FUNDAMENTALS OF

SPECTRUM ANALYSIS
Christoph Rauscher
 Christoph Rauscher
Volker Janssen, Roland Minihold

Fundamentals of Spectrum Analysis

 © Rohde & Schwarz GmbH & Co. KG, 2001
Mühldorfstrasse 15
81671 München
Germany
www.rohde-schwarz.com

All copyrights are reserved, particularly those of translation, reprinting

(photo-copying), and reproduction. Also, any further use of this book,
particularly record-ing it digitally, recording it in microform, and distributing it,
e.g. via online data-bases, the filming of it, and the transmission of it are
prohibited. However, the use of excerpts for instructional purposes is permitted
provided that the source and proprietorship are indicated.

Even though the contents of this book were developed with utmost care, no lia-
bility shall be assumed for the correctness and completeness of this
information. Neither the author nor the publisher shall be liable under any
circumstances for any direct or indirect damage that may result from the use of
the information in this book.

The circuits, equipment, and methods described in this book may also be
pro-tected by patents, utility models, or design patterns even if not expressly
indi-cated. Any commercial use without the approval of possible licensees
represents an infringement of an industrial property right and may result in
claims for dam-ages. The same applies to the product names, company names,
and logos men-tioned in this book; thus, the absence of the ® or ™ symbol
cannot be assumed to indicate that the specified names or logos are free from
trademark protection.

If this book directly or indirectly refers to laws, regulations, guidelines, and stan-
dards (DIN, VDE, IEEE, ISO, IEC, PTB, etc), or cites information from them, neither
the author nor the publisher makes any guarantees as to the correctness, com-
pleteness, or up-to-dateness of the information. The reader is advised to consult
the applicable version of the corresponding documentation if necessary.
 Table of Contents

Table of Contents

1 Introduction 7

2 Signals 8

2.1 Signals displayed in time domain 8

2.2 Relationship between time and frequency domain 9

3 Configuration and Control Elements of a

Spectrum Analyzer 17

3.1 Fourier analyzer (FFT analyzer) 17

3.2 Analyzers operating in accordance

with the heterodyne principle 27

3.3 Main setting parameters 30

4 Practical Realization of an Analyzer Operating on the

­Heterodyne Principle 32

4.1 RF input section (frontend) 32

4.2 IF signal processing 44

4.3 Determination of video voltage and video filters 55

4.4 Detectors 61

4.5 Trace processing 73

4.6 Parameter dependencies 76

4.6.1 Sweep time, span, resolution and video bandwidths 76
4.6.2 Reference level and RF attenuation 80
4.6.3 Overdriving 86
 Fundamentals of S p e cctt r u m A n a l y s i s

5 Performance Features of Spectrum Analyzers 95

5.1 Inherent noise 95

5.2 Nonlinearities 102

5.3 Phase noise (spectral purity) 114

5.4 1 dB compression point and maximum input level 120

5.5 Dynamic range 125

5.6 Immunity to interference 135

5.7 LO feedthrough 138

5.8 Filter characteristics 139

5.9 Frequency accuracy 140

5.10 Level measurement accuracy 141

5.10.1 Uncertainty components 142
5.10.2 Calculation of total measurement uncertainty 148
5.10.3 Measurement error due to low signal-to-noise ratio 156

5.11 Sweep time and update rate 159

6 Frequent Measurements and Enhanced Functionality 162

6.1 Phase noise measurements 162

6.1.1 Measurement procedure 162
6.1.2 Selection of resolution bandwidth 165
6.1.3 Dynamic range 167

6.2 Measurements on pulsed signals 172

6.2.1 Fundamentals 173
6.2.2 Line and envelope spectrum 177
6.2.3 Resolution filters for pulse measurements 182
6.2.4 Analyzer parameters 184
6.2.5 Pulse weighting in spurious signal measurements 185
 Table of Contents

6.2.5.1 Detectors, time constants 186

6.2.5.2 Measurement bandwidths 190

6.3 Channel and adjacent-channel power measurement 190

6.3.1 Introduction 190
6.3.2 Key parameters for adjacent-channel
power measurement 193
6.3.3 Dynamic range in adjacent-channel
power measurements 194
6.3.4 Methods for adjacent-channel power measurement
195
195

198
200
201

204

206

Measurement Tips

Measurements in 75 W system 33
Measurement on signals with DC component 37
Maximum sensitivity 101
Identification of intermodulation products 112
Improvement of input matching 147
Fundamentals of Spectrum Analysis
 I n t r o d u ct i o n

1 Introduction

One of the most frequent measurement tasks in radiocommunications

is the examination of signals in the frequency domain. Spectrum analyz-
ers required for this purpose are therefore among the most versatile and
widely used RF measuring instruments. Covering frequency ranges of
up to 40 GHz and beyond, they are used in practically all applications of
wireless and wired communication in development, production, instal-
lation and maintenance efforts. With the growth of mobile communica-
tions, parameters such as displayed average noise level, dynamic range
and frequency range, and other exacting requirements regarding func-
tionality and measurement speed come to the fore. Moreover, spectrum
analyzers are also used for measurements in the time domain, such as
measuring the transmitter output power of time multiplex systems as a
function of time.

This book is intended to familiarize the uninitiated reader with the field
of spectrum analysis. To understand complex measuring instruments
it is useful to know the theoretical background of spectrum analysis.
Even for the experienced user of spectrum analyzers it may be helpful
to recall some background information in order to avoid measurement
errors that are likely to be made in practice.

In addition to dealing with the fundamentals, this book provides an

insight into typical applications such as phase noise and channel power
measurements.

For further discussions of this topic, refer also to Engelson [1-1] and
[1-2].

7
 Signals

2 Signals

2.1 Signals displayed in time domain

In the time domain the amplitude of electrical signals is plotted versus

time – a display mode that is customary with oscilloscopes. To clearly
illustrate these waveforms, it is advantageous to use vector projection.
The relationship between the two display modes is shown in Fig. 2-1 by
way of a simple sinusoidal signal.
jlm

1
0.8
0.6
ω0 t
0.4
0.2
0
Re
–0.2
–0.4
–0.6
–0.8
–1
0 0.5 T0 T0 1.5 T0 2 T0
t
Fig. 2-1  S
 inusoidal signal displayed by projecting a complex rotating vector on
the imaginary axis

The amplitude plotted on the time axis corresponds to the vector pro-
jected on the imaginary axis (jIm). The angular frequency of the vector
is obtained as:

w 0 = 2 ⋅ p ⋅ f0  (Equation 2-1)

where w0 angular frequency

f0 signal frequency

A sinusoidal signal with x(t)= A  · sin(2 · p · f0 · t ) can be described as

x(t)=A · Im{e j · 2p·f 0 ·t }.

8
 Relationship Between Time and Frequency Domain

2.2 Relationship between time and frequency domain

Electrical signals may be examined in the time domain with the aid of
an oscilloscope and in the frequency domain with the aid of a spectrum
analyzer (see Fig. 2-2).

A
0
t
ma in
e do
Tim A
t

0
Frequ
ency d
omain f

Fig. 2-2  Signals examined in time and frequency domain

The two display modes are related to each other by the Fourier trans-
form (denoted F), so each signal variable in the time domain has a char-
acteristic frequency spectrum. The following applies:
+∞

{ }
X f ( f ) = F x (t ) = ∫ x (t ) ⋅ e- j 2 pft dt  (Equation 2-2)
-∞

and
+∞

{ }
x (t ) = F -1 X f ( f ) = ∫ X f ( f ) ⋅ e j 2 pft dt  (Equation 2-3)
-∞

where F{x (t)} Fourier transform of x(t)

F –1 {X(f )} inverse Fourier transform of X(f )
x (t ) signal in time domain
Xf(f ) complex signal in frequency domain

To illustrate this relationship, only signals with periodic response in the

time domain will be examined first.

9
 Signals

Periodic signals
According to the Fourier theorem, any signal that is periodic in the time
­domain can be derived from the sum of sine and cosine signals of dif-
ferent frequency and amplitude. Such a sum is referred to as a Fourier
­series. The following applies:

A0 ∞ ∞
x (t ) = + ∑ An ⋅ sin(n ⋅ w0 ⋅ t ) + ∑ B0 ⋅ cos(n ⋅ w0 ⋅ t )  (Equation 2-4)
2 n=1 n=1

The Fourier coefficients A0, An and Bn depend on the waveform of ­signal

x(t ) and can be calculated as follows:
T0
2
A0 =
T0
∫ x (t )dt  (Equation 2-5)
0

    T0
2
An =
T0 ∫ x (t ) ⋅ sin(n ⋅ w0 ⋅ t )dt  (Equation 2-6)
0     

T0
2
Bn =
T0
∫ x (t ) ⋅ cos(n ⋅ w0 ⋅ t ) dt  (Equation 2-7)
0

where A0 DC component
2
x(t ) signal in time domain
n order of harmonic oscillation
T0 period
w 0 angular frequency

Fig. 2-3b shows a rectangular signal approximated by a Fourier series.

The individual components are shown in Fig. 2-3a Approximation of a
rectangular signal by summation of various sinusoidal oscillationsAp-
proximation of a rectangular signal by summation of various sinusoidal
oscillations. The greater the number of these components, the closer the
signal approaches the ideal rectangular pulse.

10
 Relationship Between Time and Frequency Domain

a) Harmonics

x(t)

n=1
n=3
n=5 n=7

t
b) Sum of harmonics

x(t)

Fig. 2-3  
Approximation of a
rectangular signal by
summation of various
sinusoidal oscillations
t
In the case of a sine or cosine signal a closed-form solution can be found
for Equation 2-2 so that the following relationships are obtained for the
complex spectrum display:

{ (
F sin 2 ⋅ p ⋅ f0 ⋅ t )} = 1j ⋅ d ( f -f ) = -j ⋅ d ( f -f ) 
0 0
(Equation 2-8)

and

{ (
F cos 2 ⋅ p ⋅ f0 ⋅ t )} = d ( f -f ) 
0 (Equation 2-9)

where d(f -f0) is a Dirac function d(f -f0) = ∞ if f -f0 = 0, and f  = f0
d(f -f0) = 0, otherwise
+∞

∫ d ( f - f ) df
0 = 1
-∞

11
 Signals

It can be seen that the frequency spectrum both of the sine signal and
cosine signal is a Dirac function at f0 (see also Fig. 2-5a). The Fourier
transforms of sine and cosine signal are identical in magnitude, so that
the two signals exhibit an identical magnitude spectrum at the same
frequency f0 .
To calculate the frequency spectrum of a periodic signal whose time
characteristic is described by a Fourier series in accordance with Equa-
tion 2-4, each component of the series has to be transformed. Each of
these elements leads to a Dirac function, that is a discrete component
in the frequency domain. Periodic signals therefore always exhibit dis-
crete spectra which are also referred to as line spectra. Accordingly, the
spectrum shown in Fig. 2-4 is obtained for the approximated rectangu-
lar signal of Fig. 2-3.

|X(f )|

Fig. 2-4   
Magnitude spectrum of
approximated rectan-
gular signal shown in
Fig. 2-3
f0 3f0 5f0 7f0 f

Fig. 2-5 shows some further examples of periodic signals in the time and
frequency domain.

Non-periodic signals
Signals with a non-periodic characteristic in the time domain cannot
be described by a Fourier series. Therefore the frequency spectrum of
such signals is not composed of discrete spectral components. Non-peri-
odic signals exhibit a continuous frequency spectrum with a frequency-
dependent spectral density. The signal in the frequency domain is calcu-
lated by means of a Fourier transform (Equation 2-2).
Similar to the sine and cosine signals, a closed-form solution can be
found for Equation 2-2 for many signals. Tables with such transform
pairs can be found in [2-1].
For signals with random characteristics in the time domain, such as
noise or random bit sequences, a closed-form solution is rarely found.

12
 Relationship Between Time and Frequency Domain

The frequency spectrum can in this case be determined more easily by a

numeric solution of Equation 2-2.
Fig. 2-6 shows some non-periodic signals in the time and frequency
domain.

a)
Time domain Frequency domain

A |A|
–

0
T0 t 1 f
f0 = ––
Sinusoidal signal T0

b)

A |A|
–

0
t fT – fS fT fT + fS f
Amplitude-modulated signal

c)
Âp sin x
A |A| Envelope si(x ) = –––
– x
τ
τ
sin n · ––
Tp (
·π
)
Ân· fp = Âp · –– · 2 · –––––––––
τ
Tp n · –– · π
Tp

0
0
Tp τ t _1 _2 _3 f
τ τ τ
Periodic rectangular signal
1
––
Tp

Fig. 2-5  Periodic signals in time and frequency domain (magnitude spectra)

13
 Signals

a) Time domain Frequency domain

A |A|
–

0
t f
Band-limited noise
b)

sin x
Envelope si(x) = _____
A |A| x
–
1

0
TBit t 1/TBit 2/TBit 3/TBit f
Random bit sequence
c)
I
A lg| –A|
0

t
Q
A
0

t fC f
QPSK signal

Fig. 2-6  Non-periodic signals in time and frequency domain

Depending on the measurement to be performed, examination may be

useful either in the time or in the frequency domain. Digital data trans-
mission jitter measurements, for example, require an oscilloscope. For
determining the harmonic content, it is more useful to examine the sig-
nal in the frequency domain:

14
 Relationship Between Time and Frequency Domain

The signal shown in Fig. 2-7 seems to be a purely sinusoidal signal with
a frequency of 20 MHz. Based on the above considerations one would
expect the frequency spectrum to consist of a single component at
20 MHz.
On examining the signal in the frequency domain with the aid of a
spectrum analyzer, however, it becomes evident that the fundamental
(1st order harmonic) is superimposed by several higher-order harmon-
ics i.e.multiples of 20 MHz (Fig. 2-8). This information cannot be easily
obtained by examining the signal in the time domain. A practical quan-
titative assessment of the higher-order harmonics is not feasible. It is
much easier to examine the short-term stability of frequency and ampli-
tude of a sinusoidal signal in the frequency domain compared to the
time domain (see also chapter 6.1 Phase noise measurement).

