Building and Designing Transistor Radios

In his characteristic accessible and clear style, R.H. Warring offers a

guide for the radio amateur to many of the things one ought to know
when tackling a radio project. Warring describes and illustrates the
design and working characteristics of the basic electronic ‘blocks’
from which radio receivers can be constructed and shows the reader
how to incorporate additional simple circuits, such as volume and
tone controls, and automatic gain control, to build upon these basic
blocks. Straightforward calculations are provided to determine the
values of components like resistors, capacitors and inductances to
optimise your results.
Building and Designing Transistor Radios will equip you with the
skills and knowledge properly to understand transistor radios. Special
attention is paid to the transistor, the most important component,
which the author provides insightful advice on. By understanding
the operating characteristics of the radio’s different elements, the
reader can learn how to devise circuits so that they perform most
efficiently. With 80 diagrams, no excessively complicated circuitry,
and minimal mathematical calculations included, this book first
published in 1977 remains perfect for amateurs and enthusiasts.

R.H. Warring (1920-1984) was one of the best-known writers in

the field of practical radio and electronics. He was renowned for Building and Designing
R.H. Warring

writing technical books that could be readily understood by all; his

vast experience meant he was keenly aware of the possible pitfalls
which could befall the beginner. Transistor Radios
The Lutterworth Press A Beginner’s Guide
P.O. Box 60
Cambridge
CB1 2NT
www.lutterworth.com
[email protected] R.H. Warring
 Building and Designing
Transistor Radios
A Beginner’s Guide

R.H. Warring
Paperback ISBN: 978 0 7188 9731 4
PDF ISBN: 978 0 7188 9732 1
ePub ISBN: 978 0 7188 9733 8

The Lutterworth Press

L
Click on the link above to see our full catalogue
for more excellent titles in
Hardback, Paperback, PDF, ePub and Kindle!

Would you like to join our Mailing List?

Click here!

Enjoyed this book?

Review it on Amazon so others can too!

Click here!
Building and Designing
Transistor Radios
Building and Designing
Transistor Radios
A Beginner’s Guide

Written and Illustrated by

R.H. Warring

The Lutterworth Press

 The Lutterworth Press
P.O. Box 60
Cambridge
CB1 2NT
United Kingdom
www.lutterworth.com.
[email protected]

Paperback ISBN: 978 0 7188 9731 4

PDF ISBN: 978 0 7188 9732 1
ePub ISBN: 978 0 7188 9733 8

British Library Cataloguing in Publication Data

A record is available from the British Library

First published by The Lutterworth Press, 1977

Reprinted 2023

Copyright © R.H. Warring, 1977

All rights reserved. No part of this edition may be reproduced,

stored electronically or in any retrieval system, or transmitted
in any form or by any means, electronic, mechanical,
photocopying, recording, or otherwise, without
prior written permission from the Publisher
([email protected])
 Contents

Introduction 7
1 Aerial Circuits 13
2 Aerials and Aerial Coupling 25
3 Semi-Conductor Diodes and their Characteristics 32
4 The Diode Detector 39
5 Transistors 48
6 Understanding Transistor Characteristics 57
7 Audio Amplifiers 75
8 Regenerative Receivers 87
9 The Superhet 93
10 Interstage Connections (Coupling) 103
11 Field Effect Transistors 111
12 Miscellaneous Circuits 118
13 Checking Radio Circuits 125

Index 127
 Introduction

Any radio receiver consists of a number of separate ‘stages’, suitably

interconnected, each stage representing the equivalent of an electronic
‘building block’. The three basic ‘blocks’ are a tuned circuit, to extract
radio signals from the ether; a detector, to turn these radio signals into
audio frequency signals (or af); and an electro-mechanical device to
turn these af signals into sound (either headphones or a loudspeaker).
Thus the simplest combination of ‘blocks’ is:

This combination has the particular limitation of providing only

very low strength af signals – so weak, in fact, that they would certainly
not work a speaker, and only give very weak signal strength in phones.
Without going to more ‘blocks’ the only way to improve ‘listening
strength’ is to add an external aerial.

Much better results are possible by adding further ‘blocks’, especially

one which magnifies or amplifies the signal output from the detector.
8 Building and Designing Transistor Radios

This four-block combination can work quite well, and an external

aerial may not be necessary (although an advantage in some cases).
Also it can be made to operate a speaker.

There are limits to what can be done with af amplifiers as these will
magnify ‘noise’ and distortion, as well as the required af signals. Thus
a further step, is to add another block to improve the signal quality in
some way or another before detection and amplification. At this point
it should also be possible to dispense with an external aerial.

A simple solution to the ‘improving’ block is a regenerative circuit,

which may also be designed to work as a detector as well. The more
or less universal solution in modern radio receiver design, however, is
the adoption of a superhet front end. This has the effect of extracting a
Introduction 9

signal frequency intermediate between the radio frequency (rf) picked

up by the tuned circuit, which intermediate frequency (if) can itself be
amplified before being passed to the detector stage.

That, in fact, is as far as it is necessary to go for satisfactory reception

of long wave and medium wave frequencies – or amplitude modulated
(AM) broadcasts. Very High Frequency (VHF) or Frequency Modulated
(FM) broadcasts, however, pose further problems for the nature of the
rf signal is different, as well as the signal frequency being much higher.
The latter means that the conventional form of tuned circuit for AM
(usually comprising a ferrite rod aerial) is no longer suitable. Instead
an external dipole aerial is necessary. Superhet working is more or less
obligatory, so the building blocks now become:
10 Building and Designing Transistor Radios

Because of the higher signal frequency at which FM is broadcast, it is

also desirable to preamplify the incoming rf signal, so one more block
can be added with advantage:

Finally, of course, you may want to build an AM/FM radio, when

some of the blocks have to be duplicated, the appropriate ‘blocks’ being
selected by a wavechange switch:
Introduction 11

It is the purpose of this book to describe and illustrate the design

and working characteristics of the various ‘blocks’ from which radio
receivers can be constructed. In some cases there are more or less
standard designs which have evolved, e.g. for tuned circuits, AM and
FM detectors, superhet mixer-oscillators, if amplifiers and af amplifiers.
It would be difficult to improve on them with simple constructions.
The main field for experiment is possible combinations of these various
‘blocks’ – and in particular the manner in which they should be con-
nected together or coupled for most efficient working.
By treating all the major subjects separately it is hoped that the various
design principles will have been made clear – and also how ‘standard’
types of ‘boxes’ can be connected together to make a working unit. It
is also shown how additional simple circuits can be incorporated to
improve the performance of any basic design.
 Chapter 1
Aerial Circuits

The conventional aerial circuit or tuned circuit comprises a variable

capacitor in parallel with an inductance (physically, a coil). The
frequency (f) at which such a circuit is resonant is given by:

f = frequency in Hz
L = inductance in henrys
C = capacitance in farads
A practical version of this formula is :

f = frequency in kHz
where L = inductance in microhenrys (μH)
C = capacitance in picofarads (pF)
It can be noted here that a practical inductance will also have a
certain amount of resistance, and so the equivalent circuit is as shown
in the second diagram in Fig. 1. The presence of such resistance does
not affect the resonant frequency of the circuit but only the sharpness
of the resonance of the circuit. This controls the quality factor of the
tuned circuit (see later). The practical capacitance also has a certain
amount of resistance, but this is normally negligible except at very high
frequencies (30 MHz and above).
By fixing one component value (e.g. inductance) and making the
other variable (e.g. capacitance) it is possible to adjust or tune the
circuit over a range of resonant frequencies. Theoretically, on this
basis, it is possible to design a tuned circuit to cover the whole range
of broadcast frequencies from the ‘top’ (wavelength) end of the long
wave band (30 kHz) to the ‘bottom’ (wavelength) end of the VHF band
14 Building and Designing Transistor Radios

Fig. 1 Capacitance (C) and inductance (L) forming a tuned circuit.

(300 MHz). This is not a realistic solution, and so tuned circuits are
designed separately to cover individual broadcast bands, e.g.

Long wave (or low frequency) – 30-300 kHz

Medium wave (or medium frequency) – 300-3000 kHz
Short wave (or high frequency) – 3-30 MHz
VHF – 30-300 MHz

In practice, tuned circuits are designed to cover the actual spread of

broadcast stations operating in these bands, e.g.

Long wave – 50-150 kHz

Medium wave – 500-1500 kHz
Short wave – 18-28 MHz
VHF – 88-100 MHz

A significant fact is that the actual frequency range covered increases

considerably with decreasing wavelength of these broadcast bands, e.g.

Long wave – range covered 100,000 Hz.

Medium wave – range covered 1,000,000 Hz.
Short wave – range covered 26,000,000 Hz.
VHF – range covered 12,000,000 Hz.
Aerial Circuits 15

This makes the design of aerial circuits increasingly critical from

long wave upwards (in frequency). Again, in practice, this means
that home-made coils are seldom suitable for other than simple long
wave and medium wave receivers. Even then, proprietary coils almost
invariably give better results because of the better quality factors (or Q)
achieved. Nevertheless it is interesting to cover the design of simple
aerial coils.
The simplest type of inductance is an open coil of insulated wire
wound on a former of insulating material, or it can even be self-
supporting if the wire is thick enough. In the latter case the coil is wound
on a mandrel and then slid off, being mounted on the wire ends (Fig. 2).

Fig. 2 Air-cored tuning coils. Coil on left is wound on a former.

The inductance of such a coil is found as follows:

where R is the mean radius of the coil in inches

L is the length of the coil in inches
N is the number of turns
The effect of wire diameter is not significant, provided the coil
diameter is reasonably large, i.e. 1 in. (25 mm) or more. It is therefore
logical to use quite thick wire (18 s.w.g. or 16 s.w.g.) in order to
minimize coil resistance.
Suppose such a coil is to be designed as the inductive component in
a medium wave tuned circuit. The resonant frequency range required
is 500 to 1500 kHz. Considering the requirements, first in terms of the
product of L and C, from the resonant frequency formula:
16 Building and Designing Transistor Radios

Thus at f = 500 kH2,

at f = 1400 kHz,

For a fixed value of inductance, maximum capacity will be required

to tune to the lowest frequency, i.e. the highest calculated value of LC
required. Typically available variable capacities offer a range of 0-200
pf or 0-500 pf. Choosing the 0–500 pf size, at maximum capacity:
L × 500 = 100,000
or inductance required = 200 microhenrys.
Using the same inductance, the minimum capacitance required to
tune to the other end of the band (1500 kHz) would be:
200 × C min = 11,000
or C min = 55 microhenrys
Thus a 200 μH inductance would be a suitable match to a 0-500 pf
capacitor to cover the range required.
To simplify the coil design we can ‘guesstimate’ a length of 1 in.
and a coil diameter of 1 in. Inserting these values in the appropriate
formula:

Inserting the value of inductance required (200 μH) and solving for
number of turns :
Aerial Circuits 17

Close winding 100 turns of wire the length of 1 in. would permit
the use of a maximum wire size of 0·01 in., say 36 s.w.g. To use a larger
wire size it would be necessary to increase the length of the coil and
recalculate the number of turns required accordingly.
A long-wave coil would require more turns; and a short-wave coil
less turns (perhaps only one or two turns).

Q Factor
The effect of resistance in the tuned circuit is shown in simple diagram-
matic form in Fig. 3, representative of a resonant circuit. The current

Fig. 3 The effect of resistance and Q factor on sharpness of tuning.

18 Building and Designing Transistor Radios

flowing in a resonant circuit peaks at the resonant frequency and falls

off sharply on either side. The lower the resistance present, the higher
the peak (more current flowing) and the sharper it is (the sharper the
tuning). Resistance values shown are nominal only to illustrate this
effect.
This can be put another way. The shape of the resonant curve is
dependent on the respective values of the reactance on either the coil
or capacitor and the resistance present. The ratio of the two is known
as the Q factor, when

Reactance, in ohms, can be calculated from the following formulas:

In the case of an inductance, Xl = 2π fL

In the case of a capacitor, Xc = 1/2π fL

At resonant frequency the reactance of a coil and capacitor are the

same, so it does not matter which is considered. A simple calculation
will prove this, taking the previous values calculated and a resonant
frequency of 500 kHz, namely:
inductance = 200 μH
capacitance = 500 pF
Thus Xl = 2π × 500 × 103 × 200 × 10-6 = 628 ohms
Xc = 1/2π × 500 × 103 × 500 × 10-12 = 628 ohms
The resistance refers to the dynamic resistance to rf currents in the
circuit, not the dc resistance. Dynamic resistance is generally known
as impedance. In the case of a simple air-cored coil, dynamic resistance
may rise up to 100 ohms or more, yielding a Q of less than 10. Very
much more efficient coils can be produced with Q factors ranging up
to 100 (or very much higher in certain cases). These are invariably
wound on a non-conducting magnetic core, either of ferrite or iron
dust bound together with an insulator. The actual value of Q achieved
has the same effect as that illustrated in Fig. 3. The higher the Q value the
sharper, and higher, the peak of the curve. With decreasing Q value
the tuning becomes broader and the peak value is reduced as shown
in Fig. 3. Sharpness of tuning is always desirable in radio receivers as it
Aerial Circuits 19

Fig. 4 Basic forms of aerial coils wound on a ferrite rod or ferrite slab.

gives good selectivity, or the ability to separate one station from another
when the two are closely spaced on the frequency band, but see later
under Modulated Signals.

Ferrite Rod Aerials

In simple terms introduction of a ferrite or similar magnetic material
core to a coil greatly increases its inductance. This means that the coil
can be made much more compact thus requiring less wire length and
less resistance. A smaller wire size can also be used without introducing
excessive resistance. Unfortunately no simple formulas apply for the
design of such coils, for the size and number of turns required are related
to the size and type of magnetic core material used. They are therefore
designed on empirical or semi-empirical lines, the latter using charts
related to the specific material properties.
Simple coils of this type are wound on standard sizes of ferrite rod or
ferrite slab, either as tapped coils or inductively coupled coils – Fig. 4.
Some design data are given in Table 1 (end of chapter).
One other advantage offered by ‘cored’ coils is that their inductance
can be varied, if necessary, by altering the position of the coil on the core.
This can be a very useful feature for adjusting the resonant frequency
range of a tuned circuit independent of the variable capacitor, e.g. for
setting up or ‘trimming’ purposes. Once adjusted in this way, the coil
is then usually locked to the core (e.g. with adhesive or hard wax) to
ensure that it remains at a fixed inductance.
The ‘Q’ of a coil can be further influenced by special forms of
20 Building and Designing Transistor Radios

Fig. 5 The tapping point on the aerial coil affects selectivity and sensitivity.

windings, which can only be tackled successfully by coil winding

machines. This is another reason why proprietary aerial coils are
(almost) invariably better than home-made coils.

Selectivity and Sensitivity

Selectivity or the ability to tune sharply has already been explained.
Sensitivity is the ability to amplify the very weak rf signals received
in the aerial circuit to a practical output level, both as regards signal
strength and depth of modulation. It is thus just as important as
selectivity for satisfactory receiver performance. Both selectivity and
sensitivity increase with increasing Q, but the two are not necessarily
compatible.
As a basic example take the simple form of tuned circuit with a
tapping point on the coil for connecting directly to the detector stage,
as in Fig. 5. The end of the coil connected to the earth of the circuit is
referred to as the ‘earthy’ end. The tapping point on a coil normally
comes about one third the length (number of turns) from this end of
the coil. If this tapping point is moved up towards the other or ‘hot’
end of the coil, this will have the effect of increasing selectivity but
reducing signal strength or sensitivity. Conversely, moving the tapping
point towards the ‘earthy’ end of the coil will reduce selectivity, but
increase sensitivity. This, in fact, is one way of adjusting the selectivity
and sensitivity of a simple tuned circuit of this type, the aim being to
Aerial Circuits 21

arrive at an optimum tapping point which gives the best possible

compromise between selectivity and sensitivity. In practice this does
not necessarily mean physically altering the tapping point. With a
simple coil it is more practical to add turns at one end and remove
turns from the other to ‘shift’ the tapping point.

Modulated Signals
It is possible to increase the Q attainable from coils to very high
levels, using some form of positive feedback to neutralize, or partially
neutralize, resistance losses. This would seem an ideal arrangement to
get very sharp tuning. However, tuning can be made too sharp for a
radio receiver as, it has to accept not a single frequency represented by
the rf carrier wave, but a whole band of frequencies representing the
modulated signal, otherwise some of the af content may be ‘tuned out’
or cut off. This is particularly true in the case of FM receivers, where
the signal represents a bandwidth rather than a specific rf frequency. It
is in this respect, both for AM and FM radio working, that the superhet
receiver scores since it changes the rf carrier and its sidebands to a
single fixed frequency, known as the intermediate frequency, and so
selectivity can be sharply peaked.

Hearing Two Stations at once

Even the superhet receiver is not immune from interference, however,
and simpler receivers considerably less so, due to what is known as
cross-modulation. This is caused by the presence of a strong signal
near to, but not at, the frequency to which the receiver is tuned. If
very much stronger than the actual ‘tuned’ signal, it can effectively
modulate that signal. In other words, although the set is tuned to a
particular carrier wave frequency which has its own modulation, this
carrier is now subject to modulation from the spurious signal. Thus the
carrier frequency to which the set is tuned is actually carrying two af
programmes. Hence two stations are heard simultaneously.
This is something which can readily happen on the medium wave
band, especially when the set is tuned to a relatively weak station
and has a much stronger station signal present close to that particu-
lar frequency. The cause may not lie in improving the selectivity of a
tuned circuit so much as improving the linearity of a detector, for it is
at this stage that the trouble will show up. Sharper tuning (i.e. better
22 Building and Designing Transistor Radios

selectivity) should help, however, for the sharper the tuning the more
a tuned circuit will automatically tend to reject any cross-modulation.
Cross modulation does not occur as a problem in the design of VHF
receivers, which is why the FM band gives better reception.

Loft Aerials
The loft is the logical place to put an external aerial to improve reception
in areas where radio reception tends to be poor. It can be a single
dipole or a folded dipole, preferably arranged vertically so as to be
non-directional, or horizontally lined up with a particular transmitting
station. Its performance can be improved further with the addition of a
reflector and directors. These are short lengths of wire or tube which act
as ‘false’ aerials (or parasitic aerials, as they are called), with beneficial
effects if they are placed in specific positions – see Fig. 6.

