors serving as the
transmission line, and the outer conductors con¬
nected together and grounded. Such a line has
twice the impedance of its individual components.
The various types of transmission lines can be
matched to antenna systems in a wide variety of
ways, as described in Chapter Three. The more
popular methods are described below.

The “J”
Used principally as a means of feeding a
stationary vertical radiator, around which parasitic
elements are rotated, the “J” consists of a half¬ arrays.
 228 Chapter 11

Information on figuring the dimensions of the

delta may be found in Chapter Three.
The chief weakness of the delta is the likeli¬
hood of radiation from the matching section,
which may interfere with the effectiveness of a
multielement array. It is also somewhat unstable
mechanically, and quite critical in adjustment.

The T Match
The T match, shown in practical form in Fig.
11-4, provides a means of adjustment by sliding the
clips along the parallel conductors, and its rigid
construction is quite suitable for rotatable arrays.
It may be used with a pair of coaxial lines of any
impedance, or with the various other balanced
transmission lines. The position of the clips should,
of course, be adjusted for minimum standing-wave
ratio (see Chapter Three). The T system is particu¬
larly well suited for use in all-metal “plumbing”
arrays.

Fig. 11-9 — A "bazooka" line balancer is used to

feed a balanced center-fed dipole with a coaxial
line. In A it takes the form of a quarter-wave
SHO section of tubing the same size as the coaxial line.
HMHZ) In B it is a metal sleeve connected to the outer
conductor at the far end from the antenna.

Fig. 11-7 — Details and the point of connection on the driven element
of a folded dipole should be adjusted for minimum standing-wave
for vhf use.
ratio. The capacitance required will be under 75 pF
for 50 MHz, or 25 pF for 144 MHz. The rf voltage
at this point is low, so a receiving-type variable can
be used for C. It should be provided with a
The Gamma Match weatherproof housing, which can be of metal,
grounded to the boom of the array. The length of
The gamma match, also described in Chapter the matching section, D, will be about 10 inches
Three, is well adapted to feeding all-metal arrays for 50 MHz or 4 inches for 144 MHz.
with coaxial line, as the outer conductor may be
connected to the metal boom or to the center of
the driven element. The inner conductor is tapped The Q Section
out from the middle of the driven element, usually As described in Chapter Three, a Q section can
through an adjustable clip. Construction can be be used as an impedance transformer to match
similar to that shown for the T match. transmission lines to antenna systems of differing
Best operation results when a variable capaci¬ impedance. The matching section can be made of
tor, C in Fig. 11-5, is included, to tune out the two pieces of wire, rod or tubing of suitable
reactance of the matching section. The capacitor diameter, spaced to give the desired impedance. A
table giving the impedance of lines of various
dimensions may be found in Chapter Three.
Where the impedance that will be needed is not
known, a Q section can be made with one of the
members movable, as shown in Fig. 11-6. The
spacing may then be adjusted for minimum
standing-wave ratio on the transmission line.
Sections of coaxial line may be used for
matching unbalanced lines to unbalanced loads,
and some matching problems with balanced lines
and loads can be solved with Q sections of
Fig. 11-8 — Combination tuning and matching stub Twin-Lead of suitable impedance. Where lines
for vhf arrays. The sliding shorting bar is used for having other than air insulation are used as
tuning the driven element along with the stub impedance transformers, their length should be
itself. The transmission line (or balun, if coax line reduced to take the propagation factor of the line
is used) is moved along the stub until the pointât into account. This will usually mean reductions to
which the SWR is closest to unity is found. 66 and 82 percent of a full quarter wavelength, for
VH F and UHF Antenna Systems 229

mechanically. The transmission line may be open

wire or Twin-Lead tapped directly on the stub.
Coaxial line also may be used, but should be
connected to the stub through a balun.
Fig. 11-10 - A bal- As described in Chapter Three, the adjustment
un made of coaxial procedure is one of varying the position of the line
line may be used to taps and the position of the shorting taps until the
feed balanced loads lowest possible SWR on the transmission line is
from unbalanced obtained.
lines, and vice versa.
It provides an im¬
pedance step-up of USING COAXIAL LINES
4 to 1. Flexible coaxial line has many desirable fea¬
tures. It is weatherproof, and may be buried
underground or run inside a metal mast or tower
without harmful effects. However, unless it is used
properly, losses may be excessive, particularly on
high frequencies when the line is more than a few
wavelengths long. When coaxial line is used to feed
a balanced load, as at the center of a driven
element, some provision should be made for
converting from the unbalanced lino to the bal¬
anced load. Otherwise rf currents will flow on the
outer conductor of the line, destroying its effect¬
polyethylene-insulated coaxial and parallel lines, iveness.
respectively. The exact length for a matching One way of doing this is to install a detuning
section can be determined experimentally, if de¬ sleeve or “bazooka” at the point where the line
sired, by shorting one end of the line, coupling it connects to the driven element, as shown in Fig.
to a calibrated grid-dip oscillator, and then trim¬ 11-9. Both methods shown employ quarter-wave
ming the line length until the grid-dip meter shows sections of line, shorted at the bottom end,
it to be resonant at the desired frequency. presenting a high impedance to rf energy at the top
end and preventing its flow along the outer
The Folded Dipole conductor of the transmission line. The detuning
An effective means of matching various bal¬ section may be a piece of rod or tubing of a
anced lines to the wide range of antenna imped¬ diameter similar to the coaxial line, as in Fig.
ances encountered in vhf antenna work is the 11-9A, or it can be a cylindrical sleeve, shorted to
folded dipole (see Chapter Two), shown in its the outer conductor at the bottom, but insulated
simplest form in Fig. 11-7. The simple folded from it elsewhere, as in B. In either case, the length
dipole of Fig. 11-7 has a feed-point impedance of of the detuning element is a full quarter wave-
approximately 288 ohms. It may be fed with the
popular 300-ohm line without appreciable mis¬
match.
The impedance at the feed point of a folded
dipole may also be raised by making the diameter
of the conductor used for the fed portion of the
dipole smaller than the conductor used for the
parallel section. Thus, in the 50-MHz array shown
in Fig. 11-17, the relatively low center impedance Fig. 11-11 — Matching circuit for coupling bal¬
is raised to a point where it may be fed directly anced to unbalanced lines.
with 300-ohm line by making the fed portion of C1 — 100-pF variable for 50 MHz, 50-pF for 144
the dipole of 1/4-inch tubing, and the parallel C2 —MHz 35-pF per-section split-stator variable,
section of 1-inch. A 3-element array of similar 0.07-inch spacing. Reduce to 4 stator and 4
dimensions could be matched by substituting rotor plates in each section in 144-MHz coupler
3/4-inch tubing in the unbroken section. Con¬ for easier tuning; see text.
ductor ratios and spacings for other applications J1 - Coaxial fitting, chassis-mounting type.
may be obtained from the folded-dipole nomogram J2 — Ceramic crystal socket.
in Chapter Two. L1 — 50 MHz: 4 turns No. 18 tinned, 1 inch
diameter, 8 turns per inch. (B8<W No.
Stub Matching 3014.)
144 MHz: 2 turns No. 14 tinned, 1 inch
The design and adjustment of the stub match¬ diameter, 1/8-inch spacing. Slip over L2
ing system shown in Fig. 11-8 are described in before mounting.
L2 — 50 MHz: 7 turns No. 14 tinned, 1-1/2 inch
detail in Chapter Three. For experimental work the diameter, 4 turns per inch. (B&W No.
stub may be made of tubing and the connections 3017.) Tap 1-1/2 turns from each end.
to it made with sliding clips. In a permanent 144-MHz: 5 turns No. 12 tinned, 1/2 inch
installation a stub of open-wire line, with all diameter, 7/8 inch long. Tap 1-1/2 turns
connections soldered, may be more satisfactory from each end.
 230 Chapter 11

As shown in Fig. 11-12, the couplers are housed

in aluminum utility boxes, complete shielding
being desirable. These boxes are 3 by 4 by 6 inches
(76 X 102 X 152 mm), and are the two-piece
variety. All the components are mounted on one of
the pieces. With only slight modification a standard
chassis could be used, the shielding being com
plcted by adding a bottom cover.
The two units use similar components. The
main tuning capacitor, C2, is fastened to the front
wall 1-1/4 inches (32 mm) from the left side. The
series capacitor, Cl, and the coaxial fitting, JI, are
1-1/8 inches (29 mm) up from the bottom of the
rear wall and 1-1/8 and 2-3/4 inches (70 mm),
respectively, from the left edge, viewing from the
back. A ceramic crystal socket, J2, is the terminal
Fig. 11-12 — Practical construction of matching for the balanced line. It is mounted on top, one
circuits of the type shown in Fig. 11-11, for 50 inch (25 mm) from the edge.
MHz, left, and 144 MHz, right. Each is built on one The 50-MHz coils are cut from commercially
piece of a two-piece aluminum box. These views available stock inductors. The coupling winding,
are of the inside of the top piece. LI, is inserted inside the tapped coil. The plastic
strips on which the coils are wound keep the two
length; the propagation factor of the line does not coils from shorting to each other, so no mechanical
enter into the picture here. support other than that provided by the leads is
Another device for feeding balanced loads with needed. The leads to LI are brought out between
coaxial line is called a “balun,” and may take the turns of L2, and are insulated from them by
several forms. These methods also provide a 4-to-l two sleeves of spaghetti, one inside the other. Do
impedance step-up at the balanced end. A balun not use the soft vinyl type of sleeving, as it will
may be simply a folded half wavelength of coaxial melt too readily if, through an accident to the
line, connected as shown in Fig. 11-10. antenna system, either coil should run warm. In
A third method is the use of an inductively the 144-MHz unit the method of assembling of the
coupled matching circuit, as shown in Fig. 11-11. coils is reversed, the tuned circuit coil, L2, being
The coupler may be at the transmitter, anywhere inside the coupling coil.
along the transmission line, or in the antenna The components used are adequate for fairly
assembly itself. It should be tuned for minimum high power. Similar tuning capacitors are used in
standing-wave ratio on the coaxial line, and then both couplers, but some of the plates are removed
the transmitter loading should be adjusted to the from the one in the 144-MHz unit. This provides
desired value. easier tuning, but the capacitor may be left in its
Practical couplers of this type are shown in Fig. original condition if desired.
11-12, one designed for 50 MHz and the other for The method of adjusting this type of matching
144 MHz. With the taps on L2 placed as specified circuit is covered in Chapter Three. Use of an SWR
in Fig. 11-11 these circuits will match balanced indicator is highly recommended, since the proper
loads in the range 100-1600 ohms to a coaxial line settings of Cl and C2 will be those that result in
(50 or 75 ohms) and so are suitable for use with the lowest possible SWR in the coax line between
300- to 450-ohm parallel-conductor lines even the transmitter and the matching circuit.
when these lines are operating at a moderately high Loading on the transmitter final amplifier
standing-wave ratio. If a 75-ohm balanced line is should be adjusted at the transmitter, after the
connected at J2 the taps should be moved toward SWR in the coax line has been brought as close as
the center of L2. possible to 1 to 1.

ANTENNA SYSTEMS FOR 50 AND 144 MHZ

A simple dipole may be used on 50 and 144 YAGI ARRAYS FOR 50 MHZ
MHz if a more pretentious antenna cannot be
installed, but it is highly recommended that some A Yagi array is favored by many amateurs for
form of directional array be used. While any of the vhf use because it offers high gain per element and
designs that follow can be adapted to either band, its mechanical assembly is simple. Vhf Yagi arrays
practical considerations usually call for the use of usually employ wide spacing of the elements; 0.15
Yagi-type arrays on 50 MHz. Yagi configurations wavelength or more for the reflector and 0.2
are also employed on higher frequencies, but the wavelength or more for the directors are com¬
small size of the elements needed for 144 MHz and monly used. Closer spacings than these tend to
higher makes it practical to use collinear arrays, make the array tune too sharply to be useful across
stacked Yagis, corner-reflector arrays and other an appreciable portion of a vhf band. Lack of
more complex systems on these bands. sufficiently broad frequency response, even with
VH F and UHF Antenna Systems 231
the screw, solder should be sweated over the nuts
and screw ends as protection against weather
corrosion. The same treatment should be used at
each standoff. Mount a soldering lug on the
ceramic cone, wrap the end of the lug around the
wire, and solder the whole assembly together.
These joints and other portions of the array may
be sprayed with clear lacquer as an additional
protection.
The inner ends of the fed section are 1-1/2
inches apart. Slip the dipole into its aluminum
casting, and then drill through both element and
casting with a No. 36 drill, and tap with 6-32
thread. Suitable inserts for mounting the standoffs
can be made by cutting the heads off 6-32rscrews.
Taper the cut end of the screw slightly with a file
Fig. 11-13 — Lightweight 3-element 50-MHz array. and it will screw into the standoff readily.
Feeder is 52-ohm coax, with a balun for connec¬ Cut the dipole length according to Fig. 11-1 for
tion to the folded-dipole driven element. Balun the middle of the frequency range you expect to
may be coiled, as shown, or taped to supporting
use most. The reflector and director will be
pipe. Dimensions are given in Fig. 11-14.
approximately 4% longer and shorter, respectively.
The closer spacing of the parasitic elements (0.15
wide spacing, is a problem in nearly all antenna wavelength) makes this deviation from the usual
designs for 50 MHz. 5% desirable.
3-Element Lightweight Array
The 3-element 50-MHz array of Fig. 11-13
weighs only 5 pounds. It uses the closest spacing
that is practical for vhf applications, in order to
make an antenna that could be used individually or
stacked in pairs without requiring a cumbersome
support. Tire elements arc half-inch aluminum
tubing of 1/16-inch wall thickness, attached to the
1-1/4-inch dural boom with aluminum castings
made for the purpose. By limiting the element
spacing to 0.15 wavelength the boom is only 6 feet
long. Two booms for a stacked array can thus be
cut from a single 12-foot length of tubing.
The folded-dipole driven element has No. 12
wire for the fed portion. The wire is mounted on Fig. 11-15 — Typical gamma-match construction.
3/4-inch cone standoff insulators and joined to the The variable capacitor, 50 pF, should be mounted
outer ends of the main portion by means of metal in an inverted plastic cup or other device to protect
pillars and 6-32 screws and nuts. When the two it from the weather. The gamma arm is about 12
inches long for 50 MHz, 5 inches for 144 MHz.
halves are pulled up tightly and wrapped around

The folded dipole gives the single 3-element

array a feed impedance of about 200 ohms at its
resonant frequency. Thus it may be fed with a
balun of the type shown in Fig. 11-10, using
52-ohm instead of 75-ohm coax. A gamma¬
matched dipole may also be used, suggested con¬
struction being as shown in Fig. 11-15. If the
gamma match and 72-ohm coax are used, a balun
will convert to 300-ohm balanced feed, if Twin-
Lead or 300-ohm open-wire TV line feed is desired.
If the dimensions are selected for optimum per¬
formance at 50.5 MHz the array will show good
performance and a fairly low standing-wave ratio
over the range from 50 to 51.5 MHz.
Fig. 11-14 — Dimensions of the 3- A closeup of a mounting method for this or any
element antenna shown in Fig. 11-13, other array using a round boom is shown in Fig.
for working in low-frequency section of 11-16. Four TV-type U bolts clamp the horizontal
the 50-MHz band. The elements are and vertical members together. The metal plate is
1/2-inch aluminum tubing. The driven
element in Fig. 11-13 is a folded dipole about 6 inches square. If 1/4-inch sheet aluminum
using No. 12 wire for fed section, as is available it may be used alone, though the
described in the text. photograph shows a sheet of 1/16-inch stock
 232 Chapter 11

If a shorter boom is desired, the reflector

spacing can be reduced to 0.15 wavelength and
both directors spaced 0.2 wavelength, with only a
slight reduction in forward gain and bandwidth.

5-Element 50-MHz Array

As aluminum or dural tubing is usually sold in
12-foot lengths this dimension imposes a practical
limitation on the construction of a 50-MHz beam.
A 5-element array that makes optimum use of a
12-foot boom may be built according to Fig.
11-18. If the aluminum casting method of mount¬
ing elements shown for the 3-element array is
employed the weight of a 5-element beam can he
held to under 10 pounds.
The gamma match and coaxial line are recom¬
mended for feeding such an array. If it is desired to
Fig. 11-16 - Closeup photograph of the boom use 300-ohm line because of its lower losses, the
mounting for the 3-element 50 MHz array. A sheet gamma can be driven through a section of 75-ohm
of aluminum 6 inches square is backed up by a
piece of wood of the same size. TV-type U clamps line sufficiently long to provide for rotation of the
hold the boom and vertical support together at antenna, and then can be converted to a balanced
right angles. At the left of the mounting assembly 300-ohm load at the anchor point by using a balun
is one of the aluminum castings for holding the as shown in Fig. 11-10.
beam elements.

backed up by a piece of wood of the same size for

stiffening. Tempered Masonite is preferable to
wood, as it will stand up better in weather.

High-Performance 4-Element Array

The 4-element array of Fig. 1 1-17 was designed
for maximum forward gain, and for direct feed
with 300-ohm balanced transmission line. The
parasitic elements may be any diameter from 1/2
to 1 inch, but the driven element should be made
as shown in the sketch. The spacing between driven
element and reflector, and between driven element
and first director, is 0.2 wavelength. Between the
first and second directors the spacing is 0.25 Fig. 11-18 - Five-element Yagi to fit on a 12-foot
wavelength. boom. Dimensions are centered at 51 MHz for
The same general arrangement may be used for working over the 50-52 MHz range.
a 3-element array, except that the solid portion of
the dipole should be 3/4-inch tubing instead of 1 Elements should be spaced approximately 0.15
inch. wavelength, or about 35 inches. With 5 or more
With the element lengths given the array will elements, good bandwidth can be secured by
give nearly uniform response from 50 to 51.5 MHz, tapering the element lengths properly. With the
and usable gain to above 52 MHz. dimensions given in Fig. 11-18 the antenna will
work well over the first two megahertz of the
band, provided that the SWR is adjusted for
minimum at 51 MHz.

WIDE SPACED 6-ELEMENT YAGI

High gain can be obtained by extending the
length of a Yagi array to about one wavelength as,
discussed in Chapter Four. At 50 MHz an overall
length of about two wavelengths, requiring a boom
of the order of 20 feet long, is practicable
constructionally. Fig. 11-19 shows a 6-element
array of this type, and Fig. 11-20 gives the element
dimensions and spacings for a center frequency of
Fig. 11-17 — Dimensional drawing of the 4-element 50.6 MHz. The same element arrangement can be
50-MHz array. Element lengths and spacings were used for frequencies higher in the band by sub¬
derived experimentally for maximum forward gain tracting 2 inches from every dimension for each
at 50.5 MHz.
megahertz increase in center frequency. The band-
VH F and UHF Antenna Systems 233

width for 2 to 1 SWR is approximately 1 mega¬

hertz.
The elements in the antenna shown in Fig.
11-19 are halt-inch aluminum tubing, mounted on
a 1-1/4-inch diameter dural boom by the use of
castings of the type used in the 3-element array
described earlier. If a boom of the requisite length
cannot be obtained in one piece it can be made by
splicing two or more pieces together, if suspension
bracing such as is shown in Fig. 11-19 is used. The
boom in this case was made from three pieces of
tubing with a short length of the next smaller
diameter tubing inside the joint. As the fit was
loose, the joint was shimmed with flat strips of
sheet aluminum. After assembly a few sheet-metal
screws were used to make a tight joint. The
element-mounting castings, if used, should be
installed before these screws are put in place.
Joints in the boom cause no serious problem if
the suspension shown is used to provide additional
support. Steel wire can be substituted for the
tubing that was incorporated in this antenna. Fig.
11-21 shows the method of fastening the tubing to
the pipe mast A comparable clamp can be used at
the boom end of the brace.
Fig. 11-21 also shows an alternative method of
mounting the elements on the boom. The clips
should be formed so that when the bolt is
tightened the element is pulled tight against the
bottom edges of the holes in the boom. The boom Fig. 11-19 — Six-element long Yagi for 50 MHz.
of this antenna is fastened to the pipe mast by the Boom has two braces running diagonally to the
method shown in Fig. 11-16, except that the plate mast, a method of construction that permits using
was made long enough so that four U clamps could tubing of relatively small diameter and obviates the
necessity for a one-piece boom. (The antenna at
be used on the boom. the top is a 16-element broadside array for 144
The 6-element antenna is fed with 52-ohm coax MHz.)
through the gamma match shown in Fig. 11-22.
This uses a tubular capacitor of the type described
in Chapter Nine. The outside tube is the same The matching section should be adjusted by
material as that used for the elements, and is using a standing-wave-ratio meter, with the meter
supported from the driven element by ceramic preferably at the antenna during the adjustment
standoff insulators one inch high. These are fast¬ process. The antenna can be temporarily mounted
ened to the tubing with clips ¡nade from sheet eight or ten feet above ground while this adjust¬
aluminum. The sliding arm is 1/4-inch rod insula¬ ment is being made if the feed point is not
ted from the outer tube by polystyrene bushings. accessible in the final location.
One of these is force-fitted on the sliding rod with
its outer diameter such as to make a sliding fit A 5-OVER-5 STACKED ARRAY FOR 50 MHZ
inside the tube. The other bushing fits tightly
inside the outer end of the tube and is drilled at Stacked 50-MHz arrays will show more than
the center so the sliding rod can move freely their theoretical improvement over single bays on
through it. The bushings can be made from poly¬ some propagation paths, and less on others. The
styrene rod. stacked system is likely to work well on most

Fig. 11-20 — Dimension drawing of

the 6-element 50-MHz antenna. The
driven element is fed through a gam¬
ma matching section constructed as
shown in Fig. 11-22.
 234 Chapter 11

in Fig. 11-25, gives some range of impedance

variation. For the system shown, two inches
center-to-center gave a good match to the 50-ohm
main transmission line.
The phasing harness and baluns are made of
RG-62A/U 93-ohm coax, each piece (4 required)
having a shield length of 77-3/4 inches. The line
and balun should be taped to the boom near to the
Fig. 11-21 — Method of mounting suspension driven element, to prevent flexing of the leads and
braces to a pipe mast is shown at the right. The the inevitable breakage that would otherwise re¬
braces are flattened at the ends and bolted to sult. The phasing lines are then run along the boom
semicircular clamps formed from aluminum sheet. and the vertical support, to a coaxial T fitting at
A method of mounting elements on a boom is the midpoint, for connection to the main line. The
shown at the left. Elements go through holes
drilled through the boom, and are held tight by the phasing lines should be taped to the vertical
pull-down clamps as shown. The clamps can be support at frequent intervals. The spacing between
made from sheet aluminum. bays, one-half wavelength, will be about 10 feet.
Greater spacing, up to about 12 feet or 5/8
circuits, and is usually a very good investment for wavelength, will give somewhat more gain. If this
the DX-oriented 6-meter man. can be handled mechanically, either of the phasing
The array consists of two 5-element Yagis lines can be made an electrical half-wavelength
stacked a half wavelength apart, and has been longer (77-3/4 inches) and the bays will still be fed
described by Edward Linde, WB2GXF. Element in phase.
lengths and spacings are given in Fig. 11-24. With
the phasing system used, the spacing could be Adjustment and Use
increased to 12 feet for slightly more gain. The With the dimensions given, checking with an
mechanical design is simple and readily duplicated. SWR indicator inserted between the coaxial T and
Principal details can be seen from the drawings and the main line showed an SWR under 1.3 to 1 from
the photograph of the mock-up, Fig. 11-23, which 51.2 MHz to the low end. The best match, about
shows how elements are mounted to the boom. 1.1 to 1, was in the most-used part of the band
from 50.1 to 50.3 MHz. Such a test should
Mechanical Details preferably be made with the array in the position
The folded-dipole driven elements allow some in which it is to be used. The best alternative is to
range of adjustment of the feed impedance of the prop up the array at ground level, with the booms
system, through the use of small-dia fed portions pointing straight up. Varying the spacing between
whose spacing from the unbroken larger portion the fed and unbroken portions of the folded-dipole
can be varied. The fed portions are 1/8-inch driven elements will provide a range of matching
aluminum rod or hard-drawn wire, with the outer adjustment, if the SWR turns out to be higher than
ends bent 90 degrees and threaded to permit you like. The two bays should have the same
fastening them in place with nuts, once the spacing for each check, if this is done.
optimum spacing from the larger portion is found.
The unbroken portion of the dipole is 1/2-inch
144-MHz PARASITIC ARRAYS
(outside diameter) hard-drawn aluminum tubing,
mounted to the boom by means of an aluminum The main features of the 50-MHz arrays previ¬
plate and a U-clamp, as shown in the mock-up. ously described can be adapted to 144-MHz anten¬
The fed portion is fastened to the lower side of nas, but the small physical size of arrays for this
the boom, using TV feed-line standoff insulators.
Should it be impossible to find suitable insulators
ready-made, the spacing between the large and
small portions of the dipole can be maintained
with blocks or pillars of insulating material. If rod
or hard-drawn wire is used, it will be stiff enough
to require little or no bracing. Softer wire may
need several insulators mounted at intervals along
the dipole. In bending the outer ends of the fed
portions, take the bending slow and easy, and
don’t try to bend at a sharp angle. Hard stiff
materials break easily.
The basic idea of the folded-dipole driven
elements in this system is to develop approxi¬
mately 400 ohms impedance in each bay, which is
then fed with 93-ohm coax and a half-wave balun Fig. 11-22 — Gamma matching section using
as shown in Fig. 11-25. The impedance at the tubular capacitor. The sheet-aluminum clip at the
center of the phasing system thus becomes approx¬ right is moved along the driven element for
imately 50 ohms, and can be fed directly with matching. The small rod can be slid in and out of
the 15-inch tube for adjustment of series capaci¬
50-ohm coaxial cable. Adjustment of the spacing, S tance. The rod should be about 14 inches long.
VHF and UHF Antenna Systems 235

dural. Elements can be 1/4- to 1/2-inch stock.

They can be fastened to the boom by the method
shown in Fig. 11-21 if the relative diameters of
elements and boom are such that the boom will
not be unduly weakened by this construction. An
alternative mounting method is shown in Fig.
11-30.
Recommended spacing for up to 6 elements is
0.2 wavelength, though this is not too critical.
Gamma-match feed is recommended for coax, or a
folded dipole and balun may be used. If balanced
line is to be used the folded dipole is recom¬
mended, the 4 to 1 ratio of conductor sizes being
about right for most designs.

LONG YAGIS FOR 144 MHZ

It becomes practicable, constructionally, at 144
Fig. 11-23 — A model showing the method of MHz and higher frequencies to build Yagi antennas
mounting the elements in the 50-MHz array. An that are several wavelengths long, resulting in
aluminum plate is shown, but suitable angle stock increased gain and directivity as described in
will provide an even stronger assembly. Elements Chapter Four. A representative design is shown in
are 1/2- or 3/8-inch hard-drawn aluminum tubing.

frequency makes it possible to use larger numbers

of elements with ease. Few 2-meter antennas have
less than 4 or 5 elements, and most stations use
more, either in a single bay or in stacked systems.

A 4-Element Array
Parasitic arrays for 144 MHz can be made
readily from TV antennas for Channels 4, 5, or 6.
The relatively close spacing normally used in TV
arrays makes it possible to approximate the recom¬
mended 0.2 wavelength at 144 MHz, though the
element spacing is not a critical factor. A 4-element Adapter
soAHaùv
array for 144 MHz that can be made from a transmission line
Channel 6 TV Yagi is shown in Fig. 11-26. It may T-Fitting
be fed with a gamma match and 52-ohm coax, as
shown. However, most TV antennas are designed
for 300-ohm feed, and the same feed system can be To similar
lower bag I
employed for the 2-meter array that Is made from
them.
If one wishes to build his own Yagi antennas Fig. 11-25 — Details of the driven elements and
from available tubing sizes, the boom of a 2-meter phasing system used in the 50-MHz stacked array.
The folded dipole is set up so that its feed
antenna should be 3/4 to 1 inch aluminum or impedance is approximately 400 ohms. Two half¬
wave baluns and half-wave phasing sections step
down to about 100 ohms. The two bays thus
connected in parallel may be fed with 50-ohm coax
directly.

Fig. 11-27. It uses 13 elements - reflector, driven

element, and eleven directors - with the director
lengths following the third (D 3) successively 1/4
inch shorter. The lengths given apply when the
type of element mounting shown in Fig. 11-28 is
used - i.e., the element is not mounted through
the boom but is fastened to it on top — and when
the elements are 3/32 inch in diameter. Steel rods
were used as elements in the original model of this
antenna. According to the designers, W2NLY and
W6QKI, the element diameter should not exceed
1/8 inch.
The driven element (which need not be of the
Fig. 11-24 — Element lengths and spacings for the same construction as the reflector and directors)
WB2GXF 5-over-5 array for 50 MHz. can be a folded dipole or can be fed by any of the
 236 Chapter 11

supporting members in back of the elements,

particularly when all-metal construction is em¬
ployed.

A 12-Element Array
Six half waves in phase, with parasitic reflec¬
tors, may be used as shown schematically in Fig.
11-29. The mechanical features of the 12-element
array are shown in Figs. 11-30 and 11-31. The
spacing of the reflectors in this array is made 0.15
wavelength, to bring down its feed impedance to
the point where it can be fed with 300-ohm line
Fig. 11-26- Four-element 144-MHz Yagi. Gamma
matching is recommended, using a gamma section without appreciable mismatch. Dimensions are
of similar construction to that shown in Fig. 11-15, given in the caption for Fig. 11-32.
with a gamma rod having a length of 6 inches. Thé
series capacitor should be a 50-pF variable; re¬ A 16-Element Array
ceiver-type plate spacing is adequate for power
levels up to a few hundred watts. Designs similar to that given for the 12-element
system may be used for eight half waves in phase,
with reflectors, as shown in Fig. 11-33. (This
matching systems discussed earlier. The feed-point
impedance with a simple dipole fed element is 15
to 20 ohms, about the same as in a 3-element Yagi
of ordinary design.
The length of the antenna can be extended with
some increase in gain (see Chapter Four) by adding
similarly spaced directors, similarly tapered in
length. It may also be made shorter, with reduced
gain, by cutting off any desired section to the right
ofD sin Fig. 11-27.
In terms of wavelengths, the optimum element
spacings for an antenna of this type are slightly
oyer 0.1 wavelength from driven element to first
director and between the first, second and third
directors, 0.2 wavelength from third to fourth
director, and 0.4 wavelength between succeeding
directors. Fig. 11-28 — Boom support and element mounting
used in the 13-element Yagi antenna of Fig. 11-27.
COLLINEAR ARRAYS FOR 144 MHZ
Excellent performance in antenna systems for antenna is the uppermost one in Fig. 11-19.)
144 MHz and higher bands is obtainable through Element dimensions are the same as for the
the use of curtains of 4, 6, 8 or more elements, 12-element array, given in the caption for Fig.
arranged in pairs and fed in phase. Parasitic 11-32.
reflectors are usually mounted 0.15 to 0.25 wave¬ The extra elements bring the feed impedance of
length in back of each driven element, though the this system down, so the reflector spacing is made
driven elements alone may be used in a bidirec¬ 0.2 wavelength for the 16-element array. The
tional array. Screen reflectors are also used with impedance is usually slightly below 300 ohms for
collinear elements. Such arrays may employ either such an arrangement, though not low enough so
horizontal or vertical polarization. Horizontal is that the efficiency is greatly affected. A 300-ohm
shown in the examples.
The supporting structure may be either wood
Fig. 11-27 — Thirteen-
or metal. If the elements are mounted at their element long Yagi antenna
for 144 MHz
(W2NLY-W6QKI). Dimen¬
sions are for optimum per-
formance in the
144-145-MHz segment of
the band. For maximum
performance in higher por¬
tions, decrease the ele¬
ment lengths 1/4 inch for
each megahertz increase in
frequency. Dimensions
shown apply only for the
type of parasitic element
construction described in
the text and shown in Fig.
11-28.
VHF and UHF Antenna Systems
237

Fig. 11-29 — Element arrangement and feed system Fig. 11-30 — Model showing the method of
of the 12-element curtain array. Reflectors are assembly for all-metal construction of phased
spaced 0.15 wavelength behind the driven ele¬ arrays. Dimensions of clamps are given in Fig.
ments. 11-31.

transmission line may be connected at the mid¬

point of the phasing line. However, the builder The clamp method of assembly results in a
may wish to experiment with an adjustable Q strong structure that will hold its alignment indef¬
section at the feed point to achieve a more precise initely. Clamps for combinations of tubing sizes
match. One feed method that has been employed other than those given may be made by bending up
with the 16-element array is to use a quarter-wave clamps experimentally out of soft metal or card¬
Q section of 300-ohm Twin-Lead, and then make board, and then using these as templates for
the main transmission line of open-wire line of 400 cutting and drilling the sheet aluminum. The stock
to 500 ohms impedance. should be 1/16 inch or heavier, and the clamps
The Q section may also be any odd multiple of should be assembled with No. 8 or larger screws.
a quarter wavelength. This suggests the use of the Lock washers should be used under all nuts.
flexible insulated Twin-Lead as the rotating por¬ Clamps and the screws and nuts should be sprayed
tion of the transmission line, bringing it to an with clear lacquer when the assembly is completed.
anchor point just below the array, where open-wire Use two coats for maximum protection.
line may comprise the balance of the run. This
method is used in the array shown in Fig. 11-19.

Very Large Arrays

Where more than 16 elements are used in a
collinear array, the system is usually broken down
into separate 12- or 16-element groups, and these
are fed in phase. This is done to achieve balanced
current distribution. A 24-element array made of
two 12-element sets is shown schematically in Fig.
11-34 (reflectors are omitted from the drawing).

All-Metal Construction
Collinear arrays may be made very light in
weight and low in wind resistance, and still have
strength to withstand the most severe weather
conditions, if all-metal design is employed in the
manner shown in Figs. 11-30, 11-31 and 11-32.
Elements, supporting arms, and vertical and hori¬
zontal supports are all of aluminum or dural
tubing, and are held together by clamps made from Fig. 11-31 - Detail drawings of the clamps used to
sheet aluminum. Dimensions for the clamps assemble the all-metal 2-meter array. A, B and C
required when the members are 1/4-inch, 3/4-inch are before bending into "U" shape. The right-angle
and 1-1/2-inch tubing are given in Fig. 11-31. A bends should be made first, along the dotted lines
model showing the method of assembling is shown as shown, then the plates may be bent around a
in Fig. 11-30, and the method of assembling a piece of pipe of the proper diameter. Sheet stock
12-element array is given in Fig. 11-32. should be 1/16-inch or heavier aluminum.
 238 Chapter 11

QUADS FOR 144 MHZ

Though it has not been used to any great extent
in vhf work, the quad antenna has interesting
possibilities. It can be built of very inexpensive
materials, yet its performance should be at least
equal to other arrays of its size. Adjustment for
resonance and impedance matching can be accom¬
plished readily. Quads can be stacked horizontally
and vertically, to provide high gain, without
sharply limiting the frequency response.

The 2-Element Quad

The basic 2-element quad array for 144 MHz is
shown in Fig. 11-35. The supporting frame is 1 by
1-inch wood, of any kind suitable for outdoor use.
Elements are No. 8 aluminum wire. The driven
element is one wavelength (83 inches) long, and
the reflector 5 percent longer, or 87 inches.
Dimensions are not particularly critical, as the
quad is relatively broad in frequency response.
The driven element is open at the bottom, its
ends fastened to a plastic block, which is mounted
at the bottom of the forward vertical support. The
top portion of the element runs through the
support and is held firmly by a screw running into
the wood and then bearing on the aluminum wire.
Feed is by means of 52-ohm coax, connected to
the driven element loop. Fig. 11-33 — Schematic drawing of a 16-element
array. A variable Q section may be inserted at the
feed point if accurate matching is desired. Re¬
flector spacing is 0.2 wavelength.

The reflector is a closed loop, its top and

bottom portions running through the rear vertical
support. It is held in position with screws, top and
bottom. The loop can be closed by fitting a length
of tubing over the element ends, or by hammering
them flat and bolting them together, as shown in
the sketch.
The elements in this model are not adjustable,
though this can easily be done by the use of stubs.
It would then be desirable to make the loops
slightly smaller, to compensate for the wire in the
adjusting stubs. The driven-element stub would be
trimmed for length and the point of connection for
the coax would be adjustable for best match. The
reflector stub could be adjusted for maximum gain
or front-to-back ratio, whichever quality the
builder wished to optimize.
In the model shown only the spacing is adjus¬
ted, and this is not particularly critical. If the
wooden supports are made as shown, the spacing
between the elements can be adjusted for best
Fig. 11-32 — Supporting framework for a 12- match, as indicated in an SWR meter connected in
element 144-MHz array of all-metal design. Dimen¬
sions are as follows: element supports (1) 3/4 by the coaxial line. The spacing has little effect on the
16 inches; horizontal members (2) 3/4 by 46 gain, from 0.15 to 0.25 wavelength, so the varia¬
inches; vertical members (3) 3/4 by 86 inches; tion in impedance with spacing can be utilized for
vertical support (4) 1-1/2-inch diameter, length as matching. This also permits use of either 52- or
required; reflector-to-driven-element spacing 12 72-ohm coax for the transmission line.
inches. Parts not shown in sketch: driven elements
1/4 by 38 inches; reflectors 1/4 by 40 inches; Stacking
phasing lines No. 18 spaced 1 inch, 80 inches long^
fanned out to 3-1/2 inches at driven elements Quads can be mounted side by side or one
(transpose each half-wave section). The elements above the other, or both, in the same general way
and phasing lines are arranged as shown in Fig as described for other antennas. Sets of driven
11-29. M
elements can also be mounted in front of a screen
VHF and UHF Antenna Systems 239
Long-wire arrays may be constructed according
to the principles given in Chapter Five, designing
them specifically for vhf use. In such instances an
effective rhombic array assumes proportions that
make it usable in many amateur locations where a
similar array for 7 or 14 MHz would be out of the
question because of its great size. By proper choice
of leg lengths a V or rhombic can be made to work
well on both 28 and 50 MHz, making it a highly
useful system where the necessary space is available
for its erection. Several examples are given in Table
11-1.
The tilt angles and leg lengths in wavelengths
for other combinations can be worked out from
the design data in Chapter Five, using the curves
for zero wave angle. The wire lengths in feet are
then given by
492 (N - 0.05)
Length (feet) =-
Freq. (MHz)
where N is the number of half waves on the leg.
The above formula need be used only where the leg
length is short in terms of wavelength. For longer
dimensions the standard half-wave formula may be
used:
Fig. 11-34 — Method of feeding the driven ele¬ 492 TV
Length (feet) =-
ments of a 24-element array. Phasing lines may be Freq. (MHz)
open-wire line with 1/2 to 1 inch spacing between
conductors. Long-wire systems for combining operation on
50 and 144 MHz are even more attractive as to
reflector. The recommended spacing between adja¬ size. Because of the nearness to third-harmonic
cent element sides is a half wavelength. Phasing and relationship which exists between these two bands,
feed methods can be similar to those employed the same matching section and feeder may be used
with other antennas described in this chapter. to feed a terminated rhombic for both bands with
a flat line. Since a Q section can be any odd
Adding Directors multiple of a quarter wavelength, the matching
section for a two-band vhf rhombic can be a
Parasitic elements ahead of the driven element quarter wavelength long at 50 MHz, in which case
work in a manner similar to those in a Yagi array. it will be approximately three quarter waves long
Closed loops can be used for directors, by making at 144 MHz. The feed impedance of a terminated
them 5 percent shorter than the driven element, or rhombic is about 800 ohms; thus, a 490-ohm Q
about 79 inches. Spacings can be similar to those
for conventional Yagis. In an experimental model
built by W8HHS the reflector was spaced 0.25
wavelength and the director 0.15. A square array
using four 3-element bays worked out extremely
well.
Workers using quads on 144 MHz have reported
reduced fading, compared with horizontal Yagis.
Possibly this is due to the presence of some vertical
polarization with the quad, making it less affected
by polarization changes that tend to occur over
long paths.

LONG-WIRE ARRAYS FOR 50 AND 144 MHz

Long-wire antenna systems such as the V or the
rhombic can usually be used on 50 or even 144
MHz with good results even though they were Fig. 11-35 - Mechanical details of a 2-element
quad for 144 MHz. Driven element, L1, is one
erected with lower-frequency operation in mind. wavelength long; reflector, L2, 5 percent longer.
The included angles in such arrays will not be Sets of elements of this type can be stacked
optimum for vhf, but the arrays will be so large, in horizontally and vertically for high gain with broad
terms of vhf wavelengths, that they will work well, frequency response. Bay spacing recommended is
particularly if the feeder systems are not too long. 1/2 wavelength between adjacent element sides.
They will show little frequency discrimination over Example shown may be fed directly with 52-ohm
an entire vhf band. coax.
 240
Chapter 11

Fig. 11-36 — A 144-MHz rhombic with an esti¬

mated 27-dB gain over a dipole. The wires are all
on the horizontal plane with the crossovers insula¬
ted.
L1 — 29.5 feet
L2 — 50.67 feet
X - 52.2°
Y - 37.7°
R1-2 - 660 ohms, total wattage should equal half
the power output of the transmitter.
Height above ground - 12.29 feet
Elevation angle — 7.5°
Vertical beamwidth - 5.5°
Horizontal beamwidth - 8.5°

are based on the assumption that the lowest

possible radiation angle is desired, in which case
one side should be a half wave longer than half the
overall length. Using the terms of Table 11-1:
B 480
A = - +-
2 Freq. (MHz)
The shape of a multiband V or rhombic may be
set up according to Table 11-1 with its width, C, at
the optimum value for the band where highest
efficiency is desired. It will be noted that the larger
the array the less difference there is in the included
angles for adjacent bands. In other words, the
section is required to match this impedance to a larger the array the better will be its capabilities for
multiband operation.
300-ohm line. Such a matching section could be
made of No. 14 wire spaced 1-3/4 inches, about 53
inches long, as a compromise for the two bands. AN IMPROVED RHOMBIC FOR 144 MHZ
The array could be fed directly with a 600-ohm A rhombic with improved performance (Fig.
line, without appreciable mismatch. Preferably 11-36) was designed by Mike Staal, K6MYC, based
such a line would be of small wire, in order to keep on specifications given by E. A. LaPort and A. C.
the spacing to relatively small dimensions. See Veldhuis. It features lower sidelobes than previous
Chapter Three for wire sizes and line impedances. types, and has an estimated gain of 27 dB over a
Laying out a rhombic antenna for the vhf bands dipole.
is somewhat less complicated than for lower The narrow beamwidth reduces the usefulness
frequencies, because it is usually possible to have of this antenna for general purposes, but for a
the vhf array high enough (in terms of wave¬ specific path or EME window it should be very
lengths) so that the effect of ground is a minor effective. It has been used by several European and
consideration. The dimensions given in Table 11-1 U. S. amateurs.

ANTENNAS FOR 220 AND 420 MHZ

The use of high-gain antenna systems is vir¬ as even high-gain systems may be of table-top
tually a necessity on 220 MHz and higher frequen¬ proportions.
cies, if communication is to be carried on over Any of the arrays already described may be
other than line-of-sight distances. Experimentation scaled down for use on 220 and 420 MHz by
with antenna arrays for these frequencies is fascin¬ reducing all dimensions in proportion to the
ating, and the size of elements and supporting wavelength, or in inverse proportion to the fre¬
structures is such that various element arrange¬ quency. Using 144-MHz designs as a base, the scale
ments and feed systems can be tried with ease. factor is 144/220, or 0.655, for converting a
Arrays for 420 MHz, particularly, are ideal for 144-MHz antenna design to 220 MHz, and
scale-model demonstrations of antenna principles, 144/420, or 0.343, for converting 144 MHz to
VH F and UHF Antenna Systems 241

420. Using the scale factor requires reducing all

dimensions, including the element diameters. How¬
ever, if a different length/diameter ratio is used in
the elements the proper lengths can be found from
Fig. 11-1. In most cases the necessary modification
will not be large.
On 220 and 420 MHz the broad frequency
response and ease of adjustment of collinear
systems make them more attractive than the more
critical Yagi configurations. The use of plane and
corner reflectors becomes practical from the stand¬
point of overall size, and even parabolic reflectors
are usable. Additional details of parabolic re¬
flectors may be found in Chapter Twelve.