Fig. 2-7  
Sinusoidal signal  
(f = 20 MHz) exam-
ined on oscilloscope
Ch1 500 mV M 10.0 ns CH1 –560 mV

15
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

*RBW 300 kHz Marker 1 [T1 CNT]

*VBW 3 kHz 14.61 dBm
Ref 20 dBm Att 50 dB SWT 175 ms 20.000 MHz
20 1 Delta 2 [T1]
–45.25 dB
10 20.08800000 MHz A
1 AP
CLRWR 0

–10
PRN
–20

–30 2

–40

–50

–60

–70

–80
Center 39 MHz 6.2 MHz/ Span 62 MHz

Fig. 2-8  T he sinusoidal signal of Fig. 2-7 examined in the frequency domain with
the aid of a spectrum analyzer

16
 F o u r i e r A n a l y z e r (FFT A n a l y z e r )

3 Configuration and Control Elements of a

Spectrum Analyzer

Depending on the kind of measurement, different requirements are

placed on the maximum input frequency of a spectrum analyzer. In view
of the various possible configurations of spectrum analyzers, the input
frequency range can be subdivided as follows:
❙ AF range up to approx. 1 MHz
❙ RF range up to approx. 3 GHz
❙ microwave range up to approx. 40 GHz
❙ millimeter-wave range above 40 GHz

The AF range up to approx. 1 MHz covers low-frequency electronics as

well as acoustics and mechanics. In the RF range, wireless communica-
tion applications are mainly found, such as mobile communications and
sound and TV broadcasting, while frequency bands in the microwave
or millimeter-wave range are utilized to an increasing extent for broad-
band applications such as digital radio links.
Various analyzer concepts can be implemented to suit the frequency
range. The two main concepts are described in detail in the following
sections.

3.1 Fourier analyzer (FFT analyzer)

As explained in chapter 2, the frequency spectrum of a signal is clearly

defined by the signal’s time characteristic. Time and frequency domain
are linked to each other by means of the Fourier transform. Equation 2-2
can therefore be used to calculate the spectrum of a signal recorded in
the time domain. For an exact calculation of the frequency spectrum
of an input signal, an infinite period of observation would be required.
Another prerequisite of Equation 2-2 is that the signal amplitude should
be known at every point in time. The result of this calculation would be a
continuous spectrum, so the frequency resolution would be unlimited.
It is obvious that such exact calculations are not possible in prac-
tice. Given certain prerequisites, the spectrum can nevertheless be deter-
mined with sufficient accuracy.

17
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

In practice, the Fourier transform is made with the aid of digital signal
processing, so the signal to be analyzed has to be sampled by an ana-
log-digital converter and quantized in amplitude. By way of sampling
the continuous input signal is converted into a time-discrete signal and
the information about the time characteristic is lost. The bandwidth of
the input signal must therefore be limited or else the higher signal fre-
quencies will cause aliasing effects due to sampling (see Fig. 3-1). Accord-
ing to Shannon’s law of sampling, the sampling frequency fS must be at
least twice as high as the bandwidth Bin of the input signal. The follow-
ing applies:

fS ≥ 2 ⋅ B in and 1  (Equation 3-1)

fS =
TS
where fS sampling rate
Bin signal bandwidth
TS sampling period

For sampling lowpass-filtered signals (referred to as lowpass signals)

the minimum sampling rate required is determined by the maximum
signal frequency fin,max . Equation 3-1 then becomes:

fS ≥ 2 ⋅ fin, max  (Equation 3-2)

If fS = 2 · fin,max , it may not be possible to reconstruct the signal from the

sampled values due to unfavorable sampling conditions. Moreover, a
lowpass filter with infinite skirt selectivity would be required for band
limitation. Sampling rates that are much greater than 2 · fin, max are there-
fore used in practice.

A section of the signal is considered for the Fourier transform. That is,
only a limited number N of samples is used for calculation. This process
is called windowing. The input signal (see Fig. 3-2a) is multiplied with a
specific window function before or after sampling in the time domain. In
the example shown in Fig. 3-2, a rectangular window is used (Fig. 3-2b).
The result of multiplication is shown in Fig. 3-2c.

18
 F o u r i e r A n a l y z e r (FFT A n a l y z e r )

a)
fS
A A ––
2
Sampling with
sampling rate fS

fin t fin fS fS + fin 2 fS 3 fS f

fS – fin
b)

A A
fS
fin,max < –– fS
2 ––
2

fin,max t fin,max fS 2 fS 3 fS f

c)
Aliasing
A A
fS fS
fin,max > –– ––
2 2

fA t fin, max fS 2 fS 3 fS f
fin,max > ––
2

 ampling a lowpass signal with sampling rate fS a), b) fin, max < fS/2,  

Fig. 3-1  S
c) fin, max > fS/2, therefore ambiguity exists due to aliasing

The calculation of the signal spectrum from the samples of the signal
in the time domain is referred to as a discrete Fourier transform (DFT).
Equation 2-2 then becomes:
N -1
X (k ) = ∑ x (nT ) ⋅ e S
-j 2 pk n /N
 (Equation 3-3)
n =0

where k index of discrete frequency bins,

where k = 0, 1, 2, …
n index of samples
x(nTS) samples at the point n  · TS , where n = 0, 1, 2, …
N length of DFT, i. e. total number of samples used for
calculation of Fourier transform

19
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

The result of a discrete Fourier transform is again a discrete frequency

spectrum (see Fig. 3-2d). The calculated spectrum is made up of individ-
ual components at the frequency bins which are expressed as:
fS 1
()
f k = k⋅
N
= k⋅
N ⋅ TS
 (Equation 3-4)

where f (k) discrete frequency bin

k index of discrete frequency bins, where k = 0, 1, 2 …
fS sampling frequency
N length of DFT

It can be seen that the resolution (the minimum spacing required

between two spectral components of the input signal for the latter being
displayed at two different frequency bins f (k) and f (k + 1) depends on
the observation time N ·TS . The required observation time increases
with the desired resolution.
The spectrum of the signal is periodicized with the period fS through
sampling (see Fig. 3-1). Therefore, a component is shown at the frequency
bin f (k = 6) in the discrete frequency spectrum display in Fig. 3-2d. On
examining the frequency range from 0 to fS in Fig. 3-1a, it becomes evi-
dent that this is the component at fS-fin .

In the example shown in Fig. 3-2, an exact calculation of the signal spec-
trum was possible. There is a frequency bin in the discrete frequency
spectrum that exactly corresponds to the signal frequency. The follow-
ing requirements have to be fulfilled:
❙ the signal must be periodic (period T0)
❙ the observation time N ·TS must be an integer multiple of the
period T0 of the signal.

These requirements are usually not fulfilled in practice so that the result
of the Fourier transform deviates from the expected result. This devia-
tion is characterized by a wider signal spectrum and an amplitude error.
Both effects are described in the following.

20
 F o u r i e r A n a l y z e r (FFT A n a l y z e r )

a) Input signal x(t) Samples |–X(f)|

1

A | –A|

–1 0
TA 0 Te t fin = 1 f
b) Window w(t ) |W(f)| Tin
–
1

A | –A|

0
N · TS

0
0 t 1 0 –––
1 f
c) x (t) · w(t)
– –––
N · TS N · TS
1
N=8
A

–1
0 t
d) x (t ) · w(t ), continued periodically |X( f) · W (f)|k = 2 k= 6
1

A | –A|

–1
0 N · TLS t k = 0 k = 1 fe fA 1 f
frequency bins –– ––––
2 N · TA

Fig. 3-2  D
 FT with periodic input signal. Observation time is an integer multiple
of the period of the input signal

21
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

a) Samples |X(f
Input signal x(t) – )|
1

A |A|
–

–1 0
TS 0 Te t fin = 1 f
Tin
b) Window w(t) |W
– (f )|
1

A | –A|

0
N · TS

0
0 t 1 0 –––1 f
c) – –––
x(t) · w(t) N · T S N · TS
1
N= 8
A

–1
0 t
d) x(t) · w(t), continued periodically |X(f
– ) · –W(f )|
1

A |A|
–

N=8
–1 f
0 N · TS t k = 0 k = 1 fin fS
–– 1 f –f
–––
frequency bins 2 N · TS S in

Fig. 3-3  D
 FT with periodic input signal. Observation time is not an integer
multiple of the period of the input signal

22
 F o u r i e r A n a l y z e r (FFT A n a l y z e r )

The multiplication of input signal and window function in the time

domain corresponds to a convolution in the frequency domain (see [2-1]).
In the frequency domain the magnitude of the transfer function of the
rectangular window used in Fig. 3-2 follows a sine function:

sin (2pf ⋅ N ⋅ TS/2)

W ( f ) = N ⋅ TS⋅ si (2pf ⋅ N ⋅ TS/2) = N ⋅ TS⋅  
2pf ⋅ N ⋅ TS/2
 (Equation 3-5)

where W (f ) windowing function in frequency domain

N ·TS window width

In addition to the distinct secondary maxima, nulls are obtained at mul-

tiples of 1 / (N ·TS). Due to the convolution by means of the window
function the resulting signal spectrum is smeared, so it becomes dis-
tinctly wider. This is referred to as leakage effect.
If the input signal is periodic and the observation time N ·TS is an
integer multiple of the period, there is no leakage effect of the rectan-
gular window since, with the exception of the signal frequency, nulls
always fall within the neighboring frequency bins (see Fig. 3-2d).
If these conditions are not satisfied, which is the normal case, there
is no frequency bin that corresponds to the signal frequency. This case
is shown in Fig. 3-3. The spectrum resulting from the DFT is distinctly
wider since the actual signal frequency lies between two frequency bins
and the nulls of the windowing function no longer fall within the neigh-
boring frequency bins.
As shown in Fig. 3.3d, an amplitude error is also obtained in this case.
At constant observation time the magnitude of this amplitude error
depends on the signal frequency of the input signal (see Fig. 3-4). The
error is at its maximum if the signal frequency is exactly between two
frequency bins.

23
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

max.
amplitude error

Fig. 3-4   
Amplitude error
caused by rectangular
­windowing as a function
of signal frequency
fin f(k)

Frequency bins

By increasing the observation time it is possible to reduce the absolute

widening of the spectrum through the higher resolution obtained, but
the maximum possible amplitude error remains unchanged. The two
effects can, however, be reduced by using optimized windowing instead
of the rectangular window. Such windowing functions exhibit lower sec-
ondary maxima in the frequency domain so that the leakage effect is
reduced as shown in Fig. 3-5. Further details of the windowing functions
can be found in [3-1] and [3-2].

To obtain the high level accuracy required for spectrum analysis a flat-
top window is usually used. The maximum level error of this window-
ing function is as small as 0.05 dB. A disadvantage is its relatively wide
main lobe which reduces the frequency resolution.

24
 F o u r i e r A n a l y z e r (FFT A n a l y z e r )

Rectangular window Hann window

Amplitude error

Leakage

f f
Fig. 3-5  L eakage effect when using rectangular window or Hann window
(MatLab® simulation)

The number of computing operations required for the Fourier transform

can be reduced by using optimized algorithms. The most widely used
method is the fast Fourier transform (FFT). Spectrum analyzers operat-
ing on this principle are designated as FFT analyzers. The configuration
of such an analyzer is shown in Fig. 3-6.
Input
A
RAM FFT
D

Lowpass Memory Display

Fig. 3-6  Configuration of FFT analyzer

To adhere to the sampling theorem, the bandwidth of the input signal

is limited by an analog lowpass filter (cutoff frequency fc = fin, max) ahead
of the A/D converter. After sampling the quantized values are saved in
a memory and then used for calculating the signal in the frequency
domain. Finally, the frequency spectrum is displayed.

Quantization of the samples causes the quantization noise which causes

a limitation of the dynamic range towards its lower end. The higher the
resolution (number of bits) of the A/D converter used, the lower the
quantization noise.

25
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

Due to the limited bandwidth of the available high-resolution A/D con-

verters, a compromise between dynamic range and maximum input fre-
quency has to be found for FFT analyzers. At present, a wide dynamic
range of about 100 dB can be achieved with FFT analyzers only for low-
frequency applications up to 100 kHz. Higher bandwidths inevitably
lead to a smaller dynamic range.
In contrast to other analyzer concepts, the phase information is not
lost during the complex Fourier transform. FFT analyzers are therefore
able to determine the complex spectrum by magnitude and phase. If
they feature sufficiently high computing speed, they even allow realtime
analysis.
FFT analyzers are not suitable for the analysis of pulsed signals (see
Fig. 3-7). The result of the FFT depends on the selected section of the
time function. For correct analysis it is therefore necessary to know cer-
tain parameters of the analyzed signal, such as the triggering a specific
measurement.

A N · TS = n · T0
Window

N · TS T0
t

A A A

f 1 f 1 f
–– ––
T0 T0
Fig. 3-7  F FT of pulsed signals. The result depends on the time of the
­measurement

26
 A n a l y z e r s O p e r a t i n g A cc o r d i n g to the Heterodyne Principle

3.2 Analyzers operating in accordance with the heterodyne

principle

Due to the limited bandwidth of the available A/D converters, FFT ana-
lyzers are only suitable for measurements on low-frequency signals. To
display the spectra of high-frequency signals up to the microwave or
millimeter-wave range, analyzers with frequency conversion are used.
In this case the spectrum of the input signal is not calculated from the
time characteristic, but determined directly by analysis in the frequency
domain. For such an analysis it is necessary to break down the input
spectrum into its individual components. A tunable bandpass filter as
shown in Fig. 3-8 could be used for this purpose.