Fig. 6 The simple folded dipole aerial associated with a reflector and
directors, or ‘parasitic’ aerials.

A reflector needs to be about the same length as a dipole or slightly

longer, located a quarter of a wavelength away (e.g. about 15 in. in the
case of a VHF radio aerial). This will have the effect of strengthening
Aerial Circuits 23

the signals coming from the other side and, more important still,
reducing interference from the side on which the reflector is.
A false aerial placed on the opposite side to the reflector will have
the effect of improving the tuning characteristics of the aerial, again
provided it is the right length and in the right position. It needs to
be shorter than a dipole and about one-tenth of a wavelength away
from it. Its actual length and position will alter the phase relationship
between the voltage and current induced in the dipole aerial. In its
optimum position it will improve the tuning characteristics for a signal
coming from the director-dipole direction. Further improvement may
be obtained by adding still more directors, each one shorter than the
previous one. This also reduces the impedance of a dipole aerial, which
is an advantage for cable matching when using a folded dipole. Folded
dipoles typically have an impedance of 300 ohms, although this is
reduced by a reflector and director. They may still need a balancing
transformer or balun to match a receiver aerial circuit fed by a 75-ohm
coaxial cable.

Fig. 7 Ribbon-type folded dipole wire aerial with an effective feeder

impedance of 300 ohms.

A simple folded dipole can be made from a single length of insulated

wire arranged in the form of an outline ‘T’, as shown in Fig. 7. Spacing
between the wires should be ¼ in. The parallel ‘leg’ of the ‘T’ forms
the feeder, with an effective impedance of 300 ohms. A simple – and
inexpensive – indoor aerial of this type should be readily available from
24 Building and Designing Transistor Radios

radio shops as a ribbon aerial (a plastic ribbon with two parallel wires
incorporated in each edge of the ‘ribbon’ which forms the bar of the ‘T’,
and also in the separate ribbon forming the ‘leg’).
Quite elaborate aerial forms consisting of a dipole and reflector
and directors are used for television reception, where even higher
broadcast frequencies are involved. What may appear puzzling at first
is that all the units seem to be electrically connected, i.e. folded dipole,
reflector and directors are all mounted on a common metal rod. In fact
no electrical connection is involved as the centre of a folded dipole
aerial is at zero potential.

Table 1. Examples of Medium-wave Aerial Coil Windings Matching

Tuning Capacitor 0-500 pF*

Ferrite rod dia. † in. mm wire size no. of turns tap at alternative
s.w.g. coupling coil

6·5 36 or 38 80 60 turns 16 turns

8 36 or 38 70 55 turns 15 turns
10 28 or 32 60 50 turns 10–15 turns
Ferrite slab †
(19 mm) × (3 mm) 28 or 32 60 10 turns 10 turns
* Increase turns by approximately 25 per cent to match a 0–350 pF tuning capacitor.
† Length is not important.
 Chapter 2
Aerials and Aerial Coupling

Satisfactory performance can usually be obtained for long wave and

medium wave reception with a simple ferrite rod aerial fitted internally,
although this will be very directional. The rod is usually mounted
horizontally (or intended to be used in that attitude). It will produce
maximum signal strength when at right angles to the direction of the
transmitting station’s aerial, with falling strength at any lesser angle,
reaching a minimum when the rod is pointing towards the transmitting
aerial. This, in fact, applies to any aerial formed from a coil or loop of
wire. Maximum signal strength is received when a loop is in line with
the transmitting aerial; and minimum signal (theoretically zero signal)
when the loop is at right angles.
To render such an aerial non-directional it is necessary to connect
it with an external aerial, which in its simplest form is a length of wire
run vertically. Theoretically, at least, such an aerial would work on its
own as a tuned circuit, if it was made a resonant length of 1, ½, ¼, ⅛,
etc. of a wavelength (see also Table 2 at end of chapter). This would
necessitate a very high aerial to give any reasonable degree of signal
amplification, and also one whose length would have to be adjusted to
tune to different broadcast frequencies.
In practice, therefore, any shorter length can be used, connected to a
conventional tuned circuit as in Fig. 8. The latter is then used to adjust
the resonant frequency of the whole. Since the actual length of aerial
wire to be used is small, compared with a wavelength, the tuned circuit
can be designed to cover the normal frequency range required as in
Chapter 1. The external aerial is then regarded simply as an appendage.
Its addition will modify the performance of a tuned circuit to some
extent. The longer – and thus the more efficient it is – the greater the
modifying effect.
This is one reason why the average radio receiver fitted with an
internal ferrite rod aerial normally has no provision to plug in an
external aerial. Connecting an external aerial would simply mis-tune
26 Building and Designing Transistor Radios

the set, which is calibrated against the adjustable tuning range provided
by its own internal tuned circuit.
In other words, if an external aerial is to be used the tuning should
be calibrated with external aerial attached. If it is to be provided as
an additional facility, e.g. to improve reception on weak signals, then
the method of coupling to the existing tuned circuit needs particular
attention if mistuning is to be avoided. In the first case it is possible
to connect the external aerial directly to the tuned circuit as in Fig. 8.

Fig. 8 External aerial connected to the tuned aerial circuit.

Normally it would be connected directly to the ‘hot’ end of the coil,

as this should give maximum sensitivity. However, better selectivity
may be produced if the connection is made lower down the coil, at the
expense of sensitivity. In most practical receiver circuits, other than
elementary designs, the aerial circuit is inductively coupled to the next
stage of the receiver. This provides variable control of the coupling, and
the possibility of better matching of the aerial to the input requirements
of the following stage.
The principal trouble associated with the use of a simple wire aerial
is that it can have a considerable capacitance effect. The fact the length
of wire also has resistance, and thus effectively inductance, is why it
can work on its own as a tuned circuit. The actual modifying effect
of a wire aerial is not readily determinable, and varies both with its
Aerials and Aerial Coupling 27

length and frequency of signal. It could be high enough seriously to

interfere with the working of a tuned circuit.
The answer to this is surprisingly simple – just connect the external
aerial via a small value series capacitor, as in Fig. 9. The effective
capacitance added to the tuned circuit cannot be greater than the
value of this series capacitor and so choosing a small value of 100 pf
or less will ensure that the tuned circuit is not mis-tuned to any great
extent. Also this method of coupling prevents the tuned circuit from

Fig. 9 Coupling capacitor between an external aerial and tuned circuits

limits ‘capacitance effect’ of the external aerial, which tends to modify tuning.

being swamped by the external aerial, i.e. the two are under-coupled
rather than over-coupled. This should improve both selectivity and
sensitivity and very large aerial lengths can be used without upsetting
the working of a tuned circuit. On the other hand, if the tuning of the
receiver is to be calibrated with an external aerial attached, it would be
advantageous to use a higher value of coupling capacitor in order to
make the external aerial more effective.
As previously noted, the greater the height of a vertical aerial the
better. However the effective length can be increased by connecting
a vertical wire to a horizontal wire producing an aerial in the form
of a large ‘T’ or an ‘L’. The longer the horizontal run the greater the
signal strength generated in the vertical wire. The longer the length of
28 Building and Designing Transistor Radios

horizontal wire the better, with the advantage that this is easier to
arrange than vertical wiring. This again adds a directional quality to the
area by virtue of a horizontal length which can be eliminated by running
the horizontal wire through a 90-degree bend. This arrangement is
quite suitable for accommodating in a loft or even in a single room.
Note: the ‘directional’ characteristics of a horizontal wire aerial are
the opposite to that of a coil or loop aerial. It is most effective when
the wire is in line with the transmitting station.
There are more elaborate forms of receiving aerials, but these are
not necessary for long-wave and medium-wave reception with the
high efficiency offered from modern tuned circuits. In fact, internal
ferrite rod aerials are generally quite adequate – except for the special
requirements of short wave and VHF reception. Coverage of the former
is too specialized a subject for this present book, but dipole aerials are
commonplace for FM reception (see Chapter 1).

VHF Aerials
The conventional ferrite rod aerial is basically unsuitable for VHF/ FM
reception because of the high inductance of the magnetic core. This makes
it impossible to accommodate the frequencies covered (88-100 MHz)
associated with practical values of variable capacitors. A favourite form
of aerial is a simple (telescopic) dipole. A single (telescopic) vertical
aerial of resonant length may also be used in some cases.
A good aerial system is essential for satisfactory FM radio receivers.
Requirements are also complicated by the fact that VHF/FM signals are
both highly directional and also polarized (i.e. radiated in a horizontal
or vertical plane). The first situation is met by aligning the dipole aerial
with the broadcast station to get maximum signal strength. The second
is met by aligning the aerial in the plane of polarization. Most VHF/FM
signals are horizontally polarized, calling for the aerial to be aligned in
a horizontal plane.
The theoretical resonant length of a dipole or half-wave aerial is
calculated from the formula:

where f is the frequency in MHz.

Aerials and Aerial Coupling 29

Thus to match a nominal middle frequency in the VHF range the

value of f can be taken as 95 MHz, when

Strictly speaking theoretical length should be corrected for

‘end effect’ losses, and also the size of the wire or tube used for
conductors. However since the match is only nominal to the middle
of the range to be covered, and the resonant frequency of the tuned
circuit will be variable, the actual length used is not that critical.
Shorter lengths can also be used in multiples, e.g. 30 in. or 15 in.
The vertical VHF aerial can be of any of these lengths. A horizontal
dipole should have a full 60 in. length. (Fig. 10.) There are other
more elaborate forms of folded dipole aerial which can be used, as
mentioned in Chapter 1.

Fig. 10 The dipole aerial—the standard form of aerial used with VHF (FM)
receivers.

Telescopic aerials have the advantage that a resonant length can be

adjusted for optimum effect, if necessary. Also the vertical attitude of a
telescopic dipole can be altered by adjusting the spread. This type of
30 Building and Designing Transistor Radios

aerial can be designed either to fit on the back of a receiver or on

a separate stand. It is also usually mounted so it can be adjusted
directionally.
It is essential that dipole aerials be connected to the receiver tuned
circuit with a properly matched cable. Dipole aerials can be taken as
offering a purely resistive load for matching purposes, with low or very
low reactance. Dynamic resistance or impedance is specified for various
proprietary types and these are usually supplied with matching cables.
A typical dipole has an impedance of about 70 ohms at the centre, and
should be connected via 70–75 ohm coaxial cable, e.g. the standard type
of cable used for connecting a TV aerial to its set. Folded dipoles have
a higher impedance, (theoretically four times that of a simple dipole, or
typically about 300 ohms), but this is reduced by the effect of a reflector
and directors in the case of an aerial array of the type already described
in Chapter 1.

Fig. 11 Two methods of coupling a dipole aerial via impedance-matching

transformers.

Since a dipole aerial involves two leads (in the feeder), it cannot be
connected simply to a tuned circuit like a simple wire aerial. Instead,
it is normally inductively coupled to the receiver circuit. Two such
Aerials and Aerial Coupling 31

methods are shown in Fig. 11. In the first illustration the dipole
aerial is inductively coupled to the tuned circuit via a centre-tapped
transformer, the transformer also providing the required impedance
match. The second shows simpler transformer coupling (again with
impedance matching) to a preselector circuit, which is commonly used
with an FM superhet receiver.

Table 2. Resonant Aerial lengths

Radio Theoretical ½-wave ¼-wave ⅛-wave

frequency resonant inches inches inches
MHz length* (metres)

80 3·75 74 37 18½
82 3·66 72 36 18
84 3·57 70 35 17½
85 3·53 69 34½ 17¼
86 3·49 68 34 17
88 3·41 67 33½ 16¾
90 3·33 66 33 16½
92 3·26 64 32 16
94 3·19 62 31 15½
96 3·125 61 30½ 15¼
98 3·06 60 30 15

* The practical equivalent aerial length is slightly less than the theoretical value.
 Chapter 3
Semi-Conductor Diodes and
their Characteristics

A diode is an electronic device with two electrodes. Its basic action is

that of a rectifier, i.e. it allows current to flow in one direction but not
in the other. A semi-conductor diode also has the characteristics of a
capacitor, as will be explained later.
Solid state diodes consist of a P-type semi-conductor material joined
to an N-type semi-conductor material. As far as conduction of electricity
is concerned, P-type material conducts by positive carriers (known as
holes), and an N-type material by negative carriers, or electrons.
A simple combination of a P-type material and an N-type material
joined or connected together will not work. It has to be a single
crystalline element, one side of which is given P-type characteristics
by ‘doping’ with a suitable impurity, and the other side given N-type
characteristics by doping with another impurity. The two zones are
bounded by a P-N junction. Mobile charges in the two materials will
then tend to diffuse across this junction, cancelling each other out and
forming what is in effect a neutral zone at the boundary, known as a
depletion layer – see Fig. 12.

Fig. 12 Diagrammatic representation of the production of a

semi-conductor diode.
Semi-Conductor Diodes and their Characteristics 33

This depletion layer, which is extremely thin, effectively isolates

the fixed charges present in the crystal – positive on the N side and
negative on the P side – and also prevents migration of further mobile
charges across the junction, unless the potential barrier at the junction
is upset by applying an external electromotive force (emf) or voltage to
the device.

Fig. 13 The two modes of working a diode.

A ‘demonstration’ circuit is shown in Fig. 13, using a battery for the

source of emf and a means of varying the voltage. In the first diagram
the positive side of the battery is connected to the P side of the diode.
Starting from zero and gradually increasing the voltage the first effect
is to push holes from the P material towards the junction and attract
electrons from the N materials towards the junction until the potential
barrier is neutralized. If the emf is then increased further the diode
will become fully conducting, with current flowing through it. In other
words, applying positive bias to the P side of the diode has turned it
into a conductor.
If reverse bias is applied (negative to the P side) as in the second
diagram, both holes and electrons are repelled from the junction,
increasing the extent of the depletion layer. In effect the junction sets
up a back voltage equal to the applied voltage and a diode behaves
34 Building and Designing Transistor Radios

as a non-conductor. In practice there will be a small leakage current,

although this can be negligible for a majority of applications.
The typical working characteristics of a semi-conductor diode are
shown in Fig. 14. Note that this presentation, which is conventional, is
the opposite way round to the two diagrams in Fig. 13. In the forward

Fig. 14 Characteristic performance of a diode in forward working

(conducting) and reverse working (non-conducting).

direction, increasing voltage at first has no effect until the potential

barrier is neutralized. Typically this will occur at about 0·2 volts with
a germanium diode and 0·6 volts with a silicon diode. From then
on, increasing voltage increases the current flow, limited only by the
specific resistance of a diode material and any other resistance in the
circuit. Ultimately, if the current rises to too high a figure, the diode
would become overheated and burn out.
In the reverse direction the effect of increasing voltage is a very slight
increase in leakage current up to the saturation point of the diode.
In practice the leakage current can usually be considered constant,
Semi-Conductor Diodes and their Characteristics 35

unless excessive voltage is applied. When this occurs the diode

becomes over-saturated and the reverse current rises sharply. If this is
not limited to a safe level by another resistance in the circuit the diode
will overheat and be destroyed.
One inherent limitation with reverse working (apart from the pres-
ence of a leakage current which makes a diode an imperfect rectifier)
is that heating effects are present due to power dissipation, and the
consequent rise in temperature will automatically result in an increase
in the leakage current itself. In the case of a germanium diode this can
be quite marked, e.g. the leakage current typically doubles with each 9°
Centigrade temperature rise. The effect is far less marked with a silicon
diode, which is often a preferred type for this reason. On the other
hand the effect of heating on current flow, when the diode is operating
in the forward direction, is usually negligible.

Types of Diodes
Two different types of diodes have already been mentioned –
germanium diodes and silicon diodes. Silicon diodes are less sensitive
to temperature and thus generally maintain a low reverse current.
Working in the forward direction, a higher initial voltage is needed to
start the current flowing.

Fig. 15 Two different types of diode construction.

Diodes may also be described by their construction, e.g. point-

contact type or junction type. With a point-contact diode a small N-type
crystal touched by a pointed wire is enclosed in a suitable envelope. It is
treated during manufacture to give P-type characteristics to the crystal
in the vicinity of the point – Fig. 15. In the case of a junction diode, its
36 Building and Designing Transistor Radios

construction is essentially similar to that of a transistor (see Chapter 5),

comprising a P-type pellet alloyed or diffused to an N-type crystal slice.
Like a transistor it is encapsulated. Point-contact diodes are almost
exclusively made from a germanium crystal. Junction-type diodes may
be germanium or silicon, but most are made from silicon.

Capacitor Characteristics
Working in the reverse direction, the depletion layer is a non-
conducting barrier between the two ends of the diode and so virtually
forms a dielectric between two electrodes. In other words, the diode
working in this direction is equivalent to a capacitor. Hence it has an
ability to pass alternating currents which may be present.
The degree of capacitance effect present depends mainly on the area
of a junction – the larger the area the greater the capacitance present
when working in the reverse direction. In practice, the capacitance
effect is not generally significant in the passage of ac up to frequencies
of about 30-50 Hz, but at very much higher frequencies capacitance
effects can be considerable. The only way in which they can be reduced
is to reduce the junction area, which is why a point-contact diode is
normally preferred for circuits carrying very high frequency currents.
There is another type of diode which uses this capacitance effect in a
different way. Beside the junction area, capacitance also depends on the
thickness of the depletion layer. This, in turn, will vary with the reverse
voltage. Increasing the reverse voltage increases the thickness of this
layer and reduces the capacitance. Reducing the reverse voltage has
the effect of increasing the capacitance. Thus a diode can be worked
as a capacitor varying with voltage, such devices being known as
variable capacity diodes or variactors or varicaps. They are often useful
for stabilizing tuning circuits.