SCREEN-REFLECTOR ARRAYS
At 220 MHz and higher, where their dimensions
become practicable, plane-reflector arrays are wide¬
ly used. Except as it affects the impedance of the
system, as shown by the curve marked 180° in Fig.
11-37, the spacing between the driven elements Fig. 11-37 — Feed impedance of the driven
element in a corner-reflector array, for various
and the reflecting plane is not particularly critical. corner angles of 180 (flat sheet), 90, 60 and 45
Maximum gain occurs around 0.1 to 0.15 wave¬ degrees.
length, which is also the region of lowest imped¬
ance. Highest impedance appears at about 0.3
the area of the driven elements, extending at least a
wavelength. With a spacing of 0.22 wavelength
quarter wavelength on all sides. Chicken wire on a
between the driven elements and the screen reflec¬
wood or metal frame makes a good plane reflector.
tor, the impedance of the elements approximates
Gosely spaced wires or rods may be substituted,
that of a dipole in free space. As the gain of a
for lower wind resistance, with the spacing be¬
plane-reflector array is nearly constant at spacings tween them running up to 0.1 wavelength without
from 0.1 to 0.25 wavelength, the spacing may be
appreciable loss in effectiveness.
varied to obtain an impedance match.
An advantage of the plane reflector is that it The Corner Reflector
may be used with two driven-element systems, one
on each side of the plane, providing either for When a single driven element is employed, the
two-band operation or the incorporation of hori¬ plane reflector may be bent to form an angle,
zontal and vertical polarization in a single struct¬ giving an improvement in the radiation pattern and
ure. The gain of a plane-reflector array is slightly gain. At 220 and 420 MHz its size assumes
higher than that of a similar number of driven practical proportions, and it can even be used at
elements backed up by parasitic reflectors. The 144, though usually at less than optimum size.
plane-reflector array also has a broader frequency The comer angle can be 90, 60, or even 45
response and higher front-to-back ratio. To achieve degrees, but the side length must be increased as
these ends, the reflecting plane must be larger than the angle is narrowed. The driven-element spacing

Fig. 11-38 — Con¬

struction of a cor¬
ner-reflector array.
Frame can be wood
or metal. Reflector
elements are stiff
wire or tubing. Di¬
mensions for three feedline
bands are given in
Table Re-
flector element spa¬ Hm(e
(optional)
cing, G, is the max¬
imum that should
be used for the fre¬
quency; closer spa¬ Center support
cings are optional. (wood or metal)
Hinge permits fold¬
ing for portable use.
 242
Chapter 11

TABLE 11-11
Dimensions of Corner-Reflector Arrays for 144, 220, and 420 MHz
Dipole
Side to Reflector Reflector Corner
Length Feed
Vertex Length Spacing Angle Im¬
Band "S" "D" “L" “G” “V"
(MHz) (Inches) (Inches) (Inches) pedance
(Inches) (Degrees) (Ohms)
144' 65 27.5 48 7-3/4
144 90 70
80 40 48 4 90 150
220' 42 18 30 5
220 90 70
52 25 30 3 90
220 100 150
25 30 screen 60 70
420 27 8-3/4 16-1/4 2-5/8
420 90 70
54 13-1/2 16-1/4 screen 60 70
slWrSon in"^ °' re,leCt °r S°meWhat below ° P"mum "

from the corner can be anything from 0.25 to 0.7 optimum for gain. The antenna should be sup¬
wavelength for a 90-degree corner, 0.35 to 0.75 for ported near its mechanical balance point, roughly 2
a 60-degree one, and 0.5 to 1 for a 45-degree feet from the reflector end. If a TV-type U clamp
corner. Feed impedance for various corner angles is used, it is well to bend up a U-shaped metal plate
and spacings is given in Fig. 11-37. Since the the width of the boom and about 3 inches long,
spacing is not critical as to gain, it may be varied to and slip it over the boom at the point where the
achieve impedance matching.
holes are to be drilled for the clamp. This protects
Gain with a 60- or 90-degree corner with the boom from crushing when the U-clamp nuts
1-wavelength sides runs about 10 dB. A 60-degree are tightened, and leaves it strong enough to stand
comer with 2-wavelength sides has about 12 dB up well without bracing. Gusset plates of wood or
gain. It will be seen that this is not outstanding for Masonite make stronger assemblies.
the size of such an array, but there are other Parasitic elements in the 432-MHz model are
advantages. A corner may be used for several made of 3/32-ineh aluminum welding rod, and
bands, for example, or perhaps for uhf television 1/8-inch rod is used for the 220-MHz model. This
reception, as well as for amateur uhf work. material can be purchased very reasonably at
A suggested arrangement for a corner reflector welding supply houses, usually in 3-foot lengths.
system is shown in Fig. 11-38. Sheet metal or wire Any stiff wire or rod up to 1/8 inch diameter will
mesh may be used with equal effectiveness for the do. Drill the boom for a hole size that will just take
reflecting plane. A series of rods, as shown, is the elements with a force fit, then run a 1/2-inch
equally good, if the space between them is kept brass or aluminum screw into the boom to bear on
under 0.06 wavelength at the highest frequency for the element and hold it in place. The screws can
which the reflector is to be used. The frame may then be bonded together and connected to ground
be made of wood, with a hinge at the corner to for lightning protection, if desired.
facilitate portable work or assembly atop a tower. The driven elements originally tried were step-
Principal dimensions for corner-reflector arrays for up folded dipoles similar to those used in the
144, 220, and 420 MHz are given in Table 11-11. 144-MHz Yagis, but it was found that these did not
These dimensions are not critical, because of the work well at 220 and 432 MHz. This is probably
broad frequency response of any plane-reflector the result of the spacing between the two parts of
system.
such a dipole being a considerable portion of a
wavelength at these frequencies. The 432-MHz
YAGIS FOR 220 AND 432 MHZ Yagi was made with a driven element of the same
material as the parasitic elements, mounted as
Moderate-size Yagis for the 220- and 420-MHz shown in Fig. 11-39A and B. Blocks of insulating
bands can be built at very low cost, and with only
material 1/4 inch thick and 1-1/2 inches square are
simple tools, if the suggestions of Figs. 11-39 and
fastened to the boom with two 1-1/2-inch brass
40 are followed. Booms are 1 x l-inch wood,
screws and nuts. The upper portion of the dipole
available in any lumberyard. (Your dealer will call
runs through the boom, just above the center, and
it “one by one” but the actual size will be more
the lower is held in place with 4-40 nuts on either
like 3/4 X 7/8 inch.) Be sure that it is straight, dry,
side of the insulating plates, as shown in the end
and free of knots. Take the man’s advice as to
view, B. The 3/32-inch rod is easily threaded for
which kind of wood will be best for outdoor use,
4-40, if this is done before the element is bent. The
as available stocks vary around the country. Prime
total length of the wire is about 25 inches. An
and paint it thoroughly, if you want the array to
last well. alternative to threading is to hammer the ends flat,
and drill for 4-40 screws.
An 11-element array is shown for 432 MHz
The antenna is matched by means of a universal
and a 7-element one for 220 MHz, both using
stub, shown in Fig. 11-39C, made of the same
element spacings and lengths that are close to
material as the elements. It should be cut about 15
VHF and UHF Antenna Systems 243

Fig. 11-39 — Details of a

6-foot 11-element Yagi for
432 MHz. The square
boom and one polystyrene
mounting block for the
driven element are shown
at A. The blocks, element,
and boom are shown in
detail in the end view, B.
Matching stub, C, fastens
to ends of the driven ele¬
ment, and is mounted
under the boom between
two poly blocks. Element
lengths and spacings for
the middle of the 420-MHz
band are shown in the side
view of the complete array.

inches long, and suspended under the boom. An tubing at the outer ends. The slope down to the
adjustable short and two sliding clips for con¬ feed point at the polystyrene blocks determines
necting the transmission line or balun are provided the impedance. With the dimensions shown the
for adjusting the matching. The ends of the stub array can be fed with 52-ohm coax and a balun,
that are connected to the dipole are pounded flat connected to the lugs at the insulating plates. The
with a hammer, and then drilled to pass the SWR is under 1.5 to 1 from 220 to 224 MHz, with
threaded ends of the dipole. These are held in place optimum match at about 221.5 MHz.
by the 4-40 nuts shown in B. A ceramic cone
standoff insulator (not shown in the drawing) is A 13-ELEMENT YAGI FOR 432 MHZ
fastened to the underside of the boom. Two pieces
of polystyrene similar to that used for the dipole A Yagi with high performance has been devel¬
mounting blocks, one above and one below the oped by K2R1W and duplicated by many uhf
matching stub, are fastened to this cone, clamping enthusiasts with excellent results. It is attractive
the stub in place. from the construction standpoint as well because
The short and the point of connection of the of the aluminum boom, which has an excellent
balun are adjusted for zero reflected power, as strength-to-weight ratio. The uncertainty of the
indicated by an SWR meter connected in the line. detuning effect from all-metal construction is
The 220-MHz Yagi can be made in the manner alleviated by mounting the elements in small
just described, using a dipole made of a single piece plastic blocks, Fig. 11-41.
of wire. The universal stub for matching should be For those who like to experiment with an¬
about 28 inches long, to assure an adequate range tennas, this method of element fastening is ideal.
of adjustment. A variation of the ratio-type folded The block and element assembly may be held in
dipole was made for the 220-MHz antenna as place on the boom by a rubber band or a nylon tie
shown at B in Fig. 11-40. Here a flat strip of of the variety used to secure cables. Changing the
aluminum comprised the fed portion of the dipole, position of the elements to check the gain or
and a 3/8-inch tube the unbroken portion. The pattern characteristics of an antenna becomes easy
aluminum strip is bolted to the underside of the with this temporary fastening. After the best

Fig. 11-40 — Seven¬

element 220-MHz Yagi on
a 6-foot wood boom. Poly
blocks located on each side
of the boom support the
modified folded-dipole
driven element. The latter
has a sloping lower por¬
tion, for matching 52-ohm
coax. A balun is connected
to the lugs shown at the
bottom of sketch B. With
element lengths and spa¬
cings given in the side view
of the array, optimum per¬
formance is obtained over
the first 3 to 4 megahertz
of the band.
244 Chapter 11

system, as shown in Fig. 11-42. The array is fed

TABLE 11-111 with a universal stub and a coaxial balun.
K2R1W 432 MHz Yagi If the individual Yagis are of the wood-boom
Element Lengths and Spacings construction, the framework for supporting them
Length Spacing from DE in 2- or 4-bay arrays can be of the same size
Element
material. For arrays of all-metal construction, the
Inches cm Inches cm

R 33.813
DE
13-5/16
12-3/8 31.432
6-1/4 15.875
supporting framework can be of the same or
DI 11-3/4 29.845 3 7.62 slightly larger size tubing as that for the boom.
D2 5-11/16 14.44
D3 9-1/16 23.01 Suitable diagonal bracing is a must for either type
D4 14-9/16 36.98 of assembly.
D5 22-1/16 56.03
D6 33-7/8 86.04 Each individual Yagi in the array should be
D7
D8
44-1/2
55-1/8
113.03
140.01
adjusted for a good match, preferably with the
D9 65-3/4 167.00 system pointing straight up at a clear area. After
DIO
Dll
76-3/8
87
193.99
220.98
the phasing harness is connected, the short on the
universal stub and the position of the balun
R — Reflector; DE — Driven Element; D — Director
connection should be moved in small increments
for best match to the feed line.

VERTICAL POLARIZATION
position is found, the plastic block can be secured
to the boom with aluminum pop rivets. OF YAGI ANTENNAS
Anyone who is not so experimentally inclined
Of considerable interest to amateurs using vhf
can duplicate the antenna with the dimensions
given in Table 11-111 and obtain good results. This fm is a vertically polarized antenna with some
directional gain. This type of antenna would be
type of Yagi has been used by several entrants in
useful for gaining access to a repeater from some
antenna gain-measuring contests where gain figures
for individual Yagis made to these dimensions were distance away or if the operator is using a
better than 15 dB over a dipole. Four-bay Yagi low-powered transmitter. It is also helpful in
arrays have been consistent winners of such con¬ communicating on the commonly accepted sim¬
tests. plex frequencies, where it is desirable to match the
It should be emphasized here that the boom polarization of the antenna on the mobile station.
diameter, element diameter, element-to-boom spac¬ Most Yagi antennas or arrays, as described
ing, and boom shape are all interrelated factors, earlier in this chapter, may be mounted with the
and a change in any one of them will require elements in a vertical plane. Individual Yagis may
complete retuning of the Yagi for best gain. If the be stacked in varied configurations, one above the
builder is not equipped to measure gain with other, side-by-side, or in groups of four or more.
sensitive and accurate equipment, then he should There is only an insignificant difference in perfor-
follow the instructions precisely.
The lengths of the elements will be found to be
more critical than the spacing. Tolerances should M0UNTIHG ARRANGEMENT, ALL ELEMENTS
be to 1/64 inch. The feed is arranged for 200
ohms, such as 52-ohm coax and a 4-to-l balun. A BOOM T 00 ALUMINUM
good balance point for mounting the Yagi should
be between directors five and six.
¿ ALUMINUM ELEMENT FORCE FIT WITH OAB OF
2-BAY AND 4-BAY ARRAYS FOR 432 MHZ 8 POP RIVET SILASTIC RTV ADHESIVE ON EACH SIDE

The 432-MHz Yagis described above can be

used effectively in stacked pairs, or in a 4-bay
DRIVEN ELEMENT

POP RIVET
— CENTER TO CENTER brass ROO
I, ... I MOUNTING BLOCK
SOLDER

WIRE

AS* BEND

- S CENTER TO REXOLITE MOUNTING BLOCKS

16 CENTER
(20 mm)

Fig. 11-41 — A 13-element Yagi for 432 MHz as

designed by K2RIW. The driven element should be
fed with a 4-to-1 balun and 50-ohm line. For
element spacings, see Table 11-111.
VH F and UHF Antenna Systems 245

Fig. 11-42 — Phasing arrange¬

ments for two and four
432-MHz 11 -element Yagis.
Bay spacing of approximately
two wavelengths is set by the
length of the phasing lines.
The universal-stub matching
device may be used with any
type of transmission line, as
well as with the coaxial line
and balun as shown.

manee of such arrays, whether the elements are material for at least 1/4-wavelength beyond the tip
vertical or horizontal. of the nearest element. Any diagonal bracing for
The one consideration that must be applied to a the boom should also be nonmetallic.
vertically polarized Yagi or array is that the Another method of mounting vertically polar¬
support must not disturb the beam pattern. For a ized Yagis is to place two of them side-by-side, as
single Yagi mounted as in Fig. 11-43A, the vertical in Fig. 11-43B, with 1-wavelength spacing between
support should be of wood or other nonconducting the two booms. The main support will be in the
material where it is in the vicinity of the active center of the boom, which allows 1/2-wavelength
elements of the beam. A good rule of thumb would of space to the nearest element. In this case, the
be to make the support from a nonmetallic support mast and the boom can both be of metal.

NONDIRECTIONAL VHF ANTENNAS

Most amateur vhf communication is carried on Ground-Plane Antennas
with directive arrays of one kind or another, but in
some types of work it is desirable to radiate power When an antenna is mounted at the top of a
equally in ail directions. For such work vertical metal mast, standing waves may develop on the
polarization is generally used. Any of the dipole mast or on the coaxial cable used to feed the
arrangements mentioned earlier in this chapter will antenna if the support is nonmetallic. When this
give essentially uniform radiation patterns when occurs, radiation from the mast combines with that
mounted in a vertical position, but there are from the antenna to raise the angle of radiation,
modifications that are better adapted to such thereby reducing the effectiveness of the system.
service. The ground-plane type of antenna has largely

( B>

Fig. 11-43 - Methods of mounting Yagi antennas

with the elements vertical. The boom-to-mast joint
may be strengthened by a gusset plate. If the arrays
are large enough to require diagonal bracing, such
bracing should be of nonconducting material
where it is close to the elements.
 246
Chapter 11

RADIATOR Fig. 11-44 — The simple 1/4-wavelength ground

plane.
/AÀ

another hole to take a self-tapping screw for

connecting the inner conductor.
The same procedure is used on six meters.
However, on this band, the coax is brought up
outside the mast in permanent installation. The
dimensions shown in Fig. 11-47 will provide a
starting point tor adjustments. The antennas shown
are used on 146.5 and 52.525 MHz.
The completed antenna can be mounted on the
top of a tower or mast, providing omnidirectional
coverage. Gain of these antennas is the same as that
of a vertical half-wave dipole. The coax and feed
overcome this problem because the horizontal points should be waterproofed with a good sealing
ground plane is an effective shield between the compound.
antenna and mast. The fundamental principles of
the ground plane are given in Chapter Two, and
matching it to commonly available feed lines is
discussed in Chapter Three.
Dimensions for ground-plane antennas for some
popular vhf bands are given in Fig. 11-44. For a
more precise match to the transmission line, a
shorted 1/4-wave stub may be connected in parallel
with the antenna at the feed point. If this is done,
a slight shortening of the driven element will be
necessary.

A COMBINATION 6- AND
2-METER J POLE
Credit for this J-Pole system goes to W5WEU
who has used these two antennas as omni¬
directional radiators with excellent results. The
overall antenna is shown in Fig. 11-45. The
antenna mast is 1-1/4-inch diameter pipe, 20 feet
long. This length can be made up from two 10-foot
lengths of TV masting. The first step is to make the
2-meter J pole as shown in Fig 11-46. Note that
the stub dimension of 19 inches is the total length
above the metal support brackets.
Once the stub is mounted, the antenna can be
temporarily supported in a vertical position. Con¬
nect an SWR indicator in the feed line, and slide
the outer braid of the coax (touching the main
mast) and the inner conductor (touching the stub)
up and down until you get a low SWR. The point
where the SWR is the lowest is where the hole will
be drilled in the main mast for the coax. (The coax
will be snaked up inside the main mast) Tem¬
porarily ground the braid with a self-tapping screw
at the point of lowest SWR. Next, carefully adjust
the inner conductor up or down the stub until a
Fig. 11-45 - Constructional details for the 2- and
match is obtained. At the point of matching, drill 6-meter J poles.
VH F and UHF Antenna Systems 247

MULTIELEMENT ARRAYS

Many of the repeater installations in the coun¬

10g"
try use either stacked half-wave dipoles or stacked TO TOP
coaxial elements. Also, most repeater owners use
SHORT metal
commercially manufactured antennas. There is no self-threading
doubt that the antenna manufacturer has an screws;
PLEXIGLAS 2 METER COAX
advantage over the home constructor simply be¬ SPACER inside
cause of the tooling and material-procurement
advantages. However, the home builder can make
gain arrays which are as good as or better than
manufactured antennas. The antennas described
next are practical arrays that are not difficult to
make and have been used in the field with
excellent results.

A 6-DB GAIN OMNIDIRECTIONAL ARRAY

The antenna shown in Figs. 11-48 and 11-49
was designed by W2EWY and WB21CP, adapting
some of the more attractive features of commercial
arrays. Excellent results have been obtained with
this antenna in use at a repeater site.
Basically, the antenna consists of four stacked,
grounded, half-wave dipoles, fed off-center, pro¬
viding a 6-dB gain (omnidirectional) or 9 dB with
an offset pattern. One problem in antenna instal¬
lations is static-type noise. The grounded-dipole
configuration has the advantage of reducing this
type of noise appreciably.
Fig. 11-49 at A shows the details for one of the
elements. The two sides and ends of the elements
are made from 1/2-inch aluminum angle stock,
available in eight-foot lengths from many hardware
Fig. 11-47 — Details of the 6-meter J pole.
l%"0IA PIPE

dealers. A single eight-foot section will make one

element. The element supports are also made from
the same type of angle stock.
Fig. 11-49 also shows the details of the phasing
harness which is made from RG-59/U. Lengths A,
B, C, and D are each 40.8 inches long from the
dipole feed to the center of the T. Lengths E and F
are each 63.8 inches long from center of T to
center of T.
If the four elements are mounted in line on the
TV-mast section, the pattern will be offset with a
null in the direction of the supporting mast. The
four dipoles can be positioned around the mast to
provide an omnidirectional pattern. No matching is
needed with this antenna. Of several units built
from the specifications given, the SWR was below
1.2 to 1.

A COLLINEAR-COAXIAL ARRAY
The antenna shown in Fig. 11-50 is an excellent
array for home station or repeater use. The
antenna will provide from 6- to 9-dB omni¬
directional gain, depending on the number of
elements used. This system is one that has been
around for years. The refinements shown here were
Fig. 11-46 - Details of the 2-meter J pole. developed by K2CBA, K1DEU, and WA1KJ1.
 248
Chapter 11

takes eight half-wave coaxial elements, connected

end-to-end, to obtain the same gain figure.
However, the coaxial-collinear antenna has cer¬
tain advantages when installation problems are
considered. The completed antenna is encased in
either Plexiglas or PVC pipe and can be mounted
above the supporting tower to get best omni¬
directional coverage without the tower interfering
with the antenna pattern.

Construction
From the formula 492 divided by the fre¬
quency in MHz, calculate a half-wavelength for the
desired frequency. This comes out to 3.4 feet, or
40.4 inches for 146 MHz. Next, select the type of
146-147 MHZ

'AX IN AIR
OK TV MASTING
20.2"

Fig. 11-48 — Spacing and arrangement of the four

phased half-wavelength dipoles. 'A X IN COAX
13.45"

The antenna is a multiple of 1/2-wavelength

elements with 1/4-wavelength sections on each end
and a 1/4-wave stub at the feed point to reduce
feed-line radiation. The dimensions shown are for rg-b/u AX IN COAX
26.9"
146 to 147 MHz, but the antenna can be made for SOLIO DIELECTRIC
(not foam)
other bands and frequencies.
In order to provide the same amount of gain as
would be obtained with stacked dipoles, a larger
number of half-wave sections are required. One of
the reasons for antenna gain is the spacing between 14 A IN COAX
antenna sections. The four stacked dipoles pre¬ 13.45"
viously described approach optimum spacing for
maximum gain. In the coaxial-collinear arrange¬
ment shown, there is always the problem that as
sections are added, the antenna current decreases 50 A COAX 'AX IN AIR
from one section to the next. In other words, one FEED LINE
20.2''

end of the antenna is not radiating as much power TO

RECEIVER
as the other end. Slightly more than twice the
number of elements arc required to obtain the BALUN

same amount of gain as with stacked dipoles. Fig. 11-50 - Basic details of the coaxial-collinear
Where four stacked dipoles as described provide antenna. See fig. 11-9A for balun details.
slightly less than 6 dB omnidirectional gain, it
coax you plan to use and get the velocity factor.
Generally, the velocity factor for the solid¬
dielectric lines is 0.66 and 0.81 for foam dielectric.
The antenna shown in Fig. 11-50 is based on the
solid-dielectric coax, 0.66 velocity factor. Using
this type of coax provides a shorter overall length
for the antenna.
The first step in fabricating the antenna is to
make a 3-element version (3 half-waves plus the
1/4-wave top element, the 1/4-wave coax section,
and the bottom 1/4-wave section). Figs. 11-51 and
11-52 show the details for making the coaxial
sections. The top section can be made from a piece
of copper tubing or No. 12 wire.
When the antenna is completed, suspend it clear
of any metallic objects. Using a low-power trans¬
Fig. 11-49 — At A, constructional details of one of mitter and SWR meter, make a check across the
the dipole elements of Fig. 11-48. The phasing band to determine if the antenna is high or low in
harness is shown at 8. frequency. The lowest SWR reading will occur at
VHF and UHF Antenna Systems 249

resonance. If this is not within ±1 MHz of the

desired frequency, trim the half-wave elements
accordingly. More than likely, this will not be
required. Also, don’t be concerned about the
specific SWR at this time. Look only for the
minimum reading.
Depending on whether resonance is too high or
too low in frequency, alter another pair of half¬
wave elements, making them 1/4- to 1/2-inch
longer if the antenna is too high in frequency, or a
like amount shorter if the antenna is too low in
frequency. Continue this operation, adding pairs of
elements until you reach the desired length. Eight
half-wave elements will provide about 6 dB of gain
and 16 elements will give approximately 9 dB of
gain.

1) Cut to desired length plus 2".

Fig. 11-52 — This shows the prepared end of one
of the coax sections and also the method of joining
two sections together.
RG-8/U SOLID DIELECTRIC
Next, tape each connection with a good grade
of electrical tape, applying several layers. This will
provide mechanical strength and weatherproofing.
2) Cut insulation back 1" on each end. Flux and With the several arrays that were built using this
tin each end, allow to cool. design, the SWR was always below 1.3 to 1 at the
design frequency.
TIN The antenna can be housed inside 1-3/4-inch
diameter PVC pipe. Also, a new type of pipe has
recently become available from plumbing supply
dealers. This is fiberglass pipe and is available in
25-foot lengths with diameters starting at 2-1/2
inches. The ends of this pipe are tapered so that it
can be joined to another section. The fiberglass
pipe is extremely flexible without danger of
3) Using tubing cutter, cut shield off 3/4" from breaking so it can be supported at one end, such as
first end. Measure off final dimension from shield
on cut end to other end, mark, and cut shield with
at the top of a tower, permitting the antenna to be
tubing cutter. in the clear.

Bibliography
Source material and more extended discussion
of topics covered in this chapter can be found in
the references given below.
Campbell, "Turnstile for Two," QST, April, 19 59.
Kasper, “Array Design with Optimum Antenna
Spacing,” QST, Nov., 1960.
Kmosko and Johnson, “Long Long Yagis,” QST,
Jan., 1956.
Kraus, “The Square Corner Reflector Beam Anten¬
na,” QST, Nov., 1940.
4) Using single-edge razor blade, trim insulation LaPort and Veldhuis, “Improved Antennas of the
leaving 1/16” to 1/8" remaining. Rhombic Class,” RCA Review, March, 1960.
Linde, “5-Over-5 for Six,” QST, Aug.. 1970.
McCoy, “Five-Element Two-Meter Beam for
$1.50,” QST, Oct., 1962.
Tilton, “Portable Antennas for 50 and 144 Me.,”
QST, Aug., 1955.
Tilton, “Six Elements on 6,” QST, Oct., 1957.
Tilton, “V.H.F. Antenna Facts and Fallacies,”
QST, Jan., Feb., Mar., 1963.
“Antenna Couplers for 50 and 144 Me.,” QST,
July, Sept., 1956.
“The World Above 50 Me.,” QST, April, 1967.
The Radio Amateur's VHF Manual, ARRL.
Fig. 11-51 — Method of element preparation. FM and Repeaters for the Radio Amateur, ARRL.
Chapter 12

Antennas for Space Communications

Space and satellite communications require

consideration of the effects of polarization and
angle of elevation, along with the azimuth of either
a transmitted or received signal. Normally, pro¬
visions for polarization are unnecessary on the hf
bands, since the original polarization direction is
lost after the signal passes through the ionosphere.
A vertical antenna will receive a signal emanating
from a horizontal one, and the converse is true
when transmitting and receiving antennas are inter¬
changed. Neither is it worth the effort to make
provisions for tilting the antenna, since the ele¬
vation angle is so unpredictable. However, with
satellite communications the polarization changes,
and a signal that would disappear into the noise on
a normal antenna might be S9 on one that is
insensitive to polarization direction. Angle of
elevation is also important from the standpoint of
tracking and avoiding indiscriminate ground re¬
flections which might cause nulls in signal strength.

CIRCULAR POLARIZATION
The ideal antenna for random polarization
would be one with a circularly polarized radiation
pattern. Two commonly used methods for ob¬
taining circular polarization are the crossed Yagi, as
shown in Fig. 12-1, and the helical antenna, as
described later in this chapter. The crossed Yagi is
mechanically simpler to construct, but harder to
adjust than its helical counterpart.
Polarization sense may also be a factor, es¬ Fig. 12-1 — This vhf crossed Yagi antenna design
pecially if the satellite uses a circularly polarized by KH6IJ was presented in January 1973 QST.
Placement of the phasing harness and T connector
antenna. In physics, clockwise rotation of an is shown in the lower half of the photograph. Note
approaching wave is called “right circular polariza¬ that the gamma match is mounted somewhat off of
tion,” but the IEEE standard uses the term element center for better balance of rf voltages on
“clockwise circular polarization” for a receding elements.
wave. Either clockwise or a counter-clockwise
sense can be selected by reversing the phasing
harness of a crossed Yagi antenna. The sense of a time) is 90 degrees, then the polarization is
helical antenna is fixed, being determined by its circular. Any combination between these two
physical construction. limits gives elliptical polarization.
Mathematically, linear and circular polarization
are special cases of elliptical polarization. Consider CROSSED LINEAR ANTENNAS
two electric-field vectors at right angles to each A dipole radiates a linearly polarized signal, and
other. The frequencies are the same, but the the polarization direction depends upon the orien¬
magnitudes and phase angles can vary. If either one tation of the antenna. Fig. 12-2 shows the electric
or the other of the magnitudes is zero, linear field patterns of horizontal and vertical dipoles at
polarization results. If the magnitudes are the same A and B. If the two outputs are combined with the
and the phase angle between the two vectors (in correct phasing (90 degrees), a circularly polarized
250
Antennas for Space Communications 251

Care should be taken when other types of coax are

considered, especially if one is unfamiliar with
them. For example, RG-lll/U, which has an
impedance of 95 ohms, might sound like a good
one to use, but since it is a twin cable, it would be
unsuitable.

Gamma-Match Tune-up Procedure

Once the antenna is constructed, a single Yagi
section at a time may be initially tuned and
matched. The procedures used at vhf are similar to
those for hf antennas, except that more care is
necessary in the selection of test instruments. For
example, an SWR indicator designed for hf may
burn out a diode if used on vhf, since the rf pickup
may be much greater with the same line dimen¬
sions. A Bird wattmeter is ideal, but if unavailable,
a homemade impedance bridge or SWR indicator
Fig. 12-2 — Radiation patterns looking head-on at
could be used.
dipoles.
The following method has proved useful in
simplifying tune-up of the gamma match and
antenna elements. One parameter should be kept
wave results, and the electric field pattern is shown fixed, while the rest of the adjustments are made,
in Fig. 12-2C. Notice that since the electric fields rather than varying all of them simultaneously. It
must be identical in magnitude, the power from was found convenient to use the length of the
the transmitter must be equally divided between
the two antennas; hence the gain of each one is
decreased by 3 dB when taken alone in the plane
of its orientation.
As previously mentioned, a 90-degree phase
shift must exist between the two antennas. The
simplest way to obtain the shift is to use two feed
lines with one section a quarter wavelength longer
than the other one. These two separate feed lines
are then paralleled to a common transmission line
which goes to either the transmitter or receiver.
Therein lies one of the headaches of this system,
since the impedance presented to the common
transmission line by the parallel combination of
the other two sections is one half that of either one
of them taken alone (normally not true when there
is interaction between loads, as in phased arrays).
Another factor to consider is the attenuation of
the cables used in the harness, along with the
connectors. Good low-loss coaxial line should be
used, and connectors such as type N are preferable
to the UHF variety.

A Practical Antenna
The crossed Yagi array shown in Fig. 12-3 is
part of the final design for a system used in
satellite transponder stations. After various kinds
of matching sections were tried, it was found that
the simplest one worked best. The 90-degree phase
shift is realized by making section “A” a quarter
wave longer electrically than section “B.” The
characteristic impedance of these sections should
be such that, when paralleled, they match the main
feed line.
RG-133/U (95 ohms), made by Consolidated
Wire Co., is ideal but is a hard item to find. More
commonly found in stock is RG-63/U (125 ohms).
There is some mismatch when using RG-63/U with
a 50-ohm main feed line, but it is not serious Fig. 12-3 — Construction details of a crossed Yagi
enough to warrant additional matching networks. antenna.
 252 Chapter 12

TEFLON PLUG TEFLON BUSHING

PRESS TIT TO GAMMA CAP. e™
00 -PRESS FIT ON SLIDING FIT TO GAMMA ROO /
32 GAMMA ROD

TYPE N PANEL
RECEPTACLE
O SOLDERED TO
BRACKET

No matter how far off either the reflector or

director lengths happen to be, the last few steps
-J- BRASS should at least get the SWR into the ball park. If
not, then look for the following problems:
. HOLE 1) Poor rf source; use a signal generator or low
power transmitter, not a grid-dip oscillator. Make
sure that harmonic content is kept as low as
HOLE
possible in order to avoid erroneous readings on
the SWR indicator.
2) Radiator length too far off, usually too long.
3) Poor Q of the gamma-match system. Use a
coaxial-capacitor type (preferably one with as
much air dielectric as possible) such as that shown
in Fig. 12-4.
Fig. 12-4 — Gamma-match construction details.
Finally, adjust the director and reflector for
maximum front-to-back ratio. This can be done by
driven element as the fixed parameter. Its approxi¬ looking for minimum pickup with the back of the
mate dimensions can be found in Table 12-1 (along beam aligned with a test dipole as far away as
with those of the other elements), and the length practical. Final touch-up of the SWR can be
should be slightly shorter than that given by the accomplished by adjusting the length of the radia¬
formula. Do not change the length, except for tor, but by no more than one percent.
some minor pruning which will be mentioned later. Unequal antenna currents can be equalized by
Once the antenna is assembled, the next step is offsetting the gamma section in the direction of
to adjust the gamma rod and capacitor for mini¬ the desired increase in antenna current, as shown in
mum SWR. Table 12-11 shows some approximate Figs. 12-1 and 12-3.
settings and lengths for the gamma rod, capacitor,
and shorting bar (see Fig. 12-4). The correct
procedure in tuning is to select an appropriate Final Tune-up
length for the gamma rod and capacitor sections The procedure for tuning up a crossed Yagi is
from the table, and then adjust the shorting bar similar to that used for a single one with the
and capacitor for a minimum SWR indication, with following additional steps.
the capacitor adjusted last. 1 ) Attach the phasing harness through a T
connector, as shown in Fig. 12-3, to the antennas
and feed line. Test for SWR and make any minor
TABLE 12-1 adjustments necessary in order to assure a mini¬
Crossed Yagi antenna element lengths and spacings. mum reflected reading.
2) Test for balance and circularity as described
in the following paragraphs.
Radiator length (Ra) in feet is given by: In a properly adjusted antenna system, equal rf
Ra = 460//(MHz) voltages should appear on all of the elements. A
Reflector length (R) equals: practical method is to see how much detuning
Ra + 0.1 Ra results from attaching a short piece of wire with an
First director length (D1 ) is given by: alligator clip to the ends of each element. A
Ra — .05 Ra properly balanced element will show the same
Successive director lengths (D2...Dn) equal- amount of detuning (SWR goes up or bridge null is
D2 - D1 — .01 D1 upset) regardless of which end has the clip.
Dn = (Dn - 1) - .01 (Dn - 1) To test for coupling between the Yagi sections
Spacing between reflector and radiator - 0.2X. (there should be none), feed power into the
Spacings between first director, radiator, and horizontal Yagi alone, and see how much detuning
between all other directors — 0.15A.
results by attaching the wire-and-clip combination
to the vertical Yagi. Repeat this procedure for each
 Antennas for Space Communications 253

TABLE 12-11
Approximate gamma-match dimensions.

Frequency (MHz) Dimensions in inches (Fig. 12-4)

W X Y Z

135.6 (ATS-1 out) 6 4 6 1

146 (OSCAR 6 in) 5-3/4 3-3/4 5 1-1/2

149.2 (ATS-1 in) 5-1/2 3-1/2 4-3/4 2

element. If there are no interactions, feed power

into the vertical Yagi and see if there is any
coupling from the horizontal one. In making any
of these tests, while near the antenna, make sure
that the power is off to avoid possible injury!
As a final test, tune in on a linearly polarized
signal from a satellite such as the ATS-1, or even a
repeater. Rotate the crossed Yagi on its axis and
note if there is any signal variation. A good
circularly polarized antenna should have no more
than 1 dB variation, as one rotates the antenna.
Fig. 12-6 — The antenna system can be assembled
AN AZ-EL MOUNT FOR CROSSED YAGIS using off-the-shelf components such as Hy-Gain
An important consideration when building a Yagis, Cornell-Dubilier or Blonder-Tongue rotators,
and a commercially made tripod.
station for satellite work is the method of mount¬
ing and rotating the antennas. The following
The basic criteria in the design of this system
information by KH6IJ appeared in the June, 1973,
were low cost and ease of assembly. In the matter
issue of QST.
of choice between a crossed Yagi system and a
helical antenna, the main factor was that Yagi
antennas can be bought off the dealer’s shelf, but
most helical antennas cannot.
Figs. 12-5 and 12-6 show the overall assembly
of the array. The antennas used are Hy-Gain Model
341 eight-element Yagis. Fig. 12-7 is a head-on
view of the array, showing the antennas mounted
at 90 degrees with respect to each other and 45
degrees with respect to the cross arm.
Coupling between the two Yagis is minimal at
90 degrees and is somewhat greater at 45 degrees.
By setting the angle at 45 degrees with respect to
the cross arm, coupling is minimized but not
eliminated.

Fig. 12-5 — Two individual Yagis are mounted with

the elements at 90° to each other. Elevation and Fig. 12-7 — An end-on view of the antennas show
azimuth movements are provided by two TV that they are mounted at 90 degrees to each other,
rotators. and at 45 degrees to the cross boom.
 254 Chapter 12

Length din Fig. 12-7 should be the minimum

necessary for the elements to clear the tripod base
when the array is pointed straight up and rotated.
In the array shown in Figs. 12-5 and 12-6, a
five-foot section of TV mast serves the purpose.

The Mounting Tripod

A mounting tripod could be made by using
aluminum railing, called “NuRail,” which comes
with all manner of swivels, crosses, and T fittings.
However, the cheapest method is to purchase a TV
tower such as Lafayette No. 18-56233W, which is a
collapsible tripod. It is made by the South River
Metal Products Company, South River, NJ 08820.
Their model number is HDT-5. This tower sells for
such a low price that there is little point in
constructing your own. Spread the legs of the
tripod more than usual to assure greater support,
but be sure that the elements of the antenna will
clear the base in the straight-up position.

Elevation-Azimuth Rotators (B)

The azimuth rotator is a Cornell-Dubilier Fig. 12-8 — The method of mounting two rotators
AR-20. The elevation rotator is a Blonder-Tongue together. A pair of PM-2 rotators may also be used.
Prism-matic PM-2. The latter is one of the few on The adapter plate (B) may be fabricated from
the market which allows the cross arm of the array to 1/4-inch-thick aluminum stock, or a ready-made
be rotated on its axis when supported at the center. plate is available from Blonder-Tongue.
Fig. 12-8 shows the detail of the method of
mounting the two rotators together. Notice that makes an adapter plate for their heavy-duty CATV
the flat portion of the AR-20 makes an ideal antennas. It is called a YSB Stacking Block. The
mounting surface for the PM-2. If you want to PM-2 rotator fits horizontally on this plate even
utilize commercially fabricated components though this was not the intended application. The
throughout, a mounting plate similar to that shown adapter plate may be used to fasten two PM-2
in Fig. 12-8B can be purchased. Blonder-Tongue rotators together.

