0% found this document useful (0 votes)
46 views82 pages

Dokumen - Tips Fundamentals of Spectrum Analyzer

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
46 views82 pages

Dokumen - Tips Fundamentals of Spectrum Analyzer

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 82

Fundamentals of Spectrum Analyzer

Mr.Wang Junfeng
Engineer of Equipment Testing Division
State Radio Monitoring Center
[email protected]
+(86)10-68368866-1807

Radio Monitoring and Spectrum Management Training


(China,23-31,May,2005)

1
Contents

Fundamentals of Spectrum Analyzer


 Why do we use a spectrum analyzer?
 What can we do with a spectrum analyzer?
 How many types are there for spectrum analyzers?
 The classic superheterodyne spectrum analyzer
Illustration of test cases

2
Fundamentals of Spectrum Analyzer

Why do we use a spectrum analyzer?


What can we do with a spectrum analyzer?
How many types are there for spectrum analyzers?
The classic superheterodyne spectrum analyzer

3
Analyzing a RF signal

Time domain
 Oscilloscope Waveform S(t)
Frequency domain
 Spectrum analyzer Spectrum F{S(t)}=S(f)
Frequency and amplitude information
Modulation domain
 Vector signal analyzer Vector information
Frequency, amplitude and phase information

4
Analysis in time domain

X(t) Complicated in time domain

5
Analysis in time domain

6
Analysis in frequency domain

F{X(t)}=X(f) Simpler in frequency domain

7
Why do we use a spectrum analyzer?

Easier to verify each frequency component of the


signal looked like complicated in time domain
Easier to check what happened in band, in out-of-
band domain and in spurious emission domain
Easier to allocate and assign the frequency for
regulatory agency with assistance of the statistics
from spectrum monitoring

8
Fundamentals of Spectrum Analyzer

Why do we use a spectrum analyzer?


What can we do with a spectrum analyzer?
How many types are there for spectrum analyzers?
The classic superheterodyne spectrum analyzer

9
What can we do with a spectrum analyzer?

Frequency and amplitude test


Channel power and spectrum power density test
Wanted signal spectrum mask test
Adjacent channel power test
Spurious emission test
Occupied bandwidth test
Spectrum monitoring
Etc.

10
Fundamentals of Spectrum Analyzer

Why do we use a spectrum analyzer?


What can we do with a spectrum analyzer?
How many types are there for spectrum analyzers?
The classic superheterodyne spectrum analyzer

11
Types of spectrum analyzers

superheterodyne spectrum analyzer


Real-time spectrum analyzer (Fourier analyzer)

12
Types of spectrum analyzers
superheterodyne spectrum analyzer

Larger analyzing frequency range


Excellent sensitivity
Larger dynamic range
Swept-tuned, not a real-time equipment, not suitable for
short term phenomena
13
Types of spectrum analyzers
Real-time spectrum analyzer(Fourier analyzer)

A
A RAM
RAM FFT
FFT
DD

Real-time equipment, characterizing short term phenomena


 Phase as well as amplitude can be tested
Limitation of the frequency range, sensitivity, and dynamic
range

14
Fundamentals of Spectrum Analyzer

Why do we use a spectrum analyzer?


What can we do with a spectrum analyzer?
How many types are there for spectrum analyzers?
The classic superheterodyne spectrum analyzer

15
The classic superheterodyne spectrum analyzer

spectrum analyzer appearance

16
The classic superheterodyne spectrum analyzer

Block diagram of a classic superheterodyne spectrum analyzer

17
RF attenuator

Protect the following circuit


Adjust the signal entering the mixer at the optimum
level

18
RF attenuator

Input impedance 50ohms


Attenuation range 0 to 50dB or more
Attenuation accuracy less than 0.5dB
Minimum Step size 5dB, 2dB or 1dB

19
RF attenuator
Attenuation should be large enough to avoid mixer overload
* RBW 300 Hz Marker 1 [T1 ] * RBW 300 Hz Marker 1 [T1 ]
VBW 3 kHz -89.85 dBm VBW 3 kHz -86.87 dBm
R ef - 10 dB m * At t 0 dB * SWT 1.2 s 500.037500000 MHz R ef 1 0 d Bm * Att 15 dB * SWT 1.2 s 500.037500000 MHz