Zener Diodes
A Zener diode is a special form of silicon junction diode with a low
and constant breakdown voltage. A typical performance curve is shown
in Fig. 16 (compare with the performance of a conventional diode,
Fig. 14). The sharp break from non-conducting to almost perfect
conductance at a particular reverse voltage is called the zener knee.
This characteristic makes a zener diode a particularly useful device for
producing voltage-stabilizing or current-stabilizing circuits. A Zener
Semi-Conductor Diodes and their Characteristics 37

Fig. 16 Working characteristics of a Zener diode.

diode is normally operated at or above breakdown voltage, with

sufficient resistance in the circuit to limit the actual current flowing in
the circuit to a safe figure for the zener diode used.
Thus in a simple demonstration circuit shown in Fig. 17, a zener
diode is connected in a series with a limiting resistor to a supply
voltage. This resistor effectively divides the supply voltage into the
breakdown voltage across the diode with the remainder dropped
across the resistor. The voltage across the zener diode remains constant,
even if the supply voltage varies. Thus tapping the circuit at this point
can provide a constant voltage output. The only thing that happens
with a variable supply voltage is that the current flow will rise or fall in
proportion. Only if the supply voltage falls so much that the working
point of a diode is pulled back past its knee point will the constant
voltage output cease.
Only one constant voltage can be tapped from the zener diodes,
corresponding to the breakdown voltage or zener voltage as it is
generally called. Typically this is of the order of 5 to 6 volts. Being a
38 Building and Designing Transistor Radios

Fig. 17 Simple Zener diode circuit capable of giving a constant

voltage output. This can be incorporated in another circuit to provide a
stabilized supply voltage.

small device it is also obvious that the power it can provide in a constant
voltage circuit is also limited, usually to the order of milliwatts only.
Zener diodes are, however, produced with much larger zener voltages
and also power capacities of the order of watts.

Specification Figures
The actual working characteristics of a diode are described by the
following:

Maximum reverse voltage – with figures normally given for peak (or
absolute maximum which the diode will tolerate) and average. Separate
values may also be given for different temperatures, i.e. ambient (usually
20° Centigrade or 25° Centigrade); and at some higher temperatures
(usually 60° Centigrade). In the absence of separate temperature
ratings ambient temperatures can be inferred.
Maximum forward current – given in milliamps. Again separate figures
may be given for peak and average; and separate values for ambient
temperature and 60° Centigrade.

Other data which may be given are:

Ambient temperature range – referring to the maximum and minimum

ambient temperatures, within which range the diode will not be harmed.
Maximum junction temperature – usually separate figures for
continuous and intermittent operations.
Thermal resistance – in ° Centigrade per millivolt.
 Chapter 4
The Diode Detector

The detector in an AM radio receiver has to work as a rectifier,

suppressing half of each cycle of an amplitude-modulated carrier
wave signal and converting the current variations in the other half
into a direct current fluctuating in the same manner as the original
modulation. In simpler terms, it extracts the audio frequency or af
content of the modulated radio signal in the form of a fluctuating dc
current.
The semi-conductor diode is a logical choice for achieving this, but
its performance is far from perfect. The ideal rectifier would have no
forward resistance and infinite backward resistance. Semi-conductor
diodes have a low forward resistance and high backward resistance,
but the loss of rectification imposed by these characteristics can be
recovered by the addition of a load resistor, and a reservoir capacitor in
parallel with it as in Fig. 18.

Fig. 18 Basic detector circuit using a load resistor and reservoir capacitor
to compensate for the imperfect rectifier characteristics of a diode.

Theoretically, the higher the value of the load resistor the greater
the rectified voltage. Its value, however, has also to be chosen to
match the input resistance of the next stage. The function of the
reservoir capacitor is to maintain the voltage during half cycles when
the signal is chopped off. Its value, together with that of the resistor,
40 Building and Designing Transistor Radios

determines the time constant of the circuit. To avoid loss of top frequency
and distortion of a signal this has to accommodate the highest modu-
lation frequency likely to be present – say 10,000 Hz for AM radio. This
implies a time constant of not more than 50 microseconds (the time
between a ‘peak’ and ‘trough’ of a 10,000 Hz signal), and preferably
rather less.
Starting point in deciding suitable component values therefore is
first to select R as a suitable match, and then determine the value of
C to give a time constant for the circuit of 50 μseconds or slightly less.
At the same time the value of the reservoir capacitor must be high
enough to maintain a suitably high reservoir action at the lowest radio
frequencies present in the signal.
In the case of a straightforward transistor radio receiver the
detector is followed by a transistorized audio amplifier stage, the input
impedance of which is typically of the order of 1,000 to 2,000 ohms. A
suitable value for the load resistor (R) would therefore be 1 k ohm up
to possibly 4 k ohm. Corresponding values of the reservoir capacitor to
give a CR or time constant of 50 μseconds would be:

R = 1k, C = 0·5 μF
R = 2k, C = 0·25 μF
R = 4k, C = 0·125 μF

It would be desirable to use rather lower values of capacitor than

this so that the time consonant is less than 50 μseconds (say 40 or 45
μseconds), e.g. 0·20 μF with R = 2k, and 0·12 μF with R = 4k.
The order of load resistance values also favours the use of a
germanium-point contact diode. Junction diodes are less efficient
as AM detectors because of their greater self capacitance, and quite
unsuitable as VHF or UHF detectors. However, in practice quite dif-
ferent values of R and C will work with a germanium diode, but if the
time constant is greater than 50 μseconds there will be a loss of top
performance and distortion. If much less than about 40 μseconds there
will be a loss of detection efficiency at all modulation frequencies.
This simple detector circuit, in fact, allied to a simple tuned circuit
(aerial coil), is the basis of a crystal set – Fig. 19. In this circuit the load
resistor R can be omitted as the high impedance phones will effec-
tively act in its place. It will also work without the reservoir capacitor,
The Diode Detector 41

although with more distortion present. By experimenting with differ-

ent values of C to adjust the time constant of the circuit, it is possible to
establish an optimum value for detection efficiency and/or distortion,
but not necessarily both together.
Incidentally, as well as virtually any germanium-point contact diode,
almost any af transistor will work in this circuit as a detector. The
collector lead is ignored and the transistor connected (and worked) as
a diode using the base and emitter connections only.

Fig. 19 Coupled to a tuned aerial circuit L and CI, a diode detector

completes a ‘crystal receiver’. R can be 2·2 k ohm or 4·7 k ohm.
C2—0·25 to 8 μF.

A significant feature of this basic detector circuit is that it must

be coupled to a high impedance output load to work properly. If a
low impedance load is connected directly to a tuned circuit (such
as low impedance phones in the case of this circuit) it will have a
‘damping’ effect which will drastically reduce the efficiency of the
tuned circuit and make its response very flat. This applies as a basic
rule.
The detector circuit can be improved by removing any unwanted
signal content. The diode will inevitably pass a small proportion of rf
signal which, if allowed to reach the next stage, will be amplified to
cause distortion, or even excite self-oscillation. This can be remedied
by the adoption of filters – in practice a series resistor and a parallel
capacitor.
There will also be a signal present consisting of a dc voltage with an
average level of the output wave form. Again it is undesirable that this
42 Building and Designing Transistor Radios

should reach the amplifier stage. It can be removed by inserting a

blocking capacitor in the output stage.
These additional circuit components are shown in Fig. 20, where
Rf and CF are the resistor and capacitor, respectively, forming the rf
filter; and CB is the blocking capacitor. Choice of values is to some
extent arbitrary. RF needs to be fairly small so as not to reduce the

Fig. 20 Detector circuit further developed with filter components and

blocking capacitor. Typical component values:
R—2·2 to 3·9 k ohm
C—0·25 to ·5 μF
RF up to 1 k ohm
CF—5 to 10 μF
CB—5 to 10 μF

signal strength unduly – say 1 k ohm or less. The value of CF is then

chosen to give the ‘RC’ combination the required characteristics to
make the filter work, i.e. eliminate all frequencies present above the
audible frequency range – say above 15 kHz.
The basic relationship involved is

Written as a solution for RC

taking f = 15 kHz
RC = 10·6 × 10–6
Thus if R is 1 k ohm
C = 10·6 × 10–6
or 10 μF, say.
The Diode Detector 43

A similar value is also usually suitable for the blocking capacitor, but
almost any value between 5 μF and 10 μF should be suitable.
The complete, practical detector circuit also normally incorporates a
volume control or variable resistor RV, as shown in Fig. 21. The volume
control can in fact be placed elsewhere in the receiver circuit – see later.
The usual value for this potentiometer in a transistorized radio circuit
is 5 k ohm. The blocking capacitor can be removed to the output side
of the volume control, where it also doubles as a coupling capacitor;
or left in the same position as in Fig. 20 if another form of coupling is
used. In the former case the value may be adjusted to suit the coupling

Fig. 21 Circuit of Fig. 20 with volume control added. RV = 5 k ohm.

Removal of the blocking capacitor to the output side reduces distortion from
the aerial circuit.

requirements – see Chapter 10. It is also possible to eliminate resistor

R in this circuit, choosing a slightly higher value for RF to compensate
(and adjusting the value of CF to match). This can prove an advantage if
the signal strength is weak and it is desirable to increase it by reducing
the load resistance in the circuit. It will be an advantage in a crystal set,
for example, and in a simple receiver with only moderate amplification
following the detector.

Input and Output Resistance

The effective input resistance offered by the tuned circuit is R/2,
ignoring the forward resistance of the diode which is negligible by
comparison. This represents the loading on the tuned circuit, and an
opportunity to re-select the value of R if necessary (and at the same
time modifying the value of C required as explained earlier).
44 Building and Designing Transistor Radios

With the usual 5 k ohm volume control the load resistance at

maximum volume will normally be of the order of 2 k ohm to 4 k ohm,
depending on the values of R and RF used.

Detector Performance
Besides being a vital link in the radio circuit, the detector is also
responsible for the quality of a signal passed to the next stage. Any
distortion or unwanted signal passed on will be amplified. Ideally
the detector should act as a perfect linear amplifier. Unfortunately to
work at all it has to include non-linear components, which means that
af currents passed by the detector circuit will not exactly follow the
modulation of the original rf signals. The more marked the non-linear
characteristics, the greater the resultant distortion.
The two main trouble spots are the load resistance and stray
capacitance. Basically a fairly low load resistance is necessary in order
to maintain enough bias on the diode throughout each working cycle,
and also to ‘match’ the detector to the following transistor stage. There
is no real answer to this other than employing an FET amplifier when
a very much higher load resistance can be used.
Stray capacitances can be generated by leads and close proximity
of other current carrying components. To reduce their effect, wires
(or printed circuit leads) should be kept as short as possible and the
diode should be positioned as far as practicable from leads and other
components, although there is no need to carry this to extremes.

FM Detectors
The FM detector has to handle and rectify a quite different type of
signal. Instead of the carrier wave being modulated by amplitude, it
is modulated in frequency. Thus for FM reception the detector has to
extract the frequency modulation and convert it into a corresponding
af signal. At the same time it must reject any amplitude modulated
signal which may be present. These functions are performed by a
discriminator and limiter respectively. Alternatively, a ratio detector
circuit can be used which combines functions of both discriminator
and limiter. This would seem the simpler solution as the two types of
circuit are very similar. However, the ratio detector tends to produce
slightly more distortion and only has about half the sensitivity of the
discriminator, although the latter fact is not necessarily significant.
The Diode Detector 45

There are also two other types of FM detectors which can be used,
the pulse counting detector based on digital ICs which requires no
tuned circuit and has excellent linearity; and a crystal discriminator
which is manufactured as a sealed unit and requires only an external
variable capacitor to set the slope of the detector (i.e. adjust the output
voltage so that it swings an equal amount positive and negative).
A basic discriminator circuit is shown in Fig. 22. This is, in effect,
two AM detectors back to back, fed with rf signal from a tuned circuit
L1C1 and also receiving half the voltage from a second tune circuit
L2C2. The same basic circuit will also work as a ratio detector (i.e.
as a limiter as well) by reversing one of the diodes and breaking the
connection between one side of L2C2 and the load resistors and taking
the output from these two points.

Fig. 22 Basic circuit for a phase discriminator which can also be modified
to work as a ratio detector.
R1 and R2 are load resistors
C3 and C4 are reservoir capacitors
C5 is a decoupling capacitor (only necessary if the circuit is worked as a ratio
detector).

In practice simple circuits are normally used based on an if trans-

former to produce FM-to-AM conversion, operating at the intermediate
frequency of a superhet receiver. Voltage developed on one side of
the secondary leads the primary voltage, and that on the other side
lags by the same phase angle when the circuits are resonant to the
46 Building and Designing Transistor Radios

unmodulated carrier frequency. This yields equal and opposite voltages

passed by the diodes. In the presence of a frequency modulated signal
there is a shift in the relative phase of the voltage components, voltage
applied to one diode increasing and the other decreasing. The difference
between the two, after rectification, is the audio or ap frequency output
of the detector.
A practical ratio detector circuit is shown in Fig. 23, the transformer
in this case having three windings. The primary and secondary are

Fig. 23 Practical ratio detector circuit. Component values:

R1 and R2—4·7 k ohm
R3—2·2 k ohm
R4 and R5—68 ohm
Ci and C2—330 μF
C3—8 or 10 μF
C4—0·01 μF
C5—330 μF
C6—0·001 μF
Diodes—point contact type

tuned to the rf frequency. The close coupled tertiary injects a voltage

into the secondary in series with the voltage across each half of the
secondary. It works essentially as a discriminator, with the reversed
diode acting as a limiter.
The Diode Detector 47

Fig. 24 Simplified practical ratio detector circuit. Component values:

R1—20 k ohm
C1—180 pF
C2—56 pF
C3 and C4—220 or 330 pF
C5—10 μF
Diodes—AA119

A more simplified ratio detector circuit is shown in Fig. 24, capable

of giving just as good results.
 Chapter 5
Transistors

To design and/or construct radio circuits it is not necessary to know

how a transistor is made – only how it works. However, different types
of transistors are often described by their construction, e.g. junction
type, silicon planar type or epitaxial. The following brief descriptions
of the different transistor manufacturing processes can therefore be
helpful in giving meaning to these descriptions. Mention of point
contact transistors is omitted since this was the original form of con-
struction, now obsolete.
The working characteristics of a transistor depend on the behaviour of
a P-type (positive type) and an N-type (negative type) semi-conductor
material in a three-element combination. Such combinations yield
either a P-N-P or N-P-N device.

Fig. 25 Construction of P-N-P and N-P-N transistors shown in elementary

diagrammatic form.

Starting point in the manufacture of a transistor is the production

of a crystal of either P-type or N-type germanium or silicon, which is
cut into slices. This is the base element of the transistor. The simplest
method of construction is then to fuse a pellet of opposite type material
Transistors 49

to each of the slices, i.e. two pellets of P-type material to an N-type

slice to make a P-N-P transistor; or two pellets of N-type materials to a
P-type slice to make an N-P-N transistor – Fig. 25.
These pellets form a collector and emitter of the transistor. Wires are
soldered to each of the three elements and the complete device is then
encapsulated – Fig. 26. Transistors made in this fashion are known as
alloy-junction types, the method being widely used for production of
germanium transistors.

Fig. 26 Basic stages in the manufacture of a junction-type transistor.

A slightly different method is used to produce a junction transistor

by alloy diffusion. Suppose a P-N-P transistor is required. A layer of
N-type material is deposited on one side of a P-type slice. On this are
deposited two pellets. One contains N-type additives only and the
other both P-type and N-type additives. This assembly is now heated in
an oven causing the two pellet materials to diffuse through the N-type
layer into the P-type slice. In this case the original slice becomes the
collector, the diffused N-type additive the base and the diffused N-type
and P-type additive the emitter.
In the case of a silicon transistor a different method of construction
is now normally used, known as a planar process. A slice of N-type
silicon (which eventually becomes the collector) is heated to form a
thin layer of silicon oxide. It is then given a thin coating of photo-
sensitive material resistant to etching, known as photoresist.
This coated slice is used for the manufacture of several transistors.
It is first covered by a photographic negative of the pattern of the base
50 Building and Designing Transistor Radios

positions of the number of transistors to be made and exposed to ultra-

violet light. This prints the base patterns on to the slice. The slice is
then washed, which removes all the unexposed photoresist, leaving a
print of several small windows in the photoresist through which the
silicon oxide is exposed. The slice is then etched to remove the silicon
oxide appearing in the windows, exposing silicon.
After cleaning, the slice is treated in an oven with some P-type doping
additive. This additive evaporates and some settles on the windows.
Further heating in a second oven in an atmosphere of oxygen diffuses
the P-type dope into the silicon at the window position. At this stage
the base of each of the transistors to be produced has been formed
in the slice.
The whole process is then repeated to print and etch a second pattern
of windows representing the emitter positions of all the transistors.
After etching and cleaning the baking process is repeated, this resulting
in the N-type doping additive being diffused into the emitter windows.
This is followed by another photoprinting stage to produce windows
over the base and emitter areas. The slice is then vacuum plated with a
fine layer of aluminium and a final printing and etching stage used to
remove aluminium from all areas except where it is required to provide
contacts with the bases and emitters. At these points the aluminium is
lightly alloyed to the bases and emitters.
The oxide layer is now removed from the collector side of the original
slice and the thickness of the slice reduced as required. After testing
each transistor so formed, the slice is then cut into separate pieces each
containing one transistor. Faulty transistors marked at the testing stage
are rejected. Every good transistor is then bonded, collector side down, to
a gold plated stem. These wires are bonded to the aluminium connections
to the emitter and base. Finally the transistors are encapsulated.
The process is considerably more complicated than the manufacture
of alloy-junction or alloy-diffused transistors, but a single slice may
yield from 200 to 7,000 or more transistors, depending on type. Stages
of manufacture are illustrated in Fig. 27.
A somewhat modified process is used in the production of epitaxial
silicon planar transistors. Basically the original slice is more heavily
doped, with the result that it is covered with a very thin layer of highly
doped silicon with the same crystal orientation as the original material,
this being known as the epitaxial layer. In the final product the transistor
Fig. 27 Stages in the construction of silicon planar transistor (diagrams by Mullard).
52 Building and Designing Transistor Radios

is effectively contained within this epitaxial layer and not in the

original slice (which now merely acts as a stiffening layer and back
contact). With an ordinary planar transistor the transistor is formed
in the original slice.
The epitaxial planar process is also employed to manufacture field
effect transistors. For an N-type field effect transistor an N-type epitaxial
layer is formed on a P-type substrate. After oxidation of the surface,
‘windows’ are etched in the oxide layer and diffused with super-rich
P-type additive (designated P+). This diffuses through the epitaxial
layer into the P-type substrate.
The N-type epitaxial layer forms the N-channel of the field effect
transistor. The P-type substrate forms the gate. Penetration of the P+
additive into the P-type substrate produces the N-channel isolating
region – Fig. 28. The assembly is then re-oxidized, a ‘window’ etched
corresponding to a second area and another P+ region produced to
form the upper part of the P-gate surrounding the N-channel. Super
N (N+) additive is then diffused into the ends of the N-channel.
Aluminium contact areas are bonded to the N+ regions, to which are
connected the source and drain lead wires. The gate lead is connected
to the P-type substrate. Finally the transistor is encapsulated.