A TWELVE-FOOT STRESSED PARABOLIC DISH

Very few antennas evoke as much interest rather flat dish, which makes it easy to surface and
among uhf amateurs as the parabolic dish, and for allows the use of recent advances in high-efficiency
good reason. First, the parabola and its cousins - feed horns. This results in greater gain for a given
Cassegrain, hog horn, and Gregorian - are prob¬ size of dish over conventional designs. Such an
ably the ultimate in high-gain antennas. The antenna is shown in Fig. 12-9. This parabolic dish
highest gain antenna in the world (148 dB) is a is lightweight, portable, easy to build, and can be
parabola. This is the 200-inch Mt. Palomar tele¬ used for 432- and 1296-MHz mountaintopping, as
scope. (The very short wavelength of light rays well as on 2300, 3300, and 5600 MHz. Dis¬
causes such a high gain to be realizable.) Second, assembled, it fits into the trunk of a car, and it can
the efficiency of the parabola does not change as it be reassembled in 45 minutes’ time.
gets larger. With collinear arrays, the loss of the The usually heavy structure which supports the
phasing harness increases as the size increases. The surface of most parabolic dish antennas has been
corresponding component of the parabola is loss¬ replaced in this design by aluminum spokes bent
less air between the feed horn and the reflecting into a near-parabolic shape by string. These strings
surface. If there are few surface errors, the effi¬ serve the triple function of guying the focal point,
ciency of the system stays constant regardless of bending the spokes, and reducing the error at the
antenna size. dish perimeter (as well as at the center) to zero. By
Some amateurs reject parabolic antennas be¬ contrast, in conventional designs, the dish peri¬
cause of the belief that these are all heavy, hard to meter, which has a greater surface area than the
construct, have large wind-loading surfaces, and center, is furthest from the supporting center hub
require precise surface accuracies. However, with so it often has the greatest error. This error is
modern construction techniques, a prudent choice pronounced when the wind blows. Here, each of
of materials, and an understanding of accuracy the spokes is basically a cantilevered beam with
requirements, these disadvantages can be largely end loading. The equations of beam bending
overcome. A parabola may be constructed with a predict a near-perfect parabolic curve for extreme¬
0.6 f/d (focal length/diameter) ratio, producing a ly small deflections. Unfortunately the deflections
Antennas for Space Communications 255

The front and back center plates constitute an

I-beam-like structure, which gives the dish center
considerable rigidity. Fig. 12-11 shows a side view
of the complete antenna. Aluminum alloy
(6061-T6) is used for the spokes, while 2024-T3
aluminum alloy sheet, 1/8 inch thick, serves for the
center plates. Aluminum has approximately three
times the strength-to-weight ratio of wood used in
other designs. Additionally, aluminum does not
become water-logged or warped. The end of each
of the 18 spokes has an eyebolt facing the dish
focus point which serves a double purpose: to
accept the No. 9 galvanized fence wire which is
routed through the screw eyes to define the dish
perimeter, and to facilitate rapid assembly by
accepting the S hooks which are tied to the end of
each of the lengths of 130-pound test Dacron
fishing string. The string bends the spokes into a
parabolic curve; the dish may be adapted for many
focal lengths by tightening the strings. Dacron was
chosen because it has the same chemical formula as
Fig.12-9 — A stressed parabolic dish designed by
K2RIW (shown at the right) set up for reception of Mylar. This is a low-stretch material which keeps
Apollo or Skylab signals near 2280 MHz. A the dish from changing shape. The galvanized
preamplifier is shown taped below the feed horn. perimeter wire has a five-inch overlap area which is
From QST, August, 1972. bound together with baling wire after the spokes
have been hooked to the strings.
in this dish are not that small, and the loading is The aluminum window screening is bent over
not perpendicular. For these reasons, mathematical the perimeter wire to hold it in place on the back
prediction of the resultant curve is quite difficult. of the spokes. It was thought originally that the
A much better solution is to measure the surface spokes in front of the screening might cause
error with a template and make the necessary surface perturbations and decrease the gain. How¬
correction by bending each of the spokes to fit. ever, the total spoke area is small. Placing the
The uncorrected surface is accurate enough for aluminum screening in front of the spokes requires
432- and 1296 MHz use. Trophies taken by this the use of 200 pieces of baling wire to hold the
parabola in antenna-gain contests were won using a screening in place. This procedure increases the
completely natural surface with no error cor¬ assembly time by at least an hour. For contest and
rection. mountaintop operating (when the screening is on
By placing the transmission line inside the the back of the spokes) no other fastening tech¬
central pipe which supports the feed horn, the area nique is required than bending the screen overlap
of the shadows or blockages on the reflector around the wire perimeter.
surface is much smaller than in other feeding and
supporting systems, thus increasing gain. For 1296
MHz a backfire feed horn may be constructed to
take full advantage of this feature. On 432 MHz a
dipole and reflector assembly produces 1.5 dB TABLE 12 III
additional gain over a corner-reflector type feed
horn. Since the preamplifier is located right at the Materials List for the 12-foot Dish
horn on 2300 MHz, a conventional feed horn may 1. Aluminum tubing, 12-feet x 1/2-inch OD
be used. X ,049-inch wall, 6061-T6 alloy, 9 required to
make 18 spokes.
Construction 2. Octagonal mounting plates 12 x 12 X 1/8
Table 12-III is a list of materials required for inches, 2024-T3 alloy, 2 required.
construction. Care must be exercised when drilling 3. 1-1/4-inch ID pipe flange with setscrews.
holes in the connecting center plates so that 4. 1-1/4 inches x 8 feet, TV mast tubing, 2
assembly difficulty will not be experienced later. required.
See Fig. 12-10. A notch in each plate will allow 5. Aluminum window screening, 4 x 50 feet.
them to be assembled in the same relative position. 6. 130-pound test Dacron trolling line
The two plates should be clamped together and (available from Finney Sports, 2910 Glansman
drilled at the same time. Each of the 18 half-inch¬ Rd., Toledo, OH 43614.)
diameter aluminum spokes has two No. 28 holes 7. 38 feet of No. 9 galvanized fence wire
drilled at its root to accept 6-32 machine screws (perimeter), Montgomery Ward Farm and
which go through the center plates. The 6-foot- Garden Catalog.
long spokes are created by cutting standard 12-foot 8. Two hose clamps, 1-1/2 inch; two U bolts;
lengths of tubing in half. A fixture built from a 1/2 X 14-inch Bakelite rod or dowel; water-pipe
block of aluminum assures that the holes are grounding clamp; 18 eye bolts; 18 S hooks.
drilled in exactly the same position in each spoke.
 256 Chapter 12

The Horn and Support Structure

The feed horn is supported by the 1-1/4-inch
aluminum television mast. The transmission line
which is inserted into this tubing is connected first
to the front of the feed horn which then slides
back into the tubing for support. A setscrew
assures that no further movement of the feed horn
occurs. During antenna-gain competition the set¬
screw is omitted, allowing the 1/2-inch semirigid
(CATV cable) transmission line to move in or out
while adjusting the focal length for maximum gain.
The TV mast is held firmly at the center plates by
two setscrews attached to the pipe flange which is
mounted on the rear plate. On 2300 MHz the dish
is focused for best gain by loosening these set¬
screws on the pipe flange and sliding the dish along
the TV mast tubing (the dish is moved instead of
the feed horn).
All of the fishing strings are held in position by
attaching them to a hose clamp which is perman¬
ently connected to the TV tubing. A piece of
rubber sheet under the hose clamp prevents slip¬
Fig. 12-10 — Center plate details. Two center page and keeps the hose clamp from cutting the
plates are bolted together to hold the spokes in fishing string. A second hose clamp is mounted
place. below the first as double protection against slip¬
page.
The high-efficiency 1296-MHz dual-mode feed
Surface horn, detailed in Fig. 12-12, weighs 5-3/4 pounds.
This weight causes some bending of the mast
A four-foot-wide roll of aluminum screening 50
tubing; however this is corrected by a 1/2-inch-
feet long is cut into appropriate lengths and laid
diameter Bakelite support. It is mounted to a pipe
parallel with a 3-inch overlap between the top of
grounding clamp with an 8-32 screw inserted in the
the unbent spokes and hub assembly. The overlap
end of the rod. The Bakelite rod and grounding
seams are sewn together on one half of the dish
using heavy Dacron thread and a sailmaker’s curved clamp are mounted midway between the hose
needle. Every seam is sewn twice, once on each clamp and the center plates on the mast. A double
run of fishing string slipped over the notched upper
edge of the overlapped area. The seams on the
end of the Bakelite rod counteracts bending.
other half are left open to accommodate the
increased overlap which occurs when the spokes The success of high-efficiency parabolic an¬
are bent into a parabola. The perimeter of the tennas is determined primarily by the feed-horn
screening then is trimmed. Notches are cut in the effectiveness. The multiple diameter of this feed
three-inch overlap to accept the screw eyes and S horn may seem unusual. This newly designed and
hooks. The first time this dish was assembled, the patented dual-mode feed, by Dick Turrin, W2IMU,
achieves efficiency by launching two different
screening strips were anchored to the inside surface
of the dish and the seams sewn in this position. Itkinds of waveguide modes simultaneously, which
is easier to fabricate the surface by placing the causes the dish illumination to be more constant
screen on the back of the dish frame with the
structure inverted. The spokes are sufficiently
strong to support the complete weight of the dish
when the perimeter is resting on the ground.
The 4-foot wide strips of aluminum screening
conform to the compound bend of the parabolic
shape very easily. If the seams are placed parallel to
the /¡-field polarization of the feed horn, minimum
feedthrough will occur. This feedthrough, even if
the seams are placed perpendicular to the /¡-field, is
so small that it is negligible. Some constructors
may be tempted to cut the screening into pie¬
shaped sections. This procedure will increase the
seam area and construction time considerably. The
dish surface appears most pleasing from the front
when the screening perimeter is slipped between
the spokes and the perimeter wire, and is then
folded back over the perimeter wire. When dis¬
assembly is desired, the screening is removed in one Fig. 12-11 — Side view of the stressed parabolic
piece, folded in half, and rolled. dish.
Antennas for Space Communications 257

Fig. 12-12 — Backfire type 1296-MHz feed horn, linear polarization only. The small can is a Quaker State
oil container; the large can is a 50-pound shortening container (obtained from a restaurant, "Gold Crisp"
brand). Brass tubing, 1/2-inch OD, extends from UG-23/U connector to dipole. Center conductor and
dielectric are obtained from 3/8-inch Alumafoam coaxial cable. The dipole is made from 3/32-inch copper
rod. The sceptum and 3O-degree section are made from galvanized sheet metal. Styrofoam is used to hold
the sceptum in position. The primary gain is 12.2 dB over isotropic.

than conventional designs. The illumination drops eye faces forward. Machine screws, lock washers,
off rapidly at the perimeter, reducing spillover. The and nuts are used to mount all 18 spokes.
feedback lobes are reduced by at least 35 dB 3) The 6-32 nuts are tightened using a nut
because the current at the feed perimeter is almost driver.
zero; the phase center of the feed system stays 4) The mast tubing is attached to the spoke
constant across the angles of the dish reflector. The assembly, positioned properly, and locked down
larger diameter section is a phase corrector and with the setscrews on the pipe flange at the rear
should not be changed in length. Theory predicts center plate. The S hooks of the 18 Dacron strings
that almost no increase in dish efficiency can be are attached to the screw eyes of the spokes.
achieved without increasing the feed size in a way 5) The ends of two pieces of fishing string
that would increase complexity, as well as block¬ (which go over the Bakelite rod support) are tied
age. The feed is optimized for a 0.6 fid dish. The to a screw eye at the forward center plate.
dimensions of the feeds are slightly modified from 6) The dish is laid on the ground in an upright
the original design in order to accommodate the position and No. 9 gauge galvanized wire is
cans. Either feed type can be constructed for other threaded through the eyebolts. The overlapping
frequencies by changing the scale of all dimensions. ends are lashed together with bailing wire.
7) The dish is placed on the ground in an
Multiband Use inverted position with the focus downward. The
screening is placed on the back of the dish and the
Many amateurs construct multiple-band an¬
screening perimeter is fastened as previously des¬
tennas by putting two dishes back to back on the
cribed.
same tower. This is inefficient. The parabolic 8) The extension mast tubing (with counter¬
reflector is a completely frequency-independent
weight) is connected to the center plate with U
surface and studies have shown that a 0.6 fid
bolts.
surface can be steered seven beamwidths by mov¬ 9) The dish is mounted on a support (if one is
ing only the feed horn from side to side before the used) and the transmission line is routed through
gain diminishes one dB. Therefore, the best dual¬ the tubing and attached to the horn.
band antenna can be built by mounting separate
horns side by side. At worst the antenna may have
to be moved a few degrees (usually less than a Parabola Gain Versus Errors
beamwidth) when switching between horns, and “How accurate must a parabolic surface be?” is
the unused horn increases the shadow area slightly. a frequently asked question. According to the
In fact, the same surface can function simul¬ Rayleigh limit for telescopes, little gain increase
taneously on two frequencies, making cross-band is realized by making the mirror accuracy greater
operation possible with the same dish. than ±1/8 wavelength peak error. John Ruze of the
M.I.T. Lincoln Laboratory, among others, has
derived an equation for parabolic antennas and
Assembly Order built models to prove it. The tests show that the
1) A single spoke is held upright behind the rear tolerance loss can be predicted within a fraction of
center plate with the screw eye facing forward. a dB, and less than 1 dB of gain is sacrificed with a
Two 6-32 machine screws are pushed through the surface error of ±1/8 wavelength. An eighth of a
holes in the rear center plate, through the two wavelength is 3.4 inches at 432 MHz, 1.1 inches at
holes of the spoke, and into the corresponding 1296 MHz and 0.64 inch at 2300 MHz.
holes of the front center plate. Lock washers and Some confusion about requirements of greater
nuts are placed on the machine screws and hand than 1/8-wavelength accuracy may be the result of
tightened. technical literature describing highly accurate sur¬
2) The remaining spokes are placed between the faces for reasons of low side-lobe levels. We are
machine screw holes. Make sure that each screw concerned more with forward gain than with low
 258 Chapter 12

side-lobe levels; therefore, these stringent require¬

ments do not apply. When a template is held up
against a surface, positive and negative (±) peak
errors can be measured. The graphs of dish¬
accuracy requirements are frequently plotted in
terms of rms error, which is a mathematically
derived function much smaller than ± peak error
(typically 1/3). These small rms accuracy require¬
ments have discouraged many constructors who
confuse them with ± peak errors.
Fig. 12-13 may be used to predict the resultant
gain of various dish sizes with typical errors. There
are a couple of surprises, as shown in Fig. 12-14.
As the frequency is increased for a given dish, the
gain increases 6 dB per octave until the tolerance
errors become significant. Then gain deterioration
occurs rapidly. Maximum gain is realized at the
frequency where the tolerance loss is 4.3 dB. Fig. 12-14 — Parabolic-antenna gain versus size,
Notice that at 2304 MHz, a 24-foot dish with ± frequency, and surface errors. All curves assume
2-inch peak errors has the same gain as a 6-foot 60-percent aperture efficiency and 10-dB power
dish with ± 1-inch peak errors. Quite startling, taper. Reference: J. Ruze, British IEE.
when it is realized that a 24-foot dish has 16 times
the area of a 6-foot dish. Each time the diameter or tions will be under control and the curves will
frequency is doubled or halved, the gain changes 6 represent an upper limit to the gain that can be
dB. Each time all the errors are halved, the realized. If a 24-foot dish with ± 2-inch peak errors
frequency of maximum gain is doubled. With this is being used with 432-MHz and 1296-MHz mul¬
information, the gain of other dish sizes with other tiple feed horns, the constructor might be dis¬
tolerances may be predicted. couraged from trying a 2300-MHz feed because
These curves are adequate to predict gain, there is 15 dB of gain degradation. However, the
assuming a high-efficiency feed horn is used (as dish will have 29 dB of gain remaining on 2300
described earlier) which realizes 60-percent aper¬ MHz, making it worthy of consideration.
ture efficiency. At frequencies below 1296 MHz The near-field range of the 12-foot 3-inch
where the horn is large and causes considerable antenna is 703 feet at 2300 MHz. By using the sun
blockage, the curves are a little optimistic. A as a transmitter and observing receiver noise power,
properly built dipole and splasher feed will have it was discovered that the antenna had two main
about 1.5 dB less gain when used with a 0.6 f/d lobes about 4 degrees apart. The template showed
dish than will the dual-mode feed system des¬ a surface error (insufficient spoke bending at 3/4
cribed. radius), and a correction was made. A recheck
The worst kind of surface distortion is where showed one main lobe, and the sun noise was
the surface curve in the radial direction is not almost 3 dB stronger.
parabolic but gradually departs in a smooth man¬
ner from an exact parabola. The loss of gain can be Other Surfacing Materials
severe because a large area is involved. If the The choice of surface materials is a compromise
surface is checked with a template, and if reason¬ between rf reflecting properties and wind loading.
able construction techniques are employed, devia- Aluminum screening, with its very fine mesh (and
weighing 4.3 pounds per 100 square feet) is useful
beyond X-band because of its very close spacing. It
is easy to roll up and is therefore ideal for a
portable dish. However, this close spacing causes it
to be a 34-percent filled aperture, which will cause
the wind force at 60 miles per hour to be more
than 400 pounds on this 12-foot dish. Those
amateurs considering a permanent installation of
this dish should look into other surfacing materials.
One-inch hexagonal chicken wire, which is an
8-percent filled aperture, is very desirable for
432-MHz operation. It weighs 10 pounds per 100
square feet and exhibits 81 pounds of force with
60-mile-per-hour winds. However, measurement on
a large piece reveals 6 dB feedthrough at 1296
MHz. Therefore, on 1296 MHz, one fourth of the
power will feed through the surface material, but
A .1
RM» REFLECTOR ERROR (IN-)
.»
this will cause a loss of only 1.3 dB forward gain.
Since the low-wind-loading material will provide a
Fig. 12-13 — Gain loss vs. reflector error. Basic 30-dB gain potential, it is a very good trade-off.
information obtained from J. Ruze, British 1EE. Chicken wire is very poor material for 2300 MHz
Antennas for Space Communications 259
screening and passing a current through it, fasten¬
ing water-pipe heating tape behind the screening,
or soldering heavy leads to the screening perimeter
and passing current through the screening itself.

Parabolic Template
For use at 2300 MHz and higher where high
surface accuracy is required, a parabolic template
should be constructed to measure surface errors. A
simple template may be constructed (see Fig.
12-16) by taking a 12-foot 3-inch length of 4-foot
wide tar paper and drawing a parabolic shape on it
with chalk. The points for the parabolic shape were
calculated at 6-inch intervals and these points were
connected with a smooth curve. For those who
wish to use the template with the surface material
installed, the template should be cut along the
Fig. 12-15 — Surfacing material quality. chalk line and stiffened by cardboard or a wood
lattice frame.
Surface-error measurements should take place
and higher, since the hole dimensions become with all spokes installed and deflected by the
comparable to a half wavelength. As with all fishing strings, since some bending of the center
surfacing materials, minimum feedthrough will plates does take place.
occur when the ¿’-field polarization is parallel to
the longest dimension of the surfacing holes. A Feed System for 2287.5 MHz
Half-inch hardware cloth weighs 20 pounds per
100 square feet. It has a wind loading characteristic The modification of a feed horn by W2IMU,
of 162 pounds with 60-mile-per-hour winds. The shown in Fig. 12-17, launches more accurate
filled aperature is 16 percent and this material is circularly polarized waves and has greater effi¬
useful to 2300 MHz. ciency than conventional designs, since it elimin¬
A rather interesting material worthy of investi¬ ates the need for a hybrid coupler. It is optimized
gation is 1/4-inch reinforced plastic (described in for a 0.6 fid dish. When power is fed into
Montgomery Ward Farm and Garden Catalog). It connector No. 1 only, the 10-32 screws cause the
weighs only 4 pounds per 100 square feet. The rf to become a counterclockwise circularly pol¬
plastic melts with many universal solvents such as arized wave out of the horn. After bouncing off
lacquer thinner. If a careful plastic-melting job is the dish this becomes a clockwise wave on either
done, what will be left is the 1/4-inch spaced transmit or receive. Power fed into connector No.
aluminum wires with a small blob of plastic at each 2 becomes a ccw wave after bouncing off the dish.
junction to hold the matrix together. Therefore, for moonbounce work connect the
There are some general considerations to be transmitter to connector No. 1 and the receiver to
made in selecting surface materials: connector No. 2. For Apollo reception use only
1) Joints of screening do not have to make connector No. 1.
electrical contact. The horizontal wires reflect the The 1/4-20 screw prevents energy from coup¬
horizontal wave. Skew polarizations are merely a ling between connector No. 1 and connector No. 2.
combination of horizontal and vertical components If the 10-32 screws are omitted, each connector
which are thus reflected by the corresponding
wires of the screening. To a horizontally polarized
wave, the spacing and diameter of only the
horizontal wires determine the reflection co¬
efficient (see Fig. 12-15). Many amateurs have the
mistaken impression that screening materials that
do not make electrical contact at their junctions
are poor reflectors.
2) By measuring wire diameter and spacings
between the wires, a calculation of percentage of
aperture that is filled can be made. This will be one
of the major determining factors of wind pressure X I Z ITO «MIST 1') XI I
when the surfacing material is dry. Under ice and .5' i.szie" 5.s1«-s/4'
snow conditions, smaller aperture materials may 1.0' rz«" s.o'l 1S-1Z«"
t.S' 15Z16'
become clogged, which could make the surfacing 2.0 1—11Z16"
2.5' 2- 5Z8"
material act as one solid sail. The ice and snow will 5.O' 3-3Z4"
have a rather minor effect on the reflecting 3.5' 5-1Z6"
4.0' 6-nzu"
properties of the surface, however. 4.5' t-Mt"
5.0' 1O-1Z2"
3) Amateurs who live in areas where ice and
snow are prevalent should consider a de-icing Fig. 12-16 — Parabolic template for 12-foot, 3-inch
scheme such as weaving enameled wire through the dish.
260 Chapter 12

launches an ordinary linear wave. The small cans a short piece of cable. The preamplifier output
are “Scotts Oats” type from Scotland or “Camp” cable runs straight to the perimeter of the dish.
drain cleaner cans from the U.S., 3-3/4 inches ID. When no preamplifier is used, consider placing the
The large can is a one-gallon paint can, 6-1/2 inches 2287.5-MHz converter at the feed horn and run¬
ID. The 30-degree section is galvanized sheet metal. ning power to it. This will result in a lower system
Tabs on each end add strength and make soldering noise figure since all cables are quite lossy at
the cans together easier. 2287.5 MHz.
The twp UG-58A/U connectors are each fast¬ The inside and outside of the horn may be
ened to the cans with two 4-40 bolts, and 4-40 painted with spray lacquer for preservation. The
nuts are soldered to the inside. The outside of the completely painted horn had a total loss from
can is tinned in the area of each connector to connector No. 1 to radiated circular wave out the
assure good electrical contact. The ten 10-32 bolts throat of less than 0.1 dB. With this new feed,
are 1-1/4 inches long with a 11/16-inch total length greater than 9 dB of S-band sun noise was realized.
inside the can. Each bolt has a 10-32 nut soldered
to the outside of the can and a second 10-32 nut
placed on top as a locking device. The 1/4-20 bolt Possible Variations
is 1-1/2 inches long with 3/4 inch inside the can. The stressed parabolic antenna, as described, is
Outside is a 1/4-20 nut soldered to the can and a a new construction technique for which a patent
second 1/4-20 nut added for locking purposes. application has been filed. Because of its newness,
The two 1/16-inch fiber-glass mounting sheets all of its possibilities have not been explored. For
are each slotted along half of their length, slid instance, a set of fishing strings or guy wires could
together at a right angle, and epoxy cemented. All be set up behind the dish for error correction as
cemented edges of fiber glass are first roughened long as it does not permanently bend the alum¬
with coarse sandpaper. Many very small holes are inum spokes. This technique would also protect
drilled into the fiber glass in the areas of metal the dish against wind loading from the rear. An
contact. extended piece of TV mast would be an ideal place
During Apollo reception, connector No. 1 is to hang a counterweight and attach the back guys.
connected to the preamplifier (if one is used) with It would strengthen the structure.

HELICAL ANTENNAS
The helical antenna has suffered in popularity can radiate in an axial mode, along the axis of the
(but not in performance) because it is not generally helix, or in a broadside mode, perpendicular to the
well understood. With various design dimensions helix axis. The axial mode is that normally used at
and construction, that corkscrew is capable of vhf and uhf. Once the basic geometry is established
generating linear, elliptical, or circular polarization, for axial-mode radiation, we find that other
but perhaps because of its attendant problem of changes in the geometry can create either of the
matching an odd-ball feed impedance to a standard two linear polarizations, horizontal or vertical.
transmission line, it seems to find little acceptance Elliptical polarization or circular polarization can
in amateur circles. It is common in communication also be generated. To make the subject even more
with missiles and satellites, however. interesting, the antenna can also be of either
The helical antenna represents the transition right-hand or left-hand circularity, depending on
point between linear-element antennas and the how it is constructed.
loop antenna. It has several modes of operation The helix is inherently a broad-band antenna,
which are controlled strictly by its geometry. It which eliminates SWR problems over an amateur
Antennas for Space Communications 261

Construction
The quadhelix shown in Fig. 12-18 consists of
four ten-turn helices formed from No. 10 AWG
copper house wire from which the insulation has
been stripped. Construction details are shown in
Figs. 12-19 and 12-20. The helices are mounted on
booms made from 1 x 2-inch smooth lumber.
These booms are attached with wood screws and
wood glue to a frame made from the same size
lumber. The wooden portion should be painted or
stained as a weather preservative. Fastened to the
top and bottom of the frame are two 8-inch pieces
of 1 X 1-inch angle iron from your favorite
hardware store. These angle-iron pieces are drilled
to accommodate the U bolts, which are used to
fasten this antenna to the mast.
The sheet reflector is made from perforated
aluminum. This also forms the ground plane for
the matching lines. From the photograph it can be
seen that a small piece of sheet aluminum was used
to stiffen the ground plane at its center, where the
coax connector is attached. Mounted on the
antenna side of the ground plane are tapered lines,
Fig. 12-18 — A front view of an early version of a which make up the matching section. These lines
quadhelix for 1296 MHz. Later versions have the are connected together at the coax feed point, and
elements mounted on ceramic cone insulators the other ends provide the feed to the individual
which are attached to the wooden booms. This helices. The tapered lines are of such a geometry as
procedure is strongly recommended, especially for
use at 2300 MHz.
to transfer the approximately 140-ohm impedance
of the helices to a 200-ohm point at the coaxial
fitting. Strapping all four of the 200-ohm points
together provides the 50-ohm feed point required
uhf band. Bandwidth is on the order of 1.7 to 1 in
frequency. Expressed in terms of the helix circum¬
ference wavelength, this represents a range of from
approximately 0.73 to 1.22 wavelengths. Over this
range, the VSWR varies very little, remaining in the
order of 1.15 to 1. Also over this range, the input
impedance can be seen to vary on a Smith chart in
tight curls from about 120 to 160 ohms - thus the
generally quoted figure of 140 ohms. The patterm
of the antenna is a well-defined lobe in both the
vertical and horizontal planes over the antenna
bandwidth, with pattern breakup occurring at the
limits. There is a definite sharpening of the antenna
pattern when the helix is used near its upper
frequency limit.
The only difficult part of constructing and
adjusting the helical antenna is the problem of
matching the feed line to the impedance of the
helix. Several articles have described one- to
three-helix arrays, and have described how to build
the coaxial matching section required to match the
array. There is nothing wrong with the theory, but
coaxial sections are tedious to build, they do fill up
with moisture, and it takes considerable faith to be
sure that the matching section is really 1/4 or 3/4
wavelength, or some other needed value, at these
frequencies. A simple tapered line, used as the
matching section, gets around this problem.
Fig. 12-19 — Top and front views of the
A QUADHELIX FOR 1296 MHz quadhelix array. The feed lines from the individual
helices terminate at the center of the array. The
If one helix is good, two - properly matched common meeting point for these lines is at the
and phased — are approximately 3 dB better. Four center conductor of a short extension of coaxial
helices are another 3 dB better yet — and are easy cable. This extension should be of the same
to match by the method described. impedance as the cable used to feed the array.
262 Chapter 12

helix. At the coax end where the four tapered lines

join,
4/1 200
‘° gl0 ¿ W 145

4/i
— = 28.2
d

, 28.2x0.1019
h =- = 0.718 in.
4

Slight discrepancies here are nearly meaningless as

the VSWR is between the four tapered lines in
parallel and the 52-ohm coax line. For other
conditions, just substitute the appropriate num¬
bers. The important point is to keep the matching
lines as close to the ground plane as possible.
The helices are mounted (center to center) on a
square whose sides are 1.5 wavelength, or 13.7
Fig. 12-20 — The method of feeding one of the inches for 1296 MHz. The tapered lines will then
four helices in the array. The feed-line extension is
a short piece of copper pipe or tubing. The be approximately one wavelength long. In any
diameter ratio to the center conductor should be case, keep the tapered-line lengths equal. The
such that this extension is of the same impedance aluminum reflector screen should be 3.5 wave¬
as the transmission line to the station. The height lengths (minimum) on a side. The details of the
of the extension is not frequency dependent but individual helices are shown in Fig. 12-20. The
rather is adjusted to provide a match when overall layout can be seen in the photographs.
paralleling four helices. The slope of the line from Mounting of the helix turns on the wooden booms
the coax extension to the feedpoint of the individ¬ was accomplished in earlier versions by using small
ual antenna is also important in obtaining a match.
See Table 12-1V for value of d. metal horseshoe brads. A more elegant method is
to mount the turns on small standoff insulators,
and recent models using this method show in¬
creased stability with respect to weather effects.
for RG-8, -9, -14, or -17 coaxial cable. The tapered Care should be taken when adjusting the position
lines are also made from No. 10 AWG wire. of the helix coils on the boom to assure that the
If you desire to vary the impedance ratios pitch angle (spacing between the turns) is the same
because of a different feed-line impedance, the for each turn of the helix. You should visually
following calculation is typical. For a single wire
near a ground
Zo = 138 log 10 — _ TABLE 12-IV_
d Helix Design Data
where his the height to the conductor center, and
d is the wire diameter. In order to reduce any Diameter D = .35 X, and Circumference
interaction of the fields between the helices and C= 1.1X.
the tapered phasing lines, it is important to keep
the tapered lines as close to the ground plane as Pitch Angle a= 12.5°; range may be from
practical. This condition can be met by using No. 12° to 15°.
10 wire for the tapered lines. The diameter of this
size wire is 0.1019 inch. Then, assuming that the No. of Turns = minimum of 3.
impedance at a helix feed point is 138 ohms.
Spacing of Turns d= X tan a: for
2300 MHz — 130 mm x .222 = 29 mm; for
, 4/1 138 1296 MHz = 231 mm.x .222 = 51 mm.
" Hg " 1
1. The above figures will cause the helix
to perform best for gain and pattern near its
4/t
high-frequency end.
7 =io 2. The pitch angle chosen will compen¬
sate for the dielectric constant of the boom
material where the material is of good rf
quality, such as dry wood, PVC, Plexiglas,
, 10x0.1019 etc.
h =--- = 0.255 in.
4 3. Helix will operate axially over a freq,
range of 1.7 to 1.0 from a design center
where = 1.0.
This is the height at the feed-point end of each
Antennas for Space Communications 263

check to be sure that the helix circumference Bibliography

remains constant along the helix axis.
One very important consideration when build¬ Source material and more extended discussion
ing the helices is to make some provision for keep¬ of topics covered in this chapter can be found in
ing the conductivity of the driven-element material the references given below.
high. The current is quite high in some portions of
the helix, and losses there will degrade the perfor¬
mance severely. Copper is recommended and it DeMaw, “The Basic Helical Beam,” QST, Novem¬
should be cleaned very thoroughly and sprayed ber, 1965.
with a preservative such as Krylon. To really ensure Knadle, “A Twelve-Foot Stressed Parabolic Dish,”
long-time performance a coating of silver will be a QST, August, 1972.
good investment. McKnight, “Evolution of an Amateur Weather-
In recent years, bargain prices for 75-ohm cable Satellite Picture Station,” QST, April, 1968,
have led to a modification of the original feed and and Feedback, QST, July, 1968, p. 67.
matching dimensions. These are indicated in Fig. Nose, “Crossed Yagi Antennas for Circular Polar¬
ization,” QST, January, 1973.
12-20. The quadhelix is also beginning to show up Nose, “A Simple Az-El Antenna System for
in 2300-MHz circles. The reader interested in Oscar,” QST, June, 1973.
building one for that band need only apply the Scott and Banta, “Using the Helical Antenna at
correct numbers to the formulas for determining 1215 Me.,” QST, July, 1962.
pitch angle and diameter. The feed-point modi¬ Troetschel, “A Quadhelix Antenna for the
fications are not frequency dependent. 1215-Mc. Band,” QST, August, 1963.
 Chapter 13

Construction of Wire Antennas

Although wire antennas are relatively simple, run under or over public-utility (telephone or
they can constitute a potential hazard unless power) lines. Several amateurs have lost their lives
properly constructed. Antennas should never be by failing to observe this precaution.

ANTENNA MATERIALS
ANTENNA WIRE critical, although a change in length may require
some readjustment of coupling to the transmitter.)
The rf resistance of copper wire increases as the “Hard-drawn” copper wire or copper-clad steel,
size of the wire decreases. However, in most types especially the latter, is harder to handle, because it
of antennas that are commonly constructed of has a tendency to spiral when it is unrolled.
wire, the rf resistance, even for quite small sizes of However, these types are mandatory for applica¬
wire, will not be so high, compared to the radiation tions where significant stretch cannot be tolerated.
resistance, that the efficiency of the antenna will Care should be exercised in using this wire to make
suffer greatly. Wire sizes as small as No. 30, or even sure that kinks do not develop that may cause the
smaller, have been used quite successfully in the wire to break at far under normal stress. After the
construction of “invisible” antennas in areas where coil has been unwound, it is advisable to suspend
there is local objection to the erection of more the wire a few feet above ground for a day or two
conventional types. In most cases, the selection of before making use of it. The wire should not be
wire for an antenna will be based primarily on the recoiled before installing.
physical properties of the wire, since the sus¬ The size of the wire to be selected, and the
pension of wire from elevated supports places a choice between hard-drawn and copper-clad, will
strain on the wire.» depend on the length of the unsupported span, the
amount of sag that can be tolerated, the stability
Wire Types of the supports under wind pressure, and whether
Wire having an enamel-type coating is prefer¬ or not an unsupported transmission line is to be
able to bare wire, since the coating resists oxida¬ suspended from the span.
tion and corrosion. Several types of wire having
this type of coating are available, depending on the Wire Tension
strength needed. “Soft-drawn” or annealed copper Table 13-1 shows the maximum rated working
wire is easiest to handle but, unfortunately, is tensions of hard-drawn and copper-clad steel wire
subject to considerable stretch under stress. It of various sizes.
should therefore be avoided, except for applica¬ If the tension on a wire can be adjusted to a
tions where the wire will be under little or no known value, the expected sag of the wire, as
tension, or where some change in length'can be
tolerated. (For instance, the length of a horizontal
antenna fed at the center with open-wire line is not
♦ The National Electric Code of the National
Fire Protection Association contains a section on
amateur stations in which a number of recom¬
mendations are made concerning minimum size of
antenna wire and the manner of bringing the
transmission line into the station. The code in itself
does not have the force of law, but it is frequently
made a part of local building regulations, which are
enforceable. The provisions of the code may also
be written into, or referred to, in fire and liability
insurance documents. A copy of this code may be
obtained from National Fire Protection Associa¬
tion, 60 Batterymarch St., Boston, MA 02110 Fig. 13-1 — The span and sag of a long-wire
(Price $3.50.) antenna.

264
Construction of Wire Antennas 265

TABLE 13-1
Stressed An tenna Wire
American Recommended Tension^ Weight
Wire Gauge (Pound*) (Pounds per 1000 Peel)
Copper-Clad Hard-Drawn Copper-Clad Hard-Drawn
Steel 2 Copper Steel 2 Copper
4 495 214 115.8 126
6 310 130 72.9 79.5
8 195 84 45.5 50
10 120 52 28.8 31.4
12 75 32 18.1 19.8
14 50 20 11.4 12.4
16 31 13 7.1 7.8
18 19 8 4.5 4.9
20 12 5 2.8 3.1
1 Approximately one-tenth the breaking load. Might be increased 50 per cent if end supports are firm and

there is no danger of ice loading.

2 “Copperweld," 40 per cent copper.

depicted in Fig. 13 1, may be determined in Wire Splicing

advance of installation with the aid of Table 13-1 Wire antennas should preferably be made with
and the nomograph of Fig. 13-2. Even though unbroken lengths of wire. In instances where this is
there may be no convenient method of deter¬ not feasible, wire sections should be spliced as
mining the tension in pounds, calculation of the shown in Fig. 13-3. The enamel insulation should
expected sag for practicable working tensions is be removed for a distance of about 6 inches from
often desirable. If the calculated sag is greater than the end of each section by scraping with a knife or
allowable it may be reduced by any one or a rubbing with sandpaper until the copper under¬
combination of the following: neath is bright. The turns of wire should be
1) Providing additional supports thereby de¬ brought up tight around the standing part of the
creasing the span, wire by twisting with broad-nose pliers.
2) Increasing the tension in the wire if less than The crevices formed by the wire should be
recommended, completely filled with solder. Since most antenna
3) Decreasing the size of the wire. soldering must be done outdoors, the ordinary
Conversely, if the sag in a wire of a particular
installation is measured, the tension can be deter¬ SPAN TENSION WORK SAG WEIGHT
tfeet) (pound«) AXIS <<••’> (lb«/IOOOft)
mined by reversing the procedure.

g g g g S o S S g
1000 -r tooo
M W U * *»

.1
o <* o w o o
i_i_1-1-1—1—1—M

»00 - •00
•oo - •00
Instructions for Using the Nomograph 700 -
'

400
•00 -
1) From Table 13-1, find the weight (pounds/
I

300 -
1000 feet) for the particular wire size and material 200
'

400
to be used.
2) Draw a line from the value obtained above, 100
1
w

300 -
•0
plotted on the weight axis, to the desired span «0
•

(feet) on the span axis, Fig. 13-2.

M
■ ■_
o

200 -
•

3) Choose an operating tension level (pounds)

—
«

consistent with the values presented in Table 13-1

—
•

- 20
//

(preferably less than the recommended wire ten¬

¿
/ —
¡»4» .

1—1—1—III.'

sion). 100 - - 10
m

4) Construct a line from the tension value •0 - - •

.

•0- «
chosen, plotted on the tension axis, through the 70 •
W*««***.
«
i

crossover point of the work axis and the original •0

- 4
i

line constructed from Step 2, above, and continue 30

<a
.

2
this new line to the sag axis. 40
■

5) Read the sag (feet) on the sag axis.

a

30
Example:
i

- 0.«
0.«
Weight =11 pounds/1000 feet.
1-1
.

20 - 0.4
Span = 210 feet.
—

Tension = 50 pounds.
•

0.2
1-1

Answer:
Sag = 4.7 feet.
o

- 0.1
_

to
•

Of course, these calculations do not take the

weight of a feed line into account, if it is supported Fig. 13-2 — Nomograph for determining wire sag.
by the antenna wire. (K1 AFR.)
 266 Chapter 13

stress without danger of breakage. Long elastic

bands or lengths of nylon fishing line provide long
leakage paths and make satisfactory insulators
within their limits to resist mechanical strain. They
are often used in antennas of the “invisible” type
mentioned earlier.
For low-power work with short antennas not
subject to appreciable stress, almost any small glass
or glazed-porcelain insulator will do. Homemade
insulators of Lucite rod or sheet will also be
Fig. 13-3 - Correct method of splicing antenna satisfactory. More care is required in the selection
wi re.
of insulators for longer spans and higher trans¬
mitter power.
soldering iron or gun may not provide sufficient
For the same material, the breaking tension of
heat, and the use of a propane torch may become
an insulator will be proportional to its cross-
desirable. The joint should be heated sufficiently
sectional area. It should be remembered, however,
so that the solder will flow freely into the joint
that the wire hole at the end of the insulator
when the source of heat is removed momentarily.
decreases the effective cross-sectional area. For this
After the joint has cooled completely, it should be
reason, insulators designed to carry heavy strains
wiped clean with a cloth, and then sprayed
are fitted with heavy metal end caps, the eyes
generously with acrylic to discourage corrosion.
being formed in the metal cap, rather than in the
ANTENNA INSULATION insulating material itself. The following stress
ratings of several antenna insulators made by E. F.
To prevent loss of power, the antenna should Johnson are typical:
be well insulated from ground, particularly at the 5/8 inch square by 4 inches long - 400 lbs.
outer end or ends, since these points are always at 1 inch diameter by 7 or 12 inches long - 800
a comparatively high rf potential. If an antenna is lbs.
to be installed indoors (in an attic, for instance) 1-1/2 inches diameter by 8, 12 or 20 inches
the antenna may be suspended directly from the long, with special metal end caps - 5000
wood rafters without additional insulation, if the lbs.
wood is permanently dry. However, when the These are rated breaking tensions. The actual
antenna is located outside, where it is exposed to working tensions should be limited to not more
wet weather, much greater care should be given to than 25% of the breaking rating.
the selection of proper insulators.

Insulator Leakage
The insulators should be of material that will
not absorb moisture. Most insulators designed
specifically for antenna use are made of glass or
glazed porcelain. Aside from this, the length of an
STRAIN INSULATOR
insulator in proportion to its surface area is
indicative of its comparative insulating ability. A Fig. 13-5 — Conventional manner of fastening to a
long thin insulator will have less leakage than a strain insulator. This method decreases the leakage
short thick insulator. Some antenna insulators are path, and increases capacitance, as discussed in the
deeply ribbed to increase the surface leakage path text.
without increasing the physical length of the
insulator. Shorter insulators can be used at low-
The antenna wire should be attached to the
potential points, such as at the center of a dipole.
insulators as shown in Fig. 13-4. Care should be
However, if such an antenna is to be fed with
taken to avoid sharp angular bends in the wire in
open-wire line and used on several bands, the
center insulator should be the same as those used looping it through the insulator eye. The loop
at the ends, because high rf potential will exist should be generous enough in size that it will not
across the center insulator on some bands. bind the end of the insulator tightly. If the length
of the antenna is critical, the length should be
Insulator Stress measured to the outward end of the loop, where it
passes through the eye of the insulator. The
As with the antenna wire, the insulator must soldering should be done as described earlier for
have sufficient physical strength to sustain the the wire splice.
Insulator Eye

Fig. 13-4 — In fastening antenna wire to an

insulator, the wire loop should not be made too
snug. After completion, solder should be flowed
into the turns. When the joint has cooled com¬
pletely, it should be sprayed with acrylic.
Construction of Wire Antennas 267

Strain Insulators
Strain insulators have their holes at right angles,
since they are designed to be connected as shown
in Fig. 13-5. It can be seen that this arrangement
places the insulating material under compression,
rather than tension. An insulator connected this
way can withstand much greater stress. Further¬
more, if the insulator should break, the wire will
not collapse, since the two wire loops are inter¬
locked. However, because the wire is wrapped
around the insulator, the leakage path is reduced
quite drastically, and the capacitance between the
wire loops provides an additional leakage path. For
this reason the use of the strain insulator is usually
confined to such applications as breaking up
resonances in guy wires, where high levels of stress
prevail, and where the rf insulation is of less
importance. Such insulators might, however, be
suitable for use at low-potential points on the
antenna, such as at the centers of dipoles. These Fig. 13-6 — A — Insulator for ends of folded
insulators may also be fastened in the conventional dipoles, or multiple dipoles made of 300-ohm
manner if the wire will not be under sufficient ribbon. B — A method of suspending one ribbon
tension to break the eyes out. dipole from another in a multiband dipole system.

Insulators for Ribbon-Line Antennas allowing the grooved wheel to break away under
tension.
Fig. 13-6A shows the sketch of an insulator Most good-quality clothesline pulleys are made
designed to be used at the ends of a folded dipole, of alloys which do not corrode readily. Since they
or a multiple dipole made of ribbon line. It should are designed to carry at least 50 feet of line loaded
be made approximately as shown, out of Lucite or with wet clothing in stiff winds, they should be
bakelite about 1/4-inch thick. The advantage of adequate for normal spans of 100 to 150 feet
this arrangement is that the strain of the antenna is between stable supports. One type of clothesline
shared by the conductors and the plastic webbing pulley has a 4-inch diameter plastic wheel with a
of the ribbon, which adds considerable strength. 1/4-inch shaft in bronze bearings. The sheath is of
After soldering, the screw should be sprayed with cast or forged corrosion-proof alloy. Such pulleys
acrylic. sell for about one dollar in hardware stores.
Fig. 13-6B shows a similar arrangement for Marine pulleys have good weather-resisting
suspending one dipole from another in a multiple¬ qualities, since they are usually made of bronze,
dipole system. but they arc comparatively expensive and are not
If better insulation is desired, these insulators designed to carry heavy loads. For extremely long
can be wired to a conventional insulator.

PULLEYS AND HALYARDS

Pulleys and halyards commonly used to raise
and lower the antenna are also items that must be
capable of taking the same strain as the antenna
wire and insulators. Unfortunately little specific
information on the stress ratings of most pulleys is
available. Several types of pulleys are readily
available at almost any hardware store. Among
these are small galvanized pulleys designed for
awnings, and several styles and sizes of clothesline
pulleys. In judging the stress that any pulley might
handle, particular attention should be paid to the
diameter of the shaft, how securely the shaft is
fitted into the sheath, and the size and material of
which the frame is made. Heavier and stronger Fig. 13-7 — This is one type of knot that will hold
pulleys are those used in marine work. with smooth rope, such as nylon. A shows the knot
for splicing two ends. B shows the use of a similar
Another important factor to be considered in knot in forming a loop, as might be needed for
the selection of a pulley is its ability to resist attaching an insulator to a halyard. Knot A is first
corrosion. Galvanized awning pulleys are probably formed loosely 10 or 12 inches from the end of the
the most susceptible to corrosion. While the frame rope; then the end is passed through the eye of the
or sheath usually stands up well, these pulleys insulator and knot A. Knot B is then formed and
usually fail at the shaft, which eventually rusts out. both knots pulled tight. (K7HDB.)
 268 Chapter 13

spans, the wood-sheathed pulleys used in “block

and tackle” devices, and for sail hoisting should fill
the requirements. TABLE 13-11
Halyards
Approximate Safe Working Tension (lbs.)
Table 13-11 shows recommended maximum for Various Halyard Materials
tensions for various sizes and types of line and rope
suitable for hoisting halyards. Probably the best Manila Rope
type for general amateur use for spans up to 150 or X" - 120 X" — 270 X" — 530 X" - 800
200 feet is 1/4-inch nylon rope. It is somewhat Polypropylene Rope
more expensive than ordinary rope of the same Ji" - 270 X" - 530 X" - 840
size, but it weathers much better. Furthermore, it
has a certain amount of elasticity to accommodate Nylon Rope
gusts of wind, and is particularly recommended for X" - 300 X" - 660 X" - 1140
antennas using trees as supports.
Most types of synthetic rope are slippery, and 7x11 Galvanized Sash Cord
some types of knots ordinarily used for rope will Me"-30 X" — 125 : M«" - 250 «"-450
not hold well. Fig. 13-7 shows a knot that should
hold well, even with nylon rope or plastic line.
High-Strength Stranded
Galvanized Guy Wire
For exceptionally long spans, stranded galvan¬ X" - 400 % 6" - 700 «" - 1200
ized steel sash cord is suitable. Cable advertised as
“wire rope” usually does not weather well. Rayon-filled Plastic Clothesline
A convenience in antenna hoisting (usually a %2” - 60 to 70
necessity with metal halyards) is the boat winch
sold at marinas, and also at such places as Sears.

INSTALLING TRANSMISSION LINES

CONNECTING LINE TO ANTENNA slack tie wire should then be used between the
feeder conductor and the antenna, as shown in Fig.
In connecting coaxial cable or 300-ohm ribbon
13-10. (The tie wires may be extensions of the line
line to a dipole that does not have a support at the
conductors themselves.)
center, it is essential that the conductors of the line
When using TV-type open-wire line, the ten¬
be relieved of the weight of the cable or ribbon.
dency of the line to twist and short out close to
Fig. 13-8 shows a method of accomplishing this
the antenna can be counteracted by making the
with coaxial cable. The cable is looped around the
center insulator of the antenna longer than the
center antenna insulator, and clamped before
spacing of the Une, as shown in Fig. 13-10. In this
making connections to the antenna. In Fig. 13-9,
case, a heavier spreader insulator should be added
the weight of the ribbon line is removed from the
just below the antenna insulator to prevent side
conductor by threading the line through a sheet of
stress from pulling the conductors away from the
insulating material. The sheet is suspended from light plastic feeder spreaders.
the antenna by threading the antenna through the
sheet. This arrangement is particularly suited to
folded dipoles made of 300-ohm ribbon. RUNNING LINE FROM
In connecting an open-wire line to an antenna,
the conductors of the line should be anchored to ANTENNA TO STATION
the insulator by threading them through the eyes Coaxial cable requires no particular care in
of the insulator two or three times, and twisting running from the antenna to the station entrance,
the wire back on itself before soldering fast. A except to protect it from mechanical damage. If

Fig. 13-8 — Method of relieving strain on conduc¬

tors of coaxial cable in feeding a dipole.
Construction of Wire Antennas 269

point of entrance. The points of anchorage and

entrance should be chosen so as to permit the
vertical drop without crossing windows.
If the station is located in a room on the
ground floor, one way of bringing coax trans¬
mission line in is to go through the outside wall
below floor level, feed it through the basement,
and then up to the station through a hole in the
floor. In making the entrance hole in the side of
the building, suitable measurements should be
made in advance to make sure that the hole will go
through the sill 2 or 3 inches above the foundation
line (and between joists if the bore is parallel to the
joists). The line should be allowed to sag below the
entrance-hole level to allow rain water to drip off,
Fig. 13-9 — Strain reliever for conductors of and not follow the line into the building.
300-ohm ribbon line in a folded dipole. The piece Open-wire line can be fed in a similar manner,
can be made from 1/4-inch Lucite sheet. although it will require a separate hole for each
conductor. The hole should be insulated with
the antenna is not supported at the center, the line
should bo fastened to a post more than head high
located under the center of the antenna, allowing
enough slack between the post and the antenna to
take care of any movement of the antenna in the
wind. If the antenna feed point is supported by a
tower or mast, the cable can be taped at intervals
to the mast, or to one leg of the tower.
If desired, coaxial cable can be buried a few
inches in the ground in making the run from the
antenna to the station. A deep slit can be cut by
pushing a square-end spade full depth into the
ground, and moving the handle back and forth to
widen the slit before removing the spade. After the
cable has been pushed into the slit with a piece of
1-inch board 3 or 4 inches wide, the slit can be
closed by tamping.
Ribbon line should be kept reasonably well
spaced from other conductors running parallel to it
for more than a few feet. TV-type standoff
insulators with strap-clamp mountings can be used
in running this type of line down a mast or tower
leg. Similar insulators of the screw type can be
used in supporting the line on poles for a long run.
Open-wire lines, especially TV types, require Fig. 13-10 - Method of connecting open-wire line
frequent supports to keep the line from twisting to center antenna insulator. The Lucite strip keeps
the feed-line conductors from pulling away from
and shorting out, as well as to relieve the strain. the spreaders when TV open-wire line is used.
One method of supporting a long run of heavy
open-wire line is shown in Fig. 13-11. The line
must be securely anchored at a point under the
feed point of the antenna. TV-type line can be lengths of polystyrene or Lucite tubing, and should
supported similarly by means of wire links fastened be drilled with a slight downward slant toward the
to the insulators. Fig. 13-12 shows a method of outside of the building to prevent rain seepage.
supporting an open-wire line from a tower. With TV-type line, it will be necessary to remove a
To keep the line clear of pedestrians and few of the spreader insulators, cut the line before
vehicles, it is usually desirable to anchor the feed passing through the holes (allowing enough length
line at the eave or rafter line of the station building to reach the inside), and splice the remainder on
(see Fig. 13-13), and then drop it vertically to the the inside.