Tunable
bandpass filter Amplifier Detector Display
Input
y

Sawtooth

Tunable bandpass filter

A

Fig. 3-8  
Block diagram of
spectrum analyzer
with tunable
bandpass filter
fin

The filter bandwidth corresponds to the resolution bandwidth (RBW) of

the analyzer. The smaller the resolution bandwidth, the higher the spec-
tral resolution of the analyzer.
Narrowband filters tunable throughout the input frequency range
of modern spectrum analyzers are however technically hardly feasi-
ble. Moreover, tunable filters have a constant relative bandwidth with

27
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

respect to the center frequency. The absolute bandwidth therefore

increases with increasing center frequency so that this concept is not
suitable for spectrum analysis.
Spectrum analyzers for high input frequency ranges therefore usu-
ally operate in accordance with the principle of a heterodyne receiver.
The block diagram of such a analyzer is shown in Fig. 3-9.

Logarithmic Envelope
Mixer IF amplifier IF filter amplifier detector Video filter
Input

Local oscillator
y
x

Sawtooth
Fig. 3-9  Block diagram of spectrum analyzer operating on heterodyne principle

The heterodyne receiver converts the input signal with the aid of a mixer
and a local oscillator (LO) to an intermediate frequency (IF). If the local
oscillator frequency is tunable (a requirement that is technically fea-
sible), the complete input frequency range can be converted to a con-
stant intermediate frequency by varying the LO frequency. The resolu-
tion of the analyzer is then given by a filter at the IF with fixed center
frequency.
In contrast to the concept described above, where the resolution fil-
ter as a dynamic component is swept over the spectrum of the input sig-
nal, the input signal is now swept past a fixed-tuned filter.
The converted signal is amplified before it is applied to the IF fil-
ter which determines the resolution bandwidth. This IF filter has a con-
stant center frequency so that problems associated with tunable filters
can be avoided.

28
 A n a l y z e r s O p e r a t i n g A cc o r d i n g to the Heterodyne Principle

To allow signals in a wide level range to be simultaneously displayed

on the screen, the IF signal is compressed using of a logarithmic ampli-
fier and the envelope determined. The resulting signal is referred to as
the video signal. This signal can be averaged with the aid of an adjust-
able lowpass filter called a video filter. The signal is thus freed from
noise and smoothed for display. The video signal is applied to the verti-
cal deflection of a cathode-ray tube. Since it is to be displayed as a func-
tion of frequency, a sawtooth signal is used for the horizontal deflection
of the electron beam as well as for tuning the local oscillator. Both the
IF and the LO frequency are known. The input signal can thus be clearly
assigned to the displayed spectrum.

A IF filter

Input signal
converted to IF

fIF f

A IF filter

Input signal
converted to IF

Fig. 3-10 
Signal “swept past”
resolution filter in
heterodyne receiver
fIF f

29
 Configuration and Control Elements of a S p e ct r u m A n a l y z e r

In modern spectrum analyzers practically all processes are controlled by

one or several microprocessors, giving a large variety of new functions
which otherwise would not be feasible. One application in this respect is
the remote control of the spectrum analyzer via interfaces such as the
IEEE bus.
Modern analyzers use fast digital signal processing where the input
signal is sampled at a suitable point with the aid of an A/D converter and
further processed by a digital signal processor. With the rapid advances
made in digital signal processing, sampling modules are moved further
ahead in the signal path. Previously, the video signal was sampled after
the analog envelope detector and video filter, whereas with modern spec-
trum analyzers the signal is often digitized at the last low IF. The enve-
lope of the IF signal is then determined from the samples.
Likewise, the first LO is no longer tuned with the aid of an analog
sawtooth signal as with previous heterodyne receivers. Instead, the LO
is locked to a reference frequency via a phase-locked loop (PLL) and
tuned by varying the division factors. The benefit of the PLL technique
is a considerably higher frequency accuracy than achievable with ana-
log tuning.
An LC display can be used instead of the cathode-ray tube, which
leads to more compact designs.

3.3 Main setting parameters

Spectrum analyzers usually provide the following elementary setting

parameters (see Fig. 3-11):

Frequency display range

The frequency range to be displayed can be set by the start and stop fre-
quency (that is the minimum and maximum frequency to be displayed),
or by the center frequency and the span centered about the center fre-
quency. The latter setting mode is shown in Fig. 3-11. Modern spectrum
analyzers feature both setting modes.

Level display range

This range is set with the aid of the maximum level to be displayed (the
reference level), and the span. In the example shown in Fig. 3-11, a refer-
ence level of 0 dBm and a span of 100 dB is set. As will be described later,
the attenuation of an input RF attenuator also depends on this setting.

30
 M a i n S e tt i n g P a r a m e t e r s

Frequency resolution
For analyzers operating on the heterodyne principle, the frequency reso-
lution is set via the bandwidth of the IF filter. The frequency resolution
is therefore referred to as the resolution bandwidth (RBW).

Sweep time (only for analyzers operating on the heterodyne principle)

The time required to record the whole frequency spectrum that is of
interest is described as sweep time.

Some of these parameters are dependent on each other. Very small res-
olution bandwidths, for instance, call for a correspondingly long sweep
time. The precise relationships are described in detail in chapter 4.6.

Fig. 3-11  Graphic display of recorded spectrum

31
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

4 Practical Realization of an Analyzer Operating on the

­Heterodyne Principle

This chapter provides a detailed description of the individual compo-

nents of an analyzer operating on the heterodyne principle as well as
of the practical implementation of a modern spectrum analyzer for the
frequency range 9 kHz to 3 GHz/7 GHz. A detailed block diagram can be
found on the fold-out page at the end of the book. The individual blocks
are numbered and combined in functional units.

4.1 RF input section (frontend)

Like most measuring instruments used in modern telecommunications,

spectrum analyzers usually feature an RF input impedance of 50 W. To
enable measurements in 75 W systems such as cable television (CATV),
some analyzers are alternatively provided with a 75 W input impedance.
With the aid of impedance transformers, analyzers with 50 W input may
also be used (see measurement tip: Measurements in 75 W system).

A quality criterion of the spectrum analyzer is the input VSWR, which is

highly influenced by the frontend components, such as the attenuator,
input filter and first mixer. These components form the RF input section
whose functionality and realization will be examined in detail in the fol-
lowing.
A step attenuator (2)* is provided at the input of the spectrum ana-
lyzer for the measurement of high-level signals. Using this attenuator,
the signal level at the input of the first mixer can be set.
The RF attenuation of this attenuator is normally adjustable in 10 dB
steps. For measurement applications calling for a wide dynamic range,
attenuators with finer step adjustment of 5 dB or 1 dB are used in some
analyzers (see chapter 5.5: Dynamic range).

* The colored code numbers in parentheses refer to the block diagram at the end of the
book.

32
 RF I n p u t S e ct i o n (F r o n t e n d )

  Measurements in 75 W system

In sound and TV broadcasting, an impedance of 75 W is more
common than the widely used 50 W. To carry out measurements
in such systems with the aid of spectrum analyzers that usually
feature an input impedance of 50 W, appropriate matching pads
are required. Otherwise, measurement errors would occur due to
mismatch between the device under test and spectrum analyzer.
The simplest way of transforming 50 W to 75 W is by means
of a 25 W series resistor. While the latter renders for low inser-
tion loss (approx. 1.8 dB), only the 75 W input is matched, however,
the output that is connected to the RF input of the spectrum ana-
lyzer is mismatched (see Fig. 4-1a). Since the input impedance of
the spectrum analyzer deviates from the ideal 50 W value, mea-
surement errors due to multiple reflection may occur especially
with mismatched DUTs.
Therefore it is recommendable to use matching pads that
are matched at both ends (e. g. Π or L pads). The insertion loss
through the attenuator may be higher in this case.

a)
Spectrum
Source 75 Ω 100 Ω analyzer
25 Ω

Zout = 75 Ω Zin = 50 Ω

b)
Spectrum
Fig. 4-1    Source 75 Ω 50 Ω analyzer
Input matching
Matching
to 75 W using pad
external
matching pads Zout = 75 Ω Zin = 50 Ω

33
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

The heterodyne receiver converts the input signal with the aid of a mixer
(4) and a local oscillator (5) to an intermediate frequency (IF). This type
of frequency conversion can generally be expressed as:

m ⋅ fLO ± n ⋅ fin = fIF  (Equation 4-1)

where m, n 1, 2, …
fLO frequency of local oscillator
fin frequency of input signal to be converted
fIF intermediate frequency

If the fundamentals of the input and LO signal are considered (m, n = 1),
Equation 4-1 is simplified to:

fLO ± fin = fIF  (Equation 4-2)

or solved for fin

fin = fLO ± fIF  (Equation 4-3)

With a continuously tunable local oscillator, a further input frequency

range can be implemented at constant frequency. For specific LO and
intermediate frequencies, Equation 4-3 shows that there are always two
receive frequencies for which the criterion set by Equation 4-2 is ful-
filled (see Fig. 4-2). This means that in addition to the wanted receive
frequency there are also image frequencies. To ensure unambiguity of
this concept, input signals at such unwanted image frequencies have to
be rejected with the aid of suitable filters ahead of the RF input of the
mixer.

34
 RF I n p u t S e ct i o n (F r o n t e n d )

Conversion
A

Input filter
Image frequency
reponse

∆ f = fIF

fIF fin, u fLO fin, o f

Fig. 4-2  Ambiguity of heterodyne principle

Conversion
A Overlap of
input and image
frequency range

LO frequency range

Input frequency Image frequency

range range

fIF fin, min fLO,min fim,min fe,max fLO, max fim,max f

Fig. 4-3  Input and image frequency ranges (overlapping)

Fig. 4-3 illustrates the input and image frequency ranges for a tunable
receiver with low first IF. If the input frequency range is greater than
2 · fIF , the two ranges are overlapping, so an input filter must be imple-
mented as a tunable bandpass for image frequency rejection without
affecting the wanted input signal.
To cover the frequency range from 9 kHz to 3 GHz, which is typical of
modern spectrum analyzers, this filter concept would be extremely com-
plex because of the wide tuning range (several decades). Much less com-
plex is the principle of a high first IF (see Fig. 4-4).

35
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

Conversion
A

Input filter

Image frequency
Input frequency range range
fIF = fLO – fin LO frequency fIF = fim – fLO
range

fIF f
Fig. 4-4  Principle of high intermediate frequency

In this configuration, image frequency range lies above the input fre-
quency range. Since the two frequency ranges do not overlap, the image
frequency can be rejected by a fixed-tuned lowpass filter. The following
relationships hold for the conversion of the input signal:

fIF = fLO - fin  (Equation 4-4)

and for the image frequency response:

fIF = fim - fLO  (Equation 4-5)

Frontend for frequencies up to 3 GHz

The analyzer described here uses the principle of high intermediate fre-
quency to cover the frequency range from 9 kHz to 3 GHz. The input
attenuator (2) is therefore followed by a lowpass filter (3) for rejection of
the image frequencies. Due to the limited isolation between RF and IF
port as well as between LO and RF port of the first mixer, this lowpass
filter also serves for minimizing the IF feedthrough and LO ­reradiation
at the RF input.

In our example the first IF is 3476.4 MHz. For converting the input fre-
quency range from 9 kHz to 3 GHz to an upper frequency of 3476.4 MHz,
the LO signal (5) must be tunable in the frequency range from
3476.40 MHz to 6476.4 MHz. According to Equation 4-5, an image fre-
quency range from 6952.809 MHz to 9952.8 MHz is then obtained.

36
 RF I n p u t S e ct i o n (F r o n t e n d )

  Measurement on signals with DC component

Many spectrum analyzers, in particular those featuring a very
low input frequency at their lower end (such as 20 Hz), are DC-
coupled, so there are no coupling capacitors in the signal path
between RF input and first mixer.
A DC voltage may not be applied to the input of a mixer
because it usually damages the mixer diodes. For measurements
of signals with DC components, an external coupling capacitor
(DC block) is used with DC-coupled spectrum analyzers. It should
be noted that the input signal is attenuated by the insertion loss
of this DC block. This insertion loss has to be taken into account
in absolute level measurements.
Some spectrum analyzers have an integrated coupling capac-
itor to prevent damage to the first mixer. The lower end of the fre-
quency range is thus raised. AC-coupled analyzers therefore have
a higher input frequency at the lower end, such as 9 kHz.

Due to the wide tuning range and low phase noise far from the carrier
(see chapter 5.3: Phase noise) a YIG oscillator is often used as local oscil-
lator. This technology uses a magnetic field for tuning the frequency of
a resonator.
Some spectrum analyzers use voltage-controlled oscillators (VCO)
as local oscillators. Although such oscillators feature a smaller tuning
range than the YIG oscillators, they can be tuned much faster than YIG
oscillators.

To increase the frequency accuracy of the recorded spectrum, the LO sig-

nal is synthesized. That is, the local oscillator is locked to a reference sig-
nal (26) via a phase-locked loop (6). In contrast to analog spectrum ana-
lyzers, the LO frequency is not tuned continuously, but in many small
steps. The step size depends on the resolution bandwidth. Small reso-
lution bandwidths call for small tuning steps. Otherwise, the input sig-
nal may not be fully recorded or level errors could occur. To illustrate
this effect, a filter tuned in steps throughout the input frequency range
is shown in Fig. 4-5. To avoid such errors, a step size that is much lower
than the resolution bandwidth (such as 0.1 · BN) is selected in p ­ ractice.