Fig. 28 Elemental construction of a field effect transistor or FET.

With a P-channel field effect transistor, construction is exactly the

same, substituting P-type for N-type elements, and vice versa. See also
Chapter 11 for further information on field effect transistors.

Selecting Transistors
There are now literally thousands of different designs of transistors made
by various specialist manufacturers. In the case of circuit diagrams,
Transistors 53

or component catalogues, they are identified only by a type number,

which is the manufacturer’s coding. This normally consists of letters
and numbers, the letters usually identifying the manufacturer or
general class of the device with the associated numbers indicating the
particular design or development. Thus ‘O’ indicates a semiconductor
device manufactured by Mullard, and an appended ‘C’ that the device
is a transistor. Mullard transistors are thus coded ‘OC’ followed by a
serial number identifying the particular type. This does not help at
all unless such a manufacturer’s specific type is quoted on a circuit
diagram, which it usually is. The reason why a type number should be
quoted for a specific circuit design is quite simple. Possible variations
in the design performance of a transistor are virtually infinite, so even
if different manufacturers produce transistors for a similar purpose,
the actual characteristics of different makes can vary quite widely. This
can affect choice of matching components which govern the proper
working of the circuits.
In certain cases, particularly in the field of more or less ‘standard’
radio circuits, transistor types from different manufacturers may have
sufficiently close characteristics to be regarded as equivalents or near-
equivalents. This implies they could be substituted for a specified type in
a particular circuit. In many more cases – usually with very elementary
circuits – almost any type of transistor of the same basic type, e.g.
germanium or silicon, or functional group and polarity, may produce
similar working results.
It is particularly useful to the designer/experimenter to have access
to information both on transistor equivalents and functional groupings.
The former information is well catered for in separate publications on
the subject (complete books on their own!). Information on functional
grouping is much harder to come by. Manufacturers’ data sheets
are not readily available to the amateur designer/constructor and a
comprehensive coverage would need a complete filing system to deal
with. Retail catalogues are not usually particularly helpful, either, as
they normally list types stocked merely by type number, although
there are exceptions. The latter can provide a basis of compiling a list
of transistor types readily available by functional grouping, which can
be further expanded by reference to transistor equivalent data.
There is a welcome trend to provide more information about
transistor types and manufacturers’ coding, an example being the
54 Building and Designing Transistor Radios

Pro-Electron coding (now adopted for all new Mullard transistors,

for example). This consists of a two letter code followed by a serial
number. Restricting the code used only to diodes and transistors:

the first letter indicates the semi-conductor material, i.e.

A = germanium
B = silicon
the second letter indicates the general function of the device, i.e.
A – detector diode or mixer diode
B – variable capacitance diodes
C – af transistor (excluding power types)
D – af power transistor
F – rf transistor (excluding power types)
L – rf power transistor
S – switching transistor (excluding power types)
V – power type switching transistor
Y – rectifier diode
Z – zener diode

The serial number following then consists of three figures if the

device is intended for consumer applications, e.g. radio receivers,
audio amplifiers, television receivers, etc. A serial number consisting
of a single letter followed by two figures indicates the device is intended
for industrial or specialized applications.
For the average amateur radio enthusiast, one of the most useful
functional groupings published is that of the Electrovalue Catalogue,
albeit listing only a limited number of individual types in each group.
This covers the following:

Germanium Transistors
(i) small, medium currents switching services.
(ii) medium current switching, low power output.
(iii) small, medium current amplifiers.
(iv) af amplifiers, low power output.
(v) complementary pairs.
(vi) high power output (power transistors).
Silicon Transistors
(i) af amplifiers, small signal, general purpose.
(ii) af amplifiers, low level, low noise.
Transistors 55

(iii) small signal amplifiers.

(iv) rf amplifiers and oscillators.
(v) medium current switching, low power output.
(vi) high frequency, medium powers.
(vii) general purpose switching.
(ix) power transistors.

Any reference to ‘power’ grouping is largely arbitrary since there is

no universal agreement on the range of power levels (referring to the
maximum power rating of the particular transistor). Thus low power
may generally be taken to cover 100–250 mW, but such a grouping may
include transistors with power ratings up to 1 watt. Similarly medium
power implies a possible power range of 250 mW to 1 W (but may
extend up to 5 watts). Any transistor with a power rating of greater
than 5 watts is classified as a power transistor.

Fig. 29 Representative transistor outline shapes with dimensions and lead

identification. Dimensions are in inches.
56 Building and Designing Transistor Radios

It is also useful to be able to identify transistors by their size and shape,

and essential to be able to identify base, collector and emitter leads
.

correctly. Representative outline shapes, with dimensions in inches, are

given in Fig. 29, the coding given being more or less universal but not
always quoted in transistor specifications and still less so in catalogues.
 Chapter 6
Understanding Transistor
Characteristics

Physically a transistor is a three-element device (although it is more

correctly known as a bipolar device). It consists of a combination of
P-N-P or N-P-N substances. The first two letters in these designations
indicate the respective polarities of the voltages applied to the emitter
and collector, in normal operations, i.e:

P-N-P – emitter is positive with respect to both collector and base;

and the collector is made negative with respect to both the emitter
and base.
N-P-N – emitter is negative with respect to both collector and base;
and the collector is made positive with respect to both the emitter
and base.

More simply –
P-N-P – the collector must be Negative.
N-P-N – the collector must be Positive.

Fig. 30 Forward and reverse diode circuits.

58 Building and Designing Transistor Radios

These correct polarities are obtained by applying bias of the right

polarity to each ‘side’ of the transistor. To get a clear picture of what
these ‘sides’ are, consider the characteristics of a diode (Chapter 4).
Depending on the polarity applied to it, it works on the reversed
bias mode with no current flow, or in the forward direction with the
characteristics of a conductor allowing current to flow – Fig. 30.
Now consider a pair of diodes joined with a common centre to make
a three-element device, either N-P-N or P-N-P – Fig. 31. (The diagram
shows a N-P-N combination.) This is basically what a transistor is, but
with the centre layer made very thin.

Fig. 31 Two ‘diodes’ joined ‘back-to-back’ to make a transistor.

Fig. 32 Supply for N-P-N transistor (A) with bias voltage added (B).

Now consider the effect of applying a voltage to each ‘half ’, taking

a N-P-N transistor as an example. With the voltage connected to the
‘top and bottom’ as in Fig. 32 A, the ‘top’ half is in the condition of
Understanding Transistor Characteristics 59

a reversed-biased diode, so no current will flow (other than a small

leakage current). This voltage so applied across the whole of the tran-
sistor is known as the supply voltage.
Applying a second source of emf or bias voltage to the middle layer
and the ‘bottom’ as in Fig. 32 (B) will immediately alter the situation.
The bottom ‘half ’ is now working as a forward diode. This will excite a
flow of electrons into the upper part (the reason for making the middle
layer very thin is to make this flow easy to establish), so that the whole
of the transistor becomes conductive as far as the supply is concerned.
In other words application of the correct bias produces a small current
flowing through the left hand side of the circuit, with a very much
larger current flowing through the right hand part of the circuit. These
circuit currents can readily be measured by meters at meter 1 and
meter 2 positions, respectively. Fig. 33 shows this mode of working for
both a N-P-N and P-N-P transistor.

Fig. 33 Supply and bias connections to N-P-N and P-N-P transistors.

Still visualizing the transistor in block form, with the supply and bias
connected as in Fig. 34, the basic working conditions are:

(i) The bottom ‘half ’ is a source of free electrons when correct

bias is applied, hence the name emitter.
(ii) The top ‘half ’ readily accepts or collects these electrons when
the transistor is working, and is known as the collector.
60 Building and Designing Transistor Radios

(iii) The middle layer is a point of application of the bias. It could

have been called the ‘bias’, but that description is reserved for
voltages. Alternatively it could have been called the grid, for it
works like the control grid of a triode valve. However, for no
particular reason which is clear, it is called the base.

Fig. 34 Input and output circuits of a transistor. Polarity shown is for an

N-P-N transistor.

Fig. 35 Transistor symbols.

Thus the three elements of a transistor are called the emitter, col-
lector and base, usually designated by the lower case letters e, c and
b respectively. They are represented symbolically as shown in Fig. 35
(sometimes the enclosing circle is omitted):
Understanding Transistor Characteristics 61

(i) the base by a thick black line.

(ii) the collector by a thin line.
(iii) the emitter by a thin line with an arrow. The arrowhead points
in the direction of positive current flow, so it is the opposite
way round for P-N-P and N-P-N transistors.

Basic connections, with appropriate polarity, for P-N-P and N-P-N

transistors, are shown in Fig. 36.

Fig. 36 Polarity of bias and supply for working of P-N-P and N-P-N
transistors in common emitter configuration.

We can now draw a ‘demonstration’ circuit for a transistor in which

its working characteristics can be measured and defined – Fig. 37.
Provision is included to vary both the supply voltage and bias voltage
and meters are included in the circuit where shown to read:

Meter 1 (microammeter) – base current or IB

Meter 2 (voltmeter) – base voltage or VB
Meter 3 (milliammeter) – emitter current or IE
Meter 4 (milliammeter) – collector current or IC
Meter 5 (voltmeter) – collector voltage or VC

With switch 2 closed and switch 1 open the collector/base of the

transistor acts as a reverse (diode) junction. The only current that can
62 Building and Designing Transistor Radios

flow through the reverse junction is a very small leakage current. There
will, in fact, be no current flow unless the bias circuit is completed, but
this only requires a very small bias voltage or bias current, virtually
equivalent to VB = 0 or IB = 0 with switch 1 closed. For most practical
purposes this leakage current can be ignored.

Fig. 37 ‘Demonstration’ circuit for the measurement of

transistor characteristics.

Parameters of particular importance in determining (or specifying),

the performance of a transistor are:

(i) Output Characteristics – or how collector current varies with

bias and collector voltage (supply voltage). Bias is usually
expressed in terms of base current.
(ii) Input Characteristics – or how base current varies with base
voltage.
(iii) Transfer Characteristics – or the relationship between collector
current and base voltage.
(iv) Current Amplification – or the ratio of collector current to base
current or signal current gained.
(v) Ratio of Collector to Emitter Current – This was originally
known as the alpha (a) of a transistor, now designated hfb.

The output characteristics (IC/VC) of a transistor can be measured in

terms of how the collector current (IC) varies with collector voltage (VC)
for different values of base current (IB). Typical curves are shown in
Fig. 38 where it will seem that above the transistor ‘knee’ point, collector
Understanding Transistor Characteristics 63

current is fairly constant with increasing collector voltage for all values
of bias (base current). It also indicates that the output resistance of a
transistor is quite low.

Fig. 38 Characteristic curves for small transistors.

The input characteristic (IB/NB) can be plotted by measuring base

current (IB) against base voltage (VB) for a constant collector voltage.
In this case a typical form of the curve is as shown in Fig. 39. The input
resistance at any point in this case is represented by the gradient of the
curve or the ratio of VB to IB.
64 Building and Designing Transistor Radios

Fig. 39 Base current plotted against base voltage.

The transfer characteristics (IC/VB) yield a similar shape of curve –

Fig. 40. These characteristics are determined at a constant collector
voltage but since collector current does not vary much with collector
voltage (see Fig. 38), one curve is usually representative of transfer
characteristics and all values of collector voltage. This is particularly
true of germanium junction transistors. With silicon transistors the
characteristic curves tend to depart from the generalized form shown.
Also the collector current (leakage current) and zero base current (or
zero base voltage) is very much smaller. But the main difference in the
characteristics is that a higher base voltage (bias) is needed to cause an
appreciable amount of collector current to flow in the case of a silicon
transistor, i.e. to operate the transistor above the ‘knee’ point. This bias
is about 0·6–0·7 volts for a silicon transistor but only 0·1–0·2 volts for
a germanium transistor.
The current amplification factor (IC/IB) can be derived by plotting
collector current against base current. Typically this gives a straight
line relationship – Fig. 41. This was originally called the beta (β) of a
Understanding Transistor Characteristics 65

transistor but is now designated hf or hfe. Actual values may range

from as low as 10 up to several hundreds, depending on the type of
transistor, also the slope of the IC/IB curve is not always constant, i.e.,
the value of hf may have a spread of perhaps 50 per cent on either side
of a nominal value.

Fig. 40 Collector currents plotted against base voltages.

The ratio of collector and emitter currents (IC/IB) is not of partic-

ular significance. Emitter current must equal IC + IB, but since IB is
small in comparison with IC, emitter current (IB) is roughly equal to
IC. The smaller the base current the closer the collector current comes
to being equal to the emitter current, and thus the closer the ratio of
hfb comes to unity. In practice, typical values of hfb lie between 0·92
and 0·98.
66 Building and Designing Transistor Radios

Fig. 41 Collector current plotted against base current.

Ratings
Individual types of transistors are also given ratings, representing
maximum values the transistor can handle in a circuit. These are:

Maximum collector – base voltage with emitter open circuit, designated

Vcbo max.
Maximum collector – emitter voltage, designated Vce max.
Maximum base – emitter reverse voltage or bias, designated Veb max.
Maximum collector current, designated IC max.
Understanding Transistor Characteristics 67

Maximum collector – base current with emitter open circuit, designated

Icbo max.
Maximum total power dissipation, designated Pt max.

Circuit Considerations
The three elements of a transistor represent separate resistance in a
simple direct circuit as shown in Fig. 42. The input circuit embraces
the base resistance (rb) and the emitter resistance (re). The output cir-
cuit embraces the collector resistance (rc) and any load resistance (RL),
and the emitter resistance (re). In other words the emitter resistance is
common to both input and output circuits.

Fig. 42 Resistive elements of a transistor.

The input resistance to a simple direct voltage is represented by the

base resistance (rb) and the emitter resistance (re) in series. The emitter
resistance is small compared with the base resistance – typically 10
ohms as compared with 1,000–3,000 ohms for a small transistor.
Hence the base resistance is more or less equal to rb, with the following
relationship applying

However, when the transistor is actually working the presence of

an input excites an amplified current in the output, this current being
68 Building and Designing Transistor Radios

equal to IB × hfe. The effective emitter resistance is thus modified

by feedback and becomes re + re × hfe. Thus the total effective input
resistance is:
rb + re (1 + hfe)
The effect of re can be ignored as it is working as a reversed junction
where its resistance is high enough not to have a shunting effect on the
input circuit.
This relationship only holds for as long as the output load is zero.
Under realistic working conditions with a positive load the collector
current (IC) is less than IB × hfe and so the feedback is less. In other
words the effective input resistance decreases with increasing loads.
The amount of which this resistance is reduced is directly related to
the conductance in the circuit, i.e. the intrinsic inductance represented
by 1/rc and the load inductance represented by 1/RL.
To simplify calculation both the feedback and conductance effects
can be ignored, using the original equation:

Both the current gain and voltage gain are also affected by feedback,
but again ignoring this for simple calculation

These formulas will exaggerate both current gain and voltage gain,
but not to any great extent under normal working conditions. Certainly
the difference will usually be within the spread of characteristics of
individual transistors of the same type.

Bias Circuits
The basic circuit of Fig. 37 derives base bias from a separate battery.
In any practical circuit (other than a test or demonstration circuit),
base bias voltage is normally taken from the collector supply. There
are various ways in which this can be done, the simplest being to ‘tap’
Understanding Transistor Characteristics 69

Fig. 43 Simple current biasing.

the collector supply via a bias resistor (RB) as in Fig. 43. RL is the
collector load resistance. The value of the bias resistor can be calculated
from the supply voltage (Vs), as well as the bias required.

VBE is the base-to-emitter voltage. In practical circuits this is usually

of the same order as the ‘starting’ voltage of a transistor, e.g. 0·2 volts
for a germanium transistor and 0·7 volts for a silicon transistor with a
possible spread of 0·05 volts on either side.
Knowing the bias current required (from a transistor’s characteris-
tics), which will usually be in microamps for small transistors, and using
numerical values for the voltages, the value of the bias resistor is given
in megohms (M ohms). Not knowing the transistor characteristics, a
typical value could be used, e.g. 30μA for a small transistor. Calculation
can be simplified further by ignoring VBE.
More accurately, the bias resistance should be calculated from the
desired value of the emitter current (IE)
70 Building and Designing Transistor Radios

Again VBE can be ignored to simplify calculation, when

For example, using a silicon transistor with a nominal hfe of 100, a

supply voltage of 9 volts, and a desired emitter current of 1 mA,

The value calculated from the ‘complete’ formula will be

.
Nearest preferred value (i.e. actual resistance values obtainable)
would be 820 k ohms in either case. In other words, the two differ-
ent calculations both specify the same (practical) resistor value. The
difference between the simplified formula calculation and the complete
formula calculation will be even less in the case of a germanium
transistor.
Simple current biasing by the method just described has definite
limitations, particularly in the case of germanium transistors. The more
usual method adopted is voltage biasing with emitter feedback, this cir-
cuit being shown in Fig. 44. Here there are two effective bias resistors

Fig. 44 Voltage biasing with emitter feedback.

Understanding Transistor Characteristics 71

R1 and R2, which work as a potential divider. There is also a resistor

(R3) in the emitter lead.
In this case the bias voltage developed (VB) is given by:

This ignores the voltage developed across R2 by the base current, but
the effect of this is negligible for normal calculation. The voltage at the
emitter (VE) is given by

The emitter current (bias current) is given by

To accommodate spread in transistor characteristics VE should be

large compared with changes in VBE. Also RE must be large to ‘stabilize’
the emitter current against variations in the supply voltage (Vs).
In practice a voltage drop (VE) of about 1 volt should be allowed in
the case of germanium transistors, and a voltage drop up to 3 volts
allowed in the case of silicon transistors, to stabilize the emitter circuit
effectively.
On the other hand the values of R1 and R2 should not be so high
that the base voltage is changed to a large extent by variations in base
current. At the same time the values of R1 and R2 must be high enough
not to waste power or drain power from the input signals. Logically the
combined parallel resistance of R2 and Re should be less than the value
of R1. However, this would result in a considerable loss of input signal.
This can be avoided by making R2 about four times the impedance of
the transistor.
A great advantage of this type of circuit is that provided the above
conditions are met the operating current IE is independent of hfe. Thus
it can adjust automatically to any spread of transistor characteristics
and maintain the desired collector current.
72 Building and Designing Transistor Radios

Fig. 45 Simple current biasing with collector feedback.