Fig. 13-11 - A support for

open-wire line. The sup¬
port at the antenna end of
the line must be suffi¬
ciently rigid to stand the
tension of the line.
 270
Chapter 13

through insulators can be used for open-wire line.

Ribbon line can be run through clearance holes in
the panel, and secured by a winding of tape on
either side of the panel, or by cutting the retaining
rings and insulators from a pair of TV standoff
insulators, and clamping one on each side of the
panel.

LIGHTNING PROTECTION
Two or three types of lightning arresters for
coaxial cable are available on the market. These are
designed to join two lengths of coax cable. If the
antenna feed point is at the top of a well-grounded
tower, the arrester can be fastened securely to the
top of the tower for grounding purposes. A short
length of cable, terminated in a coaxial plug, is
then run from the antenna feed point to one
receptacle of the arrester, while the transmission
line is run from the other arrester receptacle to the
station. Such arresters may also be placed at the
entrance point to the station, if a suitable ground
connection is available at that point (or arresters
may be placed at both points for added insurance).
The construction of a homemade arrester for
open-wire line is shown in Fig. 13-16. This type of
arrester can be adapted to ribbon line, as shown in
Fig. 13-17. The two TV standoff insulators should
elevate the ribbon line an inch or so away from the
center member of the arrester. Sufficient insulation
should be removed from the line where it crosses
the arrester to permit soldering the arrester con¬
necting leads.
Fig. 13-12 - A board fitted with standoff insula¬
tors and clamped to the tower with U bolts keeps Lightning Grounds
open-wire line suitably spaced from a tower
(W4NML.) Lightning-ground connecting leads should be of
conductor equivalent to at least No. 10 wire. The
No. 8 aluminum wire used for TV-antenna grounds
If the station is located above ground level, or is satisfactory. Copper braid 3/4-inch wide (Belden
there is other objection to the procedure described 8662-10) is also suitable. The conductor should
above, entrance can be made at a window, using run in a straight line to the grounding point. The
the arrangement shown in Fig. 13-14. An Am¬ ground connection may be made to a water-piping
phenol type 83-1F (UG-363/U) connector can be system, the grounded metal frame of a building, or
used as a feedthrough for coaxial line, or one can to one or more 5/8-inch ground rods driven to a
be made as shown in Fig. 13-15; ceramic feed- depth of at least 8 feet.

SUPPORTS FOR WIRE ANTENNAS

A prime consideration in the selection of a with the safety of life or property. The courts
support for an antenna is that of structural safety. generally have recognized, however, that municipal
Building regulations in many localities require that authority does not extend to esthetic questions;
a permit be secured in advance of the erection of i.e., the fact that someone may object to the mere
structures of certain types, often including antenna presence of a pole or tower, or an antenna
poles or towers. In general, localities having such structure, because in his opinion it detracts from
requirements also will have building safety codes the beauty of the neighborhood, is not grounds for
that must be observed. Such regulations may refusing to issue a permit for a safe structure to be
govern the method and materials used in con¬ erected.
struction of, for example, a self-supporting tower. But, even where such regulations do not exist
Checking with your local government building or are not enforced, the amateur should be careful
department before putting up a tower may save a to select a type of support and a location for it
good deal of difficulty later, since a tower would that will minimize the chances of collapse and, if
have to be taken down or modified if not approved collapse does occur, will minimize the chances that
by the building inspector on safety grounds. someone will be injured or property damaged. A
Municipalities have the right and duty to single injury can be far more costly than the price
enforce any reasonable regulations having to do of a more rugged support.
Construction of Wire Antennas 271

Fig. 13-15 — Feedthrough con¬

nector for coax line. An
Amphenol 83-1J (PL-258) con¬
nector, the type used to splice
sections of coax line together, is
soldered into a hole cut in a brass
COAX SPLICE mounting flange. Amphenol bulk
CONNECTOR adapter 83-1 F may be used in¬
stead.

wind, unless the tree is a very large one, and the

antenna is suspended from a point well down on
the tree trunk. As a result, the antenna must be
constructed much more sturdily than would be
necessary with stable supports. Even with rugged
construction, it is unlikely that an antenna sus¬
pended from a tree, or between trees, will stand up
indefinitely, and occasional repair or replacement
Fig. 13-13 — Anchorage for open-wire line at the usually must be expected.
station end. The springs are especially desirable if There are two general methods of securing a
the line is not supported between the antenna and pulley to a tree. If the tree can be climbed safely to
the anchoring point. the desired level, a pulley can be wired to the trunk
of the tree, as shown in Fig. 13-18. If, after passing
TREES AS ANTENNA SUPPORTS the halyard through the pulley, both ends of the
halyard are simply brought back down to ground
From the beginning of amateur radio, trees along the trunk of the tree, there may be difficulty
have been used widely for supporting wire anten¬
nas. Trees cost nothing, of course, and will often
provide a means of supporting a wire antenna at
considerable height. However, as an antenna sup¬
port, a tree is highly unstable in the presence of

Fig. 13-16 - A simple lightning arrester for

open-wire line made from three standoff or feed-
through insulators and sections of 1/8 X 1/2-inch
brass or copper strap. It should be installed in the
line at the point where the line enters the station.
The heavy ground lead should be as short and
direct as possible. The gap setting should be
adjusted to the minimum width that will prohibit
arcing when the transmitter is operated.

in bringing the antenna end of the halyard out

where it will be clear of branches. To avoid this,
one end of the halyard can be tied temporarily to
the tree at the pulley level, while the remainder of
the halyard is coiled up, and the coil thrown out
horizontally from this level, in the direction in
which the antenna will run. It may help to have the
antenna end of the halyard weighted. Then, after
attaching the antenna to the halyard, the other end
is untied from the tree, passed through the pulley,
Fig. 13-14 — An adjustable window lead-in panel and brought to ground along the tree trunk in as
made of two sheets of Lucite. A feedthrough
connector for coax line can be made as shown in straight a line as possible. The halyard need be only
Fig. 13-15. Ceramic feedthrough insulators are long enough to reach the ground after the antenna
suitable for open-wire line. (W1 RVE.) has been hauled up, since additional rope can be
 272 Chapter 13

Several trials may be necessary to determine the

optimum size of the weight for the line selected,
the distance between the weight and the hand
before throwing, and the point in the arc of the
swing where the line is released. The weight,
however, must be sufficiently large to assure that it
will carry the pilot line back to ground after
passing over the tree. Flipping the end of the line
up and down so as to put a traveling wave on the
line often helps to induce the weight to drop down
if the weight is marginal. The higher the tree, the
lighter the weight and the pilot line must be. A
glove should be worn on the throwing hand,
because a line running swiftly through the bare
Fig. 13-17 - The lightning arrester of Fig. 13-16 hand can cause a severe burn.
may be used with 300-ohm ribbon line in the If there is a clear line of sight between ground
manner shown here. The TV standoffs support the and a particularly desirable crotch in the tree, it
line an inch or so away from the grounded center may be possible to hit the crotch eventually after a
member of the arrester. sufficient number of tries. Otherwise, it is best to
try to heave the pilot line completely over the tree,
tied to the halyard when it becomes necessary to as close to the center line of the tree as possible. If
lower the antenna. it is necessary to retrieve the line and start over
The other method consists of passing a line over again, the line should be drawn back very slowly,
the tree from ground level, and using this line to otherwise the swinging weight may wrap the line
haul a pulley up into the tree and hold it there. around a small limb, making retrieval impossible.
Several ingenious methods have been used to
accomplish this. The simplest method employs a
weighted pilot line, such as fishing line or mason’s
chalk line. Grasping the line about two feet from
the weight, the weight is swung back and forth,
pendulum style, and then heaved with an under¬
hand motion in the direction of the tree top.

Fig. 13-19 - In connecting the halyard to the pilot

line, a large knot that might snag in the crotch of a
tree should be avoided, as shown.

Stretching the line out in a straight line on the

ground before throwing may help to keep the line
from snarling, but it places extra drag on the line,
and the line may snag on obstructions overhanging
the line when it is thrown. Another method is to
make a stationary reel by driving eight nails,
arranged in a circle, through a 1-inch board. After
winding the line around the circle formed by the
nails, the line should reel off readily when the
weighted end of the line is thrown. The board
should be tilted at approximately right angles to
the path of the throw.
Other devices that have been used successfully
to pass a pilot line over a tree are the bow and
arrow with heavy thread tied to the arrow, and the
short casting rod and spinning reel used by
fishermen. Still another method that has been used
Fig. 13-18 — Methods of counterweighting to where sufficient space is available is to fly a kite.
minimize antenna movement. The method at A After the kite has reached sufficient altitude,
limits the fall of the counterweight should the simply walk around the tree until the kite string
antenna break. It also has a 2 to 1 mechanical lines up with the center of the tree. Then pay out
advantage, as indicated. The method at B has the string until the kite falls to the earth. This method
disadvantage that the point of support in the tree has been used successfully to pass a line over a
must be higher than the end of the antenna. patch of woods between two higher supports,
Construction of Wire Antennas 273

which would have been impossible using any other

method.
The pilot line can be used to pull successively
heavier lines over the tree until one of adequate
size to take the strain of the antenna has been
reached. This line is then used to haul a pulley up
into the tree after the antenna halyard has been
threaded through the pulley. The line that holds
the pulley must be capable of withstanding con¬
siderable chafing where it passes through the
crotch, and at points where lower branches may
rub against the standing part. For this reason, it
may be advisable to use galvanized sash cord or
stranded guy wire for raising the pulley.
Especially with larger sizes of line or cable, care
must be taken when splicing the pilot line to the
heavier line to use a splice that will minimize the
chances that the splice cannot be coaxed through
the tree crotch. One type of splice is shown in Fig.
13-19.

Fig. 13-21 — The cleat avoids the necessity of

having to untie a knot that may have been
weather-hardened.

Wind Compensation
The movement of an antenna suspended be¬
tween supports that are not stable in wind can be
reduced materially by the use of heavy springs,
such as screen-door springs under tension, or by a
counterweight at the end of one halyard, as shown
in Fig. 13-18. The weight, which may be made up
of junk-yard metal, window sash weights, or a
galvanized pail filled with stand or stone, should be
adjusted experimentally for best results under
Fig. 13-20 — A weighted line thrown over the
existing conditions. Fig. 13-21 shows a convenient
antenna can be used to pull the antenna to one side way of fastening the counterweight to the halyard.
to avoid overhanging obstructions, such as It avoids the necessity for untying a knot in the
branches of trees in the path of the antenna, as the halyard which may have hardened under tension
antenna is pulled up. When the obstruction has and exposure to the weather.
been cleared, the line can be removed by releasing
one end. Trees as Supports for Vertical Wire Antennas
Trees can often be used to support vertical as
well as horizontal antennas. If the tree is a tall one
The crotch which the line first comes to rest in with overhanging branches, the scheme of Fig.
may not be sufficiently strong to stand up under 13-22 may be used. The top end of the antenna is
the tension of the antenna. However, if the line has secured to a halyard passed over the limb, brought
been passed over, or close to, the center line of the back to ground level, and fastened to the trunk of
tree, it will usually break through the lighter the tree.
crotches and finally come to rest in one suffi¬ Halyard_J
ciently strong lower down on the tree. secured-*1
Needless to say, any of the suggested methods to branch a
should be used with due respect to persons or
property in the immediate vicinity. A child’s U— Antenna.
sponge-rubber ball (baseball size) makes a safe I

weight for heaving a heavy thread line or fishing I

line.
If the antenna wire becomes snagged in lower
branches of the tree when the wire is pulled up, or Post— \ I
if branches of other trees in the vicinity interfere -Pultoy
with raising the antenna, a weighted line thrown ±~—Gounterweight
over the antenna and slid along to the appropriate
point is often helpful in pulling the antenna wire to
one side to clear the interference as the antenna is Fig. 13-22 — Counterweight for a vertical antenna
being raised, as shown in Fig. 13-20. suspended from an overhanging tree branch.
 274 Chapter 13

MASTS
Where suitable trees are not available, or a more
stable form of support is desired, masts are suitable
for wire antennas of reasonable span length. At one
time, most amateur masts were constructed of
lumber, but the TV industry has brought out metal
masting that is inexpensive and much more durable
than wood. However, there are some applications
where wood is necessary or desirable.

The “A-Frame” Mast

A light and inexpensive mast is shown in Fig.
13-23. In lengths up to 40 feet it is very easy to
erect and will stand without difficulty the puli of
ordinary wire antenna systems. The lumber used is
Fig. 13-24 — Method of assembling the "A-frame"
2X2 straight-grained pine (which many lumber mast.
yards know as hemlock) or even fir stock. The
uprights can be each as long as 22 feet (for a mast
slightly over 40 feet high) and the crosspieces are mast two coats of “outside-white” house paint or
cut to fit. Four pieces of 2 X 2, 22 feet long, will latex.
provide enough and to spare. The only other After the second coat of paint is dry, attach the
materials required are five 1/4-inch carriage bolts guys and rig the pulley for the antenna halyard.
5-1/2 inches long, a few spikes, about 300 feet of The pulley anchorage should be at the point where
stranded or solid galvanized iron wire for the guys the top stays are attached so that the back stay will
or stays, enough glazed-porcelain compression in¬ assume the greater part of the load tension. It is
sulators (“eggs”) to break up the guys into better to use wire wrapped around the mast with a
sections, and the usual pulley and halyard rope. If small through-bolt to prevent sliding down, than to
the strain insulators are put in every 20 feet, use eye bolts.
approximately 15 of them will be enough. If the mast is to stand on the ground, a couple
After selecting and purchasing the lumber - of stakes should be driven to keep the bottom
which should be straight-grained and knot-free — from slipping. At this point the mast may be
three sawhorses or boxes should be set up and the “walked up” by a helper. If it is to go on a roof,
mast assembled in the manner indicated in Fig. first stand it up against the side of the building and
13-24. At this stage it is a good plan to give the then hoist it, from the roof, keeping it vertical. The
whole assembly is light enough for two men to
perform the complete operation - lifting the mast,
carrying it to its permanent berth, and fastening
the guys - with the mast vertical all the while. It is
therefore entirely practicable to put up this kind of
mast on a small flat area of roof that would
prohibit the erection of one that had to be raised
to the vertical in its final location.

TV Masting
TV masting is available in 5- and 10-foot
lengths, 1-1/4 inches in diameter, in both steel and
aluminum. These sections are crimped at one end
to permit sections to be joined together. However,
a form that will usually be found more convenient
is the telescoping TV mast available from many
electronic supply houses. The masts may be ob¬
tained with three, four or five 10-foot sections, and
come complete with guying rings and a means of
locking the sections in place after they have been
extended. These masts are stronger than the
nontelescoping type because the top section is
1-1/4 inches in diameter, and the diameter in¬
creases toward the bottom section which is 2-1/2
inches in diameter in the 50-foot mast.
Guy rings are provided at 10-foot intervals, but
guys may not be required in all points. Guys at the
top are essential, and at least one other set near the
center of the mast will usually be found necessary
Fig. 13-23 — The "A-frame” mast, lightweight and to keep the mast from bowing. If the mast has any
easily constructed and erected. tendency to whip in the wind, or to bow under the
Construction of Wire Antennas 275

stress of the antenna, additional guys should be

added at the obvious points.

MAST GUYING
Three guy wires in each set will usually be
adequate for a mast. These should be spaced
equally around the mast. The number of sets of Fig. 13-26 — Stranded guy wire should be attached
guys will depend on the height of the mast, its to strain insulators by means of standard cable
natural sturdiness, and the required antenna ten¬ clamps to fit the size of wire used.
sion. A 30-foot mast will usually require two sets
of guys, while a 50-foot mast will need at least To avoid electrical resonances which might
three sets. One guy of the top set should be run in
cause distortion of the normal radiation pattern of
a direction directly opposite to the direction in
the antenna, it is advisable to break each guy into
which the antenna will run, the other two being
sections of 19 to 20 feet by the insertion of strain
spaced 120 degrees with respect to the first, as
insulators (see Figs. 13-25 and 13-26).
shown in Fig. 13-23.
The general rule is that the top guys should be Guy Material
anchored at distances from the base of the mast
equal to not less than 60% of the height of the Within their stress ratings, any of the halyard
mast. At the 60% distance, the stress on the guy materials listed in Table 13-11 may be used for the
wire opposite the antenna will be approximately construction of guys. The nonmetalic materials
twice the tension on the antenna. As the distance have the advantage that they do not have to be
between the guy anchorage and the base of the broken up into sections to avoid resonances, but all
mast is decreased, the tension on the rear guy in of these materials are subject to stretching, which
proportion to the tension on the antenna rises may cause mechanical problems in permanent
rapidly, the extra tension resulting in additional installations. At rated working load tension, dry
compression on the mast, which increases the manila rope stretches about 5 percent, while nylon
tendency for the mast to buckle. rope stretches about 20 percent.
The antenna wire listed in Table 13-1 is also
suitable for guys, particularly the copper-clad steel
types. Solid galvanized steel wire is also used
widely for making guys. This wire has approxi¬
mately twice the tension ratings of similar sizes of
copper-clad wire, but it is more susceptible to
corrosion. Stranded galvanized wire sold for guying
TV masts is also suitable for light-duty applica¬
tions, but is susceptible to corrosion.

Guy Anchors
Figs. 13-27 and 13-28 show two different styles
of guy anchors. In Fig. 13-27, one or more pipes
are driven into the earth at right angles to the guy
Fig. 13-25 — Simple lever for twisting solid guy wire. If a single pipe proves to be inadequate,
wires in attaching strain insulators. another pipe can be added in tandem, as shown.
Steel fence posts may be used in the same manner.
Fig. 13-28 shows a “dead-man” type of anchor.
The function of additional sets of guys is to
The buried anchor may consist of one or more
correct for any tendency that the mast may have
pipes 5 or 6 feet long, or scrap automobile parts,
to bow or buckle under the compression imposed
such as bumpers or wheels. The anchors should be
by the top guys. To avoid possible mechanical
buried 3 or 4 feet in the ground. Some tower
resonance in the mast that might cause the mast to
manufacturers make heavy auger-type anchors that
have a tendency to vibrate, the sets of guys should
screw into the earth. These anchors are usually
not be spaced equally on the mast. A second set of
guys should be placed at approximately 60% of the
distance between the ground and the top of the
mast, while a third set should be placed at about
60% of the distance between the ground and the
second set
The additional set of guys should be anchored
at distances from the base of the mast equal to not
less than 60% of the distance between ground and
the points of attachment on the mast. In practice,
the same anchors are usually used for all sets of
guys, which means that the latter requirement is Fig. 13-27 — Driven guy anchors. One pipe will
met automatically if the top set has been anchored usually be sufficient for a small mast. For added
at the correct distance. strength, a second pipe may be added, as shown.
 276 Chapter 13

under side of the mast with a scissors rest (Fig.

13-29), should a pause in the hoisting become
necessary. The rest may also be used to assist in the
’TT raising, if a man is used on each leg.
As the mast nears the vertical position, those
holding the guy wires should be ready to make the
Fig. 13-28 — Buried "dead-man" type of guy guys fast temporarily to prevent the mast from
anchor (see text). falling in one direction or another. The guys can
then be adjusted, one at a time, until the mast is
perfectly straight.
heavier than required for guying a mast, although For a mast over 30 feet, a “gin” of some form
they may be more convenient to install. Trees and may be required, as shown in Fig. 13-29. Several
buildings may also be used as guy anchorages if turns of rope are wound around a point on the
they are located appropriately. Care should be mast above center. The ends of the rope are then
exercised, however, to make sure that the tree is of brought together and passed over the limb of a
adequate size, or that the fastening to a building tree. The rope should be pulled as the mast is
can be made sufficiently secure. walked up to keep the mast from bending at the
center. If a tree is not available, a post, such as a 2
Guy Adjustment X 4, temporarily erected and guyed, can be used.
After the mast has been erected, the assisting rope
Most troubles that are encountered in mast can be removed by walking one end around the
guying are a result of pulling the guy wires too tight. mast (inside the guy wires).
Guy-wire tension should never be more than is
necessary to correct for obvious bowing or move¬ Other Supports
ment under wind pressure. In most cases, the
tension needed will not require the use of turn¬ Much sturdier supports are telephone poles and
buckles, with the possible exception of the guy towers. These types of supports are discussed later
opposite the antenna. If any great difficulty is in reference to rotatable antennas. Such supports
experienced in eliminating bowing from the mast, may require no guying, but they are not often used
the antenna tension should be reduced. solely for the support of wire antennas because of
their relative high cost. However, for antenna
To tree or HaJuards heights in excess of 50 feet, they are usually the
gm pole _ most practical form of support.

Rear gay
(Use to check
fallpast vertical)

Fig. 13-29 — Pulling on a gin line fastened slightly

a orio above the center point of the mast and on the
turns Side guys
OJsdto halyards will assist materially in erecting a tall
control mast. The tensions should be such as to keep the
side swag) mast in as straight a line as possible. The scissors
Scissors may be used to push on the under side and to serve
as a rest in case a pause in raising becomes
necessary.
Base securely

Bibliography
ERECTING A MAST Source material and more extended discussion
The erection of a mast of 30 feet or less can of topics covered in this chapter can be found in
usually be done by simply “walking” the mast up the references given below.
after blocking the bottom end securely so that it Elengo, “Predicting Sag in Long Wire Antennas,”
can neither slip along the ground or upend when QST, Jan., 1966.
the mast is raised. A man should be stationed at Gann, “A Center-Fed 'Zepp* for 80 and 40,” QST,
each guy wire, and in the last stages of raising, May, 1966.
some assistance may be desirable by pulling on the Gordon, “Invisible Antennas,” QST, Nov. 1965.
Gue, “An 80-Meter Inverted Vee,” QST, June,
proper guy wire. Further assistance may be gained 1968.
by using the halyards in the same manner. As the McCoy, “An Easy-to-Make Coax-Fed Multiband
mast is raised, it may be helpful to follow the Trap Dipole,” ¿ST Dec., 1964.
 Chapter 14

Rotatable Antennas

Constructing an antenna which is to be rotated smaller diameter booms, the elements sometimes
requires materials which are strong, lightweight, work lose and tend to elongate the pinning holes in
and easy to obtain. Procurement is often the most both the element and the boom. After some time,
difficult portion of the project, but that can the elements shift their positions slightly (some¬
usually be overcome with some careful searching of times from day to day!) and give a rather ragged
the Yellow Page section of the telephone book for appearance to the system, even though this doesn’t
the nearest large metropolitan area. (Such tele¬ generally harm the electrical performance.
phone books may be available in the reference A three-inch diameter boom with a wall thick¬
section of your local library.) ness of .065 inch is very satisfactory for antennas
The materials required to build a suitable up in size to about a five-element 20-meter array
rotatable antenna will vary, depending on many which is spaced on a 40-foot long boom. A truss is
factors. Perhaps the most important factor which recommended for any boom longer than 24 feet.
determines the type of hardware needed is the
weather conditions which are normally encoun¬ CONSTRUCTION OF QUADS
tered. High winds usually don’t cause as much
Most of the constructional details relating to
damage to an antenna as does ice or even heavy
quads have been given in Chapter Nine. The
ice along with high winds. Aluminum sizes should
sturdiness of a quad is directly proportional to the
be selected so that the various sections of tubing quality of the material used and the care with
will telescope to provide the necessary total length. which it is constructed.
Table 14-1 gives the specifications for aluminum The size and type of wire selected for use with
which will meet the needs for most amateur a quad antenna is important because it will
installations. determine the capability of the spreaders to with¬
The boom size for a rotatable Yagi or quad stand high winds and ice. One of the more
should be selected to provide stability to the entire common problems confronting the quad owner is
system. The best diameter for the boom depends that of broken wires. A solid conductor is more apt
on several factors, but mostly the element weight, to break than stranded wire under constant flexing
number of elements, and the overall length. Tubing conditions. For this reason, copper stranded wire is
diameters of 1-1/4 inches can easily support recommended. For 20-, 15-, or 10-meter operation,
three-element 10-meter arrays and perhaps a two- wire size No. 14 or 12 is a good choice. Soldering
element 15-meter system. For larger 10-meter of the stranded wire at points where flexing is
antennas or for harsh weather conditions, and for likely to occur should be avoided.
antennas up to three elements on 20 meters or four
elements on 15 meters, a two-inch diameter boom
will be adequate. It is not recommended that
two-inch diameter booms be made any longer than
24 feet unless additional support is given to reduce
both vertical and horizontal bending forces. Suit¬
able reinforcement for a long two-inch boom can
consist of a truss or a truss and lateral support, as
shown in Fig. 14-1.
A boom length of 24 feet is about the point
where a 3-inch diameter begins to be very worth¬
while. This dimension provides a considerable
amount of improvement in overall mechanical
stability as well as increased clamping surface area
for element hardware. The latter is extremely
important if heavy icing is commonplace and Fig. 14-1 — A long boom needs both vertical and
rotation of elements around the boom is to be horizontal support. The cross bar mounted above
avoided. Pinning an element to the boom with a the boom can support a double truss which will
large bolt helps in this regard. However, on the help keep the antenna in position.

277
 278 Chapter 14

TABLE 14-1 - STANDARD SIZES OF ALUMINUM TUBING

6O61-T6 (61S-T6) ROUND ALUMINUM TUBE

In 12-Foot lengths

O.O. WALL THICKNESS 1. 0. APPROX. WEIGHT 0. D. WALL THICKNESS 1. D. APPROX. WEIGHT

Inches Inches Stubs Ga. Inches Per Foot Per Length Inches Inches Stubs Ga. Inches Per Foot Per Length
.035 (No. 20) .117 .019 lbs. .228 lbs. r .083 (No. 14) .834 .281 lbs. 3.372 lbs.
.049 (No. 18) .089 .025 lbs. .330 lbs.
U/s* .035 (No. 20) 1.055 .139 lbs. 1.668 lbs.
%* .035 (No. 20) .180 .027 lbs. .324 lbs. .058 (No. 17) 1.009 .228 lbs. 2.736 lbs.
.049 (No. 18) .152 .036 lbs. .432 lbs.
.058 (No. 17) .134 .041 lbs. .492 lbs. 1%* .035 (No. 20) 1.180 .155 lbs. 1.860 lbs.
.049 (No. 18) 1.152 .210 lbs. 2.520 lbs.
54’ .035 (No. 20) .242 .036 lbs. .432 lbs. .058 (No. 17) 1.134 .256 lbs. 3.072 lbs.
.049 (No. 18) .214 .047 lbs. .564 lbs. .065 (No. 16) 1.120 .284 lbs. 3.408 lbs.
.058 (No. 17) .196 .055 lbs. .660 lbs. .083 (No. 14) 1.084 .357 lbs. 4.284 lbs.
%' .035 (No. 20) .305 .043 lbs. .516 lbs. 1H* .035 (No. 20) 1.305 .173 lbs. 2.076 lbs.
.049 (No. 18) .277 .060 lbs. .720 lbs. .058 (No. 17) 1.259 .282 lbs. 3.384 lbs.
.058 (No. 17) .259 .068 lbs. .816 lbs.
.065 (No. 16) .245 .074 lbs. .888 lbs. 1'/1* .035 (No. 20) 1.430 .180 lbs. 2.160 lbs.
.049 (No. 18) 1.402 .260 lbs. 3.120 lbs.
Vis' .035 (No. 20) .367 .051 lbs. .612 lbs. .058 (No. 17) 1.384 .309 lbs. 3.708 lbs.
.049 (No. 18) .339 .070 lbs. .840 lbs. .065 (No. 16) 1.370 .344 lbs. 4.128 lbs.
.065 (No. 16) .307 .089 lbs. 1.068 lbs. .083 (No. 14) 1.334 .434 lbs. 5.208 lbs.
'/ï* .028 (No. 22) .444 .049 lbs. .588 lbs. ‘.125 '/s' 1.250 .630 lbs. 7.416 lbs.
.035 (No. 20) .430 .059 lbs. .708 lbs. *.250 'A' 1.000 1.150 lbs. 14.832 lbs.
.049 (No. 18) .402 .082 lbs. .984 lbs. 1H* .035 (No. 20) 1.555 .206 lbs. 2.472 lbs.
.058 (No. 17) .384 .095 lbs. 1.040 lbs. .058 (No. 17) 1.509 .336 lbs. 4.032 lbs.
.065 (No. 16) .370 .107 lbs. 1.284 lbs.
1%' .058 (No. 17) 1.634 .363 lbs. 4.356 lbs.
Vs* .028 (No. 22) .569 .061 lbs. .732 lbs. .083 (No. 14) 1.584 .510 lbs. 6.120 lbs.
.035 (No. 20) .555 .075 lbs. .900 lbs.
.049 (No. 18) .527 .106 lbs. 1.272 lbs. 17s' .058 (No. 17) 1.759 .389 lbs. 4.668 lbs.
.058 (No. 17) .509 .121 lbs. 1.452 lbs. 2' .049 (No. 18) 1.902 .350 lbs. 4.200 lbs.
.065 (No. 16) .495 .137 lbs. 1.644 lbs. .065 (No. 16) 1.870 .450 lbs. 5.400 lbs.
%’ .035 (No. 20) .680 .091 lbs. 1.092 lbs. .083 (No. 14) 1.834 .590 lbs. 7.080 lbs.
.049 (No. 18) .652 .125 lbs. 1.500 lbs. *.125 '/s' 1.750 .870 lbs. 9.960 lbs.
.058 (No. 17) .634 .148 lbs. 1.776 lbs. *.250 'A" 1.500 1.620 lbs. 19.920 lbs.
.065 (No. 16) .620 .160 lbs. 1.920 lbs. 2%' .049 (No. 18) 2.152 .398 lbs. 4.776 lbs.
.083 (No. 14) .584 .204 lbs. 2.448 lbs. .065 (No. 16) 2.120 .520 lbs. 6.240 lbs.
7s* .035 (No. 20) .805 .108 lbs. 1.308 lbs. .083 (No. 14) 2.084 .660 lbs. 7.920 lbs.
.049 (No. 18) .777 .151 lbs. 1.810 lbs. 27z* .065 (No. 16) 2.370 .587 lbs. 7.044 lbs.
.058 (No. 17) .759 .175 lbs. 2.100 lbs. .083 (No. 14) 2.334 .740 lbs. 8.880 lbs.
.065 (No. 16) .745 .199 lbs. 2.399 lbs. *.125 '/s' 2.250 1.100 lbs. 12.720 lbs.
1' .035 (No. 20) .930 .123 lbs. 1.476 lbs. *.250 'A" 2.000 2.080 lbs. 25.440 lbs.
.049 (No. 18) .902 .170 lbs. 2.040 lbs. 3' .065 (No. 16) 2.870 .710 lbs. 8.520 lbs.
.058 (No. 17) .884 .202 lbs. 2.424 lbs. *.125 '/s' 2.700 1.330 lbs. 15.600 lbs.
.065 (No. 16) .870 .220 lbs. 2.640 lbs. *.250 'A" 2.500 2.540 lbs. 31.200 lbs.

These sizes are extruded; all other sizes are drawn tubes. Shown here are standard sizes of aluminum
tubing that are stocked by most aluminum suppliers or distributors In the United States and Canada. Note
that all tubing comes in 12-foot lengths and also that any diameter tubing will fit into the next larger size
if the larger size has a 058-mch wall thickness. For example, 5/8-inch tubing has an outside diameter of
0.62 5 inches and will fit into 3/4-inch tubing with a .058-inch wall which has an inside diameter of 0 634
,
IS iPSt r '
9h ‘ f°r a slip fU or ,or slott '
n9 ‘he tubing and then using hose
■ T 6p °I’
T6 ‘VP® of aluminum is of relatively high strength and has good workability, plus being
tant tO corrosion and wl11 bend without taking a "set.” Check the Yellow Pages for aluminum
uca ier s.

Connecting the wires to the spreader arms may Dimensions for quad elements and spacing have
be accomplished in many ways. The simplest been given in texts and QST over the years. It is
method is to drill holes through the fiber glass at generally felt that quads are not very critical in
the appropriate points on the arms and route the their tuning, nor is element spacing very critical.
wires through the holes. Soldering a wire loop Table 14-11 is a collection of dimensions that will
across the spreader, as shown in Chapter Nine, is suit almost every amateur need for a quad system.
recommended. However, care should be taken to A boom diameter of two inches is recom¬
prevent solder from flowing to the corner point mended for systems having two or three elements
where flexing could break it. for 20, 15, and 10 meters. When the boom length
Rotatable Antennas 279

TABLE 14-11 - QUAD DIMENSIONS

2 element Quad (W0HTH)

Spacing (given below)
Boom length (given below)
Band SO M. 15 M. 10 M.
Reflector 144' 11^" 72' 4" 48' 8" 35' 7"
Driven Element 140' 11J4" 70' 2" 47' 4" 34' 7"
Spacing 30' 13' 10' 6' 6"
Boom length 30' 13' 10' 6' 6"
Feed method Directly with 23' Directly with 11' Directly with 7' Directly with 5'
of RG11, then any 7" RG11, then any 8J^" RG11, then any 8" RG11, then any
length of RG8 coas length RG8 coax. length RG8 coax. length RG8 coax.
(Note that a spider or boomless quad arrangement could be used for the 10/15/20 meter parts of the above di¬
mensions yielding a triband antenna)

4 element Quad* (W0AIW (20 M.) /W0HTH** /K0KKU/K0EZH/W6FXB)

Spacing: equal; 10 ft.
Boom length: 30 ft.
Band 10 M.
Phone CW
Reflector 72' 1 Ji" 72' 5" 48' 8" 35' 8J^"
Driven Element 70' 1 Ji" 70' 5" 47' 4" 34' 8H"
Director 1 69' 1" 69' 1" 46' 4" 33' 7Ji"
Director 2 69' 1" 69' 1" 46' 4" 33' 7Ji"
Feed Method Directly with Directly with Directly with
50-ohm coax. 50-ohm coax. 5' 9" RG11, then
any length RG8
coax.
* Common boom used to form a triband array.
** The 2-element 40-meter quad given above is added to form a four-band quad array.

4 element Quad (W0HTH/K8DYZ*/K8YIB*/W7EPA*)

Spacing: equal; 13' 4".
Boom length: 40 ft.
Band 20 M. 15 M. 10 M.
Reflector 72' 5" 48' 4" 35' 8Ji"
Driven Element 70' 5" 47' 0" 34' 8Ji"*
Director 1 69' 1" 46' 1" (Directors 1-3 all
Director 2 69' 1" 46' 1" 33' 7")*
Feed method Directly with Directly with 7' Directly with
50-ohm coax. 9" RG11, then any 50-ohm coax.
length 50-ohm coax.
* For the 10-meter band the driven element is placed between the 20/15 reflector and 20/15 driven element. The
10-meter reflector is placed on the same frame as the 20/15-meter reflectors and the remaining 10-meter directors
are placed on the remaining 20/15-meter frames The 10-meter portion is then a 5-element quad.

6 element Quad (W0YDM, W7UMJ)

Spacing: equal; 12 ft.
Boom length: 60 ft.
Band: 20 M.
Reflector 72' 1JÍ"
Driven Element 70' 1JÍ"
Directors 1, 2 and 3 69' 1"
Director 4 69' 4"
Feed Method Directly with 50-ohm coax.

becomes 20 feet or longer, as encountered with boom are more difficult to relieve, and the larger
four- and five-element antennas, a three-inch dia¬ size of three-inch-diameter tubing is desirable.
meter boom is highly recommended. Wind creates There are, generally speaking, three grades of
two forces on the boom, vertical and horizontal. materials which can be used for quad spreaders.
The vertical load on the boom can be reduced with The least expensive material is bamboo. Bamboo
a guy-wire truss cable. The horizontal forces on the poles are available from many rug stores and cost
280 Chapter 14

brake will prevent gear damage on windy days.

Tliis group of rotors costs from 100 to 250 dollars.
If adequate precautions are taken, this group of
rotors is capable of holding and turning stacked
monoband arrays, or up to a five-element 20-meter
system. The next step up in strength is more
expensive. This class of rotator will turn just about
anything the most demanding amateur might want
to install, and typically costs more than 350
dollars.
A description of antenna rotors would not be
complete without the mention of the prop-pitch
class. The prop-pitch rotor system consists of a
surplus aircraft propeller-blade pitch motor coup¬
led to an indicator system as well as a power
supply. There are mechanical problems of install¬
ation, however. It has been said that a prop-pitch
rotor system, properly installed, is capable of
turning a house. This is no doubt true! Perhaps in
the same class as the prop-pitch motor but with
somewhat less capability is the electric motor of
the type used for opening garage doors. These have
been used successfully in turning quite large arrays.
Proper installation of the antenna rotor can
provide many years of trouble-free service; sloppy
Fig. 14-2 — The Ham-M rotor control. installation can cause problems such as a burned-
out motor, slippage, binding, and casting breakage.
less than a dollar each. Bamboo, however, is also Most rotors are capable of accepting mast sizes of
the weakest material normally used for quad different diameters, and suitable precautions must
construction. It has a short life, typically only a be taken to shim an undersized mast to assure
few years, and will not withstand a harsh climate dead-center rotation. It is very desirable to mount
very well. Additionally, bamboo is heavy in con¬ the rotor inside and as far below the top of the
trast to fiber glass, which weighs only about a tower as possible. The mast will absorb the torsion
pound per 13-foot length. Fiber glass is the most developed by the antenna during high winds, as
popular type of spreader material, and will with¬ well as during starting and stopping, A mast length
stand normal winter climates. One step beyond the of 10 feet or more between the rotor and the
conventional fiber-glass arm is the pole-vaulting antenna will add greatly to the longevity of the
arm. For quads designed to be used on 40-meters, entire system. Some amateurs have used a long
surplus “rejected” pole-vaulting poles are highly mast from the top to the base of the tower.
recommended. Their ability to withstand large Installation and service can be accomplished at
amounts of bending is very desirable. The cost of ground level. Another benefit of mounting the
these poles is high, and they are difficult to obtain. rotor ten or more feet below the antenna is that
Those interested should check with sporting goods any misalignment between the rotor, mast, and the
manufacturers. top of the tower is less significant. A tube at the
top of the tower through which the mast protrudes
ROTATING SYSTEMS will almost completely eliminate any lateral forces
on the rotor casting. All the rotor need do is
There are not many choices when it comes to support the downward weight of the antenna
antenna rotators or rotors for the amateur antenna system and turn the array. While the normal weight
system. However, making the correct decision as to of the antenna and the mast is usually not more
how big the rotor needs to be is very important if than a couple of hundred pounds, even with a large
trouble-free operation is desired! There are basi¬ system, one can ease this strain on the rotor by
cally four grades of rotors available to the amateur. installing a thrust bearing at the top of the tower.
The lightest duty rotor is the TV type, typically The bearing is then the component which holds all
used to turn TV antennas. The cost may range of the weight, and the rotor need perform only the
from 20 to about 50 dollars. These rotors will rotating task.
handle, without much difficulty, a small three- A problem often encountered in the amateur
element tribander (20-, 15-, and 10-meter) array or installation is that of misalignment between the
a single 15- or 10-meter monoband three-element direction indication of the rotor control box and
antenna. The important consideration with a TV the heading of the antenna. This is caused by
rotor is that it lacks braking or holding capability. mechanical slippage in the system due to loose
High winds will turn the rotor motor via the gear rotor bolts or antenna boom-to-mast hardware.
train in a reverse fashion. Sometimes broken gears Many texts suggest that the boom be pinned to the
result. The next grade up from the TV class of mast with a heavy-duty bolt and, likewise, the
rotors usually includes a braking arrangement rotor be pinned to the mast. There is a trade-off
whereby the antenna is held in place when power is here. The amateur might not like to climb his
not applied to the rotor. Generally speaking, the tower and straighten out the assembly after each
Rotatable Antennas 281
monitoring of the antenna heading.) S2 in Fig. 14-3
is incorporated in the circuit for turning off the
control box. Contacts 6 and 7 of SI are used to
energize T2. This function immediately activates
the brake-release mechanism. At the same time, a
third set of switch contacts (1, 2, and 3) deter¬
mines the direction of rotation. When the antenna
reaches a certain heading the operator returns the
switch to center position. It is at this point when
the solenoid releases the brake wedge so that it can
drop back into place on the ring gear, abruptly
halting the movement of the antenna and mast.
The possibility of rotator damage can be greatly
reduced with this modification. It will hold the
brake open while the antenna coasts to a stop.
The shaded lines in Fig. 14-3 show part of the
original Ham-M circuit. The dark lines represent
the additional connections needed to perform the
brake-delay functions. When SI is activated KIA
Fig. 14-3 — Diagram of the modified Ham-M immediately closes, allowing Cl to charge through
rotor-control box. Fixed-value resistors are 1-watt R2. K1C is used to apply voltage to T2 which, in
composition. Connections must be broken at the
two places marked "X." The shaded lines indicate turn, opens the brake. Depending on which way
original wiring and the heavy lines are additions to the lever is pushed, the antenna turns either left or
the original circuit. Parts designations not listed right. When SI is released, voltage to the rotator is
below are for text reference. interrupted by SIC, but KI remains energized,
C1 — 200-pF, 450-volt electrolytic. keeping the brake open until Cl discharges through
CR1 — 1000-volt PRV, 750-mA silicon rectifier. R2, R3, and RI. The time required for the voltage
DS1, DS2 — Neon panel lamp, 117 V ac. from Cl to drop to a point where KI deenergizes is
K1 — Dpdt, 3-A contacts, 5000-ohm coil.
determined by the setting of RI. The range is from
R1 — 100,000-ohm linear taper composition. 2 to 8 seconds. Two neon lamps, DS 1 and DS2, are
used to provide visual indication of the brake
position.
heavy wind storm. However, if there is sufficient
wind to cause slippage in the couplings, perhaps Construction
the wind could break a rotor casting. The slippage No special wiring precautions are necessary.
will act as a clutch release which may prevent The control circuitry added to the rotator is
serious rotor damage. completely contained in the homemade chassis

DELAYED-ACTION BRAKING FOR

THE HAM-M ROTOR

On most rotors equipped with braking capa¬

bilities, the brake is applied almost instantly after
power is removed from the rotor motor, to stop
the array from rotating and hold it at a chosen
bearing. Because of inertia, however, the array
itself does not stop rotating instantly. The larger
and heavier the antenna, the more it will tend to
continue its travel, in which case the mast may
absorb the torsion, the entire tower may twist back
and forth, or the brake of the rotor may shear. A
more suitable system is to remove power from the
rotor motor during rotation before the desired
bearing is reached, allowing the beam to coast to a
slower speed or to a complete stop before the
brake is applied. The Ham-M rotor system may be
modified to have delayed-action braking by adding
external controls and components. A convenient Fig. 14-4 — This top view of the homemade chassis
method of installation is shown in Fig. 14-2. shows the location of the components. A bottom
The Ham-M control box has one switch that plate is permanently attached to this unit using
handles all of the switching functions, SI of Fig. 6-32 screws which also hold the rubber feet in
14-3. Contacts 4, 7, and 8 apply voltage to place. The four short interconnecting leads run
through rubber grommets on the rear of the
transformer Tl, which supplies voltage to the chassis. The wires should be secured inside the
indicator meter. (A modification listed in the cabinet to prevent them from being pulled loose.
Ham-M instruction book mentions how to change The large electrolytic capacitor is mounted to the
the switching arrangement to allow continúous bottom plate using two small terminal strips.
282 Chapter 14
box. Four short pieces of line cord interconnect by the manufacturer. If a component fails, the
the two boxes. rotor must be removed from the tower for repairs.
The chassis shown in the photograph is 6 x 6 x For this reason, two extra conductors were added
2 inches. It was designed to allow the Ham-M to the existing control cable and the brake power
control box to sit on top of it. A Bud AC-1413 components were placed in the control box.
aluminum chassis could be used if the builder The modification of the rotor is best done in
doesn’t want to construct his own. Fig. 14-4 shows the workshop. These modifications can be per¬
the internal layout. A bottom cover should be used formed while standing on the tower with the rotor
on the chassis to prevent the operator from mounted in position, but it is more difficult.
accidentally contacting ac voltage. The first step is to remove the two diodes, the
electrolytic capacitor, and the 10-ohm resistor
Operation from the rotor motor housing. The terminal strip is
Front-panel control of the delay is desirable. set up for five screw positions. Originally the two
Different time periods are required depending on diodes were connected to a sixth terminal which
prevailing winds and the size of the antenna. The will be used for brake control with the modified
operator soon gets a feel for when to let go of the system. A seventh wire will be required for the
switch to have the antenna stop at the right place. indicator system. Connect this wire to the unused
On extremely windy days the delay time should be terminal of the potentiometer and route it out
short to keep the array from windmilling. But on through the rubber bushing. It then may be spliced
reasonably calm days, the antenna will come to a into the control cable which has been modified for
full stop in less than 5 seconds; therefore the delay seven conductors. The two new wires do not
time should be set to near maximum. The antenna handle much current, and therefore need not be
us"illy drifts less than 10 degrees. greater than No. 18 for runs up to 250 feet. This
completes the modification of the rotor motor
assembly.
A DIRECTIONAL INDICATOR FOR THE
The construction of a control box is the most
HY-GAIN MODEL 400 ROTOR difficult portion of the project. For the model
The Hy-Gain rotor system is a heavy-duty shown in Figs. 14-5 and 14-7, a Minibox, sized to
rotational device which can be used to turn large house the meter and control knobs, was used. The
arrays. The control box is designed to have a overall size of the housing will be determined by
direction dialed on the front-panel compass face, the size of the meter frame. For calibration
and when power is applied to the unit the rotor purposes, a large meter-face area is desirable.
will turn to the designated heading and then shut The Hy-Gain 400 rotor is designed to have
off automatically. While the automatic feature is about 30 degrees of rotational overlap. Unfor¬
handy for casual operating, the serious DXer or tunately there are no end-of-rotation limit switches
contester might like to have an indication of the provided, and it is possible to turn this device
antenna (or rotor) position during the actual several times around in the same direction. Need¬
turning period, which the conventional control less to say, the coaxial feed line will not withstand
does not provide. The control system described this kind of treatment. In order to keep the control
below was designed to be used in place of the circuitry simple, it was decided to use two
Hy-Gain control unit, and with the exception of switches, one for turning on power to the system,
the motor starting capacitor, no parts from the and one for determining the direction of rotation.
original control are used. One additional feature is The ROTATE switch needs to be spring-loaded for
provided: The rectifying components which power the OFF position. The purpose is to assure that the
the rotor brake are located in the rotor as provided operator keeps his mind on what he is doing while
the antenna is rotating. The spring loading acts as a
dead-man switch - one can’t go away and leave it
turned on.
The circuit for the control box is given in Fig.
14-6. A conventional low-current (less than an
ampere at 6.3 V) transformer is used to provide a
regulated 8.2 volts de to drive the indicator circuit.
The value of the voltage is not critical, and any
voltage from about six to as high as ten will suffice.
Regulation is desirable to assure the calibration
won’t change under changing line-voltage con¬
ditions. RI is a linear-taper two-watt composition
control. The meter was obtained from Simpson
Electric and is their model 524. The internal
resistance is 43 ohms.
Control of the brake at the rotor is provided by
120 volts of de developed by CR2 or CR3 in the
control box. The voltage appears at terminal six
each time voltage is applied to the rotor motor. A
small pilot lamp is used to assure the operator that
Fig. 14-5 — Front view g.f the Hy-Gain control the rotor motor isn’t working against a closed
box. brake.
Rotatable Antennas 283

Fig. 14-6 — Circuit diagram

for the Hy-Gain rotor control
box. The 6.8-pF, 240-V ac
capacitor is taken from the
original Hy-Gain control head.
CR1, CR2 and CR3 are con¬
ventional power diodes, 1000
PRV at 1 A. T1 is rated for
one ampere. R1 is a linear-
taper composition two-watt
control. The indicator meter
shown in the photographs is a
model 524, manufactured by
Simpson Electric. S2 is a
spring-return type.