37
 praCTiCal realizaTion of an analyzer operaTing on The heTerodyne prinCiple

a) Input signal
A

fin
Tuning step >> resolution bandwidth

Displayed spectrum
fin

b) Input signal
A

fin Fig. 4-5 

Tuning step >> resolution bandwidth Effects of too large 
tuning steps 
A a) input signal is 
completely lost 
b) level error in display 
Displayed spectrum
of input signal
fin

The reference signal is usually generated by a temperature-controlled

crystal oscillator (TCXO). To increase the frequency accuracy and long-
term stability (see also chapter 5.9: Frequency accuracy), an oven-con-
trolled crystal oscillator (OCXO) is optionally available for most spec-
trum analyzers. For synchronization with other measuring instruments,
the reference signal (usually 10 MHz) is made available at an output
connector (28). The spectrum analyzer may also be synchronized to an
externally applied reference signal (27). If only one connector is avail-
able for coupling a reference signal in or out, the function of such con-
nector usually depends on a setting internal to the spectrum analyzer.

38
 RF I n p u t S e ct i o n (F r o n t e n d )

As shown in Fig. 3-9, the first conversion is followed by IF signal pro-

cessing and detection of the IF signal. With such a high IF, narrowband
IF filters can hardly be implemented, which means that the IF signal in
the concept described here has to be converted to a lower IF (such as
20.4 MHz in our example).

A 2nd conversion

Image rejection
filter

2nd IF Image 1st IF f

2nd LO
Fig. 4-6  Conversion of high 1st IF to low 2nd IF

With direct conversion to 20.4 MHz, the image frequency would only
be offset 2 · 20.4 MHz = 40.8 MHz from the signal to be converted at
3476.4 MHz (Fig. 4-6). Rejection of this image frequency is important
since the limited isolation between the RF and IF port of the mixers
signals may be passed to the first IF without conversion. This effect
is referred to as IF feedthrough (see chapter 5.6: Immunity to interfer-
ence). If the frequency of the input signal corresponds to the image fre-
quency of the second conversion, this effect is shown in the image fre-
quency response of the second IF. Under certain conditions, input sig-
nals may also be converted to the image frequency of the second conver-
sion. Since the conversion loss of mixers is usually much smaller than
the isolation between RF and IF port of the mixers, this kind of image
frequency response is far more critical.

Due to the high signal frequency, an extremely complex filter with high
skirt selectivity would be required for image rejection at a low IF of
20.4 MHz. It is therefore advisable to convert the input signal from the
first IF to a medium IF such as 404.4 MHz as in our example. A fixed
LO signal (10) of 3072 MHz is required for this purpose since the image
frequency for this conversion is at 2667.6 MHz. Image rejection is then

39
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

simple to realize with the aid of a suitable bandpass filter (8). The band-
width of this bandpass filter must be sufficiently large so that the sig-
nal will not be impaired even for maximum resolution bandwidths. To
reduce the total noise figure of the analyzer, the input signal is amplified
(7) prior to the second conversion.

The input signal converted to the second IF is amplified again, filtered

by an image rejection bandpass filter for the third conversion and con-
verted to the low IF of 20.4 MHz with the aid of a mixer. The signal thus
obtained can be subjected to IF signal processing.

Frontend for frequencies above 3 GHz

The principle of a high first IF calls for a high LO frequency range
(fLO, max = fin, max + f1st IF). In addition to a broadband RF input, the first
mixer must also feature an extremely broadband LO input and IF out-
put – requirements that are increasingly difficult to satisfy if the upper
input frequency limit is raised. Therefore this concept is only suitable for
input frequency ranges up to 7 GHz.

To cover the microwave range, other concepts have to be implemented

by taking the following criteria into consideration:
❙ The frequency range from 3 GHz to 40 GHz extends over more than a
decade, whereas 9 kHz to 3 GHz corresponds to approx. 5.5 decades.
❙ In the microwave range, filters tunable in a wide range and with nar-
row relative bandwidth can be implemented with the aid of YIG tech-
nology [4-1]. Tuning ranges from 3 GHz to 50 GHz are fully r­ ealizable.

Direct conversion of the input signal to a low IF calls for a tracking

bandpass filter for image rejection. In contrast to the frequency range
up to 3 GHz, such preselection can be implemented for the range above
3 GHz due to the previously mentioned criteria. Accordingly, the local
oscillator need only be tunable in a frequency range that corresponds to
the input frequency range.
In our example the frequency range of the spectrum analyzer is thus
enhanced from 3 GHz to 7 GHz. After the attenuator, the input signal
is split by a diplexer (19) into the frequency ranges 9 kHz to 3 GHz and
3 GHz to 7 GHz and applied to corresponding RF frontends.

40
 RF I n p u t S e ct i o n (F r o n t e n d )

In the high-frequency input section, the signal passes a tracking YIG fil-
ter (20) to the mixer. The center frequency of the bandpass filter corre-
sponds to the input signal frequency to be converted to the IF. Direct
conversion to a low IF (20.4 MHz, in our example) is difficult with this
concept due to the bandwidth of the YIG filter. It is therefore best to con-
vert the signal first to a medium IF (404.4 MHz) as was performed with
the low-frequency input section.
In our example, a LO frequency range from 2595.6 MHz to
6595.6 MHz would be required for converting the input signal as upper
sideband, (that is for fIF = fin- fLO). For the conversion as lower side-
band (fIF = fLO- fin), the local oscillator would have to be tunable from
3404.4 MHz to 7404.4 MHz.
If one combines the two conversions by switching between the upper
and lower sideband at the center of the input frequency band, this con-
cept can be implemented even with a limited LO frequency range of
3404.4 MHz to 6595.6 MHz (see Fig. 4-7).

A Input signal converted

Tracking preselection as lower sideband

LO frequency range
Input frequency range

fIF f

A Input signal converted

as upper sideband

LO frequency range
Input frequency range

fIF fin,min fin, max f

Input frequency range

= Tuning range of bandpass filter
Fig. 4-7  Conversion to a low IF; image rejection by tracking preselection

41
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

The signal converted to an IF of 404.4 MHz is amplified (23) and cou-

pled into the IF signal path of the low-frequency input section through
a switch (13).

Upper and lower frequency limits of this implementation are determined

by the technological constraints of the YIG filter. A maximum frequency
of about 50 GHz is feasible.

In our example, the upper limit of 7 GHz is determined by the tuning

range of the local oscillator. There are again various possibilities for
converting input signals above 7 GHz with the specified LO frequency
range:

Fundamental mixing
The input signal is converted by means of the fundamental of the LO
signal. For covering a higher frequency range with the specified LO fre-
quency range it is necessary to double, for instance, the LO signal fre-
quency by means of a multiplier before the mixer.

Harmonic mixing
The input signal is converted by a means of a harmonic of the LO signal
produced in the mixer due to the mixer’s nonlinearities.

Fundamental mixing is preferred to obtain minimal conversion loss,

thereby maintaining a low noise figure for the spectrum analyzer. The
superior characteristics attained in this way, however, require complex
processing of the LO signal. In addition to multipliers (22), filters are
required for rejecting subharmonics after multiplying. The amplifiers
required for a sufficiently high LO level must be highly broadband since
they must be designed for a frequency range that roughly corresponds
to the input frequency range of the high-frequency input section.
Conversion by means of harmonic mixing is easier to implement but
implies a higher conversion loss. A LO signal in a comparatively low fre-
quency range is required which has to be applied at a high level to the
mixer. Due to the nonlinearities of the mixer and the high LO level, har-
monics of higher order with sufficient level are used for the conversion.
Depending on the order m of the LO harmonic, the conversion loss of the
mixer compared to that in fundamental mixing mode is increased by:

42
 RF I n p u t S e ct i o n (F r o n t e n d )

DaM = 20 dB ⋅ lg m  (Equation 4-6)

where DaM increase of conversion loss compared to that in

­fundamental mixing mode
m order of LO harmonic used for conversion

The two concepts are employed in practice depending on the price class
of the analyzer. A combination of the two methods is possible. For exam-
ple, a conversion using the harmonic of the LO signal doubled by a mul-
tiplier would strike a compromise between complexity and sensitivity at
an acceptable expense.

External mixers
For measurements in the millimeter-wave range (above 40 GHz), the fre-
quency range of the spectrum analyzer can be enhanced by using exter-
nal harmonic mixers. These mixers also operate on the principle of har-
monic mixing, so that a LO signal in a frequency range that is low com-
pared to the input signal frequency range is required.
The input signal is converted to a low IF by means of a LO harmonic
and an IF input inserted at a suitable point into the IF signal path of the
low-frequency input section of the analyzer.
In the millimeter-wave range, waveguides are normally used for
conducted signal transmission. Therefore, external mixers available
for enhancing the frequency range of spectrum analyzers are usually
waveguides. These mixers do not normally have a preselection filter and
therefore do not provide for image rejection. Unwanted mixture prod-
ucts have to be identified with the aid of suitable algorithms. Further
details about frequency range extension with the aid of external har-
monic mixers can be found in [4-2].

43
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

4.2 IF signal processing

IF signal processing is performed at the last intermediate frequency

(20.4 MHz in our example).
Here the signal is amplified again and the resolution bandwidth
defined by the IF filter.
The gain at this last IF can be adjusted in defined steps (0.1 dB steps
in our example), so the maximum signal level can be kept constant in
the subsequent signal processing regardless of the attenuator setting
and mixer level. With high attenuator settings, the IF gain has to be
increased so that the dynamic range of the subsequent envelope detec-
tor and A/D converter will be fully utilized (see chapter 4.6: Parameter
dependencies).
The IF filter is used to define that section of the IF-converted input
signal that is to be displayed at a certain point on the frequency axis.
Due to the high skirt selectivity and resulting selectivity characteristics,
a rectangular filter would be desirable. The transient response, however,
of such rectangular filters is unsuitable for spectrum analysis. Since
such a filter has a long transient time, the input signal spectrum could
be converted to the IF only by varying the LO frequency very slowly
to avoid level errors from occurring. Short measurement times can be
achieved through the use of Gaussian filters optimized for transients.
The transfer function of such a filter is shown in Fig. 4-8.
0
–3
–6
lg (Hv ( f )) / dB

Fig. 4-8   
Voltage transfer function
–60 of Gaussian filter
f0 f

In contrast to rectangular filters featuring an abrupt transition from

passband to stopband, the bandwidth of Gaussian filters must be spec-
ified for filters with limited skirt selectivity. In spectrum analysis it is
common practice to specify the 3 dB bandwidth (the frequency spacing

44
 IF S i g n a l P r o c e s s i n g

between two points of the transfer function at which the insertion loss
of the filter has increased by 3 dB relative to the center frequency).

H 2V(f)
HV(f )

HV, 0 H 2V, 0
Voltage Power
transfer transfer
Pulse Noise
function function
bandwidth bandwidth
BI BN
0.5 0.5

f0 f f0 f

Fig. 4-9  Voltage and power transfer function of Gaussian filter

For many measurements on noise or noise-like signals (e. g. digitally mod-

ulated signals) the measured levels have to be referenced to the mea-
surement bandwidth, in our example the resolution bandwidth. To this
end the equivalent noise bandwidth BN of the IF filter must be known
which can be calculated from the transfer function as follows:
+∞
1
BN =
H V2, 0
()
⋅ ∫ H V2 f ⋅ df  (Equation 4-7)
0

where BN noise bandwidth

HV(f ) voltage transfer function
HV, 0 value of voltage transfer function at center of band
(at f0)

This can best be illustrated by looking at the power transfer function

(see Fig. 4-9). The noise bandwidth corresponds to the width of a rectan-
gle with the same area as the area of the transfer function HV2 (f ). The
effects of the noise bandwidth of the IF filter are dealt with in detail in
chapter 5.1 Inherent noise.

45
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

For measurements on correlated signals, as can typically be found in the

field of radar, the pulse bandwidth is also of interest. In contrast to the
noise bandwidth, the pulse bandwidth is calculated by integration of the
voltage transfer function. The following applies:
+∞
1
BI =
H V, 0
()
⋅∫ H V f ⋅ df  (Equation 4-8)
0

where BI pulse bandwidth

HV( f ) voltage transfer function
HV,0 value of voltage transfer function at center of band
(at f0)

The pulse bandwidth of Gaussian or Gaussian-like filters corresponds

approximately to the 6 dB bandwidth. In the field of interference mea-
surements, where spectral measurements on pulses are frequently car-
ried out, 6 dB bandwidths are exclusively specified. Further details of
measurements on pulsed signals can be found in chapter 6.2.
Chapter 6 concentrates on pulse and phase noise measurements.
For these and other measurement applications the exact relationships
between 3 dB, 6 dB, noise and pulse bandwidth are of particular inter-
est. Table 4-1 provides conversion factors for various filters that are
described in detail further below.

Initial value is 4 filter circuits 5 filter circuits Gaussian filter

3 dB bandwidth (analog) (analog) (digital)
6 dB bandwidth (B6 dB) 1.480 · B3 dB 1.464 · B3 dB 1.415 · B3 dB
Noise bandwidth (BN) 1.129 · B3 dB 1.114 · B3 dB 1.065 · B3 dB
Pulse bandwidth (BI) 1.806 · B3 dB 1.727 · B3 dB 1.506 · B3 dB
Initial value is
6 dB bandwidth
3 dB bandwidth (B3 dB) 0.676 · B6 dB 0.683 · B6 dB 0.707 · B6 dB
Noise bandwidth (BN) 0.763 · B6 dB 0.761 · B6 dB 0.753 · B6 dB
Pulse bandwidth (BI) 1.220 · B6 dB 1.179 · B6 dB 1.065 · B6 dB

Table 4-1  R
 elationship between 3 dB / 6 dB bandwidths and noise and pulse
bandwidths

46
 IF S i g n a l P r o c e s s i n g

If one uses an analyzer operating on the heterodyne principle to record

a purely sinusoidal signal, one would expect a single spectral line in
accordance with the Fourier theorem even when a small frequency
span about the signal frequency is taken. In fact, the display shown in
Fig. 4-10 is obtained.