A third type of bias circuit which is sometimes used with silicon

transistors is shown in Fig. 45. This is current biasing with collector
feedback. In this case the collector voltage (Vc) is given by
Vc = Vs − R I × IE
= Vs − R I × IB(I + hfe)
The base current (IB) is given by

The operating current of the transistor is then given by

It should be noted that all the circuits described are for a common
emitter configuration, which is the one normally used in radio circuits
because of its superior characteristics as an amplifier. There are other
types of bias circuits applicable to the common emitter configuration
Understanding Transistor Characteristics 73

and also two other configurations – common-collector and common-

base. The common-collector configuration automatically provides com-
pensated bias. With the common-base configuration, bias normally
requires the use of two batteries.

Formula Check: Simple current biasing – Fig. 43)

This is the simplest form of bias and it is only necessary to determine
the value required for one component – R1.
The important parameters are:

It follows that the emitter current is:

In this case the variations in working current (IE) are substantially

dependent on the spread of hfe, thus this method of bias is really only
suited to transistors which have a low hfe spread.
The emitter current is also very much influenced by changes in supply
voltage. However, this effect can be offset to a large extent by including a
resistor R2 of high value in the supply line, the higher the value used the
less dependent the emitter current will be on the supply voltage.

Formula check: Current biasing with collector

feedback – Fig. 45
With this form of bias the collector voltage is given by
VC = VS – IBRI (I + hfe)
The base current is governed by the difference between IC and VBE
and resistor R2, and can be determined as:

The emitter current is given by

74 Building and Designing Transistor Radios

It follows that variations in the operating current (IE) can be

considerably influenced by spreads in hfe unless the value of R1 is
equal to or greater than R2/(1 + hfe), which in practice is normally only
possible with a large supply voltage.

Formula check: Voltage biasing with emitter

feedback – Fig. 44
With this form of bias the bias voltage (VB) is determined by the values
of resistors R1, R2 and the supply voltage (VS), as long as the base
current does not load the potential divider R1, R2.

The emitter voltage is given by

The emitter current of the transistor is thus

In order to nullify the effect of spreads of transistor characteristics,

and changes due to temperature effects, VBE must be kept as constant
as possible, which means that VE should be large in comparison with
VBE. Equally, to nullify changes in supply voltage affecting IE, RB must
also be large. In practice this demands a value of VE of about 1 volt
for germanium transistors, and a VE of about 3 volts with silicon
transistors. This desirable mode of working is also achieved if

This is the same as saying that for the emitter current to be

independent of hfe
 Chapter 7
Audio Amplifiers

The audio amplifier employed in the radio receiver comprises one or

more stages of amplification coupled to an output stage for matching
and powering a loudspeaker. Much depends on the output power
required. An output power of about 0·005 to 0·01 watts (5 to 10
milliwatts) is high enough to operate high impedance headphones;
and about 0·05 to 0·1 watts (50 to 100 milliwatts) a very small
loudspeaker. Larger speakers to a portable transistor radio may
require 0·2 to 1 watt output. Output power for conventional domestic
radios is in the order of 2 to 5 watts; and for H-Fi, about 10 watts
upwards.
The primary requirement of an amplifier is to produce power
amplification or gain of the af signal. At the same time it should do
this with minimum frequency distortion and maximum linearity. Fre-
quency distortion occurs when the amplification is not the same on all
frequencies, thus the frequency components are not represented at
their correct relative strength (some frequencies being emphasized and
others depressed). Linearity refers to the true reproduction of the actual
waveform of the signals.
Frequency distortion is related to the gain of the amplifier. Ideally
the relationship between gain and frequency should be a straight,
parallel line. In practice this is impossible to achieve over the whole
af range. In particular there will be a marked loss of gain at each end
of a frequency range, i.e. at the lower frequencies and the higher
frequencies – Fig. 46.
Provided the rest of the curve is reasonably linear, the addition of a
tone control will normally satisfy most listening requirements. When
more exact reproduction is required as in Hi-Fi, frequency distortion,
where present, can be compensated by equalization circuits. These aim
at introducing compensation of an opposite nature to smooth out the
frequency response curve.
The type of distortion produced by non-linearity is usually more
76 Building and Designing Transistor Radios

noticeable, and therefore less acceptable. As a general rule, the greater

the gain extracted from a single stage, the greater the degree of
distortion. No practical amplifier has exactly linear characteristics, even
at low gain, but non-linearity shows up increasingly with increasing
gain. Basically, therefore, a large number of individual amplifier stages,
all operating at low or moderate gain should give less overall linear
distortion than one or two stages operating at high gain. In practice it
is necessary to adopt a compromise solution based on arriving at an

Fig. 46 Amplifiers do not amplify all frequencies by the same amount.

acceptable level of distortion in order to avoid an excessive number of

separate stages. Additional stages can, in fact, introduce other trou-
bles. In any case the question of distortion is not significant in the final
output stage, where other forms of distortion can be present. Ultimately,
too, the loudspeaker itself can be the main source of distortion in the
whole system. It is a waste of effort to design an audio amplifier circuit
with excellent linearity characteristics only to operate a poor quality
loudspeaker from it. Conversely, a Hi-Fi quality loudspeaker cannot be
expected to compensate for the distortion inherent with a poor quality
audio amplifier.
For this reason it is really necessary to design the audio amplifier
Audio Amplifiers 77

‘backwards’, i.e. start with a selection of speakers and their requirements

and work backwards to the first stage of amplification. This first
demands some knowledge of the characteristics of amplifier circuits.

Class A Amplifiers
The simplest form of amplifier circuit is a single transistor with bias
and input signal voltage such that the collector current always flows.
This is known as Class A operation. It has the advantage of producing
a low distortion as well as being simple to design and construct, but the
disadvantage of drawing a relatively high current all the time.

Fig. 47 Basic amplifier circuit using a silicon N-P-N transistor.

The simplest practical form of this type of circuit is shown in Fig. 47,
using a single resistor to provide current biasing (see Chapter 6). The
output load must be of relatively high impedance (several thousand
ohms). This can be provided by a second transistor amplifier stage,
or if used as an output stage by a step-down transformer to match
the characteristically low impedance of a loudspeaker (4–16 ohms
nominal) to the stage requirements. The same circuit can be used with
a P-N-P transistor or an N-P-N transistor, the only requirement being
that the supply voltage is opposite in polarity. Choice of an N-P-N
transistor would be preferred as this type is more stable with simple
current biasing using a single resistor.
It is more usual, and generally more desirable, to employ voltage
78 Building and Designing Transistor Radios

biasing with emitter feedback, where the basic circuit is as shown in

Fig. 48. Again the same type of load is used for the output, and the cir-
cuit is identical for a P-N-P transistor and an N-P-N transistor, with
supply voltage polarity reversed. Values of suitable bias components
can be determined from Chapter 6.
The power-gain which can be achieved from such a circuit depends
on the transistor characteristics and the load. Transistor characteristics
are given in the form of graphs of collector currents (IC) plotted against
collector voltage (VC) for different values of bias (base current, or
IB). These curves typically take the form shown in Fig. 49 (see also
Chapter 6).
The collector voltage is the supply voltage, which in theory can be
any voltage up to the maximum rating specified from the transistor
concerned. The collector current is largely determined by the load,
and the bias by the working requirements (but see also Chapter 6
again).
The power that can be dissipated safely by a transistor is represented
by the product of IC and VC, and is represented by a single value, i.e.
so many watts. This can also be represented by a load line drawn on
the graph. Any product of power below this curve is feasible to use
(provided limiting values of either IC or VC are not exceeded). Any
product of power above this line is not usable since it overloads the
transistor – Fig. 50. Component values are therefore calculated
accordingly.
Note: this also emphasizes that the maximum power rating of
a transistor is not simply the product of IC max and VC max as it is
sometimes supposedly taken to be. In other words it is never permissible
to apply both maximum VC and maximum IC to a transistor simulta-
neously as this will drastically overload it, causing it to burn out.
A simple check on individual transistor characteristics will confirm
this. For example for an OC84 transistor
Vc max = 25 volts
IC max = 500 mA
Power max = 260 mW
Erroneously employing VC max with IC max would give a power of
25 × 500 = 12,500 milliwatts – nearly 500 times the maximum power
rating of the transistor!
 Fig. 48 Basic transistor amplifier with voltage bias provided by R1
and R2. R3 and C are stabilizing components. Polarity is shown for a
P-N-P transistor.

Fig. 49 Illustrating how maximum power rating of a transistor is

related to collector current, collector voltage and base current in common
emitter configuration.
80 Building and Designing Transistor Radios

The ideal maximum efficiency that can be achieved with Class A

operation is 50 per cent, although in practice it is usually substantially
less in order to avoid too much distortion, i.e. the amplifier is best
operated at well below maximum possible gain to avoid troubles.

Fig. 50 Working envelope of a transistor as fixed by its power rating,

maximum collector current and maximum collector voltage.

Amplifier circuits of this type can form the basis of amplifier stages
connecting two or more circuits in series to derive the required
amount of gain, whilst operating each stage under reasonably linear
characteristics. Some simplification of circuitry is possible, rather than
using complete voltage biasing circuits associated with each transistor.
The output stage in a modern amplifier circuit, however, is normally a
Audio Amplifiers 81

push-pull type operating as a class B amplifier. Push-pull circuits may

also be used as a driver in multi-stage amplifiers.

Class B Amplifiers
The immediate advantage offered by a push-pull class B amplifier is
that the output power obtained is considerably greater than double the
power of a single transistor. Also the average current drain is very much
lower than with Class A operation because the transistors are biased
so that their working point is near cut-off and quiescent current is
virtually zero. (Fig. 51 shows a ‘working’ diagram of Class B operating
characteristics.)

Fig. 51 Load line superimposed on transistor characteristics for a

Class B amplifier.
82 Building and Designing Transistor Radios

Apart from much lower current consumption, making it more

suitable for battery operated circuits, the efficiency of a Class B output
can approach 80 per cent. It does have its inherent limitations, however,
and in particular a proneness to crossover distortion. This is distortion
produced at the change-over point when working is transferred from
one transistor to another (changing from ‘push’ to ‘pull’ and ‘pull’ to
‘push’).
Crossover distortion will be most marked if both transistors are
biased exactly to cut-off. It can be overcome, or at least the residual
distortion can be substantially reduced, by selecting the bias so that,
one transistor does not cut-off until the other has stopped conducting
i.e. there is a slight overlap at the changeover. Unfortunately the amount
of overlap will tend to change both with the operating temperature
of the transistors and any change in the supply voltage. Equally,
differences in the spread of characteristics of different transistors of
the same type can make design for optimum bias difficult or even
impossible without further ‘cut and try’ adjustment of values. It is pos-
sible to incorporate compensating components in the circuit design
to minimize the undesirable effects of temperature and characteristic
spread. If necessary the supply voltage can also be stabilized (e.g. by
means of a zener diode).

A Practical Class B Circuit

A basic push-pull transformerless output circuit is shown in Fig. 52,
using a complementary pair of transistors TR2 and TR3 and a driver
transistor TR1. Bias quiescent voltage for TR2 and TR3 is set by the
value of R1 (which can be a variable resistor) the value of which is
less than R2. This second resistor acts to stabilize the bias supply.
Additional very low value resistors inserted between the emitters of
TR2 and TR3 and the connecting point to C2 will improve thermal
stability. Alternatively, R1 could be paralleled with a thermistor.
Direct coupling of a transistor output to a low impedance load in
this circuit may seem a direct contradiction of previous explanations
of transistor characteristics (e.g. see Chapter 6). The difference in
this case is that the transistors employed are power transistors which
characteristically have a low output impedance and can thus be
connected directly to a low impedance speaker without the need for
an output transformer.
Audio Amplifiers 83

Fig. 52 Simple low power audio amplifier using a minimum of

components with direct coupling to a miniature 4-ohm or 8-ohm
loudspeaker. Component values:
R1—68 ohm
R2—1 k ohm
R3—1 M ohm
C1—8 μF
C2—0·001 μF
C3—50 μF
TR1—BC 109
TR2—AC 176
TR3—AC 128
Supply: 9 volts

Conventional practice is to use this type of circuit with a single

stage of initial amplification provided by a fourth transistor. This
transistor can be made to work both as an amplifier and a dc difference
amplifier, comparing the voltage offered by the potential divider with
the ‘end’ voltage in the type of circuit shown in Fig. 53. Other points
to note are that the negative feedback is taken from the output stage
to the emiter of TR1 via R5, whilst the introduction of capacitor C3
provides decoupling between these two points. A small amount of
Fig. 53 A more or less standard type of circuit for a 1-watt audio amplifier.
Component values:
R1—15 k ohm
R2—15 k ohm
R3—2·7 ohm
R4—1·5 k ohm
R5—2·2 k ohm
R6—39 ohm
R7—2·2 k ohm
R8—1 k ohm
R9—510 ohm
C1—10 μF
C2—125 μF
C3—400 μF
C4—320 μF
C5—0·005 μF
C6—320 μF
TR1—AC 127
TR2—AC 128
TR3—AC 128
TR4—AC 127
Loudspeaker: 8 ohms
Supply: 9 volts
Fig. 54 A similar circuit to Fig. 53, but designed for 3 watts output,
using higher power transistors. Component values:
R1—39 k ohm
R2—120 k ohm
R3—47 k ohm
R4—390 ohm
R5—22 ohm
R6—2·2 k ohm
R7—15 ohm
R8—560 ohm
R9—2·2 ohm
Rio—2·2 ohm
C1—222 μF
C2—25 μF
C3—250 μF
C4—3·3 μF
C5—400 μF
TR1—BC 108
TR2—AC 128
TR3—AC 176
TR4—AC 128
Loudspeaker: 8–16 ohms
Supply: 22 volts
86 Building and Designing Transistor Radios

feedback is re-introduced via R3 to reduce the effects of spread in gain,

and thus distortion.
This form of circuit is capable of excellent performance with low
distortion. Transistor types are chosen according to the output power
requirements, which in turn determine the drive current required and
the degree of preamplification necessary. Component values given in
Fig. 53, for a 1-watt amplifier, and in Fig. 54, for a 3-watt amplifier,
are typical of ‘standard’ design practice. More complex circuits are
demanded for high power audio amplifiers, especially for Hi-Fi
working.
Some designs may also incorporate the volume control potentiometer
in the amplifier stages. For minimum distortion this should be placed
in series with a flow of current, normally immediately following the
preamplifier (first stage).

Rf Amplifiers
An amplifier stage may be inserted after the tuned circuit to improve
the strength of the aerial signal, and thus the sensitivity of a receiver,
when it is known as a preamplifier. In this case, a basic amplifier circuit
is adequate, e.g. similar to Figs. 47 or 48 – but using an rf transistor. A
somewhat superior performance, again with simple circuitry, is given
by an FET amplifier – see Fig. 73, Chapter 11.
In practice, rf preamplifiers are mostly used in superhet receivers
tuning to frequencies of about 50 kHz or above (e.g. FM radio cir-
cuits) because the signal-to-noise ratio inherent at these higher
frequencies is usually poor without some amplification of the original
frequency. An FET amplifier in this case is even more desirable than
a conventional transistor amplifier because of its inherently lower
noise level. Preamplifiers used in this context are generally known as
preselectors.
 Chapter 8
Regenerative Receivers

The regenerative receiver works on a ‘supercharging’ basis, but quite

different to the principle of operation of a superhet and very much
simpler. Part of the rf reaching the detector is fed back and added as
a ‘boost’ to the incoming signal, resulting in a substantial increase of
signal strength. Effectively this increases the sensitivity of the receiver
(as incoming signals are received at greater strength); and also increases
the Q of the tuned circuit, and thus the selectivity.
This can be achieved by quite a simple circuit since the process of
detection and regenerative feedback can be accomplished using a
single rf transistor. In other words the transistor is worked both as
a detector (or rectifier) and an amplifier. There are just two basic
requirements:
(i) The degree of amplification of rf achieved is proportional to the
amount of feedback. It is limited only by the amount of feedback
the transistor can handle without being excited into oscillation.
Thus to achieve maximum gain it is necessary to introduce some
sort of adjustment or control to give the maximum amount of
feedback possible without the transistor being driven to the
point of breaking into oscillation.
(ii) It is necessary that the feedback be in phase with the incoming rf
signal, otherwise it will have a cancelling rather than amplifying
effect on this signal.
The feedback signal can be coupled to the aerial circuit either
inductively or by a capacitor. Inductive coupling is usually preferred,
when the coupling coil can consist of a few turns of wire wound on a
paper sleeve fitted to a standard tapped ferrite rod aerial coil – Fig. 55.
The complete ‘front end’ of a simple regenerative receiver is then as
shown in Fig. 56.
The choice of transistor is fairly non-critical, provided it is an rf
type (not an af type designed for use in audio frequency circuits). Also
Fig. 55 Regen. coil takes the form of a few turns of enamelled wire on a
separate sleeve free to slide on a ferrite rod aerial.

Fig. 56 Simple regenerative receiver circuit based on a P-N-P transistor.