The potentiometer in the rotor housing is gear periods (which can also be accomplished by in¬
driven to allow the antenna to turn approximately creasing transmitter power), it has the same benefi¬
380 degrees while the resistor turns through only cial effect on received signals. Therefore, a tall
about 270 degrees. The easiest way to calibrate the support for an antenna which has gain, and which
overall system is to turn the rotor motor until the can be rotated, is very desirable. This is especially
potentiometer is at midposition of its travel. This true if DX contacts are of prime interest on the 20-
can be determined with an ohmmeter by observing 15-and 10-meter bands.
the resistance between rotor wires t^vo and seven. Of the two important features of an antenna
Since the value of the potentiometer is about 5000 system, height and antenna gain, height is usually
ohms, an ohmmeter reading of 2500 ohms will considered the most important, if the antenna is
indicate midposition. Next, the rotor should be horizontally polarized. The typical amateur instal¬
installed and the antenna positioned so that its lation consists of a three-element triband beam
heading is correct for the predetermined center (tribander) for 20, 15 and 10 meters mounted on a
heading for the indicator. Then the antenna should tower which may be as low as 25 or 30 feet, or as
be rotated 180 degrees as noted by visual in¬
spection of the antenna. Do not use the control
box for the indication. After the antenna has been
rotated 180 degrees, the indicator may be labeled
for the correct position. Rotate the antenna 360
degrees in the opposite direction, again observing
the antenna (not the control box). When the
antenna is in position the proper heading may be
marked on the meter face. With the system shown
in the photographs, the center was set for north.
South appears on the meter at 15 and 85 percent
of full scale deflection.
Operation of the new control is similar to other
rotor systems. The operator should first select the
direction he wants to turn the antenna, then select
the ROTATE position of S2. Since the indicator
provides continuous indication of the antenna
heading, one can observe the heading even when
the antenna is not rotating. Selecting the OFF
position of SI completely disables the rotor
control box.

TOWER SELECTION AND INSTALLATION

Probably the most important part of any
amateur radio installation is the antenna system. It
determines the effectiveness of the signal transmit¬
ted at a particular power level. In terms of dollar
investment, the antenna provides double duty; Fig. 14-7 — Inside view of the Hy-Gain control
while it can provide gain during transmitting box.
284 Chapter 14

4) Provisions must be made to keep neighbor¬

hood children from climbing the support. Chicken
wire around the tower base will serve this need
nicely.
5) Local ordinances should be checked to
determine if any legal restrictions are on record.
Other important points are:
6) The total dollar value to be invested.
7) The size and weight of the antenna desired.
8) The overall yearly climate.
9) Ability of the owner to climb a fixed tower.
The selection of a tower support is usually
dictated more by circumstances than by desire.
The most economical system, in terms of feet-per-
dollar investment, is a guyed tower. Rohn type 25
tower is suitable to any reasonable height if guyed
properly. Details of installation are given later. The
crank-up style of tower is very handy if the
operator doesn’t wish to climb any more than 20
feet to work on his antennas. Crank-up towers
come in both guyed and free-standing configura¬
Fig. 14-8 — A gin pole is a mechanical device which
can be clamped to a tower leg to aid in the tions. The trade-off with a crank-up tower is that it
assembly of sections as well as the installation of doesn’t completely eliminate tower climbing as a
antennas. The aluminum tubing extends through necessity, and they are, to an extent, dangerous if
the clamp and may be slipped into position before mishandled or operated carelessly. The free¬
the tubing clamp is tightened. A rope should be standing fixed tower will suit many situations if
routed through the tubing and over the pulley the owner’s backyard doesn’t have sufficient space
mounted at the top. See Fig. 14-9 for installation to allow proper guying of a lighter weight system.
details.
Another type of arrangement which is very popular
is the crank-down, fold-over support. These
high as 65 or 70 feet. Some exotic systems use
large antennas on much taller towers.
The selection of a tower, it’s height, and the
type of antenna and rotor to be used all may seem
like a complicated matter, particularly for the
newcomer to amateur radio. These four aspects of
an antenna system are completely interrelated, and
one should consider the overall system before
making any decisions as to a specific component.
Perhaps the most important consideration for
many amateurs is the effect of the antenna system
on the surrounding environment. If plenty of space
is available for a tower installation and there is
little chance of the antenna irritating neighbors,
the amateur is indeed fortunate. This amateur’s
limitations will be mostly financial. But for most
amateurs, the size of the backyard, the effect of
the system on members of the family and neigh¬
bors, local ordinances, and the proximity of power
lines and poles contribute a considerable influence
on the overall selection of antenna components.
The amateur must consider the practical limita¬
tions for his installation. Some points for consider¬
ation are given below:
1) A tower should never be installed in a
position whereby it could conceivably fall onto a
neighbor’s property.
2) The antenna must be located in such a
position that it cannot possibly tangle with power
lines during normal operation (!) or if a disastrous
wind storm comes along. Fig. 14-9 — The assembly of tower sections is made
simple when agin pole is used to lift each one into
3) Sufficient yard space must be available to position. Note that the safety belts of both
position a guyed tower properly. The guy anchors climbers are fastened below the pole thereby
should be between 60 and 80 percent of the tower preventing the strap from slipping over the top
height in distance from the base of the tower. section. (Photo by K1THQ)
Rotatable Antennas 285

systems crank down to about 20 feet, and then

may be folded over to allow access from the
ground level. Strictly fold-over towers are also
available, which allow tower work to be accom¬
plished from the ground level, but do not allow the
tower to be cranked down during periods of
nonuse or during extremely bad weather.
Once a decision has been tentatively made, the
next step is to write to the manufacturer and
request a specification sheet. Meanwhile, one
should lay out any guy anchor points needed to
assure that they will fit on the assigned property.
The specification sheet for the tower should give a
wind-load capability, and an antenna can then be
selected which will not overload it. If a tentative
decision on the antenna type is made, a note to the
antenna manufacturer giving the complete set of
details for installation is not a bad idea. Be sure to
give complete details of your plans, including all Fig. 14-11 — Proper tension can be placed on the
specifications of the antenna system planned. guy wires with the aid of a block-and-tackle
system. (Photo by K1THQ)
Remember, the manufacturer will not custom
design a system directly for your needs, but he
may offer comments on any of your ideas. commit yourself. They may be able to offer a great
It is often very helpful to the novice tower deal of help. If someone in your area is planning to
installer to visit other local amateurs who have install a tower and antenna system, be sure to offer
installed towers. Look over their hardware and ask your assistance. There is no substitute for exper¬
questions. If possible, have a few local experienced ience when it comes to tower work, and your
amateurs look over your plans - before you experience there may prove invaluable to you later.
The installation of a tower is not difficult when
the proper techniques are known. A guyed tower,
in particular, is not hard to erect since each of the
individual sections are relatively light in weight and
can be handled with only a few helpers and some
good-quality rope. A gin pole is a handy device for
working with tower sections. The gin pole shown
in Fig. 14-8 is designed to fit around the leg of
Rohn No. 25 tower and clamp in place. The
tubing, which is about 12 feet long, has a pulley on
one end. A rope is routed through the tubing and
over the pulley. When the gin pole is attached to
the tower and the tubing is extended into place
and locked, the rope may be used to haul tower
sections and the antenna into place. Figs. 14-9 and
14-10 show the basic process.
When a guyed tower is used, guy anchor points
are needed. These anchor points come in two
styles, screw in and cement in. The cement anchor
types consist of a steel rod which is equipped with
a hook at the end. The hook is placed in a hole
which has been prepared in accordance with the
manufacturer’s instructions. Contrary to what
most inexperienced amateurs think, the anchors
are positioned in line with the guy cable, not at
right angles to it or up straight. (See Fig. 14-12.)
Unless many helpers are available, it will proba¬
bly be impossible to pull up the guy wires to
sufficient tension either by hand or by screwing in
the turnbuckles. It is best to use a come-along
(mechanical power hoist) or block and tackle, as
shown in Fig. 14-11. In a three-way guy anchor
system, it is important not to overtighten the first
guy wire. It is possible to bend the tower over to a
Fig. 14-10 — A gin pole is helpful in positioning point that when the other guys are tightened in
the antenna before the mounting bolts are tight¬ place, they are not able to straighten out the
ened to the mast. (Photo by K1THQ) tower.
286 Chapter 14

should be used by each person working on the

tower, such as is shown in Fig. 14-13. When
climbing the tower, if more than one person is
involved, one should climb into position before the
other begins climbing. The same procedure is
required for climbing down a tower after the job is
completed. The purpose is to have the nonclimbing
person stand relatively still so as not to drop any
tools or objects on the climbing person, or uninten¬
tionally obstruct his movements. When two per¬
sons are working on top of a tower, only one
should change position (unbelt and move) at a
time.
For most installations, a good-quality 1/2-inch-
diameter manilla hemp rope will be able to handle
adequately the work load for the hoisting tasks.
The rope must be petiodically inspected to assure
that no tearing or chafing has developed, and if the
rope should get wet from rain, it should be hung
Fig. 14-12 — A guy cable is used to assure the out to dry at the first opportunity. Safety knots
turnbuckles remain in place after they are tight¬ should be used to assure that none come loose
ened. This procedure is an absolute requirement during the hoisting of a tower section or antenna.
for a guyed tower system. (Photo by K1THQ)

After the guy wires are installed, it is necessary

that the turnbuckles be looped with an extra piece
of guy wire. The purpose is to prevent any twist in
the wire (especially when it is new) from unscrew¬
ing the turnbuckles. The wire loop should be
clamped in position as shown in Fig. 14-12.
As with any building project, the placement of
the base and first section of tower will determine
the straightness of the final product. To get the
first section true, a carpenters level of about three
or four feet in length is ideal. Spot checks as the
tower is assembled will aid in keeping the installa¬
tion from looking tilted. A simple expedient is to
stand away from the tower and sight along a string
which suspends a plumb bob, with the top end of
the string held high in the air at arm’s length.
The installation of other types of towers similar
to that shown in the photographs will use the same
basic techniques. The manufacturer’s instruction
sheet should be followed in every detail to assure
proper safety of the system.
Safety
One of the most important aspects of any
tower-installation project is the safety of all per¬
sons involved. The use of hard hats is highly
recommended for all assistants helping from the Fig. 14-13 — A good-quality safety belt is a
requirement for working on a tower. The belt
ground. Helpers should always stand clear of the should contain large steel loops for the strap snap.
tower base to prevent being hit by a dropped tool Leather loops at the rear of the belt are handy for
or hardware. A good grade of climber’s safety belt holding tools. (Photo by K1THQ)
 Chapter 15

Mobile Antennas

HF - MOBILE FUNDAMENTALS tions greatly affect operation on 160, 80, and 40

meters. From this standpoint, perhaps the opti¬
Few amateurs construct their own antennas for mum band for hf-mobile operation is 40 meters.
hf-mobile use since safety reasons dictate a very The popularity of the regional mobile nets on 7255
sound mechanical construction. However, most and 7258 kHz is perhaps the best indication of its
installations using commercially made components suitability. For local work, 10 meters is also useful
still have to be optimized for the particular since antenna efficiency is high and relatively
installation and type of operation desired. simple antennas without loading coils are possible.
The drawing of Fig. 15-1 shows a typical As the frequency of operation is lowered, an
bumper-mounted center-loaded whip suitable for antenna of fixed length looks at its base feed point
operation in the hf range. The antenna could also like a decreasing resistance in series with an
be mounted on the car body proper (such as a increasing capacitive reactance. The capacitive re¬
fender), and mounts are available for this purpose. actance must be tuned out, which necessitates the
The base spring acts as a shock absorber for the use of a series inductive reactance or loading coil.
base of the whip since the continual flexing while The amount of inductance required will be deter¬
in motion would otherwise weaken the antenna. A mined by the placement of the coil in the antenna
short heavy mast section is mounted between the system. Base loading requires the lowest value of
base spring and loading coil. Some models have a inductance for a fixed-length antenna, and as the
mechanism which allows the antenna to be tipped coil is placed farther up the whip, the necessary
over for adjustment or for fastening to the roof of value increases. This is because the capacitance of
the car when not in use. It is also advisable to the shorter antenna section (above the coil) to the
extend a couple of guy lines from the base of the car body is now lower (higher capacitive reac¬
loading coil to clips or hooks fastened to the rain tance), requiring a bigger inductance in order to
trough on the roof of the car. Nylon fishing line of tune the antenna to resonance. The advantage is
about 40-pound test strength is suitable for this that the current distribution in the whip is im¬
purpose. The guy lines act as safety cords and also proved, which increases the radiation resistance.
reduce the swaying motion of the antenna con¬ The disadvantage is that requirement of a larger
siderably. The feed line to the transmitter is coil also means the coil losses go up. Center loading
connected to the bumper and base of the antenna. has been generally accepted as a good compromise
Good low-resistance connections are important with minimal construction problems.
here.
Tune-up of the antenna is usually accomplished ADJUSTABLE
by changing the height of the adjustable whip WHIP
section above the precut loading coil. First, tune
the receiver and try to determine where the signals
seem to peak up. Once this frequency is found,
check the SWR with the transmitter on, seeking
the frequency of lowest SWR. Shortening the
adjustable section will increase the resonant fre¬
quency and making it longer will lower the
frequency. It is important that the antenna be
away from surrounding objects such as overhead
wires, ten feet or more, since considerable detuning
can occur. Once the setting is found where the
SWR is lowest at the center of the desired
operating frequency range, the length of the
adjustable section should be recorded.
Propagation conditions and ignition noise are
usually the limiting factors for mobile operation on Fig. 15-1 — A typical bumper-mounted hf-mobile
20, 15, and 10 meters, while antenna-size restric¬ antenna. Note the nylon guy lines.

287
288 Chapter 15

The difficulty in constructing suitable loading Antenna resonance is defined as the frequency
coils increases as the frequency of operation is at which the input impedance at the antenna
lowered for typical antenna lengths used in mobile terminals is a pure resistance. The shortest length
work. Since the required resonating inductance at which this occurs for a vertical antenna over a
gets larger and the radiation resistance decreases at ground plane is when the antenna is a quarter
lower frequencies, most of the power may be wavelength long at the operating frequency; the
dissipated in the coil resistance and in other ohmic impedance value for this length (neglecting losses)
losses. This is one reason why it is advisable to buy is about 36 ohms. The idea of resonance can be
a commercially made loading coil with the highest extended to antennas shorter (or longer) than a
power rating possible, even though only low-power quarter wave, and only means that the input
operation is contemplated. Percentwise, the coil impedance is purely resistive. As pointed out
losses in the higher power loading coils are usually previously, when the frequency is lowered, the
less, with subsequent improvement in radiating antenna looks like a series RC circuit, as shown in
efficiency, regardless of the power level used. Of Fig. 15-2. For the average 8-ft. whip, the reactance
course, the above philosophy also applies to of C\ may range from about 150 ohms at 21 MHz
homemade loading coils, and design considerations to as high as 8000 ohms at 1.8 MHz, while the
will be considered in a later section. radiation resistance varies from about 15 ohms
Once the antenna is tuned to resonance, the at 21 MHz to as low as 0.1 ohm at 1.8 MHz.
input impedance at the antenna terminals will look For an antenna less than 0.1 wavelength long,
like a pure resistance. Neglecting losses, this value the approximate radiation resistance may be de¬
drops from nearly 15 ohms on 15 m to 0.1 ohm on termined from the following:
160 m for an 8-foot whip. When coil and other
losses are included, the input resistance increases to = 273 (//)’ X HT*
approximately 20 ohms on 160 m and 16 ohms on where Iis the length of the whip in inches, and /is
15 m. These values are for relatively high-efficiency the frequency in megahertz.
systems. From this it can be seen that the radiating Since the resistance is low, considerable current
efficiency is much poorer on 160 m than on 15 m must flow in the circuit if any appreciable power is
under typical conditions. to be dissipated as radiation in the resistance ÆR .
Yet it is apparent that little current can be made to
flow in the circuit so long as the comparatively
high series reactance remains.

Antenna Capacitance
The capacitive reactance can be canceled out by
Fig. 15-2 — At frequencies below the resonant connecting an equivalent inductive reactance, £ L,
frequency, the whip antenna will show capacitive in series, as shown in Fig. 15-3, thus tuning the
reactance as well as resistance. Rr is the radiation system to resonance.
resistance, and Ca represents the capacitive reac¬ The capacitance of a vertical antenna shorter
tance. than a quarter wavelength is given by

Since most modern gear is designed to operate

with a 50-ohm transmission line, a matching
network may be necessary with the high-efficiency
antennas previously mentioned. This can take the
form of either a broad-band transformer, tapped where
coil, or an ¿C-matching network. With homemade C’a = capacitance of antenna in pF
or modified designs, the tapped-coil arrangement is L = antenna height in feet
perhaps the easiest one to build, while the broad¬ D = diameter of radiator in inches
band transformer requires no adjustment. As the / = operating frequency in MHz.
losses go up, so does the input resistance, and in
less efficient systems the matching network may be
eliminated. , 24£ _ . , . 24/.
m “p- - 2.3 log, o -p—
The Equivalent Circuit of a
Typical Mobile Antenna
The graph of Fig. 15-4 shows the approximate
In the previous section, some of the general capacitance of whip antennas of various average
considerations were discussed, and these will now diameters and lengths. For 1.8, 4 and 7 MHz,
be taken up in more detail. It is customary in the loading-coil inductance required (when the
solving problems involving electric and magnetic loading coil is at the base) will be approximately
fields (such as antenna systems) to try to find an the inductance required to resonate in the desired
equivalent network to replace the antenna. In band with the whip capacitance taken from the
many cases, the network may be an accurate graph. For 14 and 21 MHz, this rough calculation
representation only over a limited frequency range. will give more than the required inductance, but it
However, this is often a valuable method in will serve as a starting point for the final experi¬
matching the antenna to the transmission line. mental adjustment that must always be made.
Mobile Antennas 289

Loading Coil Design

To minimize loading-coil loss, the coil should
have a high ratio of reactance to resistance, i.e.,
high Q. A 4-MHz loading coil wound with small
wire on a small-diameter solid form of poor
quality, and enclosed in a metal protector, may
have a Q as low as 50, with a resistance of 50 ohms
or more. High-0 coils require a large conductor,
“air-wound” construction, large spacing between
turns, the best insulating material available, a
diameter not less than half the length of the coil
(not always mechanically feasible), and a minimum
of metal in the field. Such a coil for 4 MHz may
show a Q of 300 or more, with a resistance of 12
ohms or less.
Fig. 15-4 — Graph showing the approximate
The coil could then be placed in series with the capacitance of short vertical antennas for various
feed line at the base of the antenna to tune out the diameters and lengths. These values should be
unwanted capacitive reactance, as shown in Fig. approximately halved for a center-loaded antenna.
15-3. Such a method is often referred to as base
loading, and many practical mobile antenna
systems have been built in this way. tapered coaxial cable where the center conductor
Over the years, the question has come up as to remains the same diameter as the length increases,
whether or not more efficient designs are possible but with an increasing diameter of the outer
compared with simple base loading. While many conductor. The inductance per unit length of such
ideas have been tried with varying degrees of a cable would increase farther along the line while
success, only a few have been generally accepted the capacitance per unit length would decrease. In
and incorporated into actual systems. These are Fig. 15-5 the antenna is represented by a series of
center loading, continuous loading, and combina¬ LC circuits in which Cl is greater than C2, which is
tions of the latter with more conventional anten¬ greater than C3, and so on. LI is less than L2,
nas. which is less than succeeding inductances.
The net result is that most of the antenna
current returns to ground near the base of the
antenna, and very little near the top. Two things
Fig. 15-3 — The capacitive can be done to improve this distribution and make
reactance at frequencies the current more uniform. One would be to
lower than the resonant increase the capacitance of the top of the antenna
frequency of the whip can
be canceled out by adding to ground through the use of top loading or a
an equivalent inductive “capacitive hat,” as discussed in Chapter Two.
reactance in the form of a Unfortunately, the wind resistance of the hat
loading coil in series with
the antenna.

Base Loading and Center Loading

If a whip antenna is short compared to a
wavelength and the current is uniform along the
length, I, the electric field strength, E, at a distance
d away from the antenna is given approximately
by:
,. _ 120rr//
E ’

where Iis the antenna current in amperes, and dis

the wavelength in the same units as d and /. A
uniform current flowing along the length of the
whip is an idealized situation, however, since the
current is greatest at the base of the antenna and
goes to a minimum at the top. In practice, the field
strength will be less than that given by the above Fig. 15-5 — A circuit approximation of a simple
equation, being a function of the current distribu¬ whip over a perfectly conducting ground-plane.
The shunt capacitance per unit length gets smaller
tion in the whip. as the height increases, and the series inductance
The reason that the current is not uniform in a per unit length gets larger. Consequently, most of
whip antenna can be seen from the circuit approxi¬ the antenna current returns to the ground plane
mation shown in Fig. 15-5. A whip antenna over a near the base of the antenna, giving the current
ground plane is similar in many respects to a distribution shown on the right.
290 Chapter 15

Table 13-1
Approximate Value« for 8-ft. Mobile Whip

Base Loading

Loading Ac (Q50) Ac (Q300) Ar Feed R* Matching

f*.
L^. Ohm» Ohm» Ohm» Ohm»

1800 345 77 13 0.1 23 3

3800 77 37 6.1 0.35 16 1.2

7g 00 20 18 3 1.35 15 0.6

14.800 4.5 7.7 1.3 5.7 12 0.28

tl.gõO 1.25 3.4 0.5 14.8 16 0.28

89,000 36 0.23

Center Loading

1800 700 158 23 0.2 34 3.7

3800 150 72 12 0.8 22 1.4

7800 40 36 6 3 19 0.7

14.800 8.6 15 2.5 11 19 0.35

81,850 2.5 6.6 1.1 27 29 0.29

Ac — Loading-coil resistance; Ar — Radiation resistance.

• Assuming loading coil Q — 300, and including estimated ground-loss
resistance.
** For matching given feed resistance to 52 ohms.

makes it somewhat unwieldly for mobile use. The length-to-diameter ratios as high as 21. English
other method is to place the loading coil farther up experimenters have reported good results with
the whip, as shown in Fig. 15-6, rather than at the 8-foot overall lengths on the 1.8- and 3.5-MHz
base. If the coil is resonant (or nearly so) with the bands. The idea of making the entire antenna out
capacitance to ground of the section above the of one section of coil has also been tried with some
coil, the current distribution is improved as also success. Such antennas are referred to as helically
shown in Fig. 15-6. The result with both top wound. Further information on helically wound
loading and center loading is that the radiation antennas is contained in Chapter Ten.
resistance is increased, offsetting the effect of While going to extremes in trying to find a
losses and making matching easier. perfect loading arrangement may not improve
Table 15-1 shows the approximate loading-coil antenna performance very much, a poor system
inductance for the various amateur bands. Also with lossy coils and high-resistance connections is
shown in the table are approximate values of also to be avoided.
radiation resistance to be expected with an 8-ft.
whip, and the resistances of loading coils - one Matching to the Transmitter
group having a Q of 50, the other a Q of 300. A
comparison of radiation and coil resistances will Most modern transmitters require a 50-ohm
show the importance of reducing the coil resistance output load and since the feed-point impedance of
to a minimum, especially on the three lower a mobile whip is quite low, a matching network
frequency bands. may be necessary. While calculations are helpful in
the initial design, considerable experimenting is
Continuously Loaded Antennas often necessary in final tune-up. This is particularly
true for the lower bands, where the antenna is
The design of high-Q air-core inductors for rf electrically short compared with a quarter-wave
work is complicated by the number of parameters whip. The reason is that the loading coil is required
which must be optimized simultaneously. One of to tune out a very large capacitive reactance, and
these factors which affects coil Q adversely is even small changes in component values result in
radiation. Therefore, the possibility of cutting large reactance variations. Since the feed-point
down the other losses while incorporating the coil resistance is low to begin with, the problem is even
radiation into that from the rest of the antenna more aggravated. This is one reason why it is
system is an attractive one. advisable to guy the antenna, and make sure that
The general approach has been to use a coil no conductors such as overhead wires are near the
made from heavy wire (No. 14 and larger), with whip during tune-up.
Mobile Antennas 291

Transforming the low resistance of the whip up

to a value suitable for a 50-ohm system can be
accomplished with an rf transformer, or with a
shunt-feed arrangement, such as an L network. The
latter may only require one extra component at
the base of the whip, since the circuit of the
antenna itself may be used as part of the network.
The following example illustrates the calculations
involved.
Assume that a center-loaded whip antenna, 8.5
feet in overall length, is to be used on 7.2 MHz.
From Table 15-1, we see that the feed-point Fig. 15-6 — Improved current distribution resulting
resistance of the antenna will be approximately 19 from center loading.
ohms, and from Fig. 15-4 that the capacitance of
the whip, as seen at its base, is approximately 24
pF. Since the antenna is to be center loaded, the capacitor may be used as the shunt element. With
capacitance value of the section above the coil will the tune-up procedure described later, the value is
be cut approximately in half, to 12 pF. From this, not critical, and a fixed-value component may be
it may be calculated that a center-loading inductor used.
To arrive at point Pl, the value of the center
of 40.7 mH is required to resonate the antenna,
loading-coil inductance would be less than that
that is, cancel out the capacitive reactance. (This
figure agrees with the approximate value of 40 mH required for resonance. The feed-point impedance
would then appear capacitive, and an inductive
shown in Table 15-1). The resulting feedpoint
impedance would then be 19 + /0 ohms - a good shunt matching element would then be required.
match, if one happens to have a supply of 19-ohm To arrive at point P2, the center loading coil would
be more inductive than required for resonance, and
coax.
the shunt element would need to be capacitive.
Solution: The antenna can be matched to a
50-ohm line either by tuning it above or below The value of the center loading coil required for
resonance and then canceling out the undesired the shunt-matched and resonated condition may be
component with an appropriate shunt element, determined in henrys from the equation:
inductive or capacitive. The way in which the
impedance is transformed up can be seen by L =__1_ + 2^
constructing an admittance plot of the series RLC 4n 2PC 2rr/
circuit consisting of the loading coil, antenna
capacitance, and feed-point resistance. Such a plot where addition is performed if a capacitve shunt is
is shown in Fig. 15-7 for a constant feed-point to be used, or subtraction performed if the shunt is
resistance of 19 ohms. There are two points of inductive, and where / is the frequency in hertz, C,
interest, Pl and P2, where the input conductance is the capacitance of the an tenna section being
20 millimhos, which corresponds to 50 ohms.j- The matched in farads, and Xs = V^f^Zo - Rf).
undesired susceptance is shown as 1/Xp and For the example given. Xs is found to be 24.3
-1/Xp, which must be canceled with a shunt ohms, and the required loading inductance is either
element of the opposite sign but with the same
numerical value. The value of the canceling shunt
reactance, Xp, may be found from the formula:
Xd _ Rf Zo
P V Rf(Zo - Rf)
where Xp is the reactance in ohms, Rf is the
feed-point resistance, and Zo is the feed-line
impedance. For Zo = 50 ohms and Rf =19 ohms,
Xp = ±39.1 ohms. A coil or good-quality mica
TThe conductance equals the reciprocal of the
resistance, if no reactive components are present.
For a series RX circuit, the conductance is given
by:

r . *
R1 + X2
and the susceptance is given by :
Fig. 15-7 — Admittance diagram of the RLC circuit
consisting of the whip capacitance, radiation resis¬
tance and loading coil discussed in text. The
Consequently, a parallel equivalent GB circuit of horizontal axis represents conductance, and the
the series RX one can be found which makes vertical axis susceptance. The point Po is the input
computations easier. This is because conductances
and susceptances add in parallel the same way admittance with no whip loading inductance.
resistances and reactances add in series. Points P1 and P2 are described in the text.
 292 Chapter 15

40.2 mH or 41.3 mH, depending on the type of

shunt. The various matching configurations for this
example are shown in Fig. 15-8. At A is shown the
antenna as tuned to resonance with Ll, a 40.7-mH
coil, but with no provisions included for matching Fig. 15-9 — A method of
the resulting 19-ohm impedance to the 50-ohm matching the loaded whip
line. At B, Ll has been reduced to 40.2 mH to to 52-ohm coax cable. Ll
make the antenna appear capacitive, and L^, is the loading coil and L^
having a reactance of 39.1 ohms, added in shunt to the matching coil.
cancel the capacitive reactance and transform the
feed-point impedance to 50 ohms. The arrange¬
ment at C is similar to that at B except that Ll has
been increased to 41.3 mH, and CM , a shunt
capacitor having a reactance of 39.1 ohms added,
which also results in a 50-ohm nonreactive termina¬ which means that a coil of 0.86 mH will be placed
tion for the feed line. across the whip terminals to ground. With a 40-mH
The values determined for the loading coil in loading coil in place, the adjustable section above
the above example point out an important consid¬ the loading coil should be adjusted for minimum
eration concerning the matching of short antennas height. Signals in the receiver will sound weak and
- that relatively small changes in values of the the whip should be lengthened a bit at a time until
loading components will have a greatly magnified they start to peak up. Turn the transmitter on and
effect on the matching requirements. A change of check the SWR at a few frequencies to find where
less than 3 percent in loading-coil inductance value a minimum occurs. If it is below the desired
necessitates a completely different matching net¬ frequency of operation, shorten the whip slightly
work! Likewise, calculations show that a 3 percent and check again. It should be moved approxi¬
change in antenna capacitance will give similar mately 1/4 inch at a time until the SWR is
results, and the value of the precautions mentioned minimum at the center of the desired range. If the
earlier becomes clear. The sensitivity of the circuit frequency where the minimum SWR exists is above
with regard to frequency variations is also quite the desired frequency, repeat the above only
critical and an excursion around practically the lengthen the whip slightly.
entire circle in Fig. 15-7 may represent only 600 If a shunt capacitance is to be used, a value of
kHz, centered around 7.2 MHz for the above 565 pF would correspond to the needed 39.1 ohms
example. This is why tuning up a mobile antenna of reactance for a frequency of 7.2 MHz. With a
can be very frustrating unless a systematic proce¬ capacitive shunt, start with the whip in its longest
dure is followed. position and shorten it until signals peak up.

Tune-up
TAPPED-COIL MATCHING NETWORK
Assume that inductive shunt matching is to be
used with the antenna in the previous example. Some of the drawbacks of the previous circuits
can be eliminated by the use of the tapped-coil
arrangement shown in Fig. 15-9. While tune-up is
still critical, a smaller loading coil is required which
cuts down losses. This system may be initially
resonated with a dip meter while the feed line is
disconnected. Then with the transmission line
connected to a tap on L^, listen with a receiver
and adjust the tap point until the signals peak up.
Next, check the SWR with a transmitter as
described above. Adjust the tip length for minimum
SWR at the desired frequency. Slight repositioning
of the tap point on L M may be necessary to obtain
the lowest SWR.

Bibliography
Fig. 15-8 — At A, a whip antenna which is Source material and more extended discussion
resonated with a center loading coil. At B and C,
the value of the loading coil has been altered of topics covered in this chapter can be found in
slightly to make the feed-point impedance appear the references given below.
reactive, and a matching component is added in
shunt to cancel the reactance. This provides an Belcher, “RF Matching Techniques, Design and
impedance transformation to match the Zo of the Example,” QST, October, 1972.
feed line. An equally acceptable procedure, rather Belrose, “Short Antennas for Mobile Operation,”
than altering the loading-coil inductance, is to QST, September, 1953.
adjust the length of the top section above the Stephens, “Admittance Matching the Ground-Plane
loading coil for the best match, as described in the Antenna to Coaxial Transmission Line,” Tech¬
tune-up section of the text. nical Correspondence, QST, April, 1973.
 Chapter 16

Specialized Antennas

In general, most amateur antenna installations tion some of these types of antennas, including
show little departure from the standard form of loops, special long-wire type antennas and long-
either dipoles or inverted-V antennas for 80 and 40 wire arrays, as well as a variety of unique types of
meters, and some sort of trap-dipole or beam antennas which defy categorization.
antenna of a triband nature for use on 20, 15 and
10 meters. This gives the hf station five-band LOOP ANTENNAS
capability without having an unsightly amount of Small loops can be made in the form of a circle,
cable and antenna hardware strewn about the triangle, rectangle, etc., with little or no change in
premises. properties. The most convenient form, generally, is
Vhf antennas are, by their nature, small, and a square such as is shown in Fig. 16-1. So long as
one mast can easily accommodate beams for the total length of the conductor is very small
several vhf bands. Simple vertical vhf antennas to compared with the wavelength the loop acts like a
provide access to repeaters are so unobtrusive that simple inductance and can be tuned to resonance
they can be mounted almost anywhere without at the desired frequency by a capacitor, C. The
being noticed. directive pattern of such a loop is given by the
Beyond the “standard” antenna arrangements small drawing, and is the same as that of an
summarized above lies an area which includes a elementary doublet.
rather large number of interesting but seldom-used
antennas. These are worthy of note to experi¬
mentally inclined amateurs. This chapter will men-

Fig. 16-2 - Shielded

loop for direction find¬
ing. The ends of the
shielding turn are insu¬
lated from each other
to prevent shielding the
loop from magnetic
fields, although the
shielding is effective
against electric fields.

Loops of this type do not have much applica¬

tion in amateur work, although they are widely
used at frequencies below the standard broadcast
band for direction finding. They are not very
useful for this purpose at high frequencies because
waves arriving at a receiving point via the iono¬
sphere have random polarization and wave angles,
and this introduces large errors. However, they are
capable of giving good results for this purpose if
only the ground wave is present, provided the
location of the loop is such that false bearings are
not caused by reflections from nearby conductors.
Flg. 16-1 — Small loop, consisting of several turns Shielded loops have been used with considerable
of wire having a total length very much less than a
wavelength. The directional pattern of such a loop success in “hidden transmitter hunts” utilizing
is as shown in the small drawing, with maximum frequencies 28 MHz and higher. The shield, shown
response in the plane of the loop. schematically in Fig. 16-2, is used for preventing

293
 294 Chapter 16

“antenna effect” - that is, to eliminate undesired

response of the loop considered merely as a mass
of metal connected to the receiver antenna input
terminals.
The radiation resistance of a small loop is
extremely low. For this reason most of the power
supplied to the loop is wasted in conductor
resistance loss, when the loop is used for trans¬
mitting. A similar situation exists when the loop is
used for receiving; because of its small size only a
relatively small amount of energy is absorbed from
passing waves. However, at comparatively low
frequencies such as the 3.5-MHz band it can draw
energy from a fairly large area (see Chapter Two on
pickup efficiency) of the passing wave front and as
a result may establish a reasonable signal-to-noise Fig. 16-4 — The fishbone antenna provides higher
ratio at the receiver input terminals. It should be gain per acre than does a rhombic. It is essentially a
wave antenna which evolved from the Beverage
tuned to the received frequency in such case, and antenna.
when so resonated may be markedly better than
the few feet of wire often used for reception on
the lower frequencies. Also, it is sometimes possi¬ antenna terminated at each end in its characteristic
ble to take advantage of its directional effects to impedance. It need be only one wavelength long to
reduce interference. be effective, and its height can be relatively low
Loops for the lower hf bands are generally (10 to 20 feet) for a response to signals arriving at
constructed of not more than about 20 feet of wire low wave angles. Depending on the condition of
wound into a square or circle which has a the ground, the antenna height, and the number of
maximum dimension of from 12 to 20 inches. The wires comprising the antenna, the characteristic
response of this type of loop is bidirectional, as a impedance can be anywhere from 200 to 500
dipole would be, but the gain is below that of a ohms. Usually it is taken to be between 500 and
dipole. Loops can be made to have gain over a 600 ohms. It may be fed with open-wire trans¬
dipole, however, if enough wire is used so that the mission line or with coaxial cable through a
loop contains one-half wavelength or more. Load¬ suitable impedance matching device as shown in
ing coils may also be used to improve the direc¬ Fig. 16-3. If the antenna is used for transmitting,
tivity of the longer loops as shown in Fig. 2-83 and the same considerations given to termination resis¬
explained in its attendant text. tance of a rhombic should be applied. The resis¬
tance power rating should be approximately one
half the transmitter output power and the resistor
SPECIALIZED LONG WIRES should be noninductive.

The Beverage or Wave Antenna Fishbone Antennas

Nonresonant long-wire antennas were covered Another type of wave antenna is the fishbone,
in Chapter Five. One version of this type of as illustrated in Fig. 16-4. The impedance of the
antenna is called the Beverage (or wave) antenna, fishbone antenna is approximately 400 ohms. The
and is sufficiently different to warrant space here. antenna is formed of closely spaced elements
For one thing, unlike most antennas treated earlier lightly coupled capacitively to a long, terminated,
in this book, the beverage requires a poorly transmission line. The capacitors are chosen to
conducting earth beneath it. This is to facilitate have a value which will keep the velocity of
tilting of the approaching wave front in the vicinity propagation of the line more than 90% of that in
of the antenna so that a larger area of the normally air. The elements are usually spaced approximately
horizontal antenna is exposed to the wave front at 0.1 wavelength or slightly more so that an average
any given time. The beverage antenna is essentially of 7 or more elements is used for each full
a long terminated transmission line, a straight wavelength of transmission-line length. This an¬
tenna obtains low-angle response primarily as a
FAVORED DIRECTION
function of its height, and therefore, is generally
installed 60 to 120 feet above ground. If the
antenna is to be used for transmission, the capaci¬
tors should be of the transmitting type, as they will
be required to handle substantial current.
The English HAD fishbone antenna, shown in
its two-bay form in Fig. 16-5, is of less complicated
design than the one just described. It may be used
Fig. 16-3 — The Beverage (wave) antenna provides singly, of course, and may be fed with 600-ohm
a unidirectional low-angle pattern for a very low open-wire line. Installation and operational charac¬
physical height above ground. Lossy earth beneath teristics are similar to the standard fishbone an¬
the antenna enhances its efficiency as an antenna. tenna.
Specialized Antennas 295
An alternative feeding method is shown at D of
Fig. 16-6. This system can also be applied to the
3-element arrays, and will result in better sym¬
metry in any case. It is only necessary to move the
phasing line to the center of each element, making
connection to both sides of the line instead of one
only.
The free-space pattern for a 4-element array
with half-wave spacing is shown in Fig. 16-7. This
is also approximately the pattern for a 3-element
array with 3/4-wavelength spacing. The major lobe
of a 3-element array with half-wave spacing is
intermediate in sharpness between a 2-element
array (Fig. 4-18) and a 4-element array.
Larger arrays can be designed and constructed
by following the phasing principles shown in the
drawings.
No accurate figures are available for the im¬
pedances at the various feed points indicated in the
drawings. It can be estimated to be in the vicinity
of 1000 ohms when the feed point is at a junction
between the phasing line and a half-wave element,
becoming smaller as the number of elements in the
array is increased. When the feed point is midway
Fig. 16-5 — The English HAD fishbone antenna is a between end-fed elements as in Fig. 16-6C the
impedance of a 4-element array as seen by the

(A)
simplified version of the standard fishbone and
may be used as a single-bay antenna fed with transmission line is in the vicinity of 200 to 300
600-ohm open-wire line.