-10 10

A A
-20 0
OVLD

1 RM * 1 RM *
CLRWR CLRWR
-30 -10

-40 -20

-50 -30

PRN PRN

-60 -40

-70 -50

-80 -60

1
-90 -70

-100 -80

-110 -90

Center 500 MHz 7.5 kHz/ Span 75 kHz Center 500 MHz 7.5 kHz/ Span 75 kHz

Comment A: 11 Comment A: 11
Date: 27.APR.2005 16:41:38 Date: 27.APR.2005 16:42:16

20
RF attenuator

Suitable attenuation can ensure either the excellent


linear performance or the perfect noise floor of the
spectrum analyzer.

21
Pre-selector or low pass filter

Reduce the signal energy entering the mixer avoid


mixer overload
Keep unwanted signal from creating unwanted
response
22
Pre-selector or low pass filter

If there is no pre-selector or low pass filter, the


larger signal will cause mixer overload when the
smaller signal is tested. 23
Mixer and tunable LO

Frequency down conversion to IF


fIF = fLO-fsig or fIF =fsig- fLO
fsig = signal frequency, fLO = local oscillator frequency,
fIF = intermediate frequency (IF)
Tuning the spectrum analyzer to the desired range
Ramp sweep generator controls both LO and display 24
Mixer and tunable LO

Narrow IF filter Multiple mixing steps

25
Mixer

fIF = fLO-fsig or fIF =fsig- fLO


For certain fIF and fLO there are always two fsig fulfill the mixing
formula
For example:
fIF =3.9GHz, fIF = fLO-fsig , fLO =4.3GHz, fsig=400MHz
fIF =3.9GHz, fIF =fsig- fLO, fLO =4.3GHz, fsig= 8.2GHz

A pre-selector or a low pass filter is necessary to prevent


unwanted response created by image frequency from happening.

26
Mixer
Ideal mixing: fIF = fLO-fsig or fIF =fsig- fLO
Actual mixing: fIF = (fLO-fsig) +k1(fLO-fsig )2+k2(fLO-fsig )3 +--- or
fIF = (fsig-fLO) +k1(fsig- fLO)2+k2(fsig-fLO )3 +---
It is very important to reduce the non-linear components.It is
why that we need set a suitable attenuation to find a optimum
mixer level especially for harmonic measurement.

27
Mixer
How to distinguish the non-linear components?

28
Mixer

“ LO feedthrough ”

fIF = fLO-fsig
fsig =0
fIF = fLO

29
Mixer
* RBW 1 kHz Marker 1 [T1 ]
VBW 3 kHz 3.87 dBm
Ref 10 dBm * Att 10 dB SWT 20 ms 150.000000000 Hz

110

A
0

1 PK *
CLRWR
-10

-20

-30

PRN
D1 -36 dBm

-40

-50

-60
Why?
-70

-80

-90

Start 50 Hz 988 Hz/ Stop 9.93 kHz

Comment A: 11
Date: 27.APR.2005 16:52:47

30
Mixer

The LO signal is coupled into the first IF path due to its


limited isolation.

31
tunable LO
Usually tunable LO is controlled by the periodic
sawtooth signal. The scan generator runs freely.

32
tunable LO
What will happen if a pulse signal entering the spectrum analyzer?
Can we get the spectrum when spectrum analyzer run freely?
RBW 1 MHz Marker 1 [T1 ]
VBW 10 MHz -84.23 dBm

Ref - 20 dBm * Att 10 dB SWT 15 ms 2.500000 ms

-20

A
-30

1 RM * TRG -36.5 dBm


CLRWR
-40 TRG

-50

-60

PRN

-70

-80
1

-90

-100

-110 NO!
-120

Center 1 GHz 1.5 ms/

Date: 13.MAY.2005 10:56:14 33


tunable LO

How can get the spectrum of a pulse signal or a


TDMA signal?

The spectrum analyzer must be triggered by specific


signal.

34
tunable LO
Tunable LO could also be controlled by other specific signal.
The scan generator is controlled by specific condition, for instance
video signal trigger, IF signal trigger, gating trigger, and extend
signal trigger.