Test point marked for connecting high impedance phones to listen in and
adjust circuit. For coupling to an audio amplifier, the output is taken from
the collector of the transistor. Component values:
R1—1 M ohm (or higher value, as found best by test)
R2—3·3 k ohm
C1—tuning capacitor in standard aerial circuit
C2—220 pF (or larger, as found best by test)
C3—8 or 10 μF
C4—500 pF variable capacitor or trimmer
C5—8 or 10 μF
L1, L2—standard aerial coils on ferrite rod
L3—regen. coil (3–5 turns of 36 s.w.g. wire)
TR—OC45, or near equivalent
Supply: 9 volts.
Regenerative Receivers 89

the higher the performance characteristics the better. Current bias is

provided by resistor R1, the value of which is dependent on the choice
of transistor type. Alternatively, using a 1 megohm potentiometer for
R1 will enable the bias to be adjusted to match the characteristics of a
wide range of common low-power rf transistor types. L2 is the coupling
coil already referred to. R2 and C3 are the stabilizing components
for the transistor, e.g. typically 1 to 3 k ohm for typical low-power
transistors. C2 can be any value between 2 μF and 10 μF (its actual
value should not be critical). The remaining component is the 500 pF
variable capacitor (C4) included as a feedback control. Actually there
are two feedback ‘controls’ – this variable capacitor and the coupling
coil on the ferrite rod, the position of which can be altered.
The complete ‘front end’ can be coupled to a conventional af
amplifier (e.g. see Chapter 7). However, it is not necessary to build
a complete circuit to check that the front end is working. If the
detector/ regen. circuit is broken at points XX, and high impedance
phones connected here, one or two strong stations should be received
at listening strength, using a 9 volt battery supply. The original cir-
cuit shows a P-N-P transistor. Polarity must be reversed if an N-P-N
transistor is used.
Checking out this circuit will establish that it is working properly
and also shows how adjustments are made. Capacitor C4 should be
set to its mid position and an attempt made to tune into any broadcast
station, adjusting only the tuning capacitor in the aerial circuit
(C1). Then adjust the position of the coupling coil on the ferrite rod
to give maximum volume, readjusting the tuning capacitor as well,
if necessary. If this is successful, C4 should be adjusted forwards or
backwards until the transistor starts to oscillate or ‘howl’, then backed
off just enough for the ‘howl’ to disappear. Finally adjust C1 again until
maximum volume is heard. (Adjustment of C4 will affect the tuning
slightly, which is why C1 has to be readjusted).
If no signal is received at all, then the feedback signal is probably out
of phase. To correct this merely reverse the connections of the coupling
coil. Failure to receive any signal could, of course, also be due to a poor
aerial circuit used in an area where radio reception is weak.
If a signal is heard but no adjustment of C4 produces oscillation, the
coupling coil should be moved nearer the aerial coil on the ferrite rod to
tighten the coupling. Alternatively if only a howl is heard with C4 in its
90 Building and Designing Transistor Radios

mid position, move the coupling coil away from the aerial coil until it
disappears.
Once the optimum position for the coupling coil has been found
it should be cemented to the ferrite rod. The optimum setting of the
variable capacitor C4 may need readjustment when an af amplifier is
added to the ‘front end’ to complete a working radio receiver.

Fig. 57 Simple regenerative ‘front end’, not requiring a regen. coil.

Test point marked is for listening in with high impedance phones to
check circuit for working.
Component values:
R1—1 M ohm
R2—2·2 k ohm
C1—tuning capacitor in standard aerial circuit
C2—10 pF or 20 pF variable (trimmer capacitor)
C3—0·01 μF
TR—2N2926 or equivalent N-P-N type
D—germanium point-contact diode
Supply: 9 volts.

An alternative design of regenerative ‘front end’ is shown in Fig. 57.

This employs a transistor pre-amplifier followed by a conventional
diode detector, a proportion of the pre-amplifier output also being
tapped as feedback. In this case the feedback is capacitive – coupled
to the aerial circuit via a small 10 pF or 20 pF ‘trimmer’ capacitor C2.
Again the transistor used must be an rf type.
Adjustment in this case is similar, but simpler. Again high impedance
Fig. 58 Complete three-transistor regenerative receiver. A high impedance
loudspeaker can be connected directly to the output. Alternatively the
output points can be connected to a 10:1 output transformer, the secondary
of which feeds a low impedance speaker. Component values:
R1—50 k ohm potentiometer C4—100 pF
R2—560 k ohm C5—10 μF
R3—5 k ohm potentiometer C6—40 μF
R4—4·7 k ohm C7—10 μF
R5—150 k ohm C8—40 μF
R6—10 k ohm C9—100 μF
R7—2·2 k ohm C10—320 μF
R8—33 k ohm D—germanium point-contact diode
R9—4·7 k ohm L1, L2—standard aerial coils on
R10—1.5 k ohm ferrite rod
R11—470 ohm L3—regen, coupling coil on ferrite rod
R12—680 ohm L4—100 turns on 38–44 s.w.g.
R13—100 ohm enamelled wire on ¼″ polystyrene
R14—10 ohm coil former with iron dust core
C1—tuning capacitor in standard TR1—AF 117
aerial circuit TR2—OC 71
C2—0·02 μF TR3—OC 81
C3—330 pF Supply: 9 volts

Note: regen. is adjusted by R1 and also by the core of L4. R3 works as a volume
control
92 Building and Designing Transistor Radios

phones can be connected where shown for testing and the circuit tuned
to any station via the tuning capacitor C1. Capacitor C2 is adjusted
until the transistor oscillates, then backed off until the howl just dis-
appears. This will have upset the tuning slightly, so C1 is readjusted
for maximum volume. If trouble is experienced, substitute a 2 k or 5 k
ohm potentiometer for R1 to adjust the transistor bias for optimum
performance.
It is possible to dispense with capacitor C2 and simply take
connections from the aerial circuit and collector of the transistor with
short lengths of insulated wire. These wires should be cut so that they
can be twisted together over a length of about ⅜–½ inch. The bare ends
of the wire must not touch. The twisted wires will then form a coupling
capacitor and the degree of coupling can be adjusted by twisting more
tightly or untwisting. Once the optimum amount of coupling has
been established they can be cemented together. Any further slight
adjustment needed can then be made by moving the position of the
‘twist’ relative to other components. This method of coupling, which
eliminates one component, is quite critical of position, but despite its
apparent crudity can work very well and is even used on commercial
circuits.
A design for a complete regenerative receiver circuit using three
transistors is shown in Fig. 58.
 Chapter 9
The Superhet

The superheterodyne receiver or superhet works on the basis of

combining the incoming rf signal with another high frequency signal
and extracting a ‘difference’ signal at a fixed intermediate frequency (if).
At the same time the af content of the original rf signal is transferred
to the if signal frequency which now becomes a ‘carrier’. The if output
is then amplified at the intermediate frequency before being passed
to the detector. Only the ‘front end’ of the receiver is effected – the
detector and subsequent audio amplifier stages being the same as with
simpler receivers except that the detector normally operates at a higher
input because of the preceding amplifier stages.
This ‘front end’ comprises a local oscillator generating a high frequency
equal to the rf carrier plus the intermediate frequency, and a mixer. The
mixer receives rf from the aerial tuned circuit and rf plus if from the

Fig. 59 The ‘front end’ of an AM superhet receiver in box diagram form.

In the case of an FM superhet, the first ‘box’ is an if amplifier connecting
to a dipole aerial.
94 Building and Designing Transistor Radios

local oscillator. From this it extracts the modulated if as an output,

which is passed to the if amplifier – see Fig. 59. It follows that the local
oscillator circuit must tune in step with the aerial tuned circuit, hence
each incorporates a tuned circuit ganged together. The mixer receives a
combined waveform as an input, the main components of which are the
original aerial and oscillator frequencies together with their sum and
difference frequencies. The difference frequency is then preferentially
selected by the mixer tuned circuit and passed as an input to the if
amplifier.
Exactly the same working principle applies for AM and FM receivers,
except that the aerial signal for FM is normally received by a dipole aerial
and pre-amplified before being passed to the mixer, (e.g. see Chapter 2).
The other main difference is in the choice of intermediate frequency –
470 kHz for AM and 10·7 MHz for FM. Also the use of more stages of
if amplification is usually with FM (commonly three stages for FM as
compared for two stages for AM). It can also be noted that with AM the
local oscillator is (almost) invariably tuned to a frequency of 470 kHz
above the aerial signal frequency. For FM, the local oscillator may be
tuned to 10·7 MHz above or below the aerial signal.

Self Oscillating Mixers

The design and construction of practical local oscillator and mixer
circuits is considerably less complicated than a description of how
they work! A single transistor can act both as local oscillator and
mixer, or as a self-oscillating mixer. A virtually standard circuit for this
is shown in Fig. 60, together with a representative aerial tuned circuit
and inductive coupling. Correct tracking of the two tuned circuits
can be achieved using padder capacitors in parallel with the tuning
capacitors, or by the use of tuning capacitors with specially shaped
vanes.
The only real limitation of this circuit is that it is sensitive to stray
capacitance which can cause feedback from oscillator to aerial cir-
cuits. The possibility of this occurring can be eliminated by placing
a metal screen between the two sections, although this is not always
a complete solution. Stray capacitance can also arise from leads, so
all leads (particularly in the aerial section) should be kept as short as
possible. If an external aerial is coupled to the aerial circuit, then it
The Superhet 95

Fig. 60 This is a basic—and virtually standard—design for a combined

oscillatormixer circuit. Typical component values can be found in Fig. 65.
L1 and L2 are the aerial coils. L3, L4, L5 are the windings of the oscillator
transformer, L5 and its associated variable capacitor forming a tuned cir-
cuit tuned in step with the aerial tuned circuit by ganging the respective
capacitors (as indicated by the dashed line).

should feed by a series capacitor to limit the extra capacitance that the
aerial wire can add to the tuning capacitance in this circuit (see also
Chapter 2).

If Amplifiers
An if amplifier stage usually consists of a transistor working as an
amplifier with a tuned input and output circuit. The tuned circuits
are provided by if transformers, tuned to the intermediate frequency.
Thus in an AM/FM receiver, separate if transformers are required at
each stage usually with primaries connected in series, and secondaries
connected in series, ultimately feeding separate detector circuits.
A basic circuit for a single stage is shown in Fig. 61. The if transformer
normally consits of a canned coil with an iron dust core for inductive
Fig. 61 A basic if amplifier circuit comprising a single stage. Two or more
stages may be employed before the detector, e.g. see Fig. 65.

Fig. 62 Two stages of an if amplifier circuit, showing the standard

method of coupling. The single-tuned if transformers (in dashed boxes)
are adjusted by iron dust cores, to resonate at the if (intermediate
frequency) chosen.
The Superhet 97

tuning. For best results, both coils of the transformer should be

tunable.
Stage connections are straightforward and follow on the lines shown
in Fig. 62. In high quality designs, however, filters may be interposed
between stages. These may be piezoelectric crystals or ceramic filters
manufactured to be resonant at the intermediate frequency. The use of
such filters can improve if tuning and reduce noise. It should be noted
that such filters should be of bandpass type, i.e. capable of passing
the necessary af bandwidth and not single-signal crystals. A typical
double-tuned two-stage if amplifier is shown in Fig. 63.

Fig. 63 Two stages of a typical if amplifier with double-tuned if transformers.

Apart from both primary and secondary of the transformers being tuned,
this circuit is virtually identical to Fig. 62.

Automatic Gain Control

The sensitivity of a superhet receiver is inherently high in the rf and if
stages. This implies that in the presence of a strong AM signal excessive
amplification can occur in the if amplifier, resulting in ‘clipping’ of
the signal. Since rf signal strength received in the aerial circuit may
98 Building and Designing Transistor Radios

vary enormously in strength from one station to another, the possibility

of this happening is very real in practice.
Clipping, and thus distortion of the amplified signal, can be avoided
by applying automatic gain control (agc) whereby a proportion of the
if output is fed back to the front of the if amplifier to control the gain
of the first stage. Any rise in signal output then automatically tends to
reduce the gain at the input stage to compensate, and vice versa.

Fig. 64 Typical automatic gain control (age) circuit ‘tapping’ a proportion

of the output signal from the detector and feeding it back to the first
if amplifier stage.

There are a number of ways this can be done, one of the simplest being
shown in Fig. 64. A feedback signal is tapped from the detector and
applied in series with the base-emitter voltage of the first if transistor,
controlling the bias in proportion to the working voltage appearing at
the detector. The change in bias thus modifies gain (i.e. reduces it with
increasing detector voltage) in reverse proportion to the detector signal,
i.e. provides automatic gain control. The resistor forms a potential
divider feeding the base of the first transistor and can be chosen to
bring the diode to its sharpest operating point, improving its response
The Superhet 99

to weak signals. Capacitor C1 is needed to eliminate rf and af voltages

in the age signal.
The presence of af signals in the age voltage would reduce the modu-
lation depth of the incoming signal. Equally, modulation must be filtered
out so that the age signal is in the form of a dc voltage with an average
component following the strength of the carrier. This means that the
time constant of the resistance and capacitance in the age circuit must
be long enough for this to occur, but not so long that the age cannot
follow rapid variations in detector signal strength, particularly fading.
Normally this calls for a fairly large value of C1, 8 μF being typical.
A diode can be added across the first if transformer although this is
not strictly necessary. However it can substantially improve the gain
control performance by acting as a damping device to make the circuit
accommodate a wider range of rf input voltages.
Automatic gain control is not necessary with FM receivers since this
function is already performed by the limiter. Sometimes it is added,
however, to improve the performance of the if amplifier; or a single
diode is added just ahead of the limiter to serve a similar purpose. What
is virtually a ‘standard’ design of AM superhet circuit is shown in Fig. 65.

Superhet Alignment
A further problem always present with the construction of a superhet
receiver is alignment of the if amplifier. This really demands the use of
a calibrated signal generator as well as a universal meter. The signal
generator should be capable of producing audio-frequency signals and
also the intermediate and radio-frequencies it is necessary to explore,
with provision for modulation at audio-frequencies. Where af and
rf signal generators are employed the af generator should have an
impedance of the order of 600 ohms, which can be fed directly into the
top of the volume control of the receiver through a 10 K ohm resistor
and a 1-μF isolating capacitor.
The rf generator should be of low impedance (of the order of 60
ohms output impedance) and, in general, can be fed into the base of
the if transistor or mixer transistor via a 0·5 μF capacitor.
Where a signal generator is available, the if stages and signal cir-
cuit are aligned separately, in both cases using an output power meter
or ac voltmeter connected across the speech coil (speaker terminals).
 Fig. 65 This is virtually a standard design for the rf and if stages of an AM superhet receiver up to and
including the detector. It also incorporates age. Output is fed to an audio amplifier to complete the working
circuit, the 10 k ohm potentiometer acting as a volume control. Resistor values are marked on the diagram.
Other component values:
C1—10 nF C2—22 nF
C3—r6 nF C4—100 nF
C5—10 nF D—germanium point-contact diode
TR1—BF 194B, or equivalent TR2—BF 195C, or equivalent
TR3—BF 195D, or equivalent.
The three if transformers used are single-tuned. Equally, double-tuned transformers could be used, modifying
connections as in Fig. 63. The oscillator transformer is a standard type. Supply: 7·5 volts.
The Superhet 101

Before attempting alignment the volume control should be set to a

minimum in order to use the lowest signal from the signal generator
consistent with a reasonable output, e.g. 50 milliwatts or 1 volt across
the speech coil. This avoids age action.
For aligning the if stages the signal is set at the if (normally 470 kHz)
and usually applied to the base of the if transistors and mixer, in
turn, working backwards (i.e. starting with the last if transistor). The
corresponding if transformer cores are, in turn, adjusted to maximum
output. The signal is injected using a 0·5 μF capacitor and an 820-ohm
resistor in series with the generator output lead and never applied
directly.
Having peaked the if transformers, the circuit is switched to medium
wave, the generator set to a typical low frequency (e.g. 540 kHz) and
the oscillator trimmer adjusted for maximum output.
The tuning capacitor is then set to minimum capacity and the
second (mixer) gang trimmer adjusted for maximum capacity. These
two stages are then retuned, as necessary, for optimum results.
For alignment of the signal circuits no direct connection is made,
but the output or live terminal of the signal generator is connected
to a loop of wire consisting of two or three turns approximately 7 to
9 inch in diameter with a series resistor in circuit of 430 or 390 ohms.
The loop should be situated about 24 inch from the receiver ferrite rod.
Normal procedure is then:
1. Set signal generator to a low m.w. frequency (e.g. 600 kHz) and
tune gang condenser to the corresponding position on the scale. Adjust
aerial coil position or aerial trimmer for maximum output.
2. Repeat with the signal generator set to a high m.w. frequency (e.g.
1400 kHz) and adjust gang condenser trimmer for maximum output.
3. Repeat operations 1 and 2 as necessary.
4. Switch to long wave, set generator to a typical frequency in about
the middle of the l.w. band (e.g. 220 kHz) and adjust the longwave
aerial coil position or trimmer for maximum output.
Where no signal generator is available, alignment must be carried out
using broadcast stations as the source of signals. If an output meter with
an impedance of around 25 ohms is available this should be connected
in place of the loudspeaker; or alternatively a high-resistance voltmeter
capable of reading 0–2 or 0–5 volts ac can be connected across the
speech coil to give a visual indication of output signal strength. If
102 Building and Designing Transistor Radios

neither type of meter is available, the output signal strength must be

judged by ear.
1. Aerial and oscillator trimmers should be set to approximately the
mid-point of their travel and the tuning control then set to the correct
position for a local station, preferably in the middle of the m.w. band.
Turn volume control to a maximum. Adjust the oscillator core to bring
the station to the correct point on the tuning scale.
2. Turn the receiver so that it is oriented in the direction which gives
minimum signal strength. Adjust if transformer cores for maximum
output. (Note : if necessary, reduce the volume via the volume control
to a minimum audible signal before adjustment as it is easier to judge
an increase in strength of a weak signal audibly than an increase in
strength of a strong signal. A meter will give a more positive indication
regardless of the original signal strength, but at high signal levels the
true output may be modified by the effect of age action.)
3. Tune to the high-frequency end of the m.w. band, and adjust the
oscillator trimmed to bring the station into tune consistent with the
scale calibration. If this proves difficult or impossible, try adjusting
first with the aerial trimmer followed by the oscillator trimmer.
4. Tune to the low-frequency end of the m.w. band. It should be pos-
sible to adjust for maximum signal by sliding the aerial coil along the
ferrite rod. If not, it may be necessary to adjust the oscillator trimmer
to get agreement with the scale calibration.
5. Steps 3 and 4 must then be repeated until an optimum adjustment
is realized where adjustment of one end has no effect on the other.
6. Tuning on the long-wave band is not likely to be critical and it is
usually only necessary to set the tuning condenser to the calibrated
position and adjust the position of the long-wave aerial coil on the ferrite
rod to bring in the station at maximum strength (or adjust the long-wave
aerial trimmer, if fitted).
 Chapter 10
Interstage Connections
(Coupling)

Any radio receiver consists of individual stages connected together, out-

put to input, reading from the aerial to the loudspeaker (or conventionally
left to right on a circuit diagram). The method of connection or
coupling may be by direct connection (direct coupling); via a capacitor
(capacitive coupling); or via a transformer (inductive coupling) – see
Fig. 66 for diagrammatic representations. It is also necessary that

Fig. 66 Four different types of inter-stage coupling.

the output impedance of the ‘feeding’ circuit matches the input

impedance on the ‘receiving’ circuit. This may dictate the type of cou-
pling required. Alternatively impedance matching may be obtained
by adjusting the load impedance of the ‘receiving’ circuit, varying the
amount of coupling, or interposing an impedance-matching device.
104 Building and Designing Transistor Radios

The application of direct coupling is mainly limited to interstage

connections where a varying dc voltage has to be passed on, e.g. the af
output from a detector. It can also be used for interstage connections in
an audio amplifier, or any coupling where ‘matching’ is correct and a dc
as well as an ac path can be tolerated, or a dc path is necessary. Direct
coupling has the obvious advantage of simplicity.
Both capacitance coupling and inductive coupling isolate output and
input connections as far as the passage of dc is concerned. Capacitive
coupling is accomplished with a series-connected capacitor. This is
invariably associated with a load resistance and so is more correctly
referred to as RC coupling. It has the advantage of using simple,
inexpensive components and is widely used for coupling low-level
amplifiers, etc., in radio circuits. It is not particularly suitable for
coupling circuits carrying rf, when inductive coupling is preferred.
In this the transformer (inductive component) is associated with a
capacitor to ‘tune’ the coupling to the required frequency and is thus
popularly referred to as an LC coupling.
With simple RC coupling impedance matching is not particularly
critical, although the better the match the better the performance.
Optimum performance is often best arrived at empirically, trying dif-
ferent values of capacitor. Almost any value of capacitor will work as a
coupling device for ac, but the time constant of a circuit will determine
its suitability for passing ac frequencies (see also Chapter 4).
The time constant is defined as the length, of time, in seconds,
for the voltage across the capacitors to reach 63 per cent of the
applied emf. It can be calculated from the capacitance and resistance,
viz:
time constant = RC × 10−6 seconds.
where R = resistance in ohms
C = capacitance in microfarads
Specifically, the charge/discharge characteristics in terms of per-
centage voltage and time constant are presented in Fig. 67.
When it is necessary to accommodate a voltage applied at a particular
frequency, one cycle of frequency implies a working of charge/discharge
every half cycle. Thus taking a typical maximum af frequency of 10 kHz,
the time to complete a full cycle is 1/10 000 or 0·01 milliseconds.
Interstage Connections (Coupling) 105

Fig. 67 Charge and discharge times or a capacitor expressed in terms

of time constant. Thus approximately 100 per cent charge or 100 per cent
discharge is achieved in 4 times the time constant.