OTHER FORMS OF MULTIELEMENT

DRIVEN ARRAYS
Chapter Four covers the simpler forms of
broadside antennas. For those who have the
available room, multielement arrays based on the
broadside concept have something to offer. The mans line
antennas are large but of simple design and
noncritical dimensions; they are also very econ¬
omical in terms of gain per unit of cost.
Three- and four-element arrays are shown in
Fig. 16-6. In the 3-element array with half-wave
spacing, A, the array is fed at the center. This is the
most desirable point in that it tends to keep the
power distribution between elements uniform.
However, the transmission Une could be connected
at either point B or C of Fig. 16-6 A, with only
slight skewing of the radiation pattern.
When the spacing is greater than 1/2 wave¬
length, the phasing lines must be one wavelength
long and are not transposed between elements.
This is shown at B in Fig. 16-6. With this
arrangement, any element spacing up to one
wavelength can be used, if the phasing lines can be
folded as suggested in the drawing.
The 4-element array at C is fed at the center of
the system to make the power distribution be¬
tween elements as uniform as possible. However,
the transmission line could be connected at either
point B, C, D or E. In such case the section of
phasing line between B and D must be transposed
in order to make the currents flow in the same
direction in all elements. The 4-element array at C
and the 3-element array at B have approximately
the same gain when the element spacing in the Fig. 16-6 — Methods of feeding three- and four-
latter is 3/4 wavelength. element broadside arrays with parallel elements.
 296 Chapter 16

Fig. 16-7 — Free-space directive diagram of a Fig. 16-9 - Free-space pattern of the four-element
four-element broadside array using parallel ele¬ antenna shown in Fig. 16-8, in the plane perpen¬
ments. This is also the horizontal directive pattern dicular to the array axis. The pattern in the plane
at low wave angles for a vertically polarized array. containing a set of end-fire elements is the same as
Fig. 4-22 of Chapter Four.

ohms, with 600-ohm open-wire phasing lines. The half-wave spacing between broadside elements.) In
impedance at the feed point with the antenna this case the impedance at the junction is estimated
shown at D should be about 1500 ohms. to be over 10,000 ohms. With other element
spacings the impedance at the junction will be
Four-Element Broadside and End-Fire Array
reactive as well as resistive, but in any event the
The array shown in Fig. 16-8 combines parallel standing-wave ratio will be quite large. An open¬
elements in broadside and end-fire directivity. wire line can be used as a resonant line, or a
Approximate gains can be calculated by adding the matching section may be used for nonresonant
values from Chapter Four Figs. 4-16 and 4-20 for operation.
the element spacings used. The smallest (physi¬
cally) array — 3/8-wave spacing between broadside Eight-Element Driven Array
and 1/8-wave spacing between end-fire elements - The array shown in Fig. 16-11 is a combination
has an estimated gain of 6.8 dB and the largest - of collinear and parallel elements in broadside and
3/4- and 1/4-wave spacing, respectively - about end-fire directivity. The gain can be calculated as
8.5 dB. The optimum element spacings are 5/8 described earlier, using Figs. 4-9, 4-16 and 4-20.
wave broadside and 1/8 wave end-fire, giving an Common practice is to use half-wave spacing for
overall gain estimated at 9.3 dB. Directive patterns the parallel broadside elements and 1/8-wave spa¬
are given in Figs. 16-9 and 16-10. cing for the end-fire elements. This gives an
The impedance at the feed point will not be estimated gain of about 10 dB. Directive patterns
purely resistive unless the element lengths are for an array using these spacings are similar to
correct and the phasing lines are exactly a half those of Figs. 16-9 and 16-10, being somewhat
wavelength long. (This requires somewhat less than sharper.

Fig. 16-8 — Four-element

array combining both
broadside and end-fire ele¬
ments.
Specialized Antennas 297

Although even approximate figures are not

available, the SWR with this arrangement will be
high. Matching stubs are recommended for making
the line nonresonant. Their position and length can
be determined by measuring the SWR and locating
the current loop or null nearest the junction of the
transmission and phasing lines. The procedure is
described in Chapter Three.
This system can be used on two bands related
in frequency by a 2-to-l ratio, providing it is
designed for the higher of the two with 3/4-wave
spacing between the parallel broadside elements
and 1/4-wave spacing between the end-fire ele¬ Fig. 16-10 — Vertical pattern of the antenna shown
in Fig. 16-8 at a mean height of 3/4 wavelength
ments. On the lower frequency it will then operate (lowest elements 1/2 wave above ground) when the
as a four-element antenna of the type shown in antenna is horizontally polarized. For optimum
Fig. 16-8 with 3/8-wave broadside spacing and gain and low wave angle the mean height should be
1/8-wave end-fire spacing. For two-band operation at least 3/4 wavelength.
a resonant transmission line must be used.

Other Driven Systems When feeding at the point shown the imped¬
Two other types of driven antennas are worthy ance is of the order of 600 ohms. Alternatively,
of mention, although their use by amateurs has this point can be closed and the system fed
been rather limited. The Sterba array, shown at A between any two elements, as at X. In this case a
in Fig. 16-12, is a broadside radiator consisting of point near the center should be chosen so that the
both collinear and parallel elements, with 1/2-wave power distribution between elements will be as
spacing between the latter. Its distinctive feature is uniform as possible. The impedance at any such
the method of closing the ends of the system. For point will be 1000 ohms or less in systems with six
direct current and low-frequency ac, the system or more elements.
forms a closed loop, which is advantageous in that The Bruce array is shown at B in Fig. 16-12. It
heating currents can be sent through the wires to consists simply of a single wire folded so that the
melt the ice that forms in cold climates. There is vertical sections carry large currents in phase while
comparatively little radiation from the vertical the horizontal sections carry small currents flowing
connecting wires at the ends because the currents in opposite directions with respect to the center
are relatively small and are flowing in opposite (indicated by the dot). The radiation consequently
directions with respect to the center (the voltage is vertically polarized. The gain is proportional to
loop is marked with a dot in this drawing). the length of the array but is somewhat smaller,
The system obviously can be extended as far as because of the short radiating elements, than is
desired. The approximate gain is the sum of the obtainable from a broadside array of half-wave
gains of one set of collinear elements and one set parallel elements of the same overall length. The
of broadside elements, counting the two 1/4-wave array should be 2 or more wavelengths long to
sections at the ends as one element. The antenna secure a worthwhile gain. The system can be fed at
shown, for example, is about equivalent to one set any current loop; these occur at the centers ot the
of four collinear elements and one set of two vertical wires.
parallel broadside elements, so the total gain is Another form of the Bruce array is shown at C.
approximately 4.3 + 4.0 = 8.3 dB. Horizontal Because the radiating elements have twice the
polarization is the only practicable type at the height, the gain is increased. The system can be fed
lower frequencies, and the lower set of elements at the center of any of the connecting lines.
should be at least 1/2 wavelength above ground for
best results.

Fig. 16-11 — Eight-element

driven array combining col¬
linear and parallel elements
for broadside and end-fire
directivity.
 298
Chapter 16

THE DISCONE ANTENNA imately equal to the free-space 1/4-wavelength at

the lowest frequency for which the antenna is
The discone is a vertically polarized broadband built.
antenna which maintains an SWR of 1.5:1, or less, Below the design frequency, the SWR rises
over several octaves of frequency. Fig. 16-13 shows
rapidly, but within its “resonant” region the
the configuration of the antenna. Dimension L of
antenna provides an excellent match to the popular
the equilaterally skirted bottom section is approx- 50-ohm coax.
Because of its physical bulk at hf the antenna
has not enjoyed much use by amateurs working in
that part of the spectrum. However, the antenna
has much to offer at vhf and uhf. If designed for
50 MHz, for example, the antenna will also work
well on 144 and 220 MHz. Construction at hf
would best be done by simulating the skirt by a
grid of wires, and on vhf and uhf there would be
no problem in fashioning a solid skirt of some
easily workable metal, such as flashing copper.
The disk-like top-hat section should be insu¬
lated from the skirt section. This is usually done
with a block of material strong enough to support
Fig. 16-13 — The discone antenna is a wideband,
coaxially fed type best suited to vhf and uhf
coverage because of its cumbersome size at hf.
Dimension L is equal to a free-space quarter
wavelength at the lowest operating frequency. The
profile of the skirt is an equilateral triangle; the
skirt itself can be of acage type of construction,
with adjoining wires separated by not more than
■ 02X at the bottom of the cone. Dimensions varies
50-0HM COAX from 1 to 6 inches, depending on the low-
frequency cutoff of the design.
Specialized Antennas 299

Fig. 16-14 — The conical monopole antenna. At B,

top view showing dimensions for 3.5 through 14
MHz. At C side view of conical monopole at
section A-A. Note that grounding stubs, b, connect
to short radial wires, a. Wires crun up the sides of
the supporting pole, which is unguyed.

the disk. The inner conductor of the coax runs up of the band with essentially the same low SWR;
through this block and is attached to the disk; the this one will.
shield of the coax is connected to the skirt section. Like vertical antennas in general, the conical
The optimum spacing of the disk from the skirt monopole requires a ground system beneath it, to
varies as a function of the part of the spectrum for reduce ground losses and raise the radiation effi¬
which the antenna was designed. At hf this spacing ciency. At least 30 wires, each 62 feet long should
may be as much as 6 inches for 14 MHz, while at be used. Every third radial should be connected to
144 MHz the spacing may be only 1 inch. It does a ground rod at its far end and all radials should be
not appear to be particularly critical. joined at their far ends. When this much wire is
The gain of the discone is essentially flat across involved, it behooves the builder to use galvanized
its useful frequency range. The angle of radiation is steel or aluminum wire instead of copper, for
very low, for the most part, rising only slightly at economy reasons.
some frequencies.
THE MULTEE ANTENNA
THE CONICAL MONOPOLE ANTENNA
Two-band operation may be obtained on 160
A trapless vertical antenna which works well and 80 meters within the confines of the average
over a four-to-one frequency range and is operated city lot by using the multee antenna shown in Fig.
at ground potential (which affords a measure of 16-15. On 160 meters the top portion will do little
lightning protection) should appeal to users of 80 radiating, and it acts merely as top loading for the
and 40 meters. Such an antenna is the conical 52-foot vertical section. On 80 meters the horizon¬
monopole shown in Fig. 16-14. Its electrical tal portion radiates and the vertical section report¬
representation is shown in Fig. 16-14C, and dimen¬ edly acts as a matching stub to transform the high
sions for operation on 80, 40 and 20 meters are feed-point impedance down to a respectable match
given at B and C. for coaxial cable.
Whereas the length of a 1/4-wave vertical Fig. 16-15 also gives dimensions for an 80 and
antenna for 80 meters would be in the order of 66 40 meter version of the multee.
feet or so, the conical monopole need be only 0.17 Since the antenna must work against ground on
wavelength, or 43 feet high. The foreshortening is its lower frequency band, it is necessary that a
the result of increased diameter, which also results properly installed ground system be employed.
in broadbanding. Few antennas on 80 meters will Minimum requirements in this regard would in¬
allow an operator to use both the low and high end clude 20 radials, each 60 feet long. If not much
 300 Chapter 16

The tuning network is resonant at the operating

frequency. The L/C ratio should be fairly low to
provide good loading characteristics. As a starting
point, a maximum value of 75 to 150 pF is
recommended and the inductor value is determined
by C and the operating frequency. The network is
first tuned to resonance and then the tap point is
adjusted for the best match. A slight readjustment
of C may be necessary. As an alternative method of
feeding the antenna, a link coil consisting of a few
turns can be used.

Fig. 16-15 — Two-band operation in limited space

may be obtained with the multee antenna. The
feed-line portion should remain as vertical as
possible, as it does the radiating on the lower
frequency band.

area is available for the radial system, wires as short

as 25 feet long may be used if many are installed, Fig. 16-16 — Bobtail curtain is an excellent
with some reduction in efficiency. low-angle radiator having broadside bidirectional
With suitable corrections in length to account characteristics. Current distribution is represented
for the velocity factor of 300-ohm TV Twin-Lead, by the arrows. Dimensions A and 8 (in feet) can be
determined by the formulas. An extensive ground
such line may be substituted for the open-wire line, system is not necessary.
thereby eliminating a potential shock hazard to
children and pets. The velocity factor should be
taken into account for both the vertical and Bibliography
horizontal portions, to preserve the impedance Source material and more extended discussion
relationships. of topics covered in this chapter can be found in
the references given below.

BOBTAIL CURTAIN Bailey, Dean and Wintringham, “The Receiving

System for Long-Wave Transatlantic Radio
The antenna system of Fig. 16-16 uses the Telephony,” The Bell System Technical Jour¬
principles of cophased verticals to produce a nal, April, 1929.
broadside, bidirectional pattern providing 7 to 10 Beverage, “Antennas,” RCA Review, July, 1939.
Boyer, “Discone - 40 to 500 Me. Skywire ” CO
dB of gain over a dipole at the same height. It is July, 1949.
most effective for low-angle signals and makes an Brown, “Directional Antennas,” Proc. I.R.E., Jan¬
excellent DX antenna for either 3.5 or 7 MHz. The uary, 1937.
three vertical elements are the actual radiating Carter, Hansell and Lindenblad, “Development of
components. The two horizontal parts, A, act as Directive Transmitting Antennas by RCA Com¬
phasing lines and contribute very little to the munications, Inc.,” Proc. I.R.E., October
radiation pattern. The vertical elements should be 1931.
as vertical as possible. The height for the horizontal Orr, The Radio Handbook, Editors and Engineers,
Indianapolis, Ind.
portion should be slightly greater than B, as shown Pappenfus, “The Conical Monopole Antenna,”
in Fig. 16-16. QST, November, 1966.
Chapter 17

Measurements

The principal quantities to be measured on always a reliable indicator. In many cases, par¬
transmission lines are line current or voltage, and ticularly with a screen-grid tube in the final stage,
standing-wave ratio. Measurements of current or minimum loaded plate current does not occur
voltage are made for the purpose of determining simultaneously with maximum power output.
the power input to the line. SWR measurements
are useful in connection with the design of RF Voltmeter
coupling circuits and the adjustment of the match A germanium diode in conjunction with a
between the antenna and transmission line, as well low-range milliammeter and a few resistors can be
as in the adjustment of matching circuits. assembled to form an rf voltmeter suitable for
For most practical purposes a relative measure¬ connecting across the two conductors of a coaxial
ment is quite sufficient. An uncalibrated indicator line, as shown in Fig. 17-1. It consists of a voltage
that shows when the largest possible amount of divider, R1-R2, having a total resistance about 100
power is being put into the line is just as useful, in times the Zo of the line (so the power consumed
most cases, as an instrument that measures the will be negligible) with a diode rectifier and
power accurately. It is seldom necessary to know milliammeter connected across part of the divider
the actual number of watts going into the line to read relative rf voltage. The purpose of R3 is to
unless the overall efficiency of the system is being make the meter readings directly proportional to
investigated. An instrument that shows when the the applied voltage, as nearly as possible, by
SWR is close to 1 to 1 is all that is needed for most "swamping” the resistance of CR1, since the latter
impedance-matching adjustments. Accurate resistance will vary with the amplitude of the
measurement of SWR is necessary only in studies current through the diode.
of antenna characteristics such as bandwidth, or The voltmeter may be constructed in a small
for the design of some types of matching systems, metal box, indicated by the dashed line in the
such as a stub match.
Quantitative measurements of reasonable accu¬
racy demand good design and careful construction
in the measuring instruments. They also require
intelligent use of the equipment, including a
knowledge not only of its limitations but also of
stray effects that often lead to false results. A
certain amount of skepticism regarding numerical
data resulting from amateur measurements with
simple equipment is justified until the complete
conditions of the measurements are known. On the
other hand, purely qualitative or relative measure¬
ments are easy to make and are quite reliable for
the purposes mentioned above.

LINE CURRENT AND VOLTAGE

A current or voltage indicator that can be used
with coaxial line is a useful piece of equipment. It Fig. 17-1 — Rf voltmeter for coaxial line.
need not be elaborate or expensive. Its principal C1, C2 - 0.005- or 0.01 -pF ceramic.
CR1 — Germanium diode, any type.
function is to show when the maximum power is J1, J2 — Coaxial fittings, chassis-mounting type.
being taken from the transmitter; for any given set MA — 0-1 milliammeter (more sensitive meter may
of line conditions (length, SWR, etc.) this will be used if desired; see text).
occur when the transmitter coupling is adjusted for R1 — 6800 ohms, composition, 1 watt for each
maximum current or voltage at the input end of 100 watts of rf power.
the line. Although the final-amplifier plate-current R2 - 680 ohms, 1/2 or 1 watt composition.
meter is frequently used for this purpose, it is not R3 — 10,000 ohms, 1/2 watt (see text).

301
 302 Chapter 17

drawing, fitted with coax receptacles. RI and R2

should be composition resistors. The power rating
for RI should be 1 watt for each 100 watts of
carrier power in the matched line; separate 1- or
2-watt resistors should be used to make up the
total power rating required, to a total resistance as
given. Any type of resistor can be used for R3; the
total resistance should be such that about 10 volts
de will be developed across it at full scale. For
example, a 0-1 milliammeter would require 10,000
ohms, a 0-500 microammeter would take 20,000
ohms, and so on. For comparative measurements
only, R3 may be a variable resistor so the
sensitivity can be adjusted for various power levels.
In constructing such a voltmeter, care should be
used to prevent inductive coupling between RI and
the loop formed by R2, CR1 and Cl, and between
the same loop and the line conductors in the
assembly. With the lower end of RI disconnected
from R2 and grounded to the enclosure, but
without changing its position with respect to the
loop, there should be no meter indication when
full power is going through the line.
If more than one resistor is used for RI, the
units should be arranged end to end with very
short leads. RI and R2 should be kept a half inch Fig. 17-2 — A convenient method of mounting an
or more from metal surfaces parallel to the body of rf ammeter for use in coaxial line. This is a
the resistor. If these precautions are observed the metal-case instrument mounted on a thin bakelite
panel, the diameter of the cut-out in the metal
voltmeter will give consistent readings at frequen¬ being such as to clear the edge of the meter by
cies up to 30 MHz. Stray capacitances and cou¬ about an eighth inch.
plings limit the accuracy at higher frequencies but
do not affect the utility of the instrument for
comparative measurements. case instrument can be mounted on a metal panel
without introducing enough shunt capacitance to
ground to cause serious error up to 30 MHz. When
Calibration installing a metal-case instrument on a metal panel
The meter may be calibrated in rf voltage by it should be mounted on a separate sheet of
comparison with a standard such as an rf ammeter. insulating material in such a way that there is an
This requires that the line be well matched so the eighth of an inch or more separation between the
impedance at the point of measurement is equal to edge of the case and the metal.
the actual Zo of the line. Since in that case P = A two-inch instrument can be mounted in a 2 X
P Zo, the power can be calculated from the 4 X 4-inch metal box as shown in Fig. 17-2. This is
current. Then E = \[PZo. By making current and a convenient arrangement for use with coaxial line.
voltage measurements at a number of different Installed as just described, a good-quality rf
power levels, enough points may be secured to ammeter will measure current with an accuracy
permit drawing a calibration curve for the volt¬ that is entirely adequate for calculating power in
meter. the line. As discussed above in connection with
calibrating rf voltmeters, the line must be closely
RF Ammeters matched by its load so the actual impedance will
be resistive and equal to Zo. The scales of such
An rf ammeter can be mounted in any con¬ instruments are cramped at the low end, however,
venient location at the input end of the trans¬ which limits the range of power that can be
mission line, the principal precaution in its mount¬ measured by a single meter. The useful current
ing being that the capacitance to ground, chassis, range is about 3 to 1, corresponding to a power
and nearby conductors should be low. A bakelite- range of about 9 to 1.

SWR MEASUREMENTS
On parallel-conductor lines it is possible to fact, seldom used with open lines, because it is not
measure the standing-wave ratio by moving a only inconvenient but sometimes impossible to
current (or voltage) indicator along the line, noting reach all parts of the line conductors. Also, the
the maximum and minimum values of current (or method is subject to considerable error from
voltage) and then computing the SWR from these antenna currents flowing on the line.
measured values. This cannot be done with coaxial Present-day SWR measurements are practically
line since it is not possible to make measurements always made with some form of “directional
of this type inside the cable. The technique is, in coupler or rf bridge circuit. The indicator circuits
Measurements 303

themselves are fundamentally simple, but con¬ when Rs = Zo, the bridge is always in balance for
siderable care is required in their construction if the incident component. Thus the voltmeter does
the measurements are to be accurate. The require¬ not respond to the incident component at any time
ments for indicators used only for the adjustment but reads only the reflected component (assuming
of impedance-matching circuits, rather than actual that R2 is very small compared with the voltmeter
SWR measurement, are not so stringent and an impedance). The incident component can be
instrument for this purpose can be made quite measured across either RI or R2, if they are equal
easily. resistances. The standing wave ratio is then
El + E2
BRIDGE CIRCUITS
^-W^E2
Two commonly used bridge circuits are shown
where El is the incident voltage and E2 is the
in Fig. 17-3. The bridges consist essentially of two
reflected voltage. It is often simpler to normalize
voltage dividers in parallel, with a voltmeter con¬
the voltages by expressing E2 as a fraction of El,
nected between the junctions of each pair of
in which case the formula becomes
“arms,” as the individual elements are called. When
the equations shown to the right of each circuit are 1 +*
satisfied there is no potential difference between SWR= —
the two junctions, and the voltmeter indicates zero
voltage. The bridge is then said to be in “balance.” where k = E2/E1.
Taking Fig. 17-3A as an illustration, if RI = R2, The operation of the circuit in Fig. 17-3B is
half the applied voltage, E, will appear across each essentially the same, although this circuit has arms
resistor. Then if Rs = Rx, 1/2 E will appear across containing reactance as well as resistance.
each of these resistors and the voltmeter reading It is not necessary that RI = R2 in Fig. 17-3A;
will be zero. Remembering that a matched trans¬ the bridge can be balanced, in theory, with any
mission line has a purely resistive input impedance, ratio of these two resistances provided Rs is
suppose that the input terminals of such a line are changed accordingly. However, the accuracy is
substituted for Rx. Then if Rs is a resistor equal to highest, in practice, when the two are equal, and
the Zo of the line, the bridge will be balanced. If this circuit is generally so used.
the line is not perfectly matched, its input imped¬
ance will not equal Zo and hence will not equal Rs,
RESISTANCE BRIDGE
since the latter is chosen to equal Zo. There will
then be a difference in potential between points X The basic bridge type shown in Fig. 17-3A is
and Y, and the voltmeter will show a reading. Such
recommended for home construction if a bridge is
a bridge therefore can be used to show the to be used for reasonably accurate SWR measure¬
presence of standing waves on the line, because the
ment. A practical circuit for such a bridge is given
line input impedance will be equal to Zo only in Fig. 17-4 and a representative layout is shown in
when there are no standing waves. Fig. 17-5. Properly built, a bridge of this design can
Considering the nature of the incident and be used for measurement of standing-wave ratios
reflected components of voltage that make up the
up to about 15 to 1 with good accuracy.
actual voltage at the input terminals of the line, as
Important constructional points to be observed
discussed in Chapter Three, it should be clear that
are:
1 ) Keep leads in the rf circuit short, to reduce
stray inductance.
2) Mount resistors two or three times their
body diameter away from metal parts, to reduce
stray capacitance.
3) Place the rf components so there is as little
inductive and capacitive coupling as possible be¬
tween the bridge arms.
In the layout shown in Fig. 17-5, the input and
line connectors, JI and J2, are mounted fairly
close together so the standard resistor, Rs, can be
supported with short leads directly between the
center terminals of the connectors. R2 is mounted
at right angles to Rs, and a shield partition is used
between these two components and the others.
The two 47,000-ohm resistors, R5 and R6 in
Fig. 17-4, are voltmeter multipliers for the 0-100
microammeter used as an indicator. This is suffi¬
cient resistance to make the voltmeter linear (that
is, the meter reading is directly proportional to the
Fig. 17-3 — Bridge circuits suitable for SWR rf voltage) and no voltage calibration curve is
measurement. A - Wheatstone type using resis¬
tance arms. B — Capacitance-resistance bridge needed. CR1 is the rectifier for the reflected
("Micromatch"). Conditions for balance are in¬ voltage and CR2 is for the incident voltage.
dependent of frequency in both types. Because of resistor tolerances and small differences
 304
Chapter 17
in diodes, the readings may differ slightly with two
multipliers of the same nominal resistance value, so
a correction resistor, R3, is included in the circuit.
Its value should be selected so that the meter
reading is the same with SI in either position,
when rf is applied to the bridge with the line
connection open. In the instrument shown, a value
of 1000 ohms was required in series with the
multiplier for reflected voltage; in other cases
different values probably would be needed and R3
might have to be put in series with the multiplier
for the incident voltage. This can be determined by
experiment.
The value used for RI and R2 is not critical,
but the two resistors should be matched to within
1 or 2 percent if possible. The resistance of Æs
should be as close as possible to the actual Zo of
the line to be used (generally 52 or 75 ohms). The
resistor should be selected by actual measurement
with an accurate resistance bridge, if one is
available.
R4 is for adjusting the incident-voltage reading
to full scale in the measurement procedure des¬
cribed below. Its use is not essential, but it offers a
convenient alternative to exact adjustment of the
rf input voltage.

Testing
Fig. 17-5 — A 2 X 4 X 4-inch aluminum box is
RI, R2, and Rs should be measured with a used to house this SWR bridge, which uses the
good ohmmeter or resistance bridge after wiring is circuit of Fig. 17-4. The variable resistor, R4, is
completed, in order to make sure their values have mounted on the side.
not changed from the heat of soldering. Disconnect The bridge components are mounted on one
one side of the microammeter and leave the input side plate of the box and a miniature chassis
and output terminals of the unit open during such formed from a piece of aluminum. The input
connector is at the top in this view. Rs is
connected directly between the two center posts of
the connectors. R2 is visible behind it and per¬
pendicular to it. One terminal of CR1 projects
through a hole in the chassis so the lead can be
connected to J2. R1 is mounted vertically to the
left of the chassis in this view, with CR2 connected
between the junction of R1-R2 and a tie point.

measurements, in order to avoid stray shunt paths

through the rectifiers.
Check the two voltmeter circuits as described
above, applying enough rf (about 10 volts) to the
input terminals to give a full-scale reading with the
line terminals open. If necessary, try different
values for R3 until the reading is the same with SI
in either position.
With J2 open, adjust the rf input voltage and/or
R4 for full-scale reading with SI in the reflected-
Fig. 17-4 -Resistance bridge for SWR measure¬ voltage position. Then short-circuit J2 by touching
ment. Capacitors are disk ceramic. Resistors are a screwdriver between the center terminal and the
1/2-watt composition except as noted below. frame of the connector to make a low-inductance
CR1, CR2 — Germaniurn diode, high back- short. Switch SI to the incident-voltage position
resistance type (1N34A, 1N270, etc.). and readjust R4 for full scale, if necessary. Then
•R- J2 — Coaxial connectors, chassis-mounting throw SI to the reflected-voltage position, keeping
type.
M1 — 0-100 de microammeter. J2 shorted, and the reading should be full scale as
R2 — 47 ohms, 1/2-watt composition (see before. If it is not, RI and R2 are not the same
text). value, or there is stray coupling between the arms
R3 — See text. of the bridge. It is necessary that the reflected
R4 — 50,000-ohm volume control. voltage read full scale with J 2 either open or
Rs — Resistance equal to line Zo (1/2 or 1 watt shorted, when the incident voltage is set to full
composition). scale in each case, in order to make accurate SWR
S1 — Spdt toggle. measurements.
Measurements 305

shown in Fig. 17-6. The lamp DS1 tends to

maintain constant current through the resistor over
a fairly wide power range, so the voltage drop
across the resistor also tends to be constant. This
voltage is applied to the bridge, and with the
constants given is in the right range for resistance¬
type bridges.
To make a measurement, connect the line to J2
Fig. 17-6 — "Power absorber" circuit for use with and apply sufficient rf voltage to JI to give a
resistance-type SWR bridges when the transmitter full-scale incident-voltage reading. R4 may be used
has no special provisions for power reduction. For to set to exactly full scale. Then throw SI to the
rf powers up to 50 watts, DS1 is a 117-volt 40-watt reflected-voltage position and note the meter read¬
incandescent lamp and DS2 is not used. For higher
powers use sufficient additional lamp capacity at ing. The SWR is then found by substituting the
DS2 to load the transmitter to about normal readings in the formula previously given.
output; for example, for 250 watts output DS2 For example, if the full-scale calibration of the
may consist of two 100-watt lamps in parallel. R1 de instrument is 100 microamperes and the reading
is made from three 1-watt 68-ohm resistors con¬ with SI in the reflected-voltage position is 40
nected in parallel. P1 and P2 are cable-mounting microamperes, the SWR is
coaxial connectors.
Leads in the circuit formed by the lamps and 100 + 40 140
R1 should be kept short, but convenient lengths of SWR = = 2.33 to 1
cable may be used between this assembly and the 100-40 60
connectors. Instead of determining the SWR value by calcula¬
tions, the voltage curve of Fig. 17-7 may be used.
The circuit should pass these tests at all In this example the ratio of reflected to forward
frequencies at which it is to be used. It is sufficient voltage is 40/100 = 0.4, and from Fig. 17-7 the
to test at the lowest and highest frequencies, SWR value is seen to be about 2.3 to 1.
usually 3.5 and 28 or 50 MHz. If RI and R2 are The meter scale may be calibrated in any
poorly matched but the bridge construction is arbitrary units so long as the scale has equal
otherwise good, discrepancies in the readings will divisions, since it is the ratios of the voltages, and
be substantially the same at all frequencies. A not the actual values, that determine the SWR.
difference in behavior at the low and high ends of
the frequency range can be attributed to stray Avoiding Errors in SWR Measurements
coupling between bridge arms, or stray inductance The principal causes of inaccuracies within the
or capacitance in the arms. bridge are differences in the resistances of RI and
To check the bridge for balance, apply rf and R2, stray inductance and capacitance in the bridge
adjust R4 for full scale with J2 open. Then connect arms, and stray coupling between arms. If the
a resistor identical with Rs (the resistance should checking procedure described above is followed
match within 1 or 2 percent) to the line terminals, through carefully, the bridge of Fig. 17-4 should be
using the shortest possible leads. It is convenient to amply accurate for practical use. The accuracy is
mount the test resistor inside a cable connector highest for low standing-wave ratios because of the
(PL-259), a method of mounting that also mini¬ nature of the SWR calculation; at high ratios the
mizes lead inductance. When the test resistor is divisor in the equation above represents the differ-
connected, the reflected-voltage reading should
drop to zero. The incident voltage should be reset
to full scale by means of R4, if necessary. The
reading should be zero at any frequency in the
range to be used. If a good null is obtained at low
frequencies but some residual current shows at the
high end, the trouble may be the inductance of the
test resistor leads, although it may also be caused
by stray coupling between the arms of the bridge
itself. If there is a constant low (but not zero)
reading at all frequencies the cause is poor match¬
ing of the resistance values. Both effects can be
present simultaneously. A good null must be
obtained at all frequencies before the bridge is
ready for use.

Bridge Operation
The rf power input to a bridge of this type
must be limited to a few watts at most, because of
the power-dissipation ratings of the resistors. If the
transmitter has no provision for reducing power Fig. 17-7 - Chart for finding voltage standing-wave
output to a very low value - less than 5 watts - a ratio when the ratio of reflected-to-forward voltage
simple “power absorber” circuit can be made up as or reflected-to-forward power is known.
 306
Chapter 17
ence between two nearly equal quantities, so a operating at a higher SWR than it should - and
small error in voltage measurement may mean a would - if the parallel-type current on the outside
considerable difference in the calculated SWR. of the coax were eliminated.
The standard resistor Rs must equal the actual
Zo of the line. The actual Zo of a sample of line Spurious Frequencies
may differ by a few percent from the nominal
figure because of manufacturing variations, but this Off-frequency components in the rf voltage
has to be tolerated. In the 50- to 75-ohm range, the applied to the bridge may cause considerable error.
rf resistance of a composition resistor of 1/2 or 1 The principal components of this type are har¬
watt rating is essentially identical with its de monics and low-frequency subharmonics that may
resistance. be fed through the final stage of the transmitter
driving the bridge. The antenna is almost always a
“Antenna” Currents fairly selective circuit, and even though the system
An explained in Chapter Three, there are two may be operating with a very low SWR at the
ways in which “parallel” or “antenna” currents can desired frequency it is practically always mis¬
be caused to flow on the outside of a coaxial line - matched at harmonic and subharmonic frequen¬
currents induced on the line because of its spatial cies. If such spurious frequencies are applied to the
relationship to the antenna, and currents that bridge in appreciable amplitude, the SWR indi¬
result from the direct connection between the coax cation will be very much in error. In particular, it
outer conductor and (usually) one side of the may not be possible to secure a null on the bridge
antenna. The induced current usually will not be with any set of adjustments of the matching
troublesome if the bridge and the transmitter (or circuit. The only remedy is to filter out the
other source of rf power for operating the bridge) unwanted components by increasing the selectivity
are shielded so that any rf currents flowing on the of the circuits between the transmitter’s final
outside of the line cannot find their way into the amplifier and the bridge.
bridge. This point can be checked by “cutting in”
an additional section of line (1/8 to 1/4 electrical REFLECTOMETERS
wavelength, preferably) of the same Zo. The SWR
Low-cost reflectometers which do not have a
indicated by the bridge should not change except
guaranteed wattmeter calibration are not ordinarily
possibly for a slight decrease because of the
reliable for accurate numerical measurement of
additional line loss, as discussed earlier in Chapter
Three. If there is a marked change, better shielding standing-wave ratio. They are, however, very useful
may be required. as aids in the adjustment of matching networks,
since the objective in such adjustment is to reduce
Parallel-type currents caused by the connection
to the antenna will cause a change in SWR with the reflected voltage or power to zero. Relatively
inexpensive devices can be used for this, since only
line length even though the bridge and transmitter
are well sheilded and the shielding is maintained good bridge balance is required, not actual cali¬
throughout the system by the use of coaxial bration. Bridges of this type are usually “fre¬
fittings. This is because the outside of the coax quency-sensitive” - that is, the meter response
tends to become part of the antenna system, being becomes greater with increasing frequency, for the
connected to the antenna at the feed point, and so same applied voltage. When matching and line
constitutes a load on the line, along with the monitoring, rather than SWR measurement, is the
principal use of the device, this is not a serious
desired load represented by the antenna itself. handicap.
The SWR on the line then is determined by the
composite load of the antenna and the outside of Various simple reflectometers, useful for
the coax, and since changing the line length matching and monitoring, have been described
from time to time in QST and in The Radio
changes one component of this composite load, the
SWR changes too. Amateur’s Handbook. Because most of these are
frequency sensitive, it is difficult to calibrate them
The remedy for such a situation is to use a good
balun or to detune the outside of the line by accurately for SWR measurement, but their low
cost and suitability for use at moderate power
proper choice of length. It is well to note that this
levels, combined with the ability to show accur¬
is not a measurement error, since what the instru¬
ately when a matching circuit has been properly
ment reads is the actual SWR on the line. However,
adjusted, make them a worthwhile addition to the
it is an undesirable condition since the line is amateur station.

AN IN-LINE RF WATTMETER
Considerable attention was devoted to the must be capable of carrying the entire power
resistance-type SWR bridge in the preceding sec¬ output of the transmitter, and should do it with
tion because it is the simplest type that is capable negligible loss. An rf wattmeter meets this require¬
of adequate accuracy in measuring voltage stand¬ ment.
ing-wave ratio. Its disadvantage is that it must be It is neither costly nor difficult to build an rf
operated at a very low power level, and thus is not wattmeter. And, if the instrument is equipped with
suitable for continuous monitoring of the SWR in a few additional components, it can be switched to
actual transmission. To do this the instrument read reflected power as well as forward power.
 307
Measurements

With this feature the instrument can be used as an

SWR meter for antenna matching and Transmatch
adjustments (see Chapters Three and Six). The
wattmeter shown in Figs. 17-8 through 17-11
meets these requirements. The instrument uses a
directional type of coupler for sampling the energy
on the transmission line. The indicator sensitivity
of this instrument is not related to frequency, as is
the case with some types of directional couplers.
This unit may be calibrated for power levels as low
as 1 watt, full scale, in any part of the hf spectrum.
With suitable calibration, it has good accuracy over
the 3-30 MHz range. It is built in two parts, an rf
head for inserting in the coaxial line leaving the
transmitter, and a control/meter box which can be
placed in any location where it can be operated
conveniently. Only direct current flows in the
cable connecting the two pieces.
Design Philosophy
Referring to the circuit of Fig. 17-9, the
transmission-line center conductor passes through
the center of a toroid core and becomes the
primary of Tl. The multiturn winding on the core
functions as the transformer secondary. Current
flowing through the line-wire primary induces a
voltage in the secondary which causes a current to WATTS MI WATTS

flow through resistors RI and R2. The voltage 100

90
200
180
1000
900
drops across these resistors are equal in amplitude, 80 170 800
but 180 degrees out of phase with respect to 70 155 700
common or ground. They are thus, for practical «0 145 600
purposes, respectively in and out of phase with the 50 125 500

line current. Capacitive voltage dividers, C1-C3 and 40 105 400

30 85 300
C2-C4, are connected across the line to obtain 20 65 200
equal-amplitude voltages in phase with the line 10 40 100
voltage, the division ratio being adjusted so that 5 20 50
these voltages match the voltage drops across RI
and R2 in amplitude. (As the current/voltage ratio
in the line depends on the load, this can be done Fig. 17-9 - Schematic diagram of the rf wattmeter.
only for a particular value of load impedance. Load A calibration scale for M1 is shown also. Fixed-
value resistors are 1/2-watt composition. Fixed-
values chosen for this standardization are pure value capacitors are disk oeramic unless otherwise
resistances that match the characteristic impedance noted. Decimal-value capacitances are in pF.
of the transmission line with which the bridge is to Others are pF. Resistance is in ohms; k = 1000.
be used, 50 or 75 ohms usually.) Under these C1, C2 — 1.3- to 6.7-pF miniature trimmer (E. F.
Johnson 189-502-4, available from Newark
Electronics, Chicago, III.).
C3-C11, incl. - Numbered for circuit-board identi¬
fication.
CR1, CR2 — Matched small-signal germanium
diodes, 1N34A, etc. (see text).
J1, J2 - Chassis-mount coax connector of builder's
choice. Type SO-239 used here.
M1 - 0 to 200-pA meter (Triplett type 330-M used
here).
R1, R2 - Matched 10-ohm resistors (see text).
R3, R4 — 5000-ohm printed-circuit carbon control
(IRC R502-B).
R5, R6 — 25,000-ohm printed-circuit carbon con¬
trol (IRC R252-B).
RFC1, RFC2 - 500-pH rf choke (Millen
34300-500 or similar).
S1 — Dpdt single-section phenolic wafer switch
(Mallory 3222J).
Fig. 17-8 - The rf wattmeter consists of two parts, S2 — Spdt phenolic wafer switch (Centralab 1460).
the rf head (left), and the control/meter box T1 — Toroidal transformer; 35 turns of No. 26
(right). The paper scale affixed to the rf head enam. wire to cover entire core of Amidon
contains the calibration information which appears T-68-2 toroid (Amidon Assoc., 12033 Otsego
in Fig. 17-9. St., N. Hollywood, CA 91607).
 308
Chapter 17

conditions, the voltages rectified by CR1 and CR2 a pair of diodes having forward de resistances
represent, in the one case, the vector sum of the within a few ohms of being the same. Similarly, the
voltages caused by the line current and voltage, and back resistances of the diodes can be matched. The
in the other, the vector difference. With respect to matched diodes will help to assure equal meter
the resistance for which the circuit has been set up, readings when the bridge is reversed. (The bridge
the sum is proportional to the forward component should be perfectly bilateral in its performance
of a traveling wave such as occurs on a transmission characteristics.) Germanium diodes should be used
line, and the difference is proportional to the to avoid misleading results when low values of
reflected component. reflected power are present during antenna adjust¬
ments. The SWR can appear to be zero when
Component Selection actually it isn’t. The germanium diodes conduct at
approximately 0.3 volt, making them more suitable
RI and R2 should be selected for the best null for low-power readings than silicon diodes.
reading when adjusting the bridge into a resistive Any meter having a full-scale reading between
50- or 75-ohm load. Normally, the value will be 50 microamperes and 1 milliampere can be used at
somewhere between 10 and 47 ohms. The 10-ohm Ml. The more sensitive the meter, the more
value worked well with the instruments shown difficult it will be to get an absolute reflected-
here. Half-watt composition resistors are suitable power reading of zero. Some residual current will
to 30 MHz. RI and R2 should be as closely flow in the bridge circuit no matter how carefully
matched in resistance as possible. Their exact value the circuit is balanced, and a sensitive instrument
is not critical, so a VTVM can be used to match will detect this current flow. Also, the more
them.
sensitive the meter, the larger will have to be the
Ideally, C3 and C4 should be matched in value. calibrating resistances, R3 through R6, to provide
Silver-mica capacitors are usually close enough in high-power readings. A 0 to 200-microampere
tolerance that special selection is not required, meter represents a good compromise for power
providing there is enough leeway in the ranges of ranges between 100 and 2000 watts.
C 1 and C2 to compensate for any difference in the
values of C3 and C4. Construction
Diodes CR1 and CR2 should also be matched
for best results. An ohmmeter can be used to select It is important that the layout of any rf bridge
be as symmetrical as possible if good balance is to
be had. The circuit-board layout for the instrument
of Fig. 17-11 meets this requirement. The input
and output ports of the equipment should be
isolated from the remainder of the circuit so that
only the sampling circuits feed voltage to the
bridge. A shield across the end of the box which
contains the input and output jacks, and the
interconnecting line between them, is necessary. If
stray rf gets into the bridge circuit it will be
impossible to obtain a complete zero reflected-
power reading on Ml even though a 1:1 SWR
exists.
All of the rf-head components except JI, J2
and the feedthrough capacitors are assembled on
the board. It is held in place by means of a
homemade aluminum L bracket at the end nearest
Tl. The circuit-board end nearest the feedthrough
capacitors is secured with a single No. 6 spade bolt.
Ils hex nut is outside the box, and is used to secure
a solder lug wliich serves as a connection point for
the ground braid in the cable which joins the
control box to the rf head.
Tl fits into a cutout area of the circuit board.
A 1-inch long piece of RG-8/U coax is stripped of
its vinyl jacket and shield braid, and is snug-fit into
the center hole of Tl. The inner conductor is
soldered to the circuit board to complete the
Fig 17-10 - Top view of the rf head for the circuit line-wire connection between J1 and J2.
of Fig. 17-9. A flashing-copper shield isolates the
primary rf line and T1 from the rest of the circuit. The upper dashed lines of Fig. 17-9 represent
The second shield (thicker) is not required and can the shield partition mentioned above. It can be
be eliminated from the circuit. If a 2000-watt scale made from flashing copper or thin brass.
IL desired, ,ixedvalue resistors of approximately The control box, a sloping-panel utility cabinet
22,000 ohms can be connected in series with measuring 4x5 inches, houses SI and S2, and the
high-range printed-circuit controls. Or, the meter, Ml. Four-conductor shielded cable - the
25,000-ohm controls shown here can be replaced shield serving as the common lead - is used to join
by 50,000-ohm units.
the two pieces. There is no reason why the entire
Measurements 309

Fiq 17-11 — Etching pattern and parts layout for the rf wattmeter, as seen from the foil side of the
board. The etched-away portions of the foil are shown as darkened areas in this drawing. The area
represented with diagonal lines is to be cut out for the mounting of T1.

instrument cannot be housed in one container, but FORWARD position, and set SI for the 100-watt
it is sometimes awkward to have coaxial cables range. An rf ammeter or calibrated power meter
attached to a unit that occupies a prominent place should be connected between J2 and the dummy
in the operating position. Built as shown, the load during the tests, to provide power calibration
two-piece instrument permits the rf pickup head to points against which to plot the scale of M1. Apply
be concealed behind the transmitter, while the transmitter output power to JI, gradually, until
control head can be mounted where it is accessible Ml begins to deflect upward. Increase transmitter
to the operator. power and adjust R4 so that a full-scale meter
reading occurs when 100 watts is indicated on the
Adjustment rf ammeter or other standard in use. Next, switch
Perhaps the most difficult task faced by the S2 to REFLECTED and turn the transmitter off.
constructor is that of calibrating the power meter Temporarily short across R3, turn the transmitter
for whatever wattage range he desires to have. The on, and gradually increase power until a meter
least difficult method is to use a commercial reading is noted. With an insulated screwdriver
wattmeter as a standard. If one is not available, the adjust C2 for a null in the meter reading.
power output of the test transmitter can be The next step is to reverse the coax connections
computed by means of an rf ammeter in series with to JI and J2. Place S2 in the REFLECTED
a 50-ohm dummy load, using the standard formula, position and apply transmitter power until the
p = PR. The calibration chart of Fig. 17-9 is meter reads full scale at 100 watts output. In this
representative, but the actual calibration of a mode the REFLECTED position actually reads
particular instrument will depend upon the diodes forward power because the bridge is reversed.
used at CR1 and CR2. Frequently, individual scales Calibrating resistance R3 is set to obtain 100 watts
are required for the two power ranges. full scale during this adjustment. Now, switch S2
Connect a noninductive 50-ohm dummy load to FORWARD and temporarily place a short across
to J2. A Heath Cantenna or similar load will serve R4. Adjust Cl for a null reading on Ml. Repeat the
nicely for adjustment purposes. Place S2 in the foregoing steps until no further improvement can
 310
Chapter 17

be obtained. It will not be necessary to repeat the reflected power could be noted on a commercially
nulling adjustments on the 1000-watt range, but built rf wattmeter. Similarly, the insertion loss was
R5 and R6 will have to be adjusted to provide a so low that it could not be measured with ordinary
full-scale meter reading at 1000 watts. If insuffi¬ instruments.
cient meter deflection is available for nulling
adjustments on the 100-watt range, it may be Operation
necessary to adjust Cl and C2 at some power level
higher than 100 watts. If the capacitors tune It should be remembered that when the bridge
through a null, but the meter will not drop all the is used in a mismatched feed line that has not been
way to zero, chances are that some rf is leaking properly matched at the antenna, a reflected-power
into the bridge circuit through stray coupling. If reading will result. The reflected power must be
so, it may be necessary to experiment with the subtracted from the forward power to obtain the
shielding of the through-line section of the rf head. actual power output. If the instrument is calibrated
If only a small residual reading is noted it will be of for, say, a 50-ohm line, the calibration will not
minor importance and can be ignored. hold for other values of line Zo.
With the component values given in 17-9, the If the instrument is to be used for determining
meter readings track for both power ranges. That SWR, the reflected/forward power ratio can easily
is, the 10-watt level on the 100-watt range, and the be converted into the corresponding voltage ratio
100-watt point on the 1000-watt range, fall at the for use in the formula given earlier. Since power is
same place on the meter scale, and so on. This no proportional to voltage squared, the normalized
doubt results from the fact that the diodes are formula becomes
conducting in the most linear portion of their
curve. Ordinarily, this desirable condition does not VSWR =
exist, making it necessary to plot separate scales 1 -JT
for the different power ranges. where kis the ratio of reflected power to forward
Tests indicate that the SWR caused by insertion power. The power curve of Fig. 17-7 is based on
of the power meter in the transmission line is the above relationship, and may be used in lieu of
negligible. It was checked at 28 MHz and no the equation to determine the SWR.