*RBW 10 kHz Marker 1 [T1]

*VBW 30 Hz –5.16 dBm
Ref 0 dBm Att 30 dB SWT 680 ms 1.00000000 GHz
0
T T2 ndB [T1] 3.00 dB
BW 9.80000000 kHz
1 AP –10 A
CLRWR Temp 1 [T1 ndB]
–20 –81.62 dBm
999.95000000 MHz
–30 Temp 2 [T2 ndB]
–8.22 dBm
PRN
–40 1.00000500 GHz

–50

–60

–70

–80

–90

–100
Center 1 GHz 10 kHz/ Span 100 kHz

Fig. 4-10  IF filter imaged by a sinusoidal input signal

The display shows the image of the IF filter. During the sweep, the input
signal converted to the IF is “swept past” the IF filter and multiplied with
the transfer function of the filter.

A schematic diagram of this process is shown in Fig. 4-11. For reasons of

simplification the filter is “swept past” a fixed-tuned signal, both kinds
of representations being equivalent.

47
 praCTiCal realizaTion of an analyzer operaTing on The heTerodyne prinCiple

Input
A
signal IF filter

A
Image of
resolution
Fig. 4-11 
bandwidth
IF filter imaged by an 
input signal “swept past” 
the filter (schematic 
representation of imag-
ing process)
f

As pointed out before, the spectral resolution of the analyzer is mainly

determined by the resolution bandwidth, that is, the bandwidth of the
IF filter. The IF bandwidth (3 dB bandwidth) corresponds to the mini-
mum frequency offset required between two signals of equal level to
make the signals distinguishable by a dip of about 3 dB in the display
when using a sample or peak detector (see chapter 4.4.). This case is
shown in Fig. 4-12a. The red trace was recorded with a resolution band-
width of 30 kHz. By reducing the resolution bandwidth, the two signals
are clearly distinguishable (Fig. 4-12a, blue trace).

48
 IF S i g n a l P r o c e s s i n g

If two neighboring signals have distinctly different levels, the weaker

signal will not be shown in the displayed spectrum at a too high resolu-
tion bandwidth setting (see Fig. 4-12b, red trace). By reducing the resolu-
tion bandwidth, the weak signal can be displayed.
In such cases, the skirt selectivity of the IF filter is also important
and is referred to as the selectivity of a filter. The skirt selectivity is spec-
ified in form of the shape factor which is calculated as follows:
B60 dB
SF60/3 =  (Equation 4-9)
B3 dB

where B3 dB 3 dB bandwidth

B60 dB 60 dB bandwidth

For 6 dB bandwidths, as is customary in EMC measurements, the shape

factor is derived from the ratio of the 60 dB bandwidth to the 6 dB band-
width.

The effects of the skirt selectivity can clearly be seen in Fig. 4-13. One
Kilohertz IF filters with different shape factors were used for the two
traces. In the blue trace (SF = 4.6), the weaker signal can still be recog-
nized by the dip, but a separation of the two signals is not possible in the
red trace (SF = 9.5) where the weaker signal does not appear at all.

49
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

*RBW 3 kHz
*VBW 3 kHz
Ref –10 dBm Att 20 dB SWT 45 ms
–10 *
A
–20

1AP –30
CLRWR
–40
PRN
–50

–60

–70

–80

–90

–100

–110
Center 100.015 MHz 20 kHz/ Span 200 kHz

*RBW 3 kHz
*VBW 1 kHz
Ref –10 dBm Att 20 dB SWT 135 ms
–10
*
–20 A

–30

–40
PRN
–50

–60

–70

–80

–90

–100

–110
Center 100 MHz 20 kHz/ Span 200 kHz

Fig. 4-12  S
 pectrum of input signal consisting of two sinusoidal carriers with
same and with different level, recorded with different resolution
bandwidths (blue traces RBW = 3 kHz, red traces RBW = 30 kHz)

50
 IF S i g n a l P r o c e s s i n g

RBW 1 kHz RF Att 20 dB

Ref Lvl VBW 200 kHz
–10 dBm SWT 300 ms Unit dBm
–10
*
–20 A
–30
SF = 9.5
–40
SF = 4.6
–50 1SA
2AP
–60

–70

–80

–90

–100

–110
Center 100 MHz 2 kHz/ Span 20 kHz

Fig. 4-13  T wo neighboring sinusoidal signals with different levels recorded with
a resolution bandwidth of 1 kHz and a shape factor of 9.5 and 4.6

If the weaker signal is to be distinguished by a filter with a lower skirt

selectivity, the resolution bandwidth has to be reduced. Due to the lon-
ger transient time of narrowband IF filters, the minimum sweep time
must be increased. For certain measurement applications, shorter sweep
times are therefore feasible with filters of high skirt selectivity.

As mentioned earlier, the highest resolution is attained with narrow-

band IF filters. These filters, however, always have a longer transient
time than broadband filters, so contemporary spectrum analyzers pro-
vide a large number of resolution bandwidths to allow resolution and
measurement speed to be adapted to specific applications. The setting
range is usually large (from 10 Hz to 10 MHz). The individual filters are
implemented in different ways. There are three different types of ­filters:
❙ analog filters
❙ digital filters
❙ FFT

51
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

Analog IF filters
Analog filters are used to realize very large resolution bandwidths. In
the spectrum analyzer described in our example, these are bandwidths
from 100 kHz to 10 MHz. Ideal Gaussian filters cannot be implemented
using analog filters. A very good approximation, however, is possible
at least within the 20 dB bandwidth so that the transient response is
almost identical to that of a Gaussian filter. The selectivity characteris-
tics depend on the number of filter circuits. Spectrum analyzers typically
have four filter circuits, but models with five filter circuits can be found,
too. Shape factors of about 14 and 10 can thus be attained, whereas an
ideal Gaussian filter exhibits a shape factor of 4.6.
The spectrum analyzer described in our example uses IF filters that
are made up of four individual circuits. Filtering is distributed so that
two filter circuits each (29 and 31) are arranged before and after the IF
amplifier (30). This configuration offers the following benefits:
❙ The filter circuits ahead of the IF amplifier provide for rejection of
mixture products outside the passband of the IF filter. Intermodula-
tion products that may be caused by such signals in the last IF ampli-
fier without prefiltering can thus be avoided (see chapter 5.2: Nonlin-
earities).
❙ The filter circuits after the IF amplifier are used to reduce the noise
bandwidth. If they were arranged ahead of the IF amplifier, the total
noise power in the subsequent envelope detection would be distinctly
higher due to the broadband noise of the IF amplifier.

Digital IF filters
Narrow bandwidths can best be implemented with the aid of digital sig-
nal processing. In contrast to analog filters, ideal Gaussian filters can
be realized. Much better selectivity can be achieved using digital filters
instead of analog filters at an acceptable circuit cost. Analog filters con-
sisting of five individual circuits, for instance, have a shape factor of
about 10, whereas a digitally implemented ideal Gaussian filter exhibits
a shape factor of 4.6. Moreover, digital filters feature temperature sta-
bility, are free of aging effects and do not require adjustment. Therefore
they feature a higher accuracy regarding bandwidth.
The transient response of digital filters is defined and known. Using
suitable correction factors, digital filters allow shorter sweep times
than analog filters of the same bandwidth (see chapter 4.6: Parameter
­dependencies).

52
 IF S i g n a l P r o c e s s i n g

In contrast to that shown in the block diagram, the IF signal after the
IF amplifier must first be sampled by an A/D converter. To comply with
the sampling theorem, the bandwidth of the IF signal must be limited
by analog prefilters prior to sampling. This band limiting takes place
before the IF amplifier so that intermodulation products can be avoided,
as was the case for analog filters. The bandwidth of the prefilter is vari-
able, so depending on the set digital resolution bandwidth, the smallest
possible bandwidth can be selected. The digital IF filter provides for lim-
iting the noise bandwidth prior to envelope detection.
The digital IF filter can be implemented by configurations as described
in [3-1] or [3-2]. In our example, the resolution bandwidths from 10 Hz to
30 kHz of the spectrum analyzer are realized by digital filters.

FFT
Very narrow IF bandwidths lead to long transient times which consid-
erably reduce the permissible sweep speed. With very high resolution it
is therefore advisable to calculate the spectrum from the time charac-
teristic – similar to the FFT analyzer described in chapter 3.1. Since very
high frequency signals (up to several GHz) cannot directly be sampled
by an A/D converter, the frequency range of interest is converted to the
IF as a block, using a fixed-tuned LO signal, and the bandpass signal is
sampled in the time domain (see Fig. 4-14). To ensure unambiguity, an
analog prefilter is required in this case.
For an IF signal with the center frequency fIF and a bandwidth B,
one would expect a minimum sampling rate of 2 · (fIF + 0.5 ·B) in accor-
dance with the sampling theorem (Equation 3-1). If the relative band-
width, however, is small (B/fIF« 1), then undersampling is permissible to
a certain extent. That is, the sampling frequency may be lower than that
resulting from the sampling theorem for baseband signals. To ensure
unambiguity, adherance to the sampling theorem for bandpass signals
must be maintained. The permissible sampling frequencies are deter-
mined by:

2 ⋅ fIF + B 2 ⋅ fIF - B
≤ fS ≤  (Equation 4-10)
k +1 k

where fS sampling frequency

fIF intermediate frequency
B bandwidth of IF signal
k 1, 2, …

53
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

The spectrum can be determined from the sampled values with the aid
of the Fourier transform.

Conversion
A

Analog
bandpass
filter

fIF fLO f

Span
A
RAM FFT
D

Display
Fig. 4-14  Spectrum analysis using FFT

The maximum span that can be analyzed at a specific resolution by

means of an FFT is limited by the sampling rate of the A/D converter
and by the memory available for saving the sampled values. Large spans
must therefore be subdivided into individual segments which are then
converted to the IF in blocks and sampled.

While analog or digital filter sweep times increase directly proportional

to the span, the observation time required for FFT depends on the
desired frequency resolution as described in chapter 3.1. To comply with
sampling principles, more samples have to be recorded for the FFT with
increasing span so that the computing time for the FFT also increases.
At sufficiently high computing speed of digital signal processing, dis-
tinctly shorter measurement times than that of conventional filters can
be attained with FFT, especially with high span/BN ratios (see chapter
4.6 Parameter dependencies).

The far-off selectivity of FFT filters is limited by the leakage effect,

depending on the windowing function used. The Hann window described
in chapter 3.1 is not suitable for spectrum analysis because of the ampli-

54
 Determination of Video Voltage and Video Filters

tude loss and the resulting level error. A flat-top window is therefore
often used to allow the leakage effect to be reduced so that a negligible
amplitude error may be maintained. This is at the expense of an observa-
tion time that is by a factor of 3.8 longer than that of a rectangular win-
dow. The flat-top window causes a wider representation of the window-
ing function in the frequency domain (corresponding to the convolution
with a Dirac function in the frequency domain). When the flat-top win-
dow is implemented, a shape factor of about 2.6 can be attained, which
means that selectivity is clearly better than when analog or digital IF fil-
ters are used.
FFT filters are unsuitable for the analysis of pulsed signals (see chap-
ter 3.1). Therefore it is important for spectrum analyzers to be provided
with both FFT and conventional filters.

4.3 Determination of video voltage and video filters

Information about the level of the input signal is contained in the level
of the IF signal, such as amplitude-modulated signals in the envelope
of the IF signal. With the use of analog and digital IF filters, the enve-
lope of the IF signal is detected after filtering the last intermediate fre-
quency (see Fig. 4-15).

Envelope
AIF AVideo

Envelope
detection

0 0
t t

1
fIF
Fig. 4-15  Detection of IF signal envelope

55
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

This functional configuration is similar to analog envelope detector cir-

cuitry used to demodulate AM signals (see Fig. 4-16). The IF signal is
detected and the high-frequency signal component eliminated by a low-
pass filter and the video voltage is available at the output of this cir-
cuit.

VVideo
VIF
0 0
t VIF VVideo t

BIF Video filter

VIF R C VVideo

V Video
^
VIF

^
Video filter
0 fc = R · C 0
fIF f fg fIF 2 fIF f

BVideo

Fig. 4-16  Detection of IF signal envelope by means of envelope detector

For digital bandwidths, the IF signal itself is sampled, i. e. the envelope is
determined from the samples after the digital IF filter. If one looks at the
IF signal represented by a complex rotating vector (cf. chapter 2.1), the
envelope corresponds to the length of the vector rotating at an angular
velocity of wIF (see Fig. 4-17). The envelope can be determined by forming
the magnitude using the Cordic algorithm [4-3].

56
 Determination of Video Voltage and Video Filters

j lm
Samples

ωIF

o e
Vid
A

Re

Fig. 4-17   
IF signal with sinusoidal input
signal, represented by complex
rotating vector

Due to envelope detection, the phase information of the input signal

gets lost, so that only the magnitude can be indicated in the display. This
is one of the primary differences between the envelope detector and the
FFT analyzer as described in chapter 3.1.

The dynamic range of the envelope detector determines the dynamic

range of a spectrum analyzer. Modern analyzers feature a dynamic range
of about 100 dB. It has no sense to simultaneously display so much dif-
ferent values in a linear scale. The level is usually displayed in a logarith-
mic scale on the spectrum analyzer. The IF signal can therefore be ampli-
fied with the aid of a log amplifier (32) ahead of the envelope detector
(33), thereby increasing the dynamic range of the display.