35
tunable LO
We can get the spectrum of the pulse signal if we set the specific
trigger condition for spectrum analyzer, for instance IF signal
trigger or extend signal trigger.
* RBW 30 kHz Marker 1 [T1 ]
VBW 300 kHz -26.09 dBm
Ref -20 dBm * Att 0 dB * SWT 145 ms 999.992000000 MHz

-20 1

A
-30

1 RM *
GAT
CLRWR
-40 TRG

-50

-60

PRN

-70

-80

-90

-100

-110

-120

Center 1 GHz 200 kHz/ Span 2 MHz

Date: 13.MAY.2005 10:57:30

36
IF processing circuit

IF amplifier variable gain amplifier, keeping


constant display
Reference level max level can be displayed
RBW filter determining the signal to be displayed
37
IF amplifier

The gain of the IF amplifier can be adjusted in small


step size, so the maximum signal level can be kept
constant in the subsequent signal processing
regardless of the attenuation setting and mixer level.
IF gain offset is coupled to the attenuator, so larger
attenuation would bring larger noise.

38
IF amplifier

Larger attenuation import larger noise


39
IF amplifier
LDAN=10log(k×t×BN,IF/1×10-3W)+NFSA-2.5dB)
LDAN : displayed average noise level
k: Boltzmann s constant, k=1.38×10-23W/Hz

T: ambient temperature, in K
BN,IF: noise bandwidth of IF filter
NFSA: noise figure of spectrum analyzer, in dB
-2.5dB: understanding of noise by sampling detector and averaging of
logarithmic level values
For the ambient temperature 290K(17C˚):
LDAN=-174dBm/Hz+(10logB N,IF/Hz)+NFSA-2.5dB
Increasing the attenuation, the noise figure of the spectrum analyzer
will get larger.

40
Reference level

Reference level is the max level can be displayed


Reference level should be large enough to avoid IF
overload
Usually reference level is coupled to the attenuation
to protect the mixer and subsequent circuit.

41
Reference level
* RBW 300 Hz Marker 1 [T1 ]
VBW 3 kHz -99.39 dBm
R ef - 10 dB m * Att 15 dB * SWT 1.2 s 500.037500000 MHz

-10

A
-20
IFOVL

1 RM *
CLRWR
-30

-40

-50

PRN

-60

-70

-80

-90

-100

-110

Center 500 MHz 7.5 kHz/ Span 75 kHz

Comment A: 11
Date: 27.APR.2005 16:42:41

IF overload Reference level:-10dBm


42
Reference level
* RBW 300 Hz Marker 1 [T1 ]
VBW 3 kHz -105.42 dBm
R ef -5 dBm * Att 15 dB * SWT 1.2 s 500.037500000 MHz

-10

1 RM * -20
CLRWR

-30

-40

PRN
-50

-60

-70

-80

-90

-100
1

Center 500 MHz 7.5 kHz/ Span 75 kHz

Comment A: 11
Date: 27.APR.2005 16:43:16

Reference level:-5dBm
43
RBW filter
RBW filter The final IF filter
 Resolution 3dB bandwidth of the IF filter
 Selectivity Filter wave shape factor
 Response time sweep time

44
RBW filter
Different RBW filter has different resolving capability
* RBW 30 kHz Marker 2 [T1 ] * RBW 300 kHz Marker 1 [T1 ]
VBW 300 kHz -31.61 dBm VBW 3 MHz -29.35 dBm
R ef - 20 d Bm * A tt 10 dB * SWT 245 ms 999.999000000 MHz R ef - 20 d Bm *A tt 10 dB * SWT 245 ms 1.000050000 GHz

-20 Marker 1 [T1 ] -20

-31.51 dBm
1
2 1 A
-30 1.000099000 GHz A
-30

1 RM*
1 RM *
CLRWR
-40 CLRWR
-40

-50
-50

-60
-60
PRN
PRN

-70
-70

-80
-80

-90

-90

-100

-100

-110

-110

-120

Center 1 GHz 50 kHz/ Span 500 kHz -120

Center 1 GHz 500 kHz/ Span 5 MHz

Date: 13.MAY.2005 10:51:17

Date: 13.MAY.2005 10:50:29

RBW=30kHz RBW=300kHz
45
RBW filter
RBW is the 3dB bandwidth of the final IF filter
* RBW 100 kHz Delta 2 [T1 ]
VBW 1 MHz -3.03 dB
R ef - 10 dB m * A tt 10 dB * SWT 195 ms -50.000000000 kHz