AN INEXPENSIVE VHF DIRECTIONAL COUPLER

Precision in-line metering devices that are capa¬ loop, not the impedance of the line section. Energy
ble of reading forward and reflected power over a picked up by the loop is rectified by the diode, and
wide range of frequencies are very useful in the resultant current is fed to a meter equipped
amateur vhf and uhf work, but their rather high with a calibration control.
cost puts them out of the reach of many vhf The principal metal parts of the device are a
enthusiasts. The device shown in Figs. 17-12 brass plumbing T, a pipe cap, short pieces of
through 17-15 is an inexpensive adaptation of their 3/4-inch ID and 5/16-inch OD copper pipe, and
basic principles. It can be made for the cost of a two coaxial fittings. Other available tubing com¬
meter, a few small parts, and bits of copper pipe binations for 50-ohm line may be usable. The ratio
and fittings that can be found in the plumbing of outer-conductor ID to inner-conductor OD
stocks at many hardware stores. should be 2.4/1. For a sampler to be used with
other impedances of transmission line, see Chapter
Construction Three for suitable ratios of conductor sizes. The
photographs and Fig. 17-14 show construction
The sampler consists of a short section of details.
hand-made coaxial line, in this instance of 50 ohms Soldering of the large parts can be done with a
impedance, with a reversible probe coupled to it. A 300-watt iron or a small torch. A neat job can be
small pickup loop built into the probe is termin¬ done if the inside of the T and the outside of the
ated with a resistor at one end and a diode at the pipe are tinned before assembling. When the pieces
other. The resistor matches the impedance of the are reheated and pushed together, a good mechan-

Fig. 17-12 — Circuit diagram for the line sampler.

C1 — 500-pF feedthrough capacitor, solder-in type.
C2 — 1000-pF feedthrough capacitor, threaded
type.
CR1 - Germanium diode 1N34, 1N60, 1N270,
1N295, or similar.
J1, J2 - Coaxial connector, type N (UG-58A/U).
L1 — Pickup loop, copper strap 1-inch long X
3/16-inch wide. Bend into "C" shape with flat
portion 5/8-inch long.
M1 — 0-100-uA meter.
R1 — Composition resistor, 82 to 100 ohms. See
text.
R3 — 50,000-ohm composition control, linear
taper.
Measurements 311
to use a similar fitting clamped into a vise as a
holding fixture, with the T assembly resting on
top, held in place by its own weight. Use the
partially prepared center conductor to assure that
the coax connector is concentric with the outer
conductor. After being sure that the ends of the
pipe are cut exactly perpendicular to the axis,
apply heat to the coax fitting, using just enough so
that a smooth fillet of solder can be formed where
the flange and pipe meet.
Before completing the center conductor, check
its length. It should clear the inner surface of the
connector by the thickness of the split ring on the
center pin. File to length; if necessary, slot as with
the other end, and solder the center pin in place.
The fitting can now be soldered onto the pipe, to
complete the 50-ohm line section.
Fig. 17-13 — Major components of the line The probe assembly is made from a 1-1/2-inch
sampler. The brass T and two end sections are at length of the copper pipe, with a pipe cap on the
the rear in this picture. A completed probe top to support the upper feedthrough capacitor,
assembly is at the right. The N connectors have C2. The coupling loop is mounted by means of
their center pins removed. The pins are shown with small Teflon standoffs on a copper disk, cut to fit
one inserted in the left end of the inner conductor inside the pipe. The disk has four small tabs around
and the other lying in the right foreground. the edge for soldering inside the pipe. The diode,
CR1, is connected between one end of the loop
ical and electrical bond will result. If a torch is and a 500-pF feedthrough capacitor, Cl, soldered
used, go easy with the heat, as an over-heated and into the disk. The terminating resistor, RI, is
discolored fitting will not accept solder well. connected between the other end of the loop and
Coaxial connectors with Teflon or other heat- ground, as directly as possible.
resistant insulation are recommended. Type N, When the disk assembly is completed, insert it
with split-ring retainers for the center conductors, into the pipe, apply heat to the outside, and solder
are preferred. Pry the split-ring washers out with a the tabs in place by melting solder into the
knife point or small screwdriver. Don’t lose them, assembly at the tabs. The position of the loop with
as they’ll be needed in the final assembly. respect to the end of the pipe will determine the
The inner conductor is prepared by making sensitivity of a given probe. For power levels up to
eight radial cuts in one end, using a coping saw 200 watts the loop should extend beyond the face
with a fine-toothed blade, to a depth of 1/2 inch. of the pipe about 5/32 inch. For use at higher
The fingers so made are then bent together, power levels the loop should protrude only 3/32
forming a tapered end, as shown in Figs. 17-13 and inch. For operation with very low power levels the
17-14. Solder the center pin of a coaxial fitting best probe position can be determined by exper¬
into this, again being careful not to overheat the iment.
work. The decoupling resistor, R2, and feedthrough
In preparation for soldering the body of the capacitor, C2, can be connected, and the pipe cap
coax connector to the copper pipe, it is convenient put in place. The threaded portion of the capacitor
extends through the cap. Put a solder lug over it
before tightening its nut in place. Fasten the cap
FEEDTHROUGH with two small screws that go into threaded holes
SOLDER LUG
CAPACITOR
in the pipe.
PIPE CAP
Calibration

114-INCH LENGTH
The sampler is very useful for many jobs, even
COPPER PIPE LOCKING SCREW
if it is not accurately calibrated, although it is
desirable to calibrate it against a wattmeter of
3/4-INCH l'A-INCH LENGTH OF
BRASS PIPE T COPPER PIPE, 3/4-INCH I

TEFLON STANDOFF TYPE N COAXIAL

Fig. 17-14 — Cross-section view of

the line sampler. The pickup loop is
\ center SPLIT-RING
supported by two Teflon standoff
SOLDER CONDUCTOR RETAINER insulators. The probe body is secured
s/i* INCH 0 0
PICKUP SOLDER in place with one or more locking
LOOP screws through holes in the brass T.
 312 Chapter 17

Variations

Rather than use one sampler for monitoring

both forward and reflected power by repeatedly
reversing the probe, it is better to make two
assemblies by mounting two T fittings end-to-end,
using one for forward and one for reflected power.
The meter can be switched between the probes, or
two meters can be used.
The sampler described was calibrated at 146
MHz, as it was intended for 2-meter repeater use.
On higher bands the meter reading will be higher
for a given power level, and it will be lower for
lower frequency bands. Calibration for two or
three adjacent bands can be achieved by making
Fig. 17-15 - Two versions of the line sampler. The the probe depth adjustable, with stops or marks to
single unit described in detail here is in the
foreground. Two sections in a single assembly aid in resetting for a given band. And, of course,
provide for monitoring forward and reflected more probes can be made, with each calibrated for
power without probe reversal. a given band, as is done in some of the commer¬
cially available units.
Other sizes of pipe and fittings can be used by
known accuracy. A good 50-ohm dummy load is making use of information given in Chapter Three
required. to select conductor sizes required for the desired
The first step is to adjust the inductance of the impedances. (Since it is occasionally possible to
loop or the value of the terminating resistor, for pick up good bargains in 72-ohm line, a sampler for
lowest reflected-power reading. The loop is the this impedance might be desirable.)
easier to change. Filing it to reduce its width will Type N fittings were used because of their
increase its impedance. Increasing the cross-section constant impedance, and their ease of assembly.
of the loop will lower it, and this can be done by Most have the split-ring retainer, which is simple to
coating it with solder. When the reflected-power use in this application. Some have a crimping
reading is reduced as far as possible, reverse the method, as do apparently all BNC connectors. If a
probe and calibrate for forward power by in¬ fitting must be used that cannot be taken apart,
creasing the transmitter power output in steps and drill a hole large enough to clear a soldering-iron
making a graph of the meter readings obtained. Use tip in the copper-pipe outer conductor. A hole of
the calibration control, R3, to set the maximum up to 3/8-inch diameter will have very little effect
reading. on the operation of the sampler.

RF IMPEDANCE BRIDGE FOR COAX LINES

The bridge shown in Fig. 17-16, dubbed the for a single fixed impedance. A variable ratio in the
“Macromatcher,” may be used to measure un¬ C1-C2 arms is provided by two identical capacitor
known complex impedances. Measured values are sections on the same frame, arranged so that when
of equivalent series form, R + ¡X. With suitable the shaft is rotated to increase the capacitance of
frequency coils, the Macromatcher can be used one section, the capacitance of the other section
throughout the frequency range 1.8 to 30 MHz. decreases. With a fixed value for Rs, the settings of
The useful impedance range of the instrument is the capacitor may be calibrated in terms of
from about 5 to 400 ohms if the unknown load is resistance at Rx.
purely resistive, or 10 to 150 ohms resistive The circuit of Fig. 17-3B is modified slightly
component in the presence of appreciable react¬ for use in the Macromatcher, as shown in Fig.
ance. The reactance range is from 0 to approxi¬ 17-17. The differential capacitor is retained for Cl
mately 100 ohms for either inductive or capacitive and C2 to measure resistance. LI and C3 have been
loads. Although the Macromatcher cannot indicate added in series in the “unknown” arm of the
impedances with the accuracy of a laboratory bridge, and it is these components which are used
bridge, its readings are quite adequate for most to measure the amount and type of reactance at
amateur uses, including the taking of line lengths the unknown load. (Both LI and C3 are adjustable
into account with a Smith Chart. By inherent in the actual bridge circuit.) The Macromatcher is
properties of the Macromatcher, its accuracy is initially balanced at the frequency of measurement
best at the centers of the dial calibration ranges. with a pure resistance in place of R2, Fig. 17-17, so
The Basic Bridge Circuit that the reactances of LI and of C3 at its
midsetting are equal. Thus, these reactances cancel
The basic circuit of the Macromatcher is that of each other in this arm of the bridge, and no
Fig. 17-3B. If Cl and C2 of that circuit are the reactance is reflected into the remaining bridge
sections of a differential capacitor, the bridge may arms. For measurement, an unknown complex¬
be used over a wide impedance range, rather than impedance load is then connected into the bridge
Measurements 313
by dividing the reactance dial reading by the
measurement frequency in megahertz.
Fig. 17-18 is the complete schematic diagram of
the Macromatcher. Cl is the resistance-measuring
capacitor, and LI and C2 the reactance-measuring
components. RI is the bridge “standard” resistor.
Aside from the INPUT and OUTPUT jacks and J2,
the connector for LI, all other parts are associated
with the null-detector metering section of the
circuit. This portion of the circuit is adopted from
that of the resistance bridge described earlier in
this chapter, and the discussion and precautions
which pertain to that instrument in general apply
here, as well.
CR 1 rectifies rf energy present when the bridge
Fig. 17-16 — An RCL bridge for measuring is unbalanced, and this energy is filtered into direct
unknown values of complex impedances. A plug-in current which is metered at Ml. The 12-kil resistor
coil is used for each frequency band. The bridge provides a high-impedance input for the metering
operates at an rf input level of about 5 volts; circuit, and the 4700-ohm resistor at JI provides a
pickup-link assemblies for use with a grid-dip return path for meter-current flow if the input
oscillator are shown. Before measurements are
made, the bridge must be balanced with a non-
source is capacitance coupled. J4 is for the
reactive load connected at its measurement ter¬ connection of an external meter, in the event it is
minals. This load consists of a resistor mounted desired to observe readings remotely. CR2, placed
inside a coaxial plug, shown in front of the directly across Ml, protects the meter from over¬
instrument at the left. The aluminum box measures current surges. Although it appears from the
4-1/4 X 10-3/4 X 6-1/8 inches and is fitted with a schematic diagram that this germanium diode will
carrying handle on the left end and self-sticking shunt out all meter current, such is not the case in
rubber feet on the right end and bottom. Dials are actual operation because approximately 250 milli¬
Millen No. 10009 with skirts reversed and calibra¬ volts must be developed across the diode before it
tions added.

in place of R2. The resistive component of the load

is balanced by varying the C1-C2 ratio. The
reactive component is balanced by varying C3
either to increase or decrease its capacitive react¬
ance, as required, to cancel any reactance present
in the load. If the load is inductive, more capacitive
reactance (less capacitance) is required from C3 to
obtain a balance. Less reactance (more capaci¬
tance) is needed from C3 if the load is capacitive.
The end result, after C3 is properly adjusted for
the particular unknown load, is that the overall R2
arm of the bridge again looks purely resistive, and a
complete null is obtained on the null detector.
The settings of C3 are calibrated in terms of the
value and type of reactance at the load terminals.
Because of the relationship of capacitive reactance
to frequency, the calibration for the dial of the
reactance-measuring capacitor is valid at only one Fig. 17-18 — Schematic diagram of the Macro¬
frequency. It is therefore convenient to calibrate matcher. Capacitance is in microfarads; resistance
this dial for equivalent reactances at 1 MHz. is in ohms, k = 1000. Resistors are 1/2-W 10-
Frequency corrections may then be made simply percent tolerance unless otherwise indicated.
C1 — Differential capacitor, 11-161 pF per section
(Millen 28801 or Jackson Bros. C702/5301
suitable).
C2 — 17.5-327 pF with straight-line capacitance
characteristic (Millen 19335 or Jackson Bros.
C9/5070 suitable).
CR1, CR2 — Germanium diode, high back resis¬
tance.
J1, J3 — Coaxial connector, chassis type.
J2 — To mate plug of L1, ceramic.
J4 — Phone jack, closed-circuit type.
L1 — See text and Table 17-1.
M1 — 0-50 pA de (Simpson Model 1223 Bold-vue,
Fig. 17-17 — The basic circuit of the Macro¬ Cat. No. 15560 used here).
matcher. In this circuit the bridge is balanced R1 — For text reference.
before measurements are made, by setting RFC1 — Subminiature rf choke (Miller 70F103AI
= Xc3- or equiv.l.
314 Chapter 17

reactances. Placement of component parts, while

not critical, must be such that lead lengths greater
than about 1/2 inch (except in the de metering
circuit) are avoided. Shorter leads are desirable,
especially for RI, the standard resistance for the
bridge. In the unit photographed, the body of this
resistor just fits between the terminals of Cl and
J2 where it is connected. Cl should be enclosed in
a shield and connections made with leads passing
through holes drilled through the shield wall. The
frames of both variable capacitors, Cl and C2,
must be insulated from the chassis, such as on
ceramic pillars, with insulated couplings used on
the shafts. As Fig. 17-19 indicates, all parts of the
bridge except the meter and the calibrated dials are
Fig. 17-19 - All components except the meter are mounted on the top panel of the box. The dials are
mounted on the top of the box. C1 is visible inside front-panel mounted on shafts with panel bearings.
the homemade shield at the left, with C2 at the Band-switching arrangements for LI complicate
right and J2 mounted between them. J1 is hidden the construction and contribute to stray reactances
beneath C1 in this view; a part of J3 may be seen in the bridge circuit. For these reasons plug-in coils
in the lower right corner of the box. Components are used at LI, one coil for each band over which
for the de metering circuit are mounted on a the instrument is used. The coils must be adjust¬
tie-point strip which is affixed to the shield wall
for C1; all other components are interconnected able, to permit initial balancing of the bridge with
with very short leads. The 4700-ohm input resistor C2 set at the zero-reactance calibration point. Coil
is connected across J1. This photograph was made data are given in Table 17-1. Millen 45004 coil
before the diode was connected across the ter¬ forms, with the coils supported inside, provide a
minals of M1. convenient method of constructing these slug-
tuned plug-in coils. A phenolic washer cut to the
begins to conduct an appreciable amount of proper diameter is epoxied to the top or open end
current. The internal resistance of a typical 50-gA of each form, giving a rigid support for mounting
meter is 1800 or 2000 ohms, and this means that of the coil by its bushing. Small knobs for 1/8-inch
more than 100 pA of current must be flowing shafts, threaded with a No. 6-32 tap, are screwed
through the meter before the diode shunting effect onto the coil slug-tuning screws to permit ease of
becomes appreciable. In operation, this diode adjustment without a tuning tool. Knobs with
prevents the meter needle from slamming against setscrews should be used to prevent slipping. A
the peg if the load is disconnected while input rf ceramic socket to mate with the pins of the coil
excitation is still applied; the needle eventually form is used for J2.
reaches full scale, but travels more slowly with the
diode in the circuit. A Nonreactive Termination
Construction For calibrating the reactance dial and for
initially balancing the Macromatcher each time it is
In any rf-bridge type of instrument, the leads used on a new frequency, a purely resistive load is
must be kept as short as possible to reduce stray required for connection at J3. A suitable load

TABLE 17-1 - COIL DATA FOR RF IMPEDANCE BRIDGE

Nominal Frequency
inductance Coverage,
Band Range, pH MHz Qoii Type or Data
160 27.5-58 1-6-2.3 Miller 42A475CBI.
80 6.5-13.8 3-2-4.8 28 turns No. 30 enam. wire close-wound on Miller
form 42A000CBI.
40 2.0-4.4 5.8-8.5 Miller 42A336CBI or 16 turns No. 22 enam. wire
close-wound on Miller form 42A000CBI.
20 0.6-1.1 11.5-16.6 8 turns No. 18 enam. wire close-wound on Miller
form 42A000CBI.
15 0.3-0.48 18.5-23.5 4 1/2 turns No. 18 enam. wire close-wound on
Miller form 42A000CBI.
10 0.18-0.28 25.8-32.0 3 turns No. 16 or 18 enam. or tinned bus wire
spaced over 1/4-inch winding length on Miller form
42A000CBI.
Measurements 315
which is essentially nonreactive can be made by
mounting a 51- or a 56-ohm 1-W composition
(carbon) resistor inside a PL-259 plug.
The body of the resistor should be inserted as
far as possible into the plug, with one resistor lead
extending through the center-conductor pin. Sol¬
der this center-pin connection, and clip off any
excess lead length. Make a l/2-in.-dia copper or
brass disk with a small hole at its center. Use a
1/16-in. or, preferably, a No. 60 drill to make this
hole. (Initially the “disk” may be a square or
rough-cut piece of metal. It may be rounded by
filing or grinding after the assembly process is
completed.) Place the shell of the plug over its
body, and then slip the disk over the grounded-end
lead of the resistor, so the resistor lead protrudes
through the small hole. First solder the disk to the
Fig. 17-20 — Calibration scale for the reactance
body of the plug and then clip off any excess lead dial used at C2 of the Macromatcher. See text.
length from the resistor. Next, solder the con¬
nection at the small hole. The disk, when assem¬
bled in this manner, completes the shielding,
reduces lead inductance, and also prevents the shell the frequency in MHz to obtain the calibration
of the plug from being removed completely. value for the dial.

Using the Impedance Bridge

Calibration
This instrument is a low-input-power device,
The resistance dial of the bridge may be and is not of the type to be excited from a
calibrated by using a number of 1/2- or 1-watt transmitter or left in the antenna line during
5-percent-tolerance composition resistors of differ¬ station operation. Sufficient sensitivity for all
ent values in the 5- to 400-ohm range as loads. For measurements results when a 5-V rms rf signal is
this calibration, the appropriate frequency coil applied at JI. This amount of voltage can be
must be inserted at J 2 and its inductance adjusted delivered by most grid-dip oscillators. In no case
for the best null reading on the meter when C2 is should the power applied to JI exceed 1 watt or
set with its plates half meshed. For each test damage to the instrument may result. The input
resistor, Cl is then adjusted for a null reading. impedance of the bridge at JI is low, in the order
Alternate adjustment of LI and Cl should be made of 50 to 100 ohms, so it is convenient to excite
for a complete null. The leads between the test the bridge through a length of 52- or 75-ohm line
resistor and J 3 should be as short as possible, and such as RG-58/U or RG-59/U. If a grid-dip
the calibration preferably should be done in the oscillator is used, a link coupling arrangement to
3.5-MHz band where stray inductance and capa¬ the oscillator coil may be used. Fig. 17-16 shows
citance will have the least effect. two pickup link assemblies. The larger coil, 10
If the constructional layout of the bridge turns of 1-1/4-inch-dia stock with turns spaced at 8
closely follows that shown in the photographs, the turns per inch (B&W 3018), is used for the 160-,
calibration scale of Fig. 17-20 may be used for the 80-, 40- and 20-meter bands. The smaller coil, 5
reactance dial. This calibration was obtained by turns of 1-inch-dia stock with turns spaced at 4
connecting various reactances, measured on a turns per inch (B&W 3013), is used for the 15- and
laboratory bridge, in series with a 47-ohm 1-W 10-meter bands. Coupling to the oscillator should
resistor connected at J3. The scale is applied so be as light as possible, while obtaining sufficient
that maximum capacitive reactance is indicated sensitivity, to prevent severe “pulling” of the
with C2 fully meshed. oscillator frequency. Overcoupling may cause the
If it is desired to obtain an individual calibra¬ oscillator to shift in frequency by a few hundred
tion for C2, known values of inductance and kilohertz, so for the most reliable measurements, a
capacitance may be used in series with a fixed receiver should be used to check the oscillator
resistor of the same approximate value as RI. For frequency.
this calibration it is very important to keep the Before measurements are made, it is necessary
leads to the test components as short as possible, to balance the bridge. Set the reactance dial at zero
and calibration should be performed in the and adjust LI and Cl for a null with a nonreactive
3.5-MHz range to minimize the effects of stray load connected at J3. This null should be com¬
reactances. Begin the calibration by setting C2 at plete; if not, reduce the signal level being applied
half mesh, marking this point as 0 ohms reactance. to the Macromatcher. The instrument must be
With a purely resistive load connected at J 3, adjust rebalanced after any appreciable change is made in
LI and Cl for the best null on Ml. From this point the measurement frequency, more than approxi¬
on during calibration, do not adjust LI except to mately 1 percent. After the bridge is balanced,
rebalance the bridge for a new calibration fre¬ connect the unknown load to J 3 and alternately
quency. The ohmic value of the known reactance adjust Cl and C2 for the best null. Measured
for the frequency of calibration is multiplied by impedances are of equivalent series form, R + jX/f,
316 Chapter 17

where R and X are the Macromatcher dial readings, one adjustable element affects primarily the resis¬
and f is the frequency in megahertz. When the tive component (the rod length of the gamma or
reactive component, X, is divided by the frequen¬ the physical length of the hairpin), while the other
cy, the result is R + ¡X in ohms. adjustment affects primarily the reactive compo¬
As shown in Fig. 17-20, the calibration of the nent (gamma-capacitor setting or driven-element
reactance dial is nonlinear, with a maximum length with the hairpin match). Of course there is
indication for capacitive reactance of 500//. The some amount of interaction in the two adjust¬
measurement range for capacitive loads may be ments, but the effects of making just one adjust¬
extended by “zeroing" the reactance dial at some ment can be seen immediately on the Macro¬
value other than 0. For example, if the bridge is matcher. Obtaining an acceptable match in a
initially balanced with the reactance dial set at 500 matter of a few minutes is simple - adjust one of
in the Xl range, the 0 dial indication is now the two variables for the proper resistance, adjust
equivalent to an Xc reading of 500//, and the total the other variable for zero reactance, perform a
range of measurement for Xc has been extended to slight touchup on these adjustments, and the job is
1000//. finished.
When the Macromatcher is used at the antenna, Of course it is not necessary to use the
excitation may be “piped” to the instrument Macromatcher at the load to determine the im¬
through the coaxial line which normally feeds the pedance. Measurements may be performed through
antenna. Unless an assistant can check the oscil¬ an electrical half wavelength of feed line. Dis¬
lator frequency during each measurement, how¬ regarding attenuation (which will be negligible if
ever, a grid-dip oscillator is unsatisfactory for this the line is only a single half wave in length) the
type of work. A more stable frequency source, impedance will be the same at the input end of the
such as a signal generator or QRP transmitter line as it is at the load, no matter what the line
capable of delivering approximately 100 to 200 impedance may be. Nor is it necessary to trim the
milliwatts, is ideal, as it can be left running during coaxial line to an exact half wavelength (good for a
the time measurements and adjustments are being single frequency only) in order to make “remote”
made. (Alternatively, the “power absorber” circuit measurements accurately. The line may be of any
of Fig. 17-6 may be used with higher power convenient physical length, but its electrical length
transmitters.) Here is where the Macromatcher can must be known. Readings taken at the input end of
really be of value for adjustment of matching the line can be converted into actual impedances at
networks such as the gamma, and hairpin, the termination point of the line by means of a
because the resistive and reactive components of Smith Chart, as described in Chapter Three. Line
the load are indicated separately. In these networks attenuation may also be taken into account.

ANTENNA MEASUREMENTS
Of all the measurements made in amateur radio between a suitable feed line and the environment
communications systems, perhaps the most diff¬ surrounding it. In addition to efficient transfer of
icult and least understood is the measurement of power from feed line to environment, an antenna
antennas. For example, it is relatively easy to at vhf or uhf is most frequently required to
measure the frequency and cw power output of a concentrate the radiated power into a particular
transmitter, the response of a filter, or the gain of region of the environment.
an amplifier. These are all what might be called In order to be consistent in comparing different
bench measurements because, when performed antennas, it is necessary that the environment
properly, all the factors which influence the surrounding the antenna be standarized. Ideally,
accuracy and success of the measurement are under measurements should be made with the measured
control. In making antenna measurements how¬ antenna so far removed from any objects causing
ever, the “bench” is now perhaps the backyard. In environmental effects that it is literally in outer
other words, the environment surrounding the space - a very impractical situation. The purpose
antenna can affect the results of the measurement. of the measurement techniques is therefore to
Control of the environment is not at all as simple simulate, under practical conditions, a controlled
as it was for the bench measurement, because now environment. At vhf and uhf, and with practical-
the work area may be rather spacious. This section size antennas, the environment can be controlled
describes antenna measurement techniques which so that successful and accurate measurements can
are closely allied to those used in an antenna be made in a reasonable amount of space.
measuring event or contest. With these procedures The electrical characteristics of an antenna
the measurements can be made successfully and which are most desirable to obtain by direct
with meaningful results. These techniques should measurement are: (1) gain (relative to an isotropic
provide a better understanding of the measurement source, which by definition has a gain of unity);
problems, resulting in a more accurate and less (2) space-radiation pattern; (3) feed-point imped¬
difficult task. (Information provided by Dick ance (mismatch) and (4) polarization.
Turrin, W2IMU.)
Polarization
SOME BASIC IDEAS In general the polarization can be assumed from
the geometry of the radiating elements. That is to
An antenna is simply a transducer or coupler say, if the antenna is made up of a number of
Measurements 317

linear elements (straight lengths of rod or wire losses may be determined from information con¬
which are resonant and connected to the feed tained in Chapter Three.
point) the polarization of the electric field will be While on the subject of feed-point impedance,
linear and polarized parallel to the elements. If the mention should be made of the use of baluns in
elements are not consistently parallel with each antennas. A balun is simply a device which permits
other, then the polarization cannot easily be a lossless transition between a balanced system -
assumed. The following techniques are directed to feed line or antenna - and an unbalanced feed line
antennas having polarization that is essentially or system. If the feed point of an antenna is
linear (in one plane), although the method can be symmetric such as with a dipole and it is desired to
extended to include all forms of elliptic polariza¬ feed this antenna with an unbalanced feed line
tion. such as coax, it is necessary to provide a balun
between the line and the feed point. Without the
balun, current will be allowed to flow on the
Feed-Point Mismatch outside of the coax. The current on the outside of
The feed-point mismatch, although affected to the feed line will cause radiation and thus the feed
some degree by the immediate environment of the line becomes part of the antenna radiation system.
antenna, does not affect the gain or radiation In the case of beam antennas where it is desired to
characteristics of an antenna. If the immediate concentrate the radiated energy in a specific
environment of the antenna does not affect the direction, this extra radiation from the feed line
feed-point impedance, then any mismatch intrinsic will be detrimental, causing distortion of the
to the antenna tuning reflects a portion of the expected antenna pattern.
incident power back to the source. In a receiving
antenna this reflected power is reradiated back into ANTENNA TEST SITE
the environment, “free space,” and can be lost SET-UP AND EVALUATION
entirely, in a transmitting antenna, the reflected
power goes back to the final amplifier of the Since an antenna is a reciprocal device, meas¬
transmitter. In general an amplifier is not a urements of gain and radiation patterns can be
matched source to the feed line, and, if the feed made with the test antenna used either as a
line has very low loss, the amplifier output controls transmitting or as a receiving antenna. In general
are customarily altered during the normal tuning and for practical reasons, the test antenna is used
procedure to obtain maximum power transfer to in the receiving mode, and the source or trans¬
the antenna. Tire power which has been reflected mitting antenna is located at a specified fixed
from the antenna combines with the source power remote site and unattended. In other words the
to travel again to the antenna. This procedure is source antenna, energized by a suitable transmitter,
called conjugate matching, and the feed line is now is simply required to illuminate or flood the
part of a resonant system consisting of the mis¬ receiving site in a controlled and constant manner.
matched antenna, feed line, and amplifier tuning As mentioned earlier, antenna measurements
circuits. It is therefore possible to use a mis¬ ideally should be made under “free-space” con¬
matched antenna to its full gain potential, provided ditions. A further restriction is that the illumina¬
the mismatch is not so severe as to cause heating tion from the source antenna be a plane wave over
losses in the system, especially the feed line and the effective aperture (capture area) of the test
matching devices. Similarly, a mismatched re¬ antenna A plane wave by definition is one in
ceiving antenna may be conjugately matched into which the magnitude and phase of the fields are
the receiver front end for maximum power trans¬ uniform, and in the test-antenna situation, uniform
fer. In any case it should be clearly kept in mind over the effective area plane of the test antenna.
that the feed-point mismatch does not affect the Since it is the nature of all radiation to expand in a
radiation characteristics of an antenna. It can only spherical manner at great distance from the source,
affect the system efficiency wherein heating losses it would seem to be most desirable to locate the
are concerned. source antenna as far from the test site as possible.
Why then do we include feed-point mismatch as However, since for practical reasons the test site
part of the antenna characteristics? The reason is and source location will have to be near the earth
that for efficient system performance most an¬ and not in outer space, the environment must
tennas are resonant transducers and present a include the effects of the ground surface and other
reasonable match over a relatively narrow frequen¬ obstacles in the vicinity of both antennas. These
cy range. It is therefore desirable to design an effects almost always dictate that the test range
antenna, whether it be a simple dipole or an array (spacing between source and test antennas) be as
of Yagis, such that the final single feed-point short as possible consistent with maintaining a
impedance be essentially resistive and of magnitude nearly error-free plane wave illuminating the test
consistent with the impedance of the feed line aperture.
which is to be used. Furthermore, in order to make A nearly error-free plane wave can be specified
accurate, absolute gain measurements, it is vital as one in which the phase and amplitude, from
that the antenna under test accept all the power center to edge of the illuminating field over the
from a matched-source generator, or that the test aperture, do not deviate by more than about 30
reflected power caused by the mismatch be meas¬ degrees and 1 decibel, respectively. These con¬
ured and a suitable error correction for heating ditions will result in a gain-measurement error of
losses be included in the gain calculations. Heating no more than a few percent less than the true gain.
 318 Chapter 17

greater than the highest gain antenna to be

measured. For gain measurements the range con¬
sists essentially of the region in the beam of the
test antenna. For radiation-pattern measurements,
the range is considerably larger and consists of all
that area illuminated by the source antenna,
especially around and behind the test site. Ideally a
Fig. 17-21 — On an antenna test range, energy site should be chosen where the test-antenna
reaching the receiving equipment may arrive after location is near the center of a large open area and
being reflected from the surface of the ground, as the source antenna located near the edge where
well as by the direct path. The two waves may tend most of the obstacles (trees, poles, fences, etc.) lie.
to cancel each other, or may reinforce one another, The primary effect of the range surface is that
depending on their phase relationship at the some of the energy from the source antenna will be
receiving point. reflected into the test antenna, while other energy
will arrive on a direct line-of-sight path. This is
illustrated in Fig. 17-21. The use of a flat, uniform
Based on the 30-degree phase error alone, it can be ground surface assures that there will be essentially
shown that the minimum range distance is approxi¬ a mirror reflection even though the reflected
mately energy may be slightly weakened (absorbed) by the
surface material (ground). In order to perform an
analysis it is necessary to realize that horizontally
polarized waves undergo a 180-degree phase rever¬
where D is the largest aperture dimension and X is sal upon reflection from the earth. The resulting
the free-space wavelength in the same units as D. illumination amplitude at any point in the test
The phase error over the aperture D for this aperture is the vector sum of the electric fields
condition is 1/16 wavelength.
arriving from the two directions, the direct path
Since aperture size and gain are related by and the reflected path. If a perfect mirror re¬
_ . 4tr/le flection is assumed from the ground (it is nearly
Gain =--- that for practical ground conditions at vhf/uhÓ
X2 and the source antenna is isotropic, radiating
where Ae is the effective aperture area, the equally in all directions, then a simple geometric
dimension D may be obtained for simple aperture analysis of the two path lengths will show that at
configurations. For a square aperture various points in the vertical plane at the test¬
antenna site the waves will combine in different
phase relationships. At some points the arriving
4k
waves will be in phase, and at other points they
which results in a minimum range distance for a will be 180 degrees out of phase. Since the field
square aperture of amplitudes are nearly equal, the resulting phase
change due to path length difference will produce
Smin = G —— an amplitude variation in the vertical test site
2 k
and for a circular aperture of direction similar to a standing wave, as shown in
Fig. 17-22.
s -- r
¿mm U—-—
2x The simplified formula relating the location of
7T h2 for maximum and minimum values of the
For apertures with a physical area that is not two-path summation in terms of hi and 5 is
well defined or is much larger in one dimension
than in other directions, such as a long thin array
for maximum directivity in one plane, it is ad¬
visable to use the maximum estimate of D from
either the expected gain or physical aperture with n= 0, 2,4 ... for minimums and
dimensions. n = 1, 3, 5 ... for maximums, and Sis
Up to this point in the range development, only much larger than either hi or/t2-
the conditions for minimum range length, Smin, The significance of this simple ground re¬
have been established, as though the ground flection formula is that it permits the approximate
surface were not present. This minimum Sis location of the source antenna to be determined to
therefore a necessary condition even under “free- achieve a nearly plane-wave amplitude distribution
space” environment. The presence of the ground in the vertical direction over a particular test
further complicates the range selection, not in the aperture size. It should be clear from examination
determination of S but in the exact location of the of the height formula that as/ti is decreased, the
source and test antennas above the earth. vertical distribution pattern of signal at the test
It is always advisable to select a range whose site, h2, expands. Also note that the signal level for
intervening terrain is essentially flat, clear of h2 equal to zero is always zero on the ground
obstructions, and of uniform surface conditions, regardless of the height of hi.
such as all grass or all pavement. The extent of the The objective in using the height formula then
range is determined by the illumination of the is, given an effective antenna aperture to be
source antenna, usually a beam, whose gain is no illuminated from which a minimum S (range
Measurements 319

An example will illustrate the procedure. As¬

sume that we wish to measure a 7-foot-diameter
parabolic reflector antenna at 1296 MHz
(X = 0.75 foot). The minimum range distance,
•5min, can be readily computed from the formula
for a circular aperture.
D2 » 49
5min - 2 - 2 X - 130 feet

Now a suitable site is selected based on the

qualitative discussion given before.
Next determine the source height, hi. The
procedure is to choose a height hi such that the
first minimum above ground (n = 2 in formula) is
Fig. 17-22 — The vertical profile, or plot of signal at least two or three times the aperture size, or
strength versus test-antenna height, for a fixed about 20 feet.
height of the signal source above ground and at a
fixed distance. See text for definitions of symbols. X .5 0.75 130
Ai = " 4v «2 2 X-X —— = 2.4 feet
4 20
length) is determined and a suitable range site
chosen, to find a value for hi (source antenna Place the source antenna at this height and probe
height). The required value is such that the first the vertical distribution over the seven-foot aper¬
maximum of vertical distribution at the test site, ture location, which will be about ten feet off the
/t 2, is at a practical distance above the ground and ground.
at the same time the signal amplitude over the
, X S - i V 0.75 130 _ . „ , -
h2 = n Th^ ”'
aperture in the vertical direction does not vary
more than about 1 dB. This last condition is not T- T4 " 10 2 feet
sacred but is closely related to the particular
antenna under test. In practice these formulas are
The measured profile of vertical signal level versus
only useful to initialize the range setup. A final
height should be plotted. From this plot, empiri¬
check of the vertical distribution at the test site cally determine whether the seven-foot aperture
must be made by direct measurement. This can be fitted in this profile such that the 1-dB
measurement should be conducted with a small variation is not exceeded. If the variation exceeds 1
low-gain but unidirectional probe antenna such as a dB over the seven-foot aperture, the source an¬
corner reflector or 2 element Yagi which is moved tenna should be lowered and h2 raised. Small
along a vertical line over the intended aperture site. changes in hi can quickly alter the distribution at
Care should be exercised to minimize the effects of the test site. Fig. 17-23 illustrates the points of the
local environment around the probe antenna and previous discussion.
that the beam of the probe be directed at the The same set-up procedure applies for either
source antenna at all times for maximum signal. A horizontal or vertical linear polarization. However,
simple dipole is undesirable as a probe antenna it is advisable to check by direct measurement at
because it is susceptible to local environmental the site for each polarization to be sure that the
effects. vertical distribution is satisfactory. Distribution
The most practical way to instrument the probing in the horizontal plane is unnecessary as
vertical distribution measurement is to construct little or no variation in amplitude should be found,
some kind of vertical track, preferably of wood, since the reflection geometry is constant. Because
with a sliding carriage or platform which may be
used to support and move the probe antenna. It is
assumed of course that a stable source transmitter
and calibrated receiver or detector are available so
that variations of the order of 1/2 dB can be
clearly distinguished.
Once these initial range measurements are
completed successfully, the range is now ready to
accommodate any aperture size less in vertical
extent than the largest for which Smin and the
vertical field distribution were selected. The test
antenna is placed with the center of its aperture at
the height /12 where maximum signal was found.
The test antenna should be tilted so that its main
beam is pointed in the direction of the source
antenna. The final tilt is found by observing the
receiver output for maximum signal. This last
process must be done empirically since the appar¬
ent location of the source is somewhere between Fig. 17-23 — Sample plot of a measured vertical
the actual source and its image, below the ground. profile.
320 Chapter 17

of this, antennas with apertures which are long and antenna in its optimum location is noted. Then the
thin, such as a stacked collinear vertical, should be test antenna is removed and the standard-gain
measured with the long dimension parallel to the antenna is placed with its aperture at the center of
ground. location where the test antenna was located. The
A particularly difficult range problem occurs in difference in signal level between the standard and
measurements of antennas which have depth as the test antennas is measured and appropriately
well as cross-sectional aperture area. Long end-fire added to or subtracted from the gain of the
antennas such as long Yagis, rhombics, V-beams, or standard-gain antenna to obtain the absolute gain
arrays of these antennas, radiate as volumetric- of the test antenna, absolute here meaning with
arrays and it is therefore even more essential that respect to a point source which has a gain of unity
the illuminating field from the source antenna be by definition. The reason for using this reference
reasonably uniform in depth as well as plane wave rather than a dipole, for instance, is that it is more
in cross section. For measuring these types of useful and convenient for system engineering. It is
antennas it is advisable to make several vertical assumed that both standard and test antennas have
profile measurements which cover the depth of the been carefully matched to the appropriate imped¬
array. A qualitative check on the integrity of the ance and an accurately calibrated and matched
illumination for long end-fire antennas can be detecting device is being used.
made by moving the array or antenna axially A standard-gain antenna may be any type of
(forward and backward) and noting the change in unidirectional, preferably planar-aperture, antenna,
received signal level. If the signal level varies less which has been calibrated either by direct measure¬
than 1 or 2 dB for an axial movement of several ment or in special cases by accurate construction
wavelengths then the field can be considered according to computed dimensions.
satisfactory for most demands on accuracy. Large A standard-gain antenna has been suggested by
variations indicate that the illuminating field is the National Bureau of Standards. It consists of
badly distorted over the array depth and sub¬ two in-phase dipoles one half wavelength apart and
sequent measurements are questionable. It is inter¬ backed up with a ground plane one wavelength
esting to note in connection with gain measure¬ square. Such an antenna is shown in Fig. 17-24.
ments that any illuminating field distortion will When constructed accurately to scale for the
always result in measurements which are lower frequency of interest, this type of standard will
than true values. have an absolute gain of 7.7 dB with an accuracy
of plus or minus .25 dB.
ABSOLUTE GAIN MEASUREMENT
Having established a suitable range, the meas¬ RADIATION-PATTERN MEASUREMENTS
urement of gain relative to an isotropic (point
source) radiator is almost always accomplished by Of all antenna measurements, the radiation
direct comparison with a calibrated standard-gain pattern is the most demanding in measurement and
antenna. That is, the signal level with the test most difficult to interpret. Any antenna radiates to
some degree in all directions into the space
surrounding it. Therefore, the radiation pattern of
an antenna is a three-dimensional representation of
the magnitude, phase and polarization. In general,
and in practical cases for amateur radio communi¬
cations, the polarization is well defined and only
the magnitude of radiation is important. Further¬
more, in many of these cases the radiation in one
particular plane is of primary interest, usually the
plane corresponding to that of the earth’s surface,
regardless of polarization.
Because of the nature of the range set-up,
measurement of radiation pattern can only be
successfully made in a plane nearly parallel to the
earth’s surface. With beam antennas it is advisable
and usually sufficient to take two radiation pattern
measurements, one in the polarization plane and
one at right angles to the plane of polarization.
These radiation patterns are referred to in antenna
literature as the principal f-plane and //-plane
patterns, respectively, E plane meaning parallel
Fig. 17-24 - Standard-gain antenna suggested by
National Bureau of Standards. When accurately to the electric field which is the polarization
constructed for the desired frequency, this antenna plane and // plane meaning parallel to the
will exhibit a gain of 7.7 dB over a half-wave dipole magnetic field. The electric field and magnetic field
radiator, plus or minus 0.25 dB. In this model, are always perpendicular to each other in a plane
constructed for 432 MHz, the elements are 3/8- wave as it propagates through space.
inch dia tubing. The phasing and support lines are The technique in obtaining these patterns is
of 5/16-inch dia tubing or rod. simple in procedure but requires more equipment
Measurements 321