The resulting video voltage depends on the input signal and the selected
resolution bandwidth. Fig. 4-18 shows some examples. The spectrum
analyzer is tuned to a fixed frequency in these examples, so the dis-
played span is 0 Hz (zero span).

57
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

a)

Ain AIF

AVideo
0 t 0 t 0 t

1
–– 1
----
fe fIF

BIF
|Ain| |AIF|

| AVideo|
0 0 0
fin f fIF f f
b)

Ain AIF

AVideo
0 t 0 t 0 t

1
__
1
__ fIF
fe
BIF
|Ain| |AIF|
| AVideo|

0 0 0
fin f fIF f f
fm fm
c)

Ain AIF
AVideo

1
––
0 t 0 t 0 fm t

1
1
__ ––
fin fIF

BIF
|Ain| |AIF|
| AVideo|

0 0 0
fin f fIF f fm f
fm fm
Fig. 4-18  V
 ideo signal (orange traces) and IF signal after IF filter (blue traces)
for various input signals (green traces) and resolution bandwidths
a) sinusoidal signal b) AM signal, resolution bandwidth smaller than
twice the modulation bandwidth c) AM signal, resolution bandwidth
greater than twice the modulation bandwidth

58
 Determination of Video Voltage and Video Filters

d)

Ain AIF

AVideo
0 t 0 t 0 t

BIF
|Ain| |AIF|

| AVideo|
0 0 0
f fIF f f
Fig. 4-18 (continued)  Video signal (orange traces) and IF signal after IF filter
(blue traces) for various input signals (green traces) and resolution
bandwidths d) noise

The envelope detector is followed by the video filter (35) which defines
the video bandwidth (BV). The video filter is a first order lowpass config-
uration used to free the video signal from noise, and to smooth the trace
that is subsequently displayed so that the display is stabilized. In the
analyzer described, the video filter is implemented digitally. Therefore,
the video signal is sampled at the output of the envelope detector with
the aid of an A/D converter (34) and its amplitude is quantized.
Similar for the resolution bandwidth, the video bandwidth also lim-
its the maximum permissible sweep speed. The minimum sweep time
required increases with decreasing video bandwidth (chapter 4.6.1).
The examples in Fig. 4-18 show that the video bandwidth has to be
set as a function of the resolution bandwidth and the specific measure-
ment application. The detector used also has be taken into account in
the video bandwidth setting (chapter 4.5). The subsequent consider-
ations do not hold true for RMS detectors (chapter 4.4 D ­ etectors).
For measurements on sinusoidal signals with sufficiently high sig-
nal-to-noise ratio a video bandwidth that is equal to the resolution band-
width is usually selected. With a low S/N ratio the display can however
be stabilized by reducing the video bandwidth. Signals with weak level
are thus shown more distinctly in the spectrum (Fig. 4-19) and the mea-
sured level values are stabilized and reproducible. In the case of a sinu-
soidal signal the displayed level is not influenced by a reduction of the
video bandwidth. This becomes quite clear when looking at the video
voltage resulting from the sinusoidal input signal in Fig. 4-18a. The video

59
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

signal is a pure DC voltage, so the video filter has no effect on the over-
all level of the video signal.

RBW 300 kHz

VBW 1 MHz
Ref –40 dBm Att 10 dB SWT 2.5 ms
–40
–50
–60 A
–70
1AP
–80
CLRWR
–90
–100
–110
PRN
–120
–130 EXT
Center 100 MHz 1 MHz/ Span 10 MHz

RBW 300 kHz

*VBW 300 Hz
Ref –40 dBm Att 10 dB SWT 280 ms
–40
–50
–60 B
–70
1AP –80
CLRWR
–90
–100
–110
–120
–130
Center 100 MHz 1 MHz/ Span 10 MHz

Fig. 4-19  S
 inusoidal signal with low S/N ratio shown for large (top) and small
(bottom half of screen) video bandwidth

To obtain stable and reproducible results of noise measurements, a nar-

row video bandwidth should be selected. The noise bandwidth is thus
reduced and high noise peaks are averaged. As described in greater
detail in chapter 4.4, the displayed average noise level will be 2.5 dB
below the signal´s RMS value.

Averaging should be avoided when making measurements on pulsed

signals. Pulses have a high peak and a low average value (depending on
mark-to-space ratio). In order to avoid too low display levels, the video
bandwidth should be selected much greater than the resolution band-
width (Fig. 4-20). This is further discussed in chapter 6.2.

60
 D e t e ct o r s

*RBW 1 MHz Marker 1 [T1]

*VBW 10 MHz –38.30 dBm
Ref –20 dBm Att 10 dB SWT 2.5 ms 1.00000000 GHz
–20
–30 1
–40
1 AP A
–50
CLRWR
–60
–70
–80
–90
–100 PRN
–110
–120 EXT
Center 1 GHz 20 MHz/ Span 200 MHz

*RBW 1 MHz Marker 1 [T1]

*VBW 100 kHz –43.50 dBm
Ref –40 dBm Att 10 dB SWT 5 ms 1.00000000 GHz
–20
–30 1
–40 B
1 AP
–50
CLRWR
–60
–70
–80
–90
–100
–110
–120
Center 1 GHz 20 MHz/ Span 200 MHz

Fig. 4-20  P
 ulsed signal recorded with large and small video bandwidth (top and
bottom half of screen); note amplitude loss with small video bandwidth
(see marker)

4.4 Detectors

Modern spectrum analyzers use LC displays instead of cathode ray

tubes for the display of the recorded spectra. Accordingly, the resolution
of both the level and the frequency display is limited.
The limited resolution of the level display range can be remedied by
using marker functions (see chapter 4.5: Trace processing). Results can
then be determined with considerably high resolution.

Particularily when large spans are displayed, one pixel contains the
spectral information of a relatively large subrange. As explained in chap-
ter 4.1, the tuning steps of the 1st local oscillator depend on the resolu-
tion bandwidth so that several measured values, referred to as samples
or as bins, fall on one pixel. Which of the samples will be represented
by the pixel depends on the selected weighting which is ­determined by

61
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

the detector. Most of the spectrum analyzers feature min peak, max
peak, auto peak and sample detectors. The principles of the detectors is
shown in Fig. 4-21.

Samples N=5
Video voltage

Max Peak
A

Max Peak
RMS
AV
RMS
AV
Auto Peak Sample

Sample Auto Peak

Min Peak
Min Peak

Pixel n Pixel n + 1

Displayed sample f

Figs 4-21  Selection of sample to be displayed as a function of detector used

62
 D e t e ct o r s

Logarithmic Max Peak

amplifier
Envelope
lin detector A/D converter
log Sample A
D Display
IF signal
Video
filter

Min Peak
Fig 4-22  Analog realization of detectors

These detectors can be implemented by analog circuits as shown in

Fig. 4-22. In this figure, the weighted video signal is sampled at the out-
put of the detector. In the spectrum analyzer described, the detectors
(36 to 39) are implemented digitally, so that the video signal is sampled
ahead of the detectors (in this case even ahead of the video filter). In
addition to the above detectors, average and RMS detectors may also be
realized. Quasi-peak detectors for interference measurements are imple-
mented in this way.

Max peak detector

The max peak detector displays the maximum value. From the samples
allocated to a pixel the one with the highest level is selected and dis-
played. Even if wide spans are displayed with very narrow resolution
bandwidth (span/RBW >> number of pixels on frequency axis), no
input ­signals are lost. Therefore this type of detector is particularly use-
ful for EMC measurements.

Min peak detector

The min peak detector selects from the samples allocated to a pixel the
one with the minimum value for display.

Auto peak detector

The auto peak detector provides for simultaneous display of maximum
and minimum value. The two values are measured and their levels dis-
played, connected by a vertical line (see Fig. 4-21).

Sample detector
The sample detector samples the IF envelope for each pixel of the trace
to be displayed only once. That is, it selects only one value from the
samples allocated to a pixel as shown in Fig. 4-21 to be displayed. If the
span to be displayed is much greater than the resolution bandwidth

63
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

(span/RBW >> number of pixels on frequency axis), input signals are

no longer reliably detected. The same unreliability applies when too
large tuning steps of the local oscillator are chosen (see Fig. 4-5). In this
case, signals may not be displayed at the correct level or may be com-
pletely lost.

RMS detector
The RMS (root mean square) detector calculates the power for each pixel
of the displayed trace from the samples allocated to a pixel. The result
corresponds to the signal power within the span represented by the
pixel. For the RMS calculation, the samples of the envelope are required
on a linear level scale. The following applies:

1 N 2
VRMS = ⋅ ∑v  (Equation 4-11)
N i=1 i

where VRMS RMS value of voltage

N number of samples allocated to the pixel concerned
vi samples of envelope

The reference resistance R can be used to calculate the power:

VRMS
2
P=  (Equation 4-12)
R

AV detector
The AV (average) detector calculates the linear average for each pixel of
the displayed trace from the samples allocated to a pixel. For this calcu-
lation the samples of the envelope are required on a linear level scale.
The following applies:

1 N
VAV = ⋅ ∑v  (Equation 4-13)
N i=1 i

where VAV average voltage

N number of samples allocated to the pixel concerned
vi samples of envelope

Like with the RMS detector, the reference resistance R can be used to
calculate the power (Equation 4-12).

64
 D e t e ct o r s

Quasi peak detector

This is a peak detector for interference measurement applications with
defined charge and discharge times. These times are laid down by CISPR
16-1 [4-4] for instruments measuring spurious emission. A detailed
description of this type of detector can be found in chapter 6.2.5.1.

With a constant sampling rate of the A/D converter, the number of sam-
ples allocated to a certain pixel increases at longer sweep times. The
effect on the displayed trace depends on the type of the input signal and
the selected detector. They are described in the following section.

Effects of detectors on the display of different types of input signals

Depending on the type of input signal, the different detectors partly pro-
vide different measurement results. Assuming that the spectrum ana-
lyzer is tuned to the frequency of the input signal (span = 0 Hz), the
envelope of the IF signal and thus the video voltage of a sinusoidal input
signal with sufficiently high signal-to-noise ratio are constant. Therefore,
the level of the displayed signal is independent of the selected detector
since all samples exhibit the same level and since the derived average
value (AV detector) and RMS value (RMS detector) correspond to the
level of the individual samples.
This is different however with random signals such as noise or noise-
like signals in which the instantaneous power varies with time. Maxi-
mum and minimum instantaneous value as well as average and RMS
value of the IF signal envelope are different in this case.

The power of a random signal is calculated as follows:

 +T 
 2 
1 1
R T →∞ T T
()
P = ⋅ lim  ⋅∫ v 2 t dt  

(Equation 4-14)
 - 
2 

or for a certain limited observation time T

T
t+
2
1 1
P = ⋅ ⋅ ∫ v 2 t dt 
R T T
() (Equation 4-15)
t-
2

65
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

In the specified observation time T, a peak value can also be found for
the instantaneous power. The relationship between the peak value and
power can be expressed by the crest factor as follows:
P 
CF = 10 dB ⋅ lg  S   (Equation 4-16)
P 

where CF crest factor

PS peak value of instantaneous power
in observation time T
P power

With noise, any voltage values may occur theoretically, so the crest factor
would be arbitrarily high. Since the probability for very high or very low
voltage values is low, a crest factor of about 12 dB is usually obtained in
practice for Gaussian noise observed over a sufficiently long period.
Digitally modulated signals often exhibit a spectrum similar to noise.
However, the crest factor usually differs from that for Gaussian noise.
Fig. 4-23 shows the peak and RMS values of Gaussian noise and of a
IS-95 CDMA signal (forward channel).

66
 D e t e ct o r s

a) Crest factor 12 dB

RBW 3 MHz
VBW 10 MHz
Ref –50 dBm Att 10 dB SWT 100 s
–50

–55 A
1 RM* SGL
VIEW
–60
2 PK*
VIEW –65

PRN
–70
EXT
–75

–80

–85

–90

–95

–100
Center 2.2 GHz 10 s/

b) Crest factor 13.8 dB

RBW 3 MHz
VBW 10 MHz
Ref –10 dBm Att 20 dB SWT 100 s
–10
*
–15 A
1 RM* SGL
CLRWR
–20
2 PK*
CLRWR
–25
PRN
–30
EXT

–35

–40

–45

–50

–55

–60
Center 2.2 GHz 10 s/

Fig. 4-23  P
 eak (red traces) and RMS values (blue traces) of Gaussian noise (a)
and of a IS-95 CDMA signal (b), recorded with max peak and RMS
detectors

67
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

The effects of the selected detector and of the sweep time on the results
of measurements on stochastic signals are described in the following.

Max peak detector

When using the max peak detector, stochastic signals are overweighted
so that the maximum level is displayed. With increasing sweep time the
dwell time in a frequency range allocated to a pixel is also increased. In
the case of Gaussian noise the probability that higher instantaneous
values will occur also rises. This means that the levels of the displayed
­pixels also become higher (see Fig. 4-24a).
With a small ratio between span and resolution bandwidth, the noise
displayed for short sweep times is equal to that displayed with a sample
detector, since only one sample is recorded per pixel.

Min peak detector

When using the min peak detector, stochastic signals are underweighted
so that the minimum level is displayed. The noise displayed on the spec-
trum analyzer is strongly suppressed. In the case of Gaussian noise the
probability that lower instantaneous values will occur increases with
increasing sweep time. This means that the levels of the displayed ­pixels
also become lower (see Fig. 4-24a).
If measurements are carried out on sinusoidal signals with low sig-
nal-to-noise ratio, the minimum of the noise superimposed on the sig-
nal will also be displayed so that the level measurements yield too low
values.
With a small ratio between span and resolution bandwidth, the noise
displayed for short sweep times is equal to that displayed with a sample
detector, since only one sample is recorded per pixel.