Since two charge/discharge functions are involved, one positive and

one negative, the true charge/discharge time is one half or 0·005
milliseconds.
Taking 1000 ohms as typical of the input resistance for a low power
transistor, i.e. base plus emitter voltage of 1000 ohms (see Chapter 6),
the time constant for RC coupling is:
106 Building and Designing Transistor Radios

time constant = 1000 × C × 10−6

= C × 10−3 seconds
The time to accommodate a complete cycle of af at 10 kHz, is 0·005
milliseconds. Thus substituting in the equation
0.005 × 10−3 = C × 10−3
or C = 0.005 microfarads or less.
Similarly, for passing an rf signal, say 10000 kHz

This is contradictory to the general rule that ‘the higher the value
of coupling capacitor the better’, typical values ranging from 500 nF
to 10 μF. However, use of a low capacity value, thus giving a lower
time constant, simply means that a smaller percentage of the af voltage
will be developed, and discharged, across the capacitor each cycle.
In other words, the efficiency of ac transmission will be lower. The
sample calculations also illustrate that for a given value of capacitor,
lower frequencies will be transmitted with more efficiency than higher
frequencies, confirming the original statement that rf coupling is
tolerable with af (provided the efficiency of coupling is not important),
but seldom used with rf because of the very low efficiency likely with
practical values of capacitors suitable for coupling duties.

Transformers
A transformer comprises, essentially, a primary winding and a
secondary winding on a core of magnetic material. The actual form
of the core governs the eddy current and hysteresis losses. The
conventional transformer wound on a closed core, giving a continuous
magnetic circuit, is not normally suitable for use with frequencies
above about 20 kHz (i.e. above audio frequencies). Different forms
of transformer are required for the inductive coupling of rf circuits if
excessive power losses are to be avoided.
The turns ratio of a transformer – or ratio of number of primary
turns to number of secondary turns – governs both the voltage transfer
ratio and the current transfer ratio.
Interstage Connections (Coupling) 107

The power transfer relationship is expressed by

Po = e × Pi
where Po = power output from secondary
where Po = power output from secondary
e = efficiency factor (which is always less than 1).

Normally a transformer is designed to have a maximum efficiency

at a particular power output rating. Efficiency can then be expected to
be lower both at lower or higher output powers. However losses will
generally decrease with lower power, and increase with higher power.
Maximum power rating is determined by the maximum temperature
the windings and/or insulation can withstand, all energy losses in the
transformer being dissipated in the form of heat.
The designation primary and secondary is one of convenience only.
A transformer can be connected to work ‘the other way round’, if
necessary, when the specified primary becomes the secondary and vice
versa. Further types of transformers can also have a third or tertiary
coil, e.g. see Chapter 4.

Impedance
The impedance relationship with a transformer is calculated on the basis
of an ideal transformer without losses. This is generally satisfactory
in practice provided the design of the transformer generates enough
inductance to work with a low magnetic current at the voltage applied
to the primary.
108 Building and Designing Transistor Radios

The basic impedance relationship then is

when Zp = impedance looking into the primary connections

Zs = impedance of load connected to secondary
For impedance matching purposes, this formula can be rewritten

In this case Zs is the same as above (i.e. impedance of load connected

to secondary); Zp becomes the primary impedance required to match;
and Np/Ns the turns ratio required to produce this match.
A typical example of the application of this formula is the matching
of an amplifier circuit requiring a high impedance output load to a
low impedance load represented by a loudspeaker, using an output
transformer. Suppose the load impedance required at the amplifier
output is 5 000 ohms. This represents the primary impedance required.
A typical loudspeaker impedance is 8 ohms thus, for matching

In other words, the primary winding of the transformer should have

25 times as many turns as the secondary winding.

Inductive Coupling
Coupling by means of the mutual inductance between two coils or
inductive coupling works on a similar principle to that of a transformer,
but since any part of the magnetic field set up by one coil cuts the second
coil the simple transformer formulas no longer apply. Also the degree
of coupling achieved depends on the proximity of the two coils – the
closer they are together the tighter the coupling, and vice versa.
The tighter the coupling the more the set-up operates as if the secondary
were simply tapped across the corresponding number of turns on the
primary – Fig. 68. If the input circuit is tuned to resonance, then the
Interstage Connections (Coupling) 109

equivalent circuit is shown on the right where the resistor represents

load resistance on the secondary coil with the coupling coils effectively
working as an autotransformer.

Fig. 68 The ‘tighter’ inductive coupling is, the more closely it approaches
direct coupling in equivalent circuit (but providing a blocking path for de)

In practice, by adjusting the number of turns on the secondary coil

and/or adjustment of the tightness of the coupling, the impedance of
the tuned circuit can be adjusted for optimum operation of the cir-
cuit for which the secondary is providing the input. For any degree of
coupling, maximum transfer of energy is obtained when the reactance
of the secondary coil is equal to the resistance of the load. Reactance
can be calculated directly from the inductance of the coil and the
frequency of the signal current, viz
Reactance (ohms) = 2πfL
where f is the frequency in Hz
and L is the inductance in henrys.

Decoupling
The object of decoupling is to prevent unwanted signals present in one
circuit or stage of a receiver being introduced in another stage. Usually
it is a preceding stage which may be effected by a feedback. Single stages
are normally commonly connected throughout to provide supply and
return lines from a single power source (i.e. the top and bottom lines
in a circuit diagram). It is obvious that there is plenty of scope for the
transmission of ‘feedback’.
Separation or decoupling of one stage from the previous stage can
be accomplished using a component offering high impedance to the
unwanted feedback. It must be a conductive device to pass dc on to the
110 Building and Designing Transistor Radios

preceding stage, but resistant ac to components at the frequencies

that have to be eliminated. Thus it could be an RFC choke, or more
normally, a resistor. This series connected component is then usually
associated with a capacitor effectively paralleled across the supply –
Fig. 69. Resistors can be surprisingly effective in this respect as ‘filters’,
with decoupling completed through the capacitor. They can also serve
simultaneously as a voltage-dropper when the preceding stage requires
a lower operating voltage than the following one. For example, the
output stage in a moderate or high power amplifier commonly requires
a higher operating voltage than the preceding stage.

Fig. 69 Standard method of decoupling a preceding stage from feedback of

unwanted signals.

In circuit designs decoupling components are often easy to identify by

the fact that the capacitor is shown with an earth symbol immediately
below it. This is done to avoid drawing a connection down to the bottom
line of the diagram crossing a number of other lines on the way. The
adjacent series connected resistor in the ‘top’ line will then be identified
as the closely associated resistor (as in Fig. 69, right).
Similar considerations apply if the circuit design is drawn the other
way up to accommodate reversed supply polarities. Conventionally the
‘top’ line in a circuit diagram is made the + line as regards the supply
polarity, and the ‘bottom’ line the common earth or negative line. This
convention cannot always conveniently be followed.
 Chapter 11
Field Effect Transistors

Field effect transistors (FETs) are another group of semi-conductor

devices which differ substantially both in construction and
characteristics from conventional transistors. In fact, about the only
thing they have in common is that their power requirements are low.
Their actual working characteristics are more like those of a thermionic
valve than a transistor.

Fig. 70 Field effect transistors shown in diagrammatic and symbolic forms.

Construction, in effect, takes the form of a channel of N-type or

P-type semi-conductor material enclosed for part of its length by a
collar of opposite type material. One end of the channel is called the
source and the other end the drain. The collar forms a gate – as shown
in Fig. 70. This diagram also annotates the respective symbols for
N-channel and P-channel FETs.
In simple terms, the FET works as follows. Application of a voltage
112 Building and Designing Transistor Radios

across each end of the channel will cause a current to flow from source
to drain (with electrons flowing from drain to source). Application of a
voltage to the gate so that the gate is negative with repect to the source
will reverse-bias the gate with respect to source, channel and drain,
causing a depletion layer to be formed (see Chapter 5). This is equivalent
to creating a constriction for the current flow from source to drain. The
higher the gate voltage the greater the constriction, until eventually the
source-drain current is reduced to almost zero. Connections of these
two sources of voltage are shown diagrammatically in Fig. 71.

Fig. 71 Supply and bias polarities associated with N-channel and

P-channel FETs.

Only a relatively small gate voltage is needed to effect a large source-

to-gain current change, so the device works as an amplifier with good
gain characteristics. Since the gate terminal is always reverse-biased,
the resistance between the gate and the rest of the device is very high,
i.e. an FET inherently has a high input resistance, usually the order of
1 to 103 megohms. In this respect it is similar in characteristics to a
triode valve with the gate equivalent to the grid, the source equivalent
to the cathode, and the drain equivalent to the anode.
It is also a characteristic of an FET that it has a very good
performance with high frequency signals, making it particularly
suitable for use in rf circuits. A further advantage for such circuits is
that it has excellent low noise characteristics.
Field Effect Transistors 113

The type of FET described is, in fact, more correctly called a junction
FET or JFET. There is another distinct type known as an insulated-
gate FET or IGFET. As the name implies, the gate ‘collar’ is insulated
from the channel by a very thin layer of insulating material. This has an
even higher input impedance, typically of the order of 106 megohms or
more. It is also a more flexible device since either a positive or negative
voltage can be applied to the gate. With a JFET, the polarity of the drain
supply is negative for a P-channel device and positive for an N-channel
device, with the gate always reverse-biased (i.e. opposite polarity to
that of the drain).
Field effect transistors are also made with two gates – a signal gate
and a control gate. These gates are effectively worked in series. Such
devices find a particular application in rf and if amplifiers, and mixers.
They may be called MOSFETs, although this description is also applied
to single-gate IGFETs.
It can also be mentioned that FETs may be classified as depletion
mode or enhancement mode devices. The difference really refers to the
effect of applying the gate voltage. Depletion mode reduces or depletes
the current flowing through the channel. Enhancement mode increases
or enhances the current flowing in the channel. Thus JFETs always
work in the depletion mode, whereas IGFETs can be made to work
either in the depletion mode or enhancement mode, depending on the
polarity of the gate voltage. Either types of FETs can be constructed
which have no channel as such, and thus no current flows with zero
gate voltage. Application of a gate voltage then causes a channel to be
formed, through which current flows. These are therefore enhancement
mode devices.

Working Characteristics
Output characteristics of an FET are given in terms of drain voltage
VP and drain current ID related to different values of gate voltage VG.
The first two descriptions are synonymous with channel voltage and
channel current respectively, for which the alternative abbreviations
are sometimes used.
VDs = channel voltage (or voltage between drain and source)
IDs = channel current (or current between drain and source)
Gate voltage (VG) is also sometimes quoted as VGS (gate-to-source
voltage).
114 Building and Designing Transistor Radios

Characteristic curves are similar to those of conventional transistors –

Fig. 72. The main difference is that the curves tend to flatten out rather
less gradually, ‘bottoming’ voltage is higher, and drain voltages used
tend to be higher than transistor collector voltages.

Fig. 72 Characteristic curves of a small FET.

Mutual conductance characteristics, as expressed in terms of channel

current/gate voltage curves, are essentially similar to those of valve
anode current/grid voltage curves.

FET Circuit Design

The following description applies to FETs working in the depletion
mode, where the gate voltage is always of opposite polarity to that
of the drain. A basic working circuit is shown in Fig. 73, where
RG is the gate bias resistor; RS is the stabilizing resistor associated
with a capacitor C2 to form a stabilizing load; and R1 is the drain
resistor or drain load.
Field Effect Transistors 115

Fig. 73 N-channel FET amplifier circuit. With a P-channel FET the supply
polarity would be reversed, but the circuit otherwise identical. Required
component values are discussed in the text.

The required (negative) gate bias can be set by selecting the value
of RS to bias the source positively to the required extent. This will
also tend to stabilize the channel current against any spread of FET
characteristics. Further stabilizing effect is produced by capacitor C2 in
parallel. Effectively, the higher the value of RS the better in providing
stability, although this will be at the expense of a loss of voltage through
the channel. Also, and even more significant, is the fact that excessive
gate bias may be developed in this way. Hence RS is generally restricted
to a moderate value, although R2 could be added as shown to form
a potential divider to limit the gate voltage to the required value.
This is not a particularly attractive solution as the value of R2 needs
to be substantially greater than RG, which itself is normally of the
order of 10 M ohm.
With negative bias, gate current is normally zero. The purpose
of RG is to hold the gate at normal zero potential under these
conditions, hence the high value. Equally, capacitor C1 is necessary
to prevent any direct dc voltage reaching the gate from the input. If
the signal source contains no such direct voltage, then neither C1 or
RG are necessary. An example of this is when an FET amplifier stage
follows a tuned aerial circuit, when it can be directly coupled to that
circuit if preferred.
An example of a high-performance circuit utilizing the favourable
Fig. 74 Mullard design for a capacitively-tuned FM tuner using a BFW10 field effect transistor in the rf
input stage. All component values are shown on the drawing.
Field Effect Transistors 117

characteristics of an FET is shown in Fig. 74. This is a Hi-Fi FM

tuner circuit (by Mullard), and also illustrates application of the basic
FET circuit design features just described. The aerial input circuit
is transformer-coupled to the input of the FET, dispensing with the
need for a coupling capacitor. Dc stabilization is provided by a voltage
divider for the gate and a source resistance of 680 ohms. The following
circuitry is a conventional mixer-oscillator, which in turn would be
followed by an if amplifier and audio amplifier.
 Chapter 12
Miscellaneous Circuits

Tone Controls
In its simplest form a tone control is an elementary tuned circuit
comprising a resistor and capacitor in series applied across the primary
of the output transformer, or the output. It is designed to be ‘tunable’
over the audio frequency range to be covered, the resistor normally
being the adjustable component.

Fig. 75 Elementary tone control circuit applied across the

output transformer.

Either a 5 k ohm or 10 k ohm potentiometer can be used for the

resistor, associated with a capacitor of about 1 pF to 10 pF. This will
give a reasonable tone control effect of audio frequency signals up
to the order of about 10 kHz – Fig. 75. Its effect on the quality of
reproduction will be fairly poor, but that is not necessarily significant
with a small, low-power AM receiver with a small loudspeaker.
Miscellaneous Circuits 119

A better form of simple tone control is shown in Fig. 76, consisting

of a series resonant circuit for treble cut and a parallel resonant circuit
for bass cut. Each has its own separate control via a potentiometer.
Mid-frequencies should be little affected by adjustment of either or
both treble and bass controls.

Fig. 76 Practical tone control circuit with separate treble and bass controls.
This would normally be positioned in front of the audio amplifier.

The most suitable position in a receiver for a tone control of this type
would be at the input of the audio amplifier (i.e. between a detector
and amplifier), and it is especially suitable for use with conventional
4-transistor audio amplifier circuits commonly used on low and
medium power receivers which employ feedback over two or more
stages. It is a ‘passive’ type of control and thus cannot introduce
instability in the following amplifier circuits.
Considerably more elaborate tone controls are used in Hi-Fi circuits,
the object being to achieve a substantially flat response curve over the
middle frequency range with increasingly progressive boost and cut
towards each end frequency. The circuits may be of ‘passive’ type (e.g.
Baxandall), or ‘active’ tye. The latter are normally based on frequency-
dependent feedback networks between the base and collector of a
transistor and therefore cannot be used in preceding circuits which
themselves incorporate feedback without risk of introducing instability.

Power Supply
It is often more convenient – or economic – to power a transistor
radio from the mains rather than dry batteries. This requires a separate
120 Building and Designing Transistor Radios

power supply circuit (which can be built into the radio circuit if
required), capable of supplying the required dc circuit voltage and
maximum current needed. In theory such a duty can be provided by
a step-down transformer with full wave rectification added, together
with a smoothing capacitor – Fig. 77. A bridge circuit is normally used,
however, as being less wasteful of input power.

Fig. 77 Simple half-wave rectified dc supply from a step-down

mains transformer.