-10 Marker 1 [T1 ]


-21.27 dBm
1
1.000000000 GHz A
-20
2 3
Delta 3 [T1 ]
1 RM *
-3.12 dB
CLRWR
-30 50.000000000 kHz

-40

-50

PRN

-60

-70

-80

-90

-100

-110

Center 1 GHz 100 kHz/ Span 1 MHz

Date: 13.MAY.2005 11:13:12

46
RBW filter
Selectivity
Wave shape factor ratio of 60dB bandwidth to
3dB bandwidth

47
RBW filter

A low factor means a better selectivity


48
RBW filter

In theory, a rectangular filter has excellent selectivity.


But such a filter has a long transient response time.
Short measurement time can be achieved through
the use of Gaussian filter optimized for transients.

49
RBW filter
Response time
Sweep time
How long does it take to complete a sweep?
 RBW filter is a band limited filter and needs some time to
charge and discharge.
 Narrow RBW filter has higher resolving capability but needs
longer charging time.
 ST=k(span)/RBW2
ST: Sweep time
k: constant factor (variable for different filter types)

There are different kinds of filter allow resolution, selectivity and


measurement speed to be adapted to specific application.

50
RBW filter
Analog IF filter
 It is used to realize very large RBW, usually from 100kHz to
10MHz.
 Ideal Gaussian filter can not be implemented using analog filter.
 It is possible for a analog filter that the transient response is
almost identical with the ideal Gaussian filter within the 20dB
bandwidth.
SF=14, four filter circuit
SF=10, five filter circuit
Whereas SF=4.6, ideal Gaussian filter

51
RBW filter

Digital IF filter
 It is used to realize narrow RBW, usually less than 100kHz.
 The ideal Gaussian filter can be implemented by digital filter.
 Much better selectivity can be achieved, SF=4.6.
 Digital filter allows shorter sweep time than analog filter of
the same bandwidth.

52
RBW filter

Usually, sweep time is automatically coupled


to span and RBW
Sweep time can be changed manually. But try
to avoid the Mea Uncal error
Sweep time must be longer than minimum
sweep time

53
RBW filter
* RBW 100 Hz Marker 1 [T1 ]
* RBW 100 Hz Marker 1 [T1 ]
VBW 1 kHz -103.71 dBm
VBW 1 kHz -107.25 dBm
R ef - 5 d Bm * A tt 15 dB SWT 6 s 500.025000000 MHz
Ref - 5 dBm * Att 1 0 dB * SWT 3 s 500.025000000 MHz

-10
*
-10
A
UNCAL A

1 RM * -20
1 RM * -20
CLRWR
CLRWR

-30
-30

-40 -40

PRN PRN
-50 -50

-60 -60

-70 -70

-80 -80

-90
-90

-100
1 -100 1

Center 500 MHz 5 kHz/ Span 50 kHz


Center 500 MHz 5 kHz/ Span 50 kHz

Comment A: 11
Comment A: 11
Date: 27.APR.2005 16:49:10
Date: 27.APR.2005 16:44:17

54
RBW filter
Different RBW imports different noise level.
LDAN=-174dBm/Hz+(10logBN,IF/Hz)+NFSA-2.5dB
LDAN : displayed average noise level, in dBm
k: Boltzmann s constant, k=1.38×10-23W/Hz

T: ambient temperature, in K
BN,IF: noise bandwidth of IF filter, in Hz
NFSA: noise figure of spectrum analyzer, in dB
-2.5dB: understanding of noise by sampling detector and averaging
of logarithmic level values.
Different RBW result in different noise level.
The RBW setting is specified in measurement.

55
RBW filter

Larger RBW imports higher noise level 56


Video processing circuit

Log amplifier, envelope detector, detector, video filter,


display screen

57
Log amplifier

Compress the larger signal and increase the small


signal
To improve display dynamic range of the spectrum
analyzer.

58
Envelope detector

Convert the IF signal to video


The envelope of the IF signal is gotten.

59
Envelope detector

The envelope of one certain frequency component


can be obtained through 0 span at this frequency.