Auto peak detector

When using the auto peak detector, the results of the max peak and min
peak detectors are displayed simultaneously, the two values being con-
nected by a line. With increasing sweep time the displayed noise band
becomes distinctly wider.
With a small ratio between span and resolution bandwidth, the noise
displayed for short sweep times is equal to that displayed with a sample
detector, since only one sample is recorded per pixel.

68
 D e t e ct o r s

a)

*RBW 1 MHz
VBW 3 MHz
Ref –50 dBm Att 10 dB *SWT 10 s
–50
*
–60 A
1 RM* SGL
CLRWR
–70
2 PK*
VIEW
–80
PRN
–90

–100

–110

–120

–130

–140

–150
Center 1.5 GHz 10 MHz/ Span 100 MHz

b)

*RBW 1 MHz
VBW 3 MHz
Ref –50 dBm Att 10 dB *SWT 10 s
–50
*
–60 A
1 MI* SGL
CLRWR
–70
2 MI*
VIEW
–80
PRN
–90

–100

–110

–120

–130

–140

–150 Center 1.5 GHz 10 MHz/ Span 100 MHz

Fig. 4-24  D
 isplayed noise varying as a function of sweep time, with max peak
detector (a) and min peak detector (b), sweep time 2.5 ms (blue trace)
and 10 s (red trace)

69
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

Sample detector
As shown in Fig. 4-21, the sample detector always displays a sample
recorded at a defined point in time. Due to the distribution of the instan-
taneous values, the trace displayed in the case of Gaussian noise there-
fore varies about the average value of the IF signal envelope resulting
from noise. This average value is 1.05 dB below the RMS value. If noise
is averaged over a narrow video bandwidth (VBW < RBW) using the log-
arithmic level scale, the displayed average value is an additional 1.45 dB
too low. The displayed noise is then 2.5 dB below the RMS value.
In contrast to the other detectors the sweep time has no effect on the
displayed trace since the number of the recorded samples is indepen-
dent of the sweep time.

RMS detector
The RMS detector allows measurement of the actual power of an input
signal irrespective of its temporal characteristic. When using a sample
or max peak detector, the relationship between RMS value and peak
value must be precisely known for determining the power of signals
with random instantaneous value. This knowledge is not required when
using an RMS detector.
The RMS value displayed by a specific pixel is calculated from all
­samples pertaining to this pixel. By increasing the sweep time, the num-
ber of samples available for the calculation is increased, thus allow-
ing smoothing of the displayed trace. Smoothing by reducing the video
bandwidth or by averaging over several traces (see chapter 4.5) is nei-
ther ­permissible nor necessary with the RMS detector. The measure-
ment results would be falsified, since the displayed values would be too
low (max. 2.51 dB). To avoid any falsification of results, the video band-
width should be at least three times the resolution bandwidth when
using the RMS detector.

AV detector
The AV detector determines the average value from the samples using
the linear level scale. The actual average value is thus obtained irrespec-
tive of the type of input signal. Averaging of logarithmic samples (log
average) would yield results that were too low since higher signal ­levels
are subject to greater compression by logarithmation. By increasing the
sweep time, several samples are available for calculating the average
value that is displayed by a specific pixel. The displayed trace can thus
be smoothed.

70
 D e t e ct o r s

A narrow video bandwidth causes averaging of the video signal. If sam-

ples of the linear level scale are applied to the input of the video filter, the
linear average of the samples is formed when reducing the video band-
width. This corresponds to the function of the AV detector so that smooth-
ing by means of narrow video bandwidths is permissible in this case.
The same holds true for the analyzer described here, since samples
with linear level scale are applied to the input of the video filter when
the AV detector is used (see block diagram).

If the video bandwidth is reduced, the displayed noise converges for max
peak, min peak, auto peak and sample detectors since the samples are
averaged by the video filter before they are weighted by the detector. If
a linear envelope detector is used to determine the IF signal envelope,
samples with linear scale are averaged by the video filter. The result-
ing display corresponds to the true average value and hence to the dis-
played noise when using an AV detector. If the IF signal is log-amplified
before the video voltage is formed, the resulting averaged samples are
lower than the true average value. In the case of Gaussian noise the dif-
ference is 1.45 dB (see Fig. 4-25a). Since the linear average of the video
voltage resulting from Gaussian noise is already 1.05 dB below the RMS
value, the samples obtained are all 2.5 dB lower than those obtained
with the RMS detector (see Fig. 4-25a). Due to this known relationship
an RMS detector is not required to determine the Gaussian noise power.
The power can be calculated from the samples collected by the sample
detector, taking into account a correction factor of 2.5 dB.
This relationship does not apply to other random signals whose
instantaneous values are not in line with the Gaussian distribution (for
example, digitally modulated signals, see Fig. 4-25b). If the crest factor
is unknown, the power of such signals can only be determined using an
RMS detector.

Averaging over several measurements

As described in the following chapter, modern analyzers feature the pos-
sibility of averaging traces over several measurements (trace average).
This method of averaging partly leads to results different from those
when using narrowband video filters.
Depending on whether the recorded trace is displayed on a linear or
logarithmic level scale, linear or logarithmic samples are used for aver-
aging. Whether the trace is falsified by averaging depends on the dis-
play mode.

71
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

a)

RBW 300 kHz

VBW 3 MHz
Ref –87 dBm Att 10 dB *SWT 5 s
–87
*
1 RM* A
–88
VIEW SGL
–89
2 AP*
VIEW
–90 RMS

3 AV* PRN
–91 1.05 dB Average (lin)
CLRWR
EXT
–92 2.50 dB
1.45 dB Auto Peak (log)
–93

–94

–95

–96

–97
Center 100 MHz 1 MHz/ Span 10 MHz

b)

RBW 3 MHz
VBW 10 MHz
Ref –24 dBm Att 10 dB SWT 1 s
–24
*
1 RM* A
VIEW –25
SGL
–26
2 AP*
VIEW
–27

3 AV*
RMS PRN
–28
CLRWR
EXT
–29 Average (lin)

–30 >2.50 dB

–31 Auto Peak (log)

–32

–33

–34
Center 2.2 GHz 100 ms/

Fig. 4-25   M
 easurement of Gaussian noise (a) and IS-95 CDMA signal (b) using
RMS and AV detectors (green and red traces) as well as auto peak
detector with averaging over narrow video bandwidth (blue trace)

72
 Trace Processing

In the case of averaging over several measurements, the displayed

noise levels do not converge for max peak, min peak and sample detec-
tors. The average is derived from the maximum and minimum values,
whereas with the use of the video filter, the samples are averaged prior
to weighting and therefore converge.
The sample detector yields the average noise level. With logarithmic
level display, the displayed average value is 1.45 dB too low, as already
explained above. With linear level display and large video bandwidth
(VBW ≥  10 · RBW) the true average is obtained, as with the AV detector.
When using the auto peak detector, averaging over several traces is
not recommended since the maximum and minimum value is displayed.
When the trace average function is activated, automatic switchover to
sample detector is often made.
For the RMS detector, trace averaging is permitted neither in the lin-
ear nor in the logarithmic level mode.

4.5 Trace processing

As was explained in chapter 4.4, linear samples are required for AV and
RMS detectors. For displaying the traces on a logarithmic level scale
when these detectors are used, the detectors are followed by a log ampli-
fier (40) which may be optionally activated.

In modern spectrum analyzers, the measurement results are digitized

before they are displayed. This allows many different methods of trace
evaluation (41).

Measured data memory

Several traces can be stored in modern analyzers and simultaneously
displayed. This function is particularly useful for comparative measure-
ments.

Trace average
With the aid of this function a displayed trace can be smoothed by aver-
aging over several measurements (sweeps). The user can enter the num-
ber of sweeps to be averaged.
Depending on the input signal and the detector used, this way of
averaging may lead to other results than averaging by reducing the
video bandwidth.

73
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

Marker functions
Marker functions are particularly useful for the evaluation of recorded
traces. They allow frequency and level to be displayed at any point of the
trace. The limited display accuracy due to the constrained screen resolu-
tion can thus be remedied. In addition to functions which set the marker
automatically to a signal with maximum level, level differences between
signals can also be directly displayed using the delta marker feature.
Modern spectrum analyzers feature enhanced marker functions
allowing, for instance, direct noise or phase noise measurements, with-
out manual setting of bandwidth or correction factors (see Fig. 4-26).

The precise frequency of a displayed signal can also be determined with

the aid of a marker and a count function (signal count). In many cases
the spectrum analyzer can thus replace a frequency counter.

Tolerance masks (limit lines)

Limit values to be adhered to by the device under test can easily be
checked with the aid of tolerance masks. To simplify use in production,
recorded traces are automatically checked for violation of the specified
limit values and the result is output in form of a “pass” or “fail” message
(see Fig. 4-27).

Channel power measurement

In the case of digitally modulated signals, power often has to be mea-
sured within one channel or within a specific frequency range. Chan-
nel power is calculated from the recorded trace, with special functions
being provided for this purpose by modern spectrum analyzers. Adja-
cent-channel power measurement with the aid of a spectrum analyzer is
described in detail in chapter 6.3.

74
 Trace Processing

Fig. 4-26  Marker functions for easy phase noise measurement of an input signal

Fig. 4-27  Evaluation of traces with the aid of limit lines

75
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

4.6 Parameter dependencies

Some of the analyzer settings are interdependent. To avoid measurement

errors, these parameters are coupled to one another in normal operat-
ing mode of modern spectrum analyzers. That is, upon varying one set-
ting all other dependent parameters will be adapted automatically. The
parameters can, however, also be set individually by the user. In such a
case it is especially important to know relationships and effects of var-
ious settings.

4.6.1 Sweep time, span, resolution and video bandwidths

Through the use of analog or digital IF filters, the maximum permissi-

ble sweep speed is limited by the transient time of the IF filter and video
filter. The transient time has no effect if the video bandwidth is larger
than the resolution bandwidth. In this case, the required transient time
increases inversely with the square of the resolution bandwidth, so with
a decrease of the resolution bandwidth by the factor n the required min-
imum sweep time becomes n2 longer. The following applies:

Df
TSweep = k ⋅  (Equation 4-17)
B IF
2

where TSweep minimum sweep time required (with specified span

and resolution bandwidth)
BIF resolution bandwidth
Df span
k proportionality factor

The proportionality factor k depends on the type of filter and the per-
missible transient response error. For analog filters made up of four
or five individual circuits, the proportionality factor k is 2.5 (maximum
transient response error approx. 0.15 dB). With digitally implemented
Gaussian filters, the transient response is known and exactly reproduc-
ible. Compared to analog filters, higher sweep speeds without amplitude
loss can be obtained through appropriate correction factors indepen-
dent of the type of input signal. A k factor of 1 can thus be attained. Fig.
4-28 shows the required sweep time for a span of 1 MHz as a function of
the resolution bandwidth.

76
 Parameter Dependencies

10 7
10 6
k=1
10 5 k = 2.5
Min. sweep time / s

10 4 FFT filter (real)

FFT filter (theoretical)
10 3
10 2
10 1
1
10 –1

10 – 2
10 – 3
10 – 4
10 – 5
10 – 6
1 Hz 10 Hz 100 Hz 1 kHz 10 kHz 100 kHz 1 MHz
Resolution bandwith

Fig. 4-28  T heoretically required sweep time as a function of resolution band-

width at a span of 1 MHz. Example of sweep times that can be attained
with FFT filters in a modern spectrum analyzer

If the video bandwidth is smaller than the resolution bandwidth, the

required minimum sweep time is influenced by the transient time of
the video filter. Similar to the IF filter, the transient time of the video fil-
ter increases with decreasing bandwidth. The video filter is usually a 1st
order lowpass, or a simple RC section if implemented in analog form.
Therefore there is a linear relationship between video bandwidth and
sweep time. Reducing the video bandwidth by a factor n results in an n
times longer sweep time.
Upon failure to attain the minimum sweep time, the IF filter or video
filter cannot reach steady state, causing an amplitude loss and distorted
signal display (frequency offset). A sinusoidal signal, for instance, would
be displayed neither at the correct level nor correct frequency (see
Fig. 4-29). Moreover, the effective resolution would be degraded due to
the widened signal display.

77
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

RBW 3 kHz RF Att 10 dB

UNCAL VBW 3 kHz
Ref Lvl –20 dBm SWT 5 ms Unit dBm
–20
*
–30 A

–40

–50
1AVG
2VIEW 1SA
–60
2SA
–70

–80

–90

–100

–110

–120
Center 1 GHz 20 kHz/ Span 200 kHz

Fig. 4-29   A
 mplitude loss if minimum sweep time required is not attained (blue
trace)

To avoid measurement errors due to short sweep times, resolution band-

width, video bandwidth, sweep time and span are coupled in normal
operating mode of modern spectrum analyzers.
Resolution bandwidth is automatically adapted to the selected span.
Long sweep times due to narrow resolution bandwidths at large spans
or poor resolution due to high resolution bandwidths at small spans are
thus avoided. Handling of a spectrum analyzer becomes much easier.
The coupling ratio between span and resolution bandwidth can often
be set by the user.
Partial coupling of the parameters is also possible. With manual set-
ting of the resolution and video bandwidths, the sweep time can, for
instance, be adapted automatically.