Fig. 78 A more practical form of dc power supply. Diodes are type BY164,
or equivalent.

A typical power supply circuit is shown in Fig. 78, the diode rectifiers
being chosen according to the actual power requirements, i.e. capable
of carrying the maximum forward current in the output circuit.
There will be some voltage multiplication in this circuit, typically of
the order of 1·3 ×, so the transformer ratio is chosen accordingly. For
Miscellaneous Circuits 121

example, if a 12-volt output is required the ‘voltage’ multiplication

characteristics imply that transformer output voltage × 1·3 = 12, or
a transformer step-down ratio should give 12/1·3 = 9·23 volts, say 10
volts. In practice the actual output voltage is not critical, provided
it or the resulting load current does not exceed the ratings of the
semiconductor devices powered by this circuit. Neither is the choice
of smoothing capacitor value particularly critical, provided it is of the
order of 200 μF, or greater per watt output.

Tuning Meter
A tuning meter or signal meter indicates the strength of the signal into
which it is being tuned. It also serves as a tuning indicator, maximum
meter reading indicating the optimum tuning for that particular
station.

Fig. 79 Tuning indicator circuit. The diode (D1) can be any type
(germanium point-contact preferred).

The obvious point to measure a signal strength is at a ‘tapping’

point in the detector stage, so arranged that it draws only a minimal
current and thus does not degrade the signal strength. A suitable high
resistance circuit is shown in Fig. 79, feeding a microammeter. A small
proportion of detector current is tapped by this circuit, rectified by the
diode D1 and further filtered by the resistor and capacitor. High circuit
resistance is provided by the 10 k ohm potentiometer, which further
acts as an initial adjustment to set the meter at zero with the set switched
on, but in the absence of a signal. The meter itself should have a 0-10
122 Building and Designing Transistor Radios

microamp range. A miniature edge-type meter is usually preferred for

compactness.
It is also possible to ‘tap’ the age line for a tuning meter current (e.g.
in superhets), only in this case it is usual to employ a transistor to
amplify the variations in age voltage and measure the changes in the
collector current.

Battery Condition Indicator

Battery condition can be indicated by a paralleled circuit comprising a
resistor and a milliammeter or microammeter – Fig. 80. The value of
the resistance and meter range is chosen to give a full range reading at

Fig. 80 Basic battery-condition indicator circuit.

maximum battery voltage, this reading falling as the condition of the

battery deteriorates. The value of R can thus be calculated from:

Thus, using say a 0-1 milliammeter associated with a 9-volt battery

This high resistance circuit, drawing only 1 milliamp at a maximum,

could be left in circuit continously. It is generally advisable, however,
Miscellaneous Circuits 123

to isolate the battery condition circuit with a separate switch S2. It

is then only operative, i.e. indicating battery condition – when S2 is
depressed. With this arrangement it is also practical to use a lower
resistance circuit, e.g. to accommodate a 0-5 milliamp meter.
It should be noted that battery voltage under load (with S1 and S2
closed) will be less than the nominal battery voltage. Once the battery
voltage falls to less than 80 per cent of the nominal voltage under load
in the case of carbon-zinc cells, the battery has deteriorated to a stage
where it should be replaced. Lowering of supply voltage in a transistor
receiver normally introduces distortion, further aggravated by turning
up the volume control to compensate for the loss of output signal
strength.
A more practical form of battery condition indicator is, therefore,
to use a potentiometer for R (e.g. a 10 k ohm potentiometer in the
example above), so that with a fresh battery the meter can be adjusted
initially to give full scale reading under load. The meter scale is then
marked at 80 per cent full scale – at 0·8 milliamps in the case of a 0-1
milliamp meter. As long as the meter reads in the range 80 per cent-
100 per cent full scale reading (i.e. between 0·8-1·0 milliamps in the
example quoted), the battery is in ‘good’ condition. Any reading below
the 80 per cent mark indicates that the battery needs replacing.

Impedance Matching Networks

The simplest type of impedance matching is by an L network. The
left hand circuit shown in Fig. 81 is applicable when the desired
impedance Rin is greater than the load resistance R. The second

Fig. 81 Impedance matching networks.

124 Building and Designing Transistor Radios

diagram shows the circuit which can be used when the desired
impedance Rin is less than the load resistance R. The following
calculations apply:
Rin greater than R

Rin less than R:

 Chapter 13
Checking Radio Circuits

A universal meter is the chief tool used for checking or testing transistor
circuits. It should be capable of reading voltages up to 25 volts and
currents up to 1 amp or 5 amps. For voltage measurement it should
have a resistance of at least 20,000 ohms per volt; and for resistance
measurements an output terminal voltage of not more than 1·5 volts,
as a higher voltage could damage transistors or electrolytic capacitors
when the meter probes are applied to the circuit.
A dc voltage check across the emitter resistor of each transistor
provides an almost complete check on the dc conditions and will show
up such faults as component failures or open-circuit joints, and also
the stage at which the fault is present by the wrong voltage reading at
that stage. The fault can then be expected to be confined to components
in this stage, which can be checked individually if the fault is not
apparent. If individual resistors need checking, it is best to isolate the
component(s) for test by unsoldering the connection at one end. If
this is not readily possible or desirable, e.g. there may be a number of
components to test in the stage, the stage transistor (and the stage) can
be isolated by unsoldering the base lead of the transistor. A heat sink
should be used in order to avoid heat damage to the transistor and only
the base lead should be removed from the circuit.
Capacitor faults will not show up on a dc check, unless the capacitor
is short-circuited. In this case the fault will disappear (as far as dc
voltage measurements are concerned) as soon as the faulty capacitor
is disconnected. By-pass capacitors can be checked for open-circuit
faults by temporarily shunting with a 0·5 μF capacitor. If this produces
an increase in volume with the set switched on, the original capacitor
is either faulty or of wrong value.
The ac working of a receiver can only be checked satisfactorily by
signal injection. Standard procedure is to feed an audio-frequency into
the audio section of the receiver, starting from the output stage and
working backwards towards the detector. If all the audio stages work
126 Building and Designing Transistor Radios

satisfactorily, the if stages are similarly tested with the signal generator
set to the intermediate frequency modulated with an audio-frequency.
Finally the radio-frequency stages can be tested by injecting a radio-
frequency signal modulated with an audio-frequency.
In other words, to test the ac working of the complete receiver an
appropriate signal is injected at each stage in turn, starting with the
last stage and working back to the front end. The faulty stage will show
up as the first one which does not produce any increase in the audio-
frequency output, or yields no final output at all. The necessary signal is
usually best injected into the base of the appropriate transistor (i.e. that
particular stage transistor), via a suitable resistor in series to ensure
current drive. A suitable value of resistor is 10 kilohms. Having found
the faulty stage it is then necessary to check the components involved,
in turn, in order to determine the source of the fault, e.g. it could be
nothing more than a faulty coupling capacitor.
 Index

ac working, 125, 126 Class A amplifiers, 77 et seq Emitter, 49, 59

Aerial coils, 15, 16 Class B amplifiers, 81 et seq. Enhancement mode, 113
coil windings, 24 Coaxial cable, 30 Epitaxial silicon planar
coupling, 27, 89 Coding, transistors, 53 transistors, 50
Aerials, VHF, 28 Coil design, 15-17 External aerial, 7, 8, 25
af amplifier, 8-11 tapping points, 20, 26
Alignment, 99 windings, regenerative, Feedback, 68, 87, 109
Alloy diff usion, 49 88 Feeder impedance, 23
AM, 9, 21 Collector, 49, 59 Feeders, 30
AM/FM radio, 11 Control gate, 113 Ferrite rod, 24, 89
Audio amplifiers, 75 et seq. Current amplification rod aerials, 19
amplifier 1-watt, 84 factor, 64 slab, 24
amplifier 3-watt, 85 amplification, FET, 111 et seq.
frequencies, 7 transistors, 64 amplifier, 44, 86, 115
Automatic gain control, 97 biasing, 69, 72 transistor, 52
Autotransformer, 20, 109 Crossover distortion, 82 FM, 9, 21
Cross-modulation, 21 aerial connections, 30
Balun, 23 Crystal discriminator, 45 detector, 44
Bandpass fi lter, 97 set, 41 Folded dipole, 23
Bandwidth, 21 Frequency distortion, 75
Base, 49, 60 Decoupling, 109 Full wave rectifier, 120
Basic detector circuit, 41 Depletion layer, 32 Functional groups,
if amplifier, 96 mode, 113 transistors, 54
Battery condition Detector, 7-11, 39 et seq., 93
indicator, 122 Diode characteristics, 34 Gate, 111
Bias, 58 et seq., 112 rectifier, 120 voltage, 113
circuits, 68 temperature effects, 35 Germanium diodes, 35
voltage, 59 Dipole, 9 transistors, 49, 54
voltage, transistors, 58 aerial, 28
Bipolar device, 57 Direct coupling, 104 Half-wave aerial, 28
Blocking capacitor, 42 Directional aerial, 25 rectifier, 120
‘Bottom’ end, 113 Director, 22 Hi-Fi FM tuner, 116
Breakdown voltage, 36, 37 Discriminator, 45 Holes, 32 et seq.
Bridge circuit, 120 Double tuning, 97 Horizontal dipole, 29
Broadcast frequencies, 14 Drain, 111 ‘Hot’ end, 20
Driver, 81
Capacitive coupling, 103 Dynamic resistance, 18 if amplifiers, 95
Capacitor faults, 125 transformers, 95
Ceramic fi lters, 97 ‘Earthy’ end, 20 IGFET, 113
Channel current, 113 Edge-type meter, 122 Impedance, 107-8
voltage, 113 Electrons, 32 et seq. matching networks, 123
128 Building and Designing Transistor Radios

Inductance, 15 P-N-P transistors, 57 et seq. Source, 111

Inductive coupling, 102, 108 Point-contact diodes, 35,40 Specification figures,
Input characteristics, Polarities of transistors, 57 diodes, 38
transistors, 63 et seq. Polarized signal, 28 transistors, 66
circuit, transistor, 60, 67 Positive bias, 33 Superhet, 10
Intermediate frequency, Potential barrier, 33 alignment, 99 et seq.
9, 93 Power gain, 78 receivers, 93 et seq.
Interstage connections, 103 supply, 119, 120 Symbols, transistors, 60
et seq. transistors 55
Iron dust core, 35 Preamplifier, 86, 90, 94 Telescopic aerial, 29
Pre-selector, 31 Temperature effects, diode,
JFET, 113 Pro-Electron coding 54 35
Junction diodes, 35, 40 Pulse counting detector, 45 Thermistor, 82
Push-pull circuits, 81 et seq. Time constant, 40, 104
L network, 123 Tone control, 118, 119
lc coupling, 104 Q factor, 17, 18, 20, 87 ‘Top’ end, 113
Leakage current, 34, 35 Quality factor, 13, 15 Transfer characteristics,
Limiter, 45-6 transistors, 64
Linearity, 21, 75 Radio frequency, 9 Transformer coupling, 102
Load line, 80, 81 Ratings, transistors, 66 Transformers, 106-7
resistor, 39 Ratio detector, 44, 45 if, 95 et seq.
Local oscillator, 9-11 RC coupling, 104 Transistor outline shapes,
oscillator (mixer), 93 Reactance, 18, 109 55
Loft aerials, 22 Rectifier, 32 parameters, 62
Long wave, 14 Reflector, 22 power, 78
wave coil, 17 Regenerative circuit, 8 ratings, 66
Loop aerial, 25 receivers, 87 et seq. Transistors, 48 et seq.
Loudspeaker power, 75 Reservoir capacitor, 39 Trimmers, 19, 102
Resistance of transistors, 67 Tuned circuit, 7, 13
Medium wave, 14 Resonance, 13 Tuning capacitor, 15, 16
Mixer, 9, 10, 11 Resonant aerial length, 31 meter, 121
Modulated signals, 21 frequency, 13 Turns ratio, 106
MOSFET, 113 Reverse bias, 33 Types of diodes, 35
rf amplifiers, 86
N-channel, 111 Ribbon aerial, 23 Universal meter, 99, 125
N-P-N transistors, 57 et seq
Secondary Current, 107 Variactor, 36
Oscillation, 87 voltage, 107 Varicap, 36
Outline shapes, transistors Self-oscillating mixer, 94 VHF, 9, 14
55 Selectivity, 19, 20 aerial, 28
Output circuits, transistors Sensitivity, 20, 97 Voltage biasing, 70
60, 67 Short wave, 14 Voltage-stabilizing, 36
wave coil, 17 Volume control, 43, 86
Parasitic aerials, 22 Signal gate, 113
P-channel, 111 generator, 99 Wire aerial, 25-8
Piezoelectric crystals, 97 Silicon diodes, 35
Planar process, 49 transistors, 49, 54 Zener diode, 36, 54
 You may also be interested in:

Handbook of Cubik Math

by Alexander H. Frey Jr and David Singmaster
In thirty years, Rubik's Cube, invented by Erno Rubik, has established
itself as a neverending source of delight, frustration and intellectual
stimulation to children and adults alike. There is large literature on
the subject, but one of the few books to have established themselves as
offering a serious contribution to the subject is Alexander H.Frey's
and David Singmaster's Handbook of Cubik Math, first published in
1982.
Frey and Singmaster were the first to offer an elegant mathematical
solution to the cube, doing it in a form that enables readers to
understand the processes that have been undertaken. As a result the
book has proved readily accessible to generations of high-school maths
students and more advanced college students of algebra.
‘The best book yet written - or likely to be written - on the
wonders and the dark, unsolved mysteries of the cube.'
– Martin Gardiner
Alexander H. Frey Jr received his PhD in Mathematics in 1960. He
worked for over thirty years for IBM, and then for fifteen years on
distributed file system technology. Now retired, he pursues his interest
in the mathematical properties of many puzzles, including Sudoku
and four and five dimensional Rubik's Cubes.
David Singmaster is recognised as the world's leading authority on
the mathematics of Rubik's Cube. An American by birth and
citizenship he was until his retirement a Lecturer in Mathematical
Sciences at South bank University, London, England. His earlier Notes
on Rubik's Magic Cube was described by Scientific American as "the
definitive treatise".
Published 2010
Paperback ISBN: 978 0 7188 9209 8
PDF ISBN: 978 0 7188 4824 8
EPub ISBN: 978 0 718 84823 1
 You may also be interested in:

Clocks and Clock Repairing

by Eric Smith
For all those who are interested in horology, whether as a hobby or
within the trade, this user-friendly guide is invalubale as a wealth of
information for the internal workings of clocks.
Eric Smith provides a fully comprehensive manual to the repair of
clocks, whether modern or antique. Lucid language accompanies
practical diagrams and photographs to bring clarity to what is often a
highly complex task. The author recgonises the curiosity which many
feel about the workings of clocks. It is for this reason that the author
has written a guide which does not depict horology as too complex for
those outside the trade, but rather as a fascinating and accessible
occupation.
The ten chapters go through the practical stages of how different
clocks work and the tools and materials needed. Pendulum clocks,
30-hour, 8-day and 400-day clocks, chiming, striking, electric and
alarm clocks are all covered in detail. This new revised edition includes
a chapter on working with more advanced tools, including the lathe.
In addition, this volume contains a glossary of terms, a bibliogrpahy
and a list of suppliers of materials. It is also fully indexed.
The author recognises that there is much in horology which cannot
be self-taught from a book, yet Clocks and Clock Repairing is both a
book to stimulate interest and a handy reference guide.

Eric Smith is renowned for his clarity of explanation within the field
of horology in which he possesses a profound depth of knowledge.

Published 2005

Paperback ISBN: 978 0 7188 9167 1

 You may also be interested in:

Time, Science and Philosophy

by Hector C. Parr
Now, as never before, the big questions of life - purpose, the mind,
physical nature - are tied together as philosophers have to take on
board the findings of scientists about the mysteries of the universe and
physicists seek to explain ideas of human destiny.
All too often this crossover results in confusion being worse
confounded. Professional philosophers are seldom equipped to
comprehend cutting-edge scientific research, and when scientists
attempt to explain their findings they mystify the ordinary reader. Yet
the public remains eager to be enlightened; hence the great bestsellers
that end up unread.
This book is different. Without ever reducing the text to a simplistic
or patronising level, Hector Parr provides a guide to modern thought
that will genuinely inform the average person and provide a basis for
understanding more advanced works. He explores in detail the true
nature of Time, and pursues the scientific and philosophical
implications more fully than other books on the subject...; at once
scholarly, humane and challenging.
‘Heavy going, perhaps, for most people? Not at all. He feels
it should all be readily understood by the average intelligent
reader.'
– Teesdale Mercury
Hector C. Parr was educated at Queen Elizabeth's Grammar School,
Darlington, and Gonville and Caius College, Cambridge. For many
years he taught Mathematics and Computing. He was a Fellow of the
Royal College of Organists and an audio engineer in his spare time.
He died in 2018.
Published 1997

Paperback ISBN: 978 0 7188 2964 3

 Building and Designing Transistor Radios
In his characteristic accessible and clear style, R.H. Warring offers a
guide for the radio amateur to many of the things one ought to know
when tackling a radio project. Warring describes and illustrates the
design and working characteristics of the basic electronic ‘blocks’
from which radio receivers can be constructed and shows the reader
how to incorporate additional simple circuits, such as volume and
tone controls, and automatic gain control, to build upon these basic
blocks. Straightforward calculations are provided to determine the
values of components like resistors, capacitors and inductances to
optimise your results.
Building and Designing Transistor Radios will equip you with the
skills and knowledge properly to understand transistor radios. Special
attention is paid to the transistor, the most important component,
which the author provides insightful advice on. By understanding
the operating characteristics of the radio’s different elements, the
reader can learn how to devise circuits so that they perform most
efficiently. With 80 diagrams, no excessively complicated circuitry,
and minimal mathematical calculations included, this book first
published in 1977 remains perfect for amateurs and enthusiasts.

R.H. Warring (1920-1984) was one of the best-known writers in

the field of practical radio and electronics. He was renowned for Building and Designing
R.H. Warring

writing technical books that could be readily understood by all; his

vast experience meant he was keenly aware of the possible pitfalls
which could befall the beginner. Transistor Radios
The Lutterworth Press A Beginner’s Guide
P.O. Box 60
Cambridge
CB1 2NT
www.lutterworth.com
[email protected] R.H. Warring