60
Envelope detector
The envelope of single time slot GSM signal
RBW 1 MHz Marker 1 [T1 ]
* RBW 30 kHz Marker 1 [T1 ]
VBW 10 MHz -84.23 dBm
VBW 300 kHz -26.09 dBm
Ref - 20 dBm * Att 1 0 dB SWT 15 ms 2.500000 ms
Ref -20 dBm * Att 0 dB * SWT 145 ms 999.992000000 MHz
-20
-20
1

A
-30
A
-30

1 RM * TRG -36.5 dBm


1 RM *
GAT CLRWR
-40 TRG
CLRWR
-40 TRG

-50
-50

-60
-60
PRN
PRN

-70
-70

-80
-80 1

-90
-90

-100
-100

-110
-110

-120 -120

Center 1 GHz 200 kHz/ Span 2 MHz Center 1 GHz 1.5 ms/

Date: 13.MAY.2005 10:57:30 Date: 13.MAY.2005 10:56:14

61
Envelope detector
The envelope of a sine wave signal

RBW 100 kHz Marker 1 [T1 ]


* RBW 100 kHz Marker 1 [T1 ]
VBW 1 MHz -21.27 dBm
VBW 1 MHz -21.26 dBm
Ref - 10 dBm * Att 10 dB SWT 2.5 ms 1.430000 ms
Ref - 10 dBm * Att 10 dB SWT 2.5 ms 1.000000000 GHz
-10
-10

1
A
1 -20
A
-20

1 RM *
1 RM *
CLRWR
-30
CLRWR
-30

-40
-40

-50
-50
PRN
PRN

-60
-60

-70
-70

-80
-80

-90
-90

-100
-100

-110 -110

Center 1 GHz 100 kHz/ Span 1 MHz Center 1 GHz 250 µs/

Date: 13.MAY.2005 11:29:33 Date: 13.MAY.2005 11:16:04

62
video filter

63
video filter
Low pass filter, reducing the impact of noise on the displayed
signal amplitude, smoothing the display

64
Detector

Digital display Analog to digital, finite display


point for a trace, normally 625 points or more
One display point represents a frequency range
What value should be displayed for each display
point among the frequency range?
Put all the data into a bucket and we need a math
formula to extract the data to be displayed
Different detector types mean different math formulae

65
Sample detector

The data point at the center of the bucket is displayed

66
Positive peak detector

The maximum data point of the bucket is displayed

67
Negative peak detector

The minimum data point of the bucket is displayed

68
Normal detector

The maximum data point of the bucket is displayed at odd display


point and the minimum data point of the bucket is displayed at
even display point

69
RMS detector

Statistic average, RMS value of all the data points in a


bucket is displayed.

N
1
  ui
2
U RMS 
N i 1

70
Average detector

Math average
N
1
U AV    ui
N i 1

71
Difference between various detectors
* RBW 100 kHz Marker 1 [T1 ]
VBW 300 kHz -77.86 dBm
R ef - 50 d Bm * Att 15 dB * SWT 720 ms 1.935508423 GHz

-50

A
-60

1 PK *
CLRWR
-70

1
2 SA *
CLRWR
-80

3 MI *
CLRWR
-90

PRN

-100

-110

-120

-130

-140

-150

Center 1.90025 GHz 10 MHz/ Span 100 MHz

Comment A: 11
Date: 27.APR.2005 12:37:31

Positive peak, sample and negative peak


72
Difference between various detectors
* RBW 100 kHz Marker 1 [T1 ]
VBW 1 MHz -77.85 dBm
R ef - 50 d Bm * Att 15 dB * SWT 1.2 s 1.905250000 GHz

-50

A
-60

1 PK *
CLRWR
-70

1
2 RM *
CLRWR
-80

3 AV *
CLRWR
-90

PRN

-100

-110

-120

-130

-140

-150

Center 1.90025 GHz 1 MHz/ Span 10 MHz

Comment A: 11
Date: 27.APR.2005 12:38:40

Positive peak, RMS and average


73
Difference between various detectors
* RBW 100 kHz Marker 1 [T1 ]
VBW 300 kHz -77.70 dBm
R ef - 50 d Bm * Att 15 dB * SWT 1.2 s 1.905250000 GHz