Fig. 17-25 — Sample plot of a measured radiation pattern, using techniques described in the text.

or patience than does making a gain measurement. use are they?” The primary answer is as a
First, a suitable mount is required which can be diagnostic tool to determine if the antenna is
rotated in the azimuth plane (horizontal) with functioning as it was intended to. A second answer
some degree of accuracy in terms of azimuth angle is to know how the antenna will discriminate
positioning. Second, a signal-level indicator cali¬ against interfering signals from various directions.
brated over at least a 20-dB dynamic range with a Consider now the diagnostic use of the radia¬
readout resolution of at least 2 dB is required. A tion patterns. If the radiation beam is well defined,
dynamic range of up to about 40 dB would be then there is an approximate formula relating the
desirable but does not add greatly to the measure¬ antenna gain to the measured half-power beam¬
ment significance. width of the E- and //-plane radiation patterns. The
With this much equipment, the procedure is to half-power beamwidth is indicated on the polar
locate first the area of maximum radiation from plot where the radiation level falls to 3 dB below
the beam antenna by carefully adjusting the the main beam 0-dB reference on either side. The
azimuth and elevation positioning. These settings formula is
are than arbitrarily assigned an azimuth angle of 40,000
zero degrees and a signal level of zero decibels. Gain « -
Next, without changing the elevation setting (tilt 0E X0H
of the rotating axis), the antenna is carefully where 6 e and 0h are the half-power beamwidths in
rotated in azimuth in small steps which permit degrees of the E- and //-plane patterns respectively.
signal-level readout of 2 or 3 dB per step. These To illustrate the use of this formula, assume
points of signal level corresponding with an azi¬ that we have a Yagi antenna with a boom length of
muth angle are recorded and plotted on polar two wavelengths. From known relations (described
coordinate paper. A sample of the results is shown in Chapter Four) the expected gain of a Yagi with
on polar coordinate paper in Fig. 17-25. a boom length of two wavelengths is about 13 dB;
On the sample radiation pattern the measured its gain, G, equals 20. Using the formula, the
points are marked with an X and a continuous line product of 0E X 0H = 2000 square degrees. Since
is drawn in, since the pattern is a continuous curve. a Yagi produces a nearly symmetric beam shape in
Radiation patterns should preferably be plotted on cross section, 0 e « 0H = 45 degrees. Now if the
a logarithmic radial scale, rather than a voltage or measured values of 0e and 0H are much larger than
power scale. The reason is that the log scale 45 degrees, then the gain will be much lower than
approximates the response of the ear to signals in the expected 13 dB.
the audio range. Also many receivers have age As another example, suppose that the same
systems that are somewhat logarithmic in response; antenna (a 2-wavelength-boom Yagi) gives a meas¬
therefore the log scale is more representative of ured gain of 9 dB but the radiation pattern half
actual system operation. power beamwidths are approximately 45 degrees.
Having completed a set of radiation-pattern This situation indicates that although the radiation
measurements, one is prompted to ask, “Of what patterns seem to be correct, the low gain shows
322 Chapter 17

inefficiency somewhere in the antenna, such as then pattern measurements may be academic.
lossy materials or poor connections. However, if the gain is lower than expected it is
Large broadside collinear antennas can be advisable to make the pattern measurements as an
checked for excessive phasing-line losses by com¬ aid in determining the possible cause of low gain.
paring the gain computed from the radiation Regarding radiation-pattern measurements, it
patterns with the direct-measured gain. It seems should be remembered that the results measured
paradoxial but it is indeed possible to build a large under proper range facilities will not necessarily be
array with a very narrow beamwidth indicating the same as observed for the same antenna at a
high gain but actually having very low gain because home-station installation. The reasons may be
of losses in the feed distribution system. obvious now in view of the preceding information
In general, and for most vhf/uhf amateur radio on the range set-up, ground reflections, and the
communications, gain is the primary attribute of vertical-field distribution profiles. For long paths
an antenna. However, radiation in other directions over rough terrain where many large obstacles may
than the main beam, called side-lobe radiation, exist, these effects of ground reflection tend to
should be examined by measurement of radiation become diffused, although they still can cause
patterns for effects such as non symmetry on unexpected results. For these reasons it is usually
either side of the main beam or excessive mag¬ unjust to compare vhf/uhf antennas over long
nitude of sidelobes (any sidelobe which is less than paths.
10 dB below the main beam reference level of 0 dB
should be considered excessive). These effects are
usually attributable to incorrect phasing of the Bibliography
radiating elements or radiation from other parts of Source material and more extended discussion
the antenna which was not intended, such as the of topics covered in this chapter can be found in
support structure or feed line. the references given below.
The interpretation of radiation patterns is
intimately related to the particular type of antenna Bruene, “An Inside Picture of Directional Watt¬
under measurement. Reference data should be meters,” QST, April, 1959.
consulted for the particular antenna type of DeMaw, “In-Line RF Power Metering,” QST,
December, 1969.
interest, to verify that the measured results are in
Hall and Kaufmann, “The Macromatcher, an RF
agreement with expected results. Impedance Bridge for Coax Lines,” QST, Jan¬
To summarize the use of pattern measurements, uary, 1972.
if a beam antenna is first checked for gain (the McMullen, “The Line Sampler, an RF Power
easier measurement to make) and it is as expected, Monitor for VHF and UHF,” QST, April, 1972.
 Chapter 18

Finding Directions

Anyone laying out a fixed directive array does or not this is the case for the street in front of or
so in order to put his signal into certain parts of parallelling your own lot. Or from such a visit it is
the world; in such cases, it is essential to be able to often possible to locate some landmark, such as
determine the bearings of the desired points. Too, a factory chimney or church spire, which lies true
the amateur with the rotatable directive array likes north with respect to your house.
to know where to aim if he is trying to pick up If you cannot get true north by such means,
certain countries. And even the amateur with the three other methods are available: compass, pole
single wire is interested in the directive pattern of star and sun.
the lobes when the wire is operated harmonically
at the higher frequencies, and often is able to vary By Compass
the direction of the wire to take advantage of the Get as large a compass as you can; it is difficult,
lobe pattern. though not impossible, to get satisfactory results
with the “pocket” type. In any event, the compass
must have not more than two degrees per division.
FINDING DIRECTION It must be remembered that the compass points
to magnetic north, not true north. The amount by
It is probably no news to most people now¬ which magnetic north differs from true north in a
adays that true direction from one place to another particular location is known as variation. Your
is not what it appears to be on the old Mercator local weather bureau or city engineer’s office can
school map. On such a map, if one starts “east” tell you the magnetic variation for your locality.
from central Kansas, he winds up in the neighbor¬ The information is also available from U.S. Geol¬
hood of Lisbon, Portugal. Actually, as a minute’s ogical Survey topographic maps for your locality,
experiment with a strip of paper on a small globe which may be on file in your local library. When
will show, a signal starting due east from Kansas correcting your “compass north,” do so opposite
never hits Europe at all but goes into the southern
part of Portuguese West Africa.
If, therefore, we want to determine the direc¬ TABLE 18-1
tion of some distant point from our own location,
the ordinary Mercator projection is utterly useless. Apply to Clock Time as Indicated by
the Sign, To Get Time of True Noon
True bearing, however, may be found in several
ways: by using a special type of world map that
Jan. 1 4- 4 min. July 10 4- 5 min.
does show true direction from a specific location 10 4- 8 “ 20 4- « “
to other parts of the world; by working directly 20 4- 11 " 30 + « '•
from a globe; or by using mathematics. 30 4- 13 "

Feb. 10 4- 14 ° Aug. 10 + S "

20 4- 14 " 20 + 3 "
DETERMINING TRUE NORTH 28 4- 13 “ 30 + 1 “

Determining the direction of distant points is of Mar. 10 4- 10 ° Sept. 10 - 3 "

- 7 “
little use to the amateur erecting a directive array 20 4- 8 “ 20
30 - 10 “
30 + 4 °
unless he can put up the array itself in the desired
direction. This, in turn, demands a knowledge of
ÏÏÏ

Oct. 10
O» O’ w

Apr. 10
4- 1 1

~ ~ CO

the direction of true north (as against magnetic 20 20

30 30
north), since all our directions from a globe or map
are worked in terms of true north. Nov. 10
...

May 10
1 1 1
1 1 1

A number of ways may be available to the 20 20

30 30
amateur for determining true north from his
w

location. Frequently, the streets of a city or town Dec. 10

tw CM CO

June 10
1 1 +
1 ++

are laid out, quite accurately, in north-south and 20 20

30 30
east-west directions. A visit to the office of your
city engineer will enable you to determine whether

323
 324
Chapter 18

Fig. 18-1 — Azimuthal map centered on Washington, D.C.

to the direction of the variation. For instance, if of course, that the compass dial is graduated in
the variation for your locality is 12 degrees west degrees.
(meaning that the compass points 12 degrees west
of north) then true north is found by counting 12
By the Pole Star
degrees east of north as shown on the compass.
When taking the bearing, make sure that the Many amateurs use the pole star, Polaris, in
compass is located well away from ironwork, determining the direction of true north. An advan¬
fencing, pipes, etc. Place the instrument on a tage is that the pole star is never more than 0.8°
wooden tripod or support of some sort, at a from true north, so that in practice no corrections
convenient height as near eye level as possible. are necessary. Disadvantages are that some people
Make yourself a sighting stick from a flat stick have difficulty identifying the pole star, and that
about two feet long with a nail driven upright in because of its comparatively high angle above the
each end (for use as “sights”) and then, after the horizon at high northerly latitides, it is not always
needle of the compass has settled down, carefully easy to “sight” on it accurately. Polaris is not
lay this stick across the face of the compass - with visible in the southern hemisphere. In any event, if
the necessary allowance for variation - to line it visible, it is a handy check on the direction secured
up on true north. Be sure you apply the variation by other means.
correctly.
This same sighting-stick and compass rig can By the Sun
also be used in laying out directions for supporting With some slight preparation, the sun can easily
poles for antennas in other directions - provided. be used for determination of true north. One of
Finding Directions 325

Fig. 18-2 - Azimuthal map centered on San Francisco, Calif.

the most satisfactory methods is described below. thus 2° 15' is equivalent to 9 minutes of time (there
The method is based on the fact that exactly at are 60 “angle” minutes to a degree, so that each
noon, local time, the sun bears due south, so that degree of longitude equals 4 minutes of time).
at that time the shadow of a vertical stick or rod Subtract this correction from noon if you are east
will bear north. The resulting shadow direction, of your time meridian; add it if you are west.
incidentally, is true north. To the resulting time, apply a further correc¬
Two corrections to your standard time must be tion for the date from Table 18-1. The resulting
made to determine the exact moment of local true time is the time, by Standard Time, when it will be
noon. true noon at your location. Put up your vertical
The first is a longitude correction. Standard stick (use a plumb bob to make sure it is actually
Time is time at some particular meridian of vertical), check your watch with Standard Time,
longitude: EST is based on the 75th meridian; CST and, at the time indicated from your calculations,
on the 90th meridian; MST on the 105th meridian; mark the position of the shadow. That is true
and PST on the 120th meridian. From an atlas, north.
determine the difference between your own longi¬ In the case of Newington, if we wanted correct
tude and the longitude of your time meridian. time for true noon on October 20: First, subtract¬
Getting this to the nearest 15 minutes of longitude ing the longitude correction - because we are east
is close enough. Example: Newington, Conn., of the time meridian - we get 11:51 A.M.; then,
which runs on 75th meridian time (EST) is at applying the further correction of -15 minutes, we
72° 45' longitude, or a difference of 2° 15'. Now, get 11:36 A.M. EST (12:36 P.M. EDST) as the
for each 15' of longitude, figure 1 minute of time; time of true noon at Newington on October 20.
 326 Chapter 18

Fig. 18-3 — Azimuthal map centered on Wichita, Kansas.

Copyright by Rand McNally & Co., Chicago. Reproduction License No. 394.

AZIMUTHAL MAPS most accurate work; simply pick the map whose
reference point is nearest you. Greatest errors will
While the Mercator projection does not show arise when your location is to one side or the other
true directions, it is possible to make up a map of a line between the reference point and the
which will show true bearings for all parts of the destination point; if your location is near or on the
world from any single point. Three such maps are resulting line, there will be little or no error.
reproduced in this chapter. Fig. 18-1 shows direc¬ By tracing the directional pattern of the
tions from Washington, D.C., Fig. 18-2 gives antenna system on a sheet of tissue paper, then
directions from San Francisco and Fig. 18-3 (a placing the paper over the azimuthal map with the
simplified version of the ARRL amateur radio map origin of the pattern at one’s location, the “cover¬
of the world) gives directions from the approxi¬ age” of the antenna will be readily evident. This is
mate center of the United States - Wichita, a particularly useful stunt when a multilobed
Kansas. antenna, such as any of the long single-wire
For anyone living in the immediate vicinity systems, is to be laid out so that the main lobes
(within 150 miles) of any of these three reference cover as many desirable directions as possible.
points, the directions as taken from the maps will Often a set of such patterns will be of considerable
have a high degree of accuracy. However, one or assistance in determining what length antenna to
the other of the three maps will suffice for any put up, as well as the direction in which it should
location in the United States for all except the run.
 327
Finding Directions

The current edition of the ARRL Amateur manner to indicate the bearing may be added with
Radio Map of the World, entirely different in India ink, as shown in Fig. 18-4, or a small
concept and design from any other radio amateurs’ protractor-like scale may be added at the top of
map, contains a wealth of information especially the globe, over your location. A distance scale can
useful to amateurs. A special type of azimuthal be affixed to the brazen meridian so that both the
projection made by Rand-McNally to ARRL spec¬ bearing and distance to other locations may be
ifications, it gives great-circle bearings from the readily determined (12,500 miles or 40,000 km to
geographical center of the United States, as well as the semicircle).
great-circle distance measurement in miles and
DIRECTION AND DISTANCE
kilometers, within an accuracy of two percent. The
map shows principal cities of the world; local time BY TRIGONOMETRY
zones; WAC divisions; more than 265 countries, The methods to be described will give the
indexed; and amateur prefixes throughout the bearing and distance as accurately as one cares to
world. The map is large enough to be easily compute them. All that is required is a table of
readable from the operating position, 30 X 40 latitude and longitude information, such as may be
inches; and is printed in six colors on heavy paper. found in an atlas or almanac, and a set of
Cost is two dollars from ARRL Headquarters, 225 trigonometry tables. For most purposes, the lati¬
Main St., Newington, Conn. 06111. tude and longitude can be taken from maps of the
The Radio Amateur's Call book also includes areas in question.
great-circle maps and tables, and another Callbook
publication, The Radio Amateur’s World Atlas Direction Calculations
(price, $2.50), features a polar-projection World
map, maps of the continents, and world amateur With this method, the bearing or direction to a
prefixes. The maps are in color. distant location may be determined without the
need to calculate the distance. This procedure is
based on information supplied by Larry Price,
WORKING FROM A GLOBE W4DQD, and on suggestions of Dennis Haarsager,
Bearings for beam-heading purposes may be WA0KKR. Two formulas are used:
determined easily from an ordinary globe with
nothing more complicated than a small school tan 0 = cos Lcot B ( 1 )
protractor of the type available in any school¬
supply or stationery store. For best results, how¬
cotC^01 ^^^^ (2)
ever, the globe should be at least eight inches in sin 0
diameter. A thin strip of paper may be used for a
straightedge to determine the great-circle path where A = your latitude in degrees
between your location and any other location on B = latitude of the other location in de¬
the earth’s surface. The bearing from your location grees
may be determined with the aid of the protractor.
For convenience, a paper-scale circle calibrated in
degrees of bearing may be made and affixed over
the point representing your location on the globe.
The 0° mark of this scale should point toward the
north pole.

A Simplified Direction Finder

A simplified direction finder may be made by
removing a globe from its brazen meridian (semi¬
circular support) and remounting it in the manner
shown in Fig. 18-4. Drill a hole that will accept the
support at your location on the globe, and another
hole directly opposite the first. This second hole
will have the same latitude as yours but will be on
the other side of the equator (north latitude vs.
south latitude). Its longitude will be opposite in
direction from yours from the Greenwich or 0°
meridian, east vs. west, and will be equal to 180
minus your longitude. For example, if your loca¬
tion is 42’ N. lat., 72° W. long., the point opposite
yours on the globe is 42° S. lat., 180 -72 or 108
E. long.
Once the holes are drilled, remount the globe
with your location in the position formerly occu¬
pied by the north pole. By rotating the globe until Fig. 18-4 — A simple direction finder made by
the distant point of interest lies beneath the brazen modifying a globe. Bearing and distance to other
meridian, this support may be used to indicate the locations from yours may be determined quickly
great-circle path. A new “equator” calibrated in a after modification, no calculations being required.
 328 Chapter 18

<t> = -5° 14' (0 is always between -90° and

TABLE 18-11 +90°)
Algebraic Signs of Functions of Angles From tables, sin 0 = sin (-5° 14') = -0.0912.
Angle Sin Cos Tan Cot Then, substituting into Eq. 2,
0° to 90° + + + + co t C = (~°-0617 X cos[47°37' +(-5° 14')]
90° to 180° + _ _ _ -0.0912
0° to -90° - + - -
-90° to -180° - - + + = (-0.0617) X cos (47° 37' -5° 14')
-0.0912

= (-0.0617) X cos 42° 23' _ (-0.0617) X 0,7387

L = your longitude minus that of the
0.0912 " -0.0912
other location (algebraic difference).
If the resultant number is outside the = +0.4998
range between ±180°, algebraically
subtract or add 360, whichever gives a C - +63° 27' (C is always between -90° and
resultant value between +180 and +90°)
-180 degrees.
0 = an intermediate angle used in the The true bearing may now be determined. In
procedure. -this case, the final resultant value of Lis negative,
C = the quadrant bearing angle. (This and the value for C is positive. From Table 18-III
quadrant angle may be converted to the true bearing is therefore C + 180° = 243°47'.
the true bearing angle in degrees In Eq. 1, values for 0 cannot be determined if
clockwise from north by applying the location B is on the equator. In such cases, 0 may
correction as indicatedin Table 18-III.) be taken as +90°, which is suitable for any values
of L and A. Little practical difference will result if
In using these equations, northerly latitudes are the location of B is taken as being a slight distance
taken as positive, and southerly latitudes are taken away from the equator.
as negative. Also, westerly longitudes are taken as
positive, and easterly longitudes are taken as
negative. In all calculations, the appropriate signs Distance Calculations
are to be retained, and all additions and subtrac¬ The distance between two locations may be
tions throughout the procedure are to be made determined from the formula:
algebraically. When using the trig tables, it should
be noted that the various functions may be positive cos D = sin A sin B + cos A cos B cos L (3)
or negative in value, depending on the angle for
which they are taken. The algebraic signs are not where the definitions for A, B, and£ are the same
included in the body of the mathematical tables, as above, and D = distance along the path in
but must be added when substituting numbers into degrees of arc, as shown in Fig. 18-5. For the
the equations. Table 18-11 may be used for ready Seattle to Sydney example:
reference in assigning algebraic signs.
The following example will show how the
formulas are used. To find the bearing from
Seattle, Washington, U.S.A., to Sydney, New
South Wales, Australia:
From the almanac or map:
Seattle - 47°37' N. lat., 122°2O' W. long.
Sydney - 33°54' S. lat., 151° 12' E. long.
Fig. 18-5 shows the nature of this example and
the various terms which are used in the equations.
The location of Seattle is represented by Point A,
and Sydney by Point B.
Values for use in the equations are as follows:
A = lat. A = +47’37'
B = lat. B = -33°54', and from tables,
cot B = -1.4882
L = long. A - long. B = +122’20' - (-151’12’) =
+273°32'. Subtracting 360 to bring this value
into the proper range, we obtain a value of
-86°28' for L. From tables, Fig. 18-5 — Showing the various terms used in the
cos L = 0.0616, and cot L = -0.0617 equations for determining bearing and distance.
North latitudes and west longitudes are taken as
Substituting into Eq. 1, positive, while south latitudes and east longitudes
tan 0 = 0.0616 X (-1.4882) = -0.0917 are taken as negative.
Finding Directions 329

Each degree along the path equals 60 nautical

TABLE 18-111 miles, and each minute equals one nautical mile.
Therefore, 112° 11' of arc is equivalent to 60 X
Corrections to Determine True Bearing Angle 112 + 11 = 6720 + 11 = 6731 nautical miles. To
Then true bearing convert to statute miles, multiply by 1.1508. If the
If And is: distance is desired in kilometers, multiply nautical
L is + C is + C + 0° miles by 1.852. Doing this, we learn that the
distance between Seattle and Sydney is 7746
L is + C is - C + 180Î statute miles, or 12,466 km. Similarly, the direc¬
L is - C is + C + 180° tion and distance between any two points on the
¿¡s - C is - C + 360° earth can be computed.
These equations give information for the great¬
circle bearing and distance for the shortest path.
For long-path work, the bearing will be 180° away
A = 47°37', and from tables, from the answers obtained.
sin>4 =0.7387, ándeos^ =0.6741

B = -33°54', and from tables,

sin B = -0.5577, and cos B = 0.8300
Bibliography
L and cos Lhave already been determined: cos L =
0.0616. Source material and more extended discussions
of topics covered in this chapter can be found in
Substituting into Eq. 3: the references given below.
cos D = 0.7387 X (-0.5577) + 0.6741 X 0.8300 Davis, “A Simplified Direction Finder,” Hints and
X 0.0616 Kinks, QST, May, 1972.
= -0.4120 + 0.0345 =-0.3775 Hall, “Bearing and Distance Calculations by Sleight
of Hand,” QST, August. 1973.
Klopf, “A Bearing and Distance Calculator,” QST,
D = 112° 11'
March, 1971.
Norton, Norton's Star Atlas and Reference Hand¬
Note: From Table 18-11 it may be seen that D, the book, Gall and Inglis, London, England; also
angle having a cosine of —0.3775, must Ue either
between 90 and 180 degrees, or between —90 and published in the U.S. by Sky Publishing Corp.,
—180 degrees. D is always taken as positive. Cambridge, Mass.
 331
INDEX
A Antenna Counterweight .272
Antenna Construction
Absorption. 17, 20, 189 Insulation.266
Adjusting Parasitic Arrays .. 156 Insulator Stress .266
Adjustment of Matching Circuits .100 Materials .264
Wire.264
Admittance ..* " * 125,291
Admittance Coordinates, Smith Chart ... 79 Antenna Current/SWR.306
Air-Insulated Lines . . .. 88 Antenna Rotator .218
Alignment of Lobes, Rhombic .177 Antenna Support, Trees .271,273
Angle, Critical . 15 Antenna Supports, Wire .270
Angles of Arrival, Vertical .18, 19 Antenna System Losses .190
Antenna Antennas
Beverage or Wave.294 Broadside.295
Bob-Tail Curtain.300 Broadside and Endfire Arrays .296
Capacitance, Lumped ... -.191 Bruce Array.297
Collinear-Coaxial. 247 Center-Fed Multiband.179
Combination J-Pole. 246 Continuously Loaded.290
Conical Monopole . 299 Corner Reflector.241
Current Distribution . 2$ Coupled .30
DDRR. 214 Crossed Linear.250
Delta Match. 227 Crossed Yagi .250,251
Elements.132 Direction Finding .293
Element Phase.132 Directly Fed.178
Five-Band, Trap Dipole.1 8$ Discone.298
Folded Dipole. 19 $ Element Lengths and Spacing, VHF . . . 225
Four-Element Yagi, 15 Meter .203 End-Fed .I 79
Gamma Match.203,228 Fishbone. 294
Grounding Systems.61 Folded Dipole.65, 66
Impedance . 27 Ground Plane .62, 63
Interlaced Yagi . 204 Ground Reflection Factor.45
Inverted V .192 Grounded.$9
Inverted-V Dipole .196 Half-Wave Loops.64
Isotropic .132 Half-Wavelength.1 9$
j .227 Harmonic Operation .31
LandT.192 Harmonic, Power Gain of.43
Length of Half-Wave .26 Helical. 2$0
Log-Periodic Dipole Array.208 Helical, VHF .260
Multee.299 Helically Wound.290
Multi-Band .170 HF, Restricted Space.211
Non-Resonant Rhombic.173 Image . 44

Off-Center Feed.1 82 Impedance of Harmonic.32

Q . 30, 148, 151 Indoor Systems .212
Quad.15$ Invisible .211
Random Height and Length Wire .... 178 Long-Wire.165
Range/Measurement .317-322 Long-Wire, Radiation Resistance .... 44
Reactance.30 Long-Wire, VHF.239
Receiving Loop, 160 Meter .193 Loop. 64, 293
Remote Switching . 201 Mobile. 287
Resonance .Il 2 Mobile HF . 287
Resonance in Helically Wound.222 Multiband.I 78 » 1 9$
Resonant Rhombic.171 Multiband Vertical.1 86
Shielded Loop.193 Multielement Array.247
Shortened Yagi . 217
Multiple Dipole .I 83
Stacked 50-MHz Array .233 Non-Resonant Long-Wire .171
Stacking . 204, 224 One-Wavelength Loops .65
Standard-Gain.320 Phased Vertical .1 97
System, Design of . 7 Physically Shortened Dipoles.212
T-Match .226,228 Quad, Directive Patterns and Gain . . .159
Three-Band Quad System .205 Quadhelix, 1296 MHz.261
Three-Element Monoband.205 Receiving, Impedance Matching . . . .317
Trap Verticals.1 87 Rotatable. 277
V .169 Rotating Systems . 280
Vertical Noise Pickup.202 Screen-Reflector Arrays.241
VHF and UHF Systems.224 Short Helically Wound Vertical.221
Voltage Distribution .25 Short Vertical.I 86 » 213
Space Communications.250
Windom . 182
Y Match.227 Specialized .293
Yagi, Helically Wound .219 Step-Up Ratio, Folded Dipole .66
15-Meter Four-Element.203 Sterba Array .297
20 and 1 5 Meter Interlaced .204 Textbooks .67
40-Meter Loop.198 Tilted-Wire .I 68
40-Meter Short Beam.220 Top Loading .61, 62
40-Meter Sloper .200 Two- and Four-Bay Arrays.244
40-Meter Vertical, 6-Foot.213 Vertical.* 9^
Antenna as a Circuit.28 Vertical Polarization .244
Antenna Capacitance, Mobile.288 VHF Non-Directional.245
 332

3.S and 7 MHz.195 Collinear Arrays .135

14, 21, and 28 MHz .202 Collinear Arrays, 144 MHz.236
SO MHz. 231, 232, 233, 239 Combination Driven Arrays.142
144 MHz . . . 234, 235, 236, 238, 239, 240 Combining Vertical and Horizontal Conductors
220 and 420 MHz .240 187
160 Meter Receiving .193 Communication, Long-Path.19
Arrester, Lightning .271 Communication, Short-Path .19
Atmospheric Ducts .13 Conductance. 79, 120, 128
Attenuation .9, 81 Conjugate Match .317
Aurora .21 Construction, All-Metal .237
Az-El Mount, Yagis .253 “Control Points” .18
Counterpoise.190
B Coupled Antennas.30
Couplers.230
Backscatter .22 Coupling Line to Antenna .104
Balun Coupling, Mutual .133
102, 103, 113, 114, 115, 221, 228, 230, 248, 317 Coupling Transmitter to Lines.96
Bandwidth.148, 151 Crank-Up Tower .284
Base Loading, Mobile .287 Critical Angle.15
Bazooka.228 Critical Frequencies.20
Beam, W8JK.143 Critical Frequency .16
Beamwidth.131,132 Crossed Yagi.250
Bearings from Globe.327 Current
Bending, Wave .10, 15 Loop.25
Beta Match.121 Node.25
Beverage Antenna.294 Current and Voltage Feed .105
Bi-Directional Array.131 Current Flow in Long Lines.69
Block and Tackle .268 Currents, Earth.190
Boom Size, Yagi or Quad.277 Curtain Array, 144-MHz.237
Braces/Suspension.234 Cycle, Sunspot .14
Bridge Circuits .303
Bridge, RF Impedance.312 D
Broadside Array .131
Broadside Arrays .137 D Region .14
Decibel .43
C Delta Matching.116,117
Design
Capacitive Hat .191 Antenna System.7
Capture Area.317 Log-Periodic Array.161
Center Loading, Calculation of Value .... 291 Designing Stub Matches with Smith Chart . .125
Center Loading, Mobile .287 Detuning Sleeves .115
Chart Dielectric Constant .10
Antenna Height vs. Wave Angle.49 Dielectric, Earth .11
Attenuation, Transmission Lines .... 93 Diffraction.9, 10
Beamwidth, Arrays.155 Direct Matching.113
Capacitance, Sphere, Disk, & Cylinder . . 62 Direct Ray.10
Characteristic Impedance .88, 89 Direction/Distance by Trigonometry .... 327
Gain, Broadside Array .138 Direction Finding.293
Gain, Collinear Array.135 Direction Finder, Simplified .327
Gain, End-Fire Array.140 Directional Antennas, Sloper .200
Gain/Number of Elements.153 Directional Indicator .282
Gain of Long-Wire Antenna .165 Directional Patterns.148
Line Loss.82 Directivity. 36,41,57,135,138
Off-Center Loaded Dipoles .212 Directivity, Long-Wire.166
Optimum Length of Array.154 Directors, VHF.239
Optimum Spacing for Stacking.156 Distance
Parabolic-Antenna Gain.258 Line-of-Sight Propagation .11, 12
Parabolic Reflector Error .258 Skip .15, 189
Radiation Angles, Long-Wire.166 Transmission, Long.18
Radiation Resistance, Collinear Array . .134 Transmission/Layer Height .17
Radiation Resistance, Variation with Height 50 Disturbances, Sudden Ionospheric.21
Radiation Resistance vs. “Diversity” Reception.20
Vertical Antenna Height.60 Driven Array.131
Reactance, Vertical Antenna vs. Height 61 Driven Arrays .134
Transmission Line Data.91 Driven Elements .131
Vertical Directivity.57 Driven Element Length .226
Wave Angle vs. Antenna Height.49 Ducts, Atmospheric.13
Yagi and Quad Compared.159
Yagi Element Length.204 E
Charts, Prediction.23
Circular Polarization.250 E Layer .17, 18
Coax-Line RF Choke .115,116 E Region .14
Coaxial Fittings.92 Earth Currents .190
Coaxial Fittings, Assembly Instructions . . 94, 95 Earth, Dielectric .11
Coaxial Lines.89 Effective Aperture .317
Coaxial Lines, Stubs.124 Electric Fields .7, 8, 9
 333

Electric Force, Lines of .8, 9 Gain Formula .321

Electrical Length .26, 92 Gain/Mutual Impedance .133
Electromagnetic Waves.7, 24, 35 Gamma Dimensions.119, 120
Element Length Adjustment .226 Gamma Match . 1 18,228,231,251
Element Lengths and Spacings .227 Gin Pole.276, 284, 285
End Effect.27 Great-Circle Path .18, 19
End-Fed Antenna.179 Ground Effects.43, 132
End-Fire Array.131 Ground Plane . . . 190,191,196,202,215,245
End-Fire Arrays.139 Ground-Plane Antennas .126, 128
Extended Double Zepp .137 Ground Reflection Factor .45
Ground Reflection/Radiation Resistance . . 47
F Ground Screens.48
Ground System.198
F Layer.17, 18 Ground Wave.10-12, 109
F Region .15 Guy Anchors. 275,284
Fading .20 Guy Material.275
Fading, Selective .20
Feed-Horn. 255,256 H
Feed Horn, Backfire Type .257
Feed Horn, 1296 MHz.257 Hair-Pin Match.121
Feed-Line Radiation.180 Half-Wave Dipoles.202
Feed-Point Mismatch .317 Halyards.267
Feed System, 2287.5 MHz.259 Harmonic Operation.25,106
Feed-Through Connectors .271 Harmonic Wire, Length of .32
Feeder-to-Transmitter Matching Networks . 180 Heat Losses .29
Feeding Long-Wires.168 Helically Wound Vertical.190
Field Intensity .9
Field Strength .9, 24, 130 I
Field-Strength Measurement .157
Fields Impedance
Electric.7, 9 Helically-Wound Vertical .222
Electric and Magnetic.9 Input.150
Magnetic.7, 9 Mutual.133, 134
Magnetic and Electric.7 Plotting, Smith Chart.77
Finding Directions .323 Step-Up .115
Finding Directions, by Sun.324 Transformation .86
Five-Band Antenna, Trap Dipole.185 Impedance, Rhombic Antenna .176
Flares, Solar .21 Impedance Bridge.312
Flexible Lines .90 Impedance Matching, Receiving Antennas . . 35
Flexible Sections for Rotatable Arrays . 126, 127 Incident Power.71
Folded Dipole .117,229 In-Line RF Wattmeter. 306-310
Folded Dipole, Step-Up Ratio.66 Input Impedance, Line.74
Formula Input Impedances, Smith Chart Determination of
Directivity .41 87
Field Strength.41 Insulator Leakage.266
Gain. 42, 132, 321 Insulators, Ribbon Line .267
Half-Wave.225 Inversion, Temperature .13
Length, Long-Wire.167 Ionosphere.14
SWR.85 Ionosphere Variations.19
Fot .17 Ionospheric
Four-Band Trap Dipole .186 Control Points.18
Four-Element Arrays .152 Predictions .23
Four-Wire Lines.89 Refraction .15
“Free” Space .7, 8, 9, 11 Scatter.21
Free-Space Radiation Resistance.29 Storms.20
Frequency Isotropic Antenna.132
Critical.16, 20 Isotropic Radiator.36
Lowest Usable.17
Maximum Usable.16 L
Optimum Working.17
Frequency Predictions.23 LaPort Rhombic .240
Frequency Range and SWR.104 Layer Characteristics .14
Front-to-Back Ratio.147, 150 Layer Height/Transmission Distance .... 17
Lazy-H Array.143
G Length/Diameter Ratio .27
Light, Speed of.7
Gain .135,138 Lightning Grounds .270
Gain Lightning Protection.270
Broadside Array.143 Line
Combination Arrays .142 Length, Determination of Electrical ... 81
Parabolic Antenna.257 Loss, Considerations with Smith Chart . . 83
Quad Antenna ..159 Radiation.105
Rhombic Antenna .175 Line Current/Voltage, Coaxial.301
Three-Element Beam.149, 150 Line-of-Sight Propagation .11, 12
Two-Element Parasitic Array.146 Line Sampler.310-312
Gain/Effective Aperture .318 Linear Polarization .250
 334
Lines as Circuit Elements.86 0
Lines
Electric and Magnetic Force.8 Off-Center Feed, Windom .182
Of Force, Electric .9 Omega Match.120, 121
Loading Coil, Mobile .2 87 “One-Way” Skip .20
Loading Coil Design.289 Open-Wire Line.268
Loading Coils.190,211,217,218 Open Wire Line/Antenna Connection .... 269
Log-Periodic Dipole Array. .160 Open-Wire Support .269
Log Periodic Radiation Pattern .163 Optimum Working Frequency.17
Long-Distance Transmission .18
“Long Path”, Great-Circle .19 P
Long Yagis.152
Longwire Antenna, Span and Sag .264 Parabolic Dish, Template.259
Losses, Trap.184 Parabolic Dish, Surfacing Materials.258
Lowest Usable Frequency .17 Parallel-Conductor Lines.90
Parallel/Series Equivalent Circuits .75
M Parasitic
Array .131
Macromatcher .212.312 Arrays .145
Magnetic Element .131, 226
Fields .7, 8, 9 Elements, Self-Resonant.147
Storms.20, 21 Patterns.148
Magnetic Force, Lines of.8 Patterns
Magnetic North.323 E and H Planes.37
Magnetic Variation .323 Graphical Construction .40
Major Lobes.131 Vertical Directivity.57
Maps, Azimuthal .326 Phase.8
Mast Erection .276 Phase and Wavelength .8
Masts.;.274 Phase Shift/Reflection.11
Masts, A-Frame.274 Phased Arrays .198
Masts, Guying .275 Phased Verticals .197
Matching by Length Adjustment.127 Phasing .144
Matching Circuit, Balanced to Unbalanced . 229 Phasing Sections .226
Matching Network, Tapped-Coil.292 Phasing System, 50-MHz.235
Matching Networks, Commercial .181 Pickup Efficiency.294
Matching Networks, Feeder-to-Transmitter . 180 Plane Wave.7, 8
Matching Sections.226 Polarization . . . 9,1 1,12,189,224,316,317
Matching Stub .114 Polarization
Matching Stubs.122,123,124 Circular. 34, 260
Matching with Lumped Constants.126 Elliptical . 34, 260
Maximum Usable Frequency .16 Linear . 34, 260
Measurement, Field Strength .157 Reciprocity, Receiving and Transmitting 34
Measurement of Line Input Current .... 101 Sky Wave.19
Measurements .301 Tropospheric Wave.13
Measurements Pole Star/Polaris .324
Absolute Gain.320 Power
Antenna .316 Density.41
Antenna Test Site .317 Gain.130
Gain.318 Power Absorber. 305, 316
Macromatcher.212 Prediction Charts.23
Radiation Pattern . 318, 320-322 Propagation .189
SWR.302 Propagation
SWR Errors.305 Line-of-Sight .11,12
Meteor Trails.22, 23 Multihop.is
Miles to Kilometer Conversion .329 Single and Multihop .16
Minor Lobes.131 Sporadic-F .22
Mobile Antenna Tuneup. 287, 292 Tropospheric .13
MUF Determination, Control Point .... 18 Velocity .26
MU F Predictions .23 Wave.7
Multiband Antenna, Traps .186 Prop-Pitch Rotor System.280
Multiband Antennas, Harmonic Radiation . .188 Pulleys .267
Multielement Directive Arrays.130
Multihop Propagation .16, 18 Q
Multiple Tuning.191
Mutual Q
Coupling.133 Antenna .128, 148, 151
Impedance .133, 134, 146 Circuit.97, 98
Impedance/Gain.133 Line Input Impedance .98
2-Section, Adjustable .227
N e vs. SWR.129
Quarter-Wave Transformer .116
National Electric Code.264 Quad Antenna .158
Nautical Miles/Statute.329 Quad Spreaders.207
Noise, Solar Radio, Flux (0) 20 Quad
Non-Reactive Termination .314 Diamond/Square.206
Numbers, Sunspot.19, 20 Directive Patterns and Gain . . . .159,160
 335

Quads, Construction.277 Skip Distance.15,189

Quads Sky Wave .189, 190
Spreader Material .279 Sloping Dipole .200
Three Band System.205 Smith-Chart .76, 125
VHF Stacking.238 Solar Flares .21
144 MHz.238 Solar Radio Noise Flux (0).20
Space Attenuation .9
R Space, “Free” .7, 8
Space Wave .11,12
Radials .190, 198, 213 Speed of Light .7
Radiation .24 Splicing Rope .267
Radiation, Feed-Line .ISO Sporadic-E Propagation .22
Radiation Patterns .' • • 36 Stacked Dipoles.248
Radiation Resistance Stacked Yagis .154
29, 54, 60, 140, 146, 147, 190 Standard Time .325
Radiation Resistance/Ground Reflection . . 47 Standing-Wave Ratio.74
Radiation Resistance, Loop.294 Standing Waves.73
Radiation Resistance, Parallel Elements . . .133 Statute Miles/Nautical.329
Radiation Resistance, Short Antennas . . . 288 Step-Up Ratio, Folded Dipole.66
Radiation Resistance, Small Loop.193 Storms
Radio Waves.7 Ionospheric.20
Range, Ground-Wave .10 Magnetic .20
Ray, Direct .10 Strain Insulators .267
RCL Bridge . .312, 313 Strain-Reliever, Ribbon Line.269
Receiving Antennas, Impedance Matching . . 35 Stressed Parabolic Dish. 254, 256
Reception, “Diversity”.20 Stub Matching .229
Reflected Power .71 Sudden Ionospheric Disturbances .21
Reflection.9, 10 Sunspot
Reflection Coefficient.72 Cycle.14, 19
Reflection/Phase Shift.11 Numbers .19, 20
Reflection/Refraction.16 Surface Wave.1°
Reflectometers.306 Surface, Parabolic Dish.256
Reflectors and Directors Defined .145 Susceptance . 79, 120, 128
Refraction.9, 10 SWR .74
Refraction SWR and Frequency Range.104
Ionospheric.15 SWR and Line Loss .82
Tropospheric .13 SWR/Antenna Currents .306
Resistance Bridge.303 SWR Bridge . 303-306
Resonance, Helically Wound .222 SWR/Reflectometers .306
Resonance in Linear Circuits .24 SWR vs. Q. 129
RF Ammeters .302
RF Impedance Bridge .312 T
RF Inline Wattmeter. 306-310
RF Voltmeter .301 T Match.118
Rhombic, LaPort.240 Table
Rhombic, Multiwire.176 Aluminum Tubing Sizes.278
Rhombic Antenna Antenna Stress.265
Front-to-Back.176 Crossed Yagi Elements .252
Gain.175 Design Data, Helical Antennas .262
Impedance .176 Dimensions, Corner Reflector .242
Lobe alignment .177 Front-to-Back Ratio .149
Tilt Angle.173 Gamma Match Dimensions, VHF .... 253
Rooftop Mounting, Vertical Section . . . .188 Power Ratio to dB .42
Rotators, Az-El.254 Quad Dimensions .279
Rotor, Delayed Action Braking .281 Quad Loop Dimensions.206
Tension, Halyard Material.268
S V and Rhombic Dimensions (VHF) . . . 240
Voltage Ratio to dB ..42
Safety/Tower Climbing .286 Yagi Arrays, Optimum Element Spacing . 153
Scatter Tapped-Coil Matching.127
Back.22 Temperature Inversion.13
Side .22 Termination, Non-Reactive.314
Transequatorial .22 Testing Coaxial Cable .96
Scatter Signals .21,22 Three-Element Parasitic Beam.149
Scattering .10, 19 Tilt-Angle, Rhombic Antenna.173
Selective Fading .20 Top Hat.213
Self-Resonant Parasitic Elements .147 Top-Loading.191
Series and Parallel Coupling.98 Tower Climbing Safety.286
Series/Parallel Equivalent Circuits.75 Tower Selection and Installation.283
“Shadow Area”.10 Transequatorial Scatter .22
Shielded Loop .194, 293 Transmatch . . . .99, 100, 180, 182, 195, 307
“Short Path”, Great-Circle .19 Transmission Distance/Layer Height .... 17
Sidescatter.22 Transmission Lines
Single-Hop Propagation .16 Antenna Currents on.110
Single-Wire Lines.92 Characteristic Impedance .70
Skip, One-Way .20 Detuning for Antenna Current.Ill
336

Installation .268 Front .8

Radiation Loss.68 Propagation.7
Textbooks .67 Sky .14-22
Velocity of Propagation.70 Travel, Direction of.8
Wavelength .69 Wave Angles .166
Transmission, Long-Distance .18 Wavelength, Formula for.8
Trap Construction.185 Wavelength, Phase and.8
Trap Dipoles.183 Waves
Trap Losses .184 Diffraction of.9, 10
Trap Verticals .187 Electromagnetic.7
Troposhpere.10, 21 Ground.10-12
Tropospheric Plane.7, 8
Propagation.13 Radio .7
Refraction .13 Reflection of .9, 10
Tropospheric Wave, Polarization of .... 13 Refraction of .9, 10
True North.323 Space .11, 12
Tuned and Untuned Lines .104 Surface.10
Tuned Feeders .202 Wide Spacing.149
Two-Element Beam, Parasitic.146 Wind Compensation.273
Window-Lead-in Panel.271
U Wire Antennas, Methods of Raising . . . .272
Wire Splicing.265
Unidirectional Array.131 Wire Tension ..264
Unidirectional End-Fire Arrays .141 Wire Types.264
Wolf Number.20
V
Y
Velocity of Propagation .26
Vertical Angles of Arrival.18 Yagi, High-Performance .243
Vertical Directivity .57 Yagis, Vertical Polarization.244
VHF Directional Coupler.310-312 Yagi, 13-Element, 432 MHz.243
Virtual Height .16 Yagi-Uda Antenna.145
Yagi, 50-MHz.230
W Yagis, 220 and 420 MHz.242
Wattmeter, RF Inline . 306-310 Z
Wave
Angle .17 Zepp, Extended Double .137
Bending .10, 15 Zepp or End Feed.108