-50

A
-60

1 AP
CLRWR
-70

-80

-90

PRN

-100

-110

-120

-130

-140

-150

Center 1.90025 GHz 1 MHz/ Span 10 MHz

Comment A: 11
Date: 27.APR.2005 12:39:19

Normal
74
Fundamentals of Spectrum Analyzer

Fundamentals of Spectrum Analyzer


 Why do we use a spectrum analyzer?
 What can we do with a spectrum analyzer?
 How many types are there for spectrum analyzers?
 The classic superheterodyne spectrum analyzer
Illustration of test cases

75
How does the RBW filter work?
* RBW 1 kHz Marker 1 [T1 ]
VBW 10 kHz -99.87 dBm
Re f - 5 dB m * Att 15 dB * SWT 1.2 s 500.037500000 MHz

-10

1 RM * -20

CLRWR

-30

-40

PRN
-50

-60

-70

-80

-90

1
-100

Center 500 MHz 7.5 kHz/ Span 75 kHz

Comment A: 11
Date: 27.APR.2005 16:43:42

Spectrum for FM signal 1kHz mod 3kHz deviation


RBW:1kHz
76
How does the RBW filter work?
* RBW 100 Hz Marker 1 [T1 ]
VBW 1 kHz -103.71 dBm
R ef -5 dB m * Att 15 dB SWT 6 s 500.025000000 MHz

-10
*
A

1 RM * -20
CLRWR

-30

-40

PRN
-50

-60

-70

-80

-90

-100 1

Center 500 MHz 5 kHz/ Span 50 kHz

Comment A: 11
Date: 27.APR.2005 16:44:17

Spectrum for FM signal 1kHz mod 3kHz deviation


RBW:100Hz
77
How to set the RBW in measurement?

Spectrum mask measurement RBW is approximately


3% to 5% necessary bandwidth.
Unwanted emission measurement RBW is set
according to the test specification.

78
Channel power measurement
* RBW 100 Hz Marker 1 [T1 ]
VBW 1 kHz -104.43 dBm
Ref - 5 dBm * Att 10 dB SWT 6 s 500.025000000 MHz

CH PWR -4.25 dBm


-10
*
A

1 RM * -20
CLRWR

-30

-40

PRN
-50

-60

-70

-80

-90

C0
-100 C0
1

Center 500 MHz 5 kHz/ Span 50 kHz

Comment A: 11
Date: 27.APR.2005 16:45:37

Integral bandwidth:16kHz
79
Adjacent channel power measurement
* RBW 100 Hz Marker 1 [T1 ]
VBW 1 kHz -101.93 dBm
Ref -5 dBm * Att 10 dB SWT 18 s 500.025000000 MHz

-10

-20
A
-30
cl2
1 RM * -40
cl2
CLRWR
-50 cl1
cl1
-60 cu1
cu1
-70
cu2
-80 cu2
C0
-90
C0 1
-100

Center 500 MHz 15 kHz/ Span 150 kHz PRN

Tx Channel
Bandwidth 16 kHz
Power -4.23 dBm
Adjacent Channel
Bandwidth 16 kHz Lower -79.32 dB
Spacing 25 kHz
Upper -79.68 dB
Alternate Channel
Bandwidth 16 kHz Lower -79.83 dB
Spacing 50 kHz
Upper -79.93 dB

Comment A: 11
Date: 27.APR.2005 16:47:51

2 adjacent channel: 25kHz 50kHz


80
Occupied bandwidth
* RBW 100 Hz Marker 1 [T1 ]
VBW 1 kHz -102.21 dBm
Ref - 5 dBm * Att 10 dB SWT 6 s 500.025000000 MHz

OBW 8.092948718 kHz


-10 Temp 1 [T1 OBW]
-25.59 dBm A

499.995913462 MHz
T2
1 RM * -20 Temp 2 [T1 OBW]
T1
CLRWR
-21.82 dBm
500.004006410 MHz
-30

-40

PRN
-50

-60

-70

-80

-90

1
-100

Center 500 MHz 5 kHz/ Span 50 kHz

Comment A: 11
Date: 27.APR.2005 16:48:29

99% power bandwidth


81
The End
Thanks for your attention!!

Any questions?

82

You might also like