When using manual settings, if the minimum sweep time is not adhered
to, a warning is usually displayed (UNCAL in Fig. 4-29 upper left corner).
With FFT filters, the transient time is replaced by the observation
time required for a specific resolution (Equation 3-4). In contrast to the
sweep time with analog or digital filters, the observation time is inde-
pendent of the span, so even if the span were increased, the observation
time would not be increased for constant resolution. The observation

78
 Parameter Dependencies

time as a function of the resolution (yellow trace) shown in Fig. 4-28 is

therefore independent of the span.
In practice, larger spans are made up of several subranges. At a spe-
cific resolution, the resulting observation time is required for each sub-
range. The total observation time is directly proportional to the number
of subranges. The attainable measurement time therefore is distinctly
longer than the theoretically expected one. Fig. 4-28 shows sweep times
that can be attained with a modern spectrum analyzer using FFT filters.
It is clearly shown that high span-to-resolution bandwidth ratios allow
greatly reduced sweep times with FFT filters, especially when using very
narrow resolution bandwidths.

In modern spectrum analyzers, the video bandwidth can be coupled to

the resolution bandwidth. When varying the IF bandwidth, the video
bandwidth is automatically adapted. The coupling ratio (the ratio
between re­solution and video bandwidth) depends on the application
mode and therefore has to be set by the user (see chapter 4.3). In addi-
tion to the user-defined entry of a numeric value, the following options
are often available:
❙ Sine BN/BV = 0.3 to 1
❙ Pulse BN/BV = 0.1
❙ Noise BN/BV = 10

In the default setting, the video bandwidth is usually selected so that

maximum averaging is achieved without increasing the required sweep
time with the video filter. With a proportionality factor k = 2.5 (Equa-
tion 4-17), the video bandwidth must be at least equal to the resolution
bandwidth (BN/BV = 1). If the IF filter is implemented digitally, a pro-
portionality factor k = 1 can be attained through appropriate compen-
sation as described above, and the minimum sweep time required can
be reduced by a factor of 2.5. To ensure steady state of the video filter
despite the reduced sweep time, the video bandwidth selected should be
about three times greater than the resolution bandwidth (BN/BV = 0.3).

79
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

4.6.2 Reference level and RF attenuation

Spectrum analyzers allow measurements in a very wide level range that

is limited by the inherent noise and the maximum permissible input
level (see chapter 5.1 and chapter 5.4). With modern analyzers this level
range may extend from -147 dBm to +30 dBm (with a resolution band-
width of 10 Hz), thus covering almost 180 dB. It is not possible however
to reach the two range limits at a time since they require different set-
tings and the dynamic range of log amplifiers, envelope detectors and
A/D converters is much smaller anyway. Within the total level range only
a certain window can be used which must be adapted by the user to the
specific measurement application by selecting the reference level (max-
imum signal level to be displayed). The RF attenuation aRF and the IF
gain g IF are to be adjusted as a function of the ­reference level.
To avoid overdriving or even damaging of the first mixer and sub­
sequent processing stages, the high-level input signals must be attenu-
ated by the analyzer’s attenuator (see Fig. 4-30). The attenuation required
for a specific reference level depends on the dynamic range of the first
mixer and subsequent stages. The level at the input of the first mixer (i.  e.
the mixer level) should be distinctly below the 1 dB compression point.
Due to nonlinearities, products are generated in the spectrum analyzer
whose levels increase over-proportionally with increasing mixer level. If
the mixer level is too high, these products may cause interference in the
displayed spectrum so that the socalled intermodulation-free range will
be reduced.

80
 Parameter Dependencies

a)
L
Strong input signal
Max.
input level

Dynamic range Dynamic range limit

limit of logarithmic amplifier /
(reference level) A/D converter
RF attenuation

aRF

IF gain
gIF

Mixer level

Dynamic range

Input 1st mixer Logarithmic amplifier

Envelope detector
A/D converter

b)
L Weak input signal
Max.
input level

Dynamic range limit of logarithmic amplifier /

A/D converter

Mixer level
Dynamic

IF gain
range

gIF
RF attenuation

Dynamic range
limit
aRF
(reference level)

Logarithmic amplifier
Input 1st mixer Envelope detector
A/D converter

Fig. 4-30  A
 daptation of RF attenuation and IF gain to maximum signal level
level to be displayed (max. signal level = reference level)

81
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

a)

*RBW 300 kHz

*VBW 3 kHz
Ref –20 dBm *Att 0 dB SWT 140 ms
–20

–30 A

–40

1 AP –50
CLRWR PRN
–60

–70

–80

–90

–100

–110

–120
Start 10 MHz 5 MHz/ Stop 60 MHz

b)

*RBW 300 kHz

*VBW 3 kHz
Ref –20 dBm *Att 20 dB SWT 140 ms
–20
A
–30

–40

1 AP –50
CLRWR PRN
–60

–70

–80

–90

–100

–110

–120
Start 10 MHz 5 MHz/ Stop 60 MHz

82
 Parameter Dependencies

c)

*RBW 300 kHz

*VBW 3 kHz
Ref –20 dBm *Att 10 dB SWT 140 ms
–20
A
–30

–40

1 AP –50
CLRWR PRN
–60

–70

–80

–90

–100

–110

–120
Start 10 MHz 5 MHz/ Stop 60 MHz

Fig. 4-31  S
 ingle-tone input: dynamic range reduced by too high (a) or too low (b)
mixer level. Dynamic range attainable with optimum mixer level (c)
shown in comparison

If the RF attenuation is too high, causing the mixer level to be too low,
the signal-to-noise ratio of the input signal will be unnecessarily reduced.
As shown in Fig. 4-32, the attainable dynamic range is then reduced by
the higher noise floor. Fig. 4-31 shows the effects of the mixer level with
single-tone input (see chapter 5.2: Nonlinearities).

L
Max.
input level
Dynamic range
limit
(reference level) qRF, 1
qRF, 2 gIF,1

Lmix, 1 gIF, 2
Mixer level
Lmix, 2
Dynamic range

Input Noise floor Logarithmic amplifier

Envelope detector
1st mixer A/D converter

Fig. 4-32  Dynamic range limited by noise floor as a function of mixer level

83
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

To obtain the total dynamic range of the log amplifier and envelope
detector (with analog IF filters) or of the A/D converter (with digital IF
filters), the signal level is appropriately amplified with the aid of the IF
amplifier at the last IF. The gain is selected so that signals attaining the
reference level cause the full drive of the log amplifier, envelope detector
(with linear level display) or A/D converter (with digital IF filters). The IF
gain is therefore set indirectly via the reference level although it is also
dependent on the selected attenuator. At a constant reference level, the
IF gain has to be increased with increasing RF attenuation (see gIF, 1 and
gIF, 2 in Fig. 4-32).
If the level of the input signal to be displayed exceeds the reference
level, this may cause overdriving. The IF gain has then to be reduced by
increasing the reference level.

Coupling of reference level and RF attenuation

In modern spectrum analyzers, the RF attenuation can be coupled to
the reference level setting. The coupling criterion is the maximum mixer
level attained by an input signal whose level corresponds to the refer-
ence level. The mixer level attained with full drive therefore results from
the difference between reference level and RF attenuation. The follow-
ing applies:

Lmix (mW) = Lin, max (mW) - a RF = LRef (mW) - a RF  (Equation 4-18)

where Lmix level at input of first mixer with full drive, relative to
1 mW
Lin, max input level causing full drive, relative to 1 mW
LRef reference level, relative to 1 mW
aRF RF attenuation set via attenuator

A compromise between low signal-to-noise ratio and low distortion has

to be found in the selection of the mixer level. To optimize the mixer
level for specific applications, some analyzers allow the user to freely
select the mixer level for a specific reference level. Predefined coupling
degrees are often provided:

Low signal-to-noise ratio

The lower the RF attenuation, the lower the reduction of the signal-to-
noise ratio before the first mixer. For low displayed noise, a high mixer
level is required (see chapter 5.1: Inherent noise).

84
 Parameter Dependencies

Low distortion
The lower the mixer level, the lower the distortion produced in a spec-
trum analyzer due to nonlinearities. In this display mode, the RF attenu-
ation is higher (see chapter 5.2: Nonlinearities).

Table 4-2 shows some typical settings of RF attenuation and IF gain at

a specified reference level for the different display modes. The example
shows that even at very low reference levels, an RF attenuation of at
least 10 dB is always set. In this way the first mixer is protected and a
good input match is realized. Thus a higher level accuracy for absolute
level measurements is achieved (see chapter 5.10.1: Error components).
In this example, the RF attenuation can be set to a maximum of 70 dB,
and the IF gain to a maximum of 50 dB.

Mixer -40 dBm -30 dBm -20 dBm

level (low distortion) (normal) (low noise)
Reference RF atten- IF gain RF atten- IF gain RF atten- IF gain
level uation uation uation

+30 dBm 70 dB 30 dB 60 dB 20 dB 50 dB 10 dB

+20 dBm 60 dB 30 dB 50 dB 20 dB 40 dB 10 dB
+10 dBm 50 dB 30 dB 40 dB 20 dB 30 dB 10 dB
0 dBm 40 dB 30 dB 30 dB 20 dB 20 dB 10 dB
-10 dBm 30 dB 30 dB 20 dB 20 dB 10 dB 10 dB
-20 dBm 20 dB 30 dB 10 dB 20 dB 10 dB 20 dB
-30 dBm 10 dB 30 dB 10 dB 30 dB 10 dB 30 dB
-40 dBm 10 dB 40 dB 10 dB 40 dB 10 dB 40 dB
-50 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB
-60 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB
-70 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB
-80 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB
-90 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB
-100 dBm 10 dB 50 dB 10 dB 50 dB 10 dB 50 dB

Table 4-2  E
 xample of setting RF attenuation and IF gain at a specified reference
level (IF gain max. 50 dB, RF attenuation max. 70 dB)

85
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

4.6.3 Overdriving

When using a spectrum analyzer care should be taken to avoid overdriv-

ing by input signals with levels too high. Overdriving may occur at sev-
eral points in the signal path. To avoid this, both the RF attenuation and
the reference level (IF gain) have to be set correctly. In the following, the
critical components and the criteria to be observed are described.

First mixer
To cover the lower frequency range (up to 3 GHz in the case of the ana-
lyzer described here), the principle of a high first intermediate frequency
is usually employed in the RF input sections. If the spectrum analyzer
does not feature a narrowband preselector ahead of the first mixer, sig-
nals may be taken to the first mixer in the total input frequency range
(up to 3 GHz in our example) irrespective of the span to be displayed.
The mixer may thus also be overdriven by signals lying far outside the
displayed span. The distortion products produced in this way (harmon-
ics of higher order) may impair the displayed spectrum depending on
the span chosen for display (Fig. 4-33 and Fig. 4-34).

Harmonics of input signal,

A produced in 1st mixer

Input signal fIF = 3475.4 MHz

A A
fin 2fin 3 fin f
Sweep
RF ZF
fin = 1 GHz f fIF = fLO – fin f = fLO – 2 fin f
1st LO f = fLO – 3 fin f = fLO – fin
fLO = 3.4 to 6.6 GHz

Displayed spectrum
A

f
Fig. 4-33  H
 igher-order harmonics of input signals which are produced in first
mixer

86
 Parameter Dependencies

a)

*RBW 3 MHz Marker 1 [T1]

*VBW 10 MHz –53.77 dBm
Ref –10 dBm *Att 0 dB SWT 5 ms 1.04000000 GHz
–10
A
–20
1 AP
CLRWR
–30

–40
PRN
–50 1

–60

–70

–80

–90

–100

–110
Start 300 MHz 100 MHz/ Stop 1.3 GHz

b)

*RBW 3 MHz Marker 1 [T1]

*VBW 10 MHz –53.74 dBm
Ref –10 dBm *Att 0 dB SWT 5 ms 1.04000000 GHz
–10
A
–20
1 AP
CLRWR
–30

–40
PRN
–50 1

–60

–70

–80

–90

–100

–110
Start 800 MHz 50 MHz/ Stop 1.3 GHz

Fig. 4-34  S
 pectrum analyzer driven by a sinusoidal signal with f = 520 MHz (a).
The second harmonic with f = 1040 MHz that is produced in the first
mixer appears even if the fundamental of the signal is not contained in
the displayed spectrum (b)

87
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

To avoid overdriving, the mixer level, i. e. the total signal level at the input
of the first mixer, should be below the mixer’s 1 dB compression point.
The latter is specified in the data sheet of the respective spectrum ana-
lyzer (see chapter 5.4). As described in chapter 4.6.2, the mixer level is set
with the aid of the attenuator. Some contemporary spectrum analyzers
feature an overload detector ahead of the first mixer, so that in case of
overdriving a warning can be displayed.
If the input section of the spectrum analyzer features a narrowband
tracking preselector, the risk of the analyzer being overdriven by sig-
nals outside the displayed spectrum is considerably reduced. The ana-
lyzer described in this chapter contains a narrowband preselector in
form of a tracking YIG filter in the signal path for the frequency range
from 3 GHz to 7 GHz. If a small span of this frequency range is displayed,
the first mixer can only be overdriven by signals within or close to the
displayed spectrum. Due to the limited skirt selectivity of the YIG filter,
input signals outside the displayed spectrum must have a certain spac-
ing from the range of interest so that they will be sufficiently suppressed
by the filter and not overdrive the mixer (Fig. 4-35).
To allow EMC measurements, which often imply a very large number
of simultaneously occurring spectral components at a high level, with
the spectrum analyzer in line with relevant standards, analyzers can
usually also be enhanced with optional narrowband tracking preselec-
tors in the lower input frequency range.

IF signal processing through to resolution filter

The first mixer is followed by analog signal processing stages such as IF
amplifiers and conversion stages. These stages can only be overdriven
by strong signals within or in the vicinity of the displayed spectrum.
Signals outside the displayed spectrum are suppressed after the first
conversion by the subsequent IF filters, provided the frequency spacing
from the range of interest is sufficiently large (Fig. 4-36). The IF filter in
the 1st and 2nd IF stage is usually extremely wideband so the required
frequency spacing for adequate attenuation may be very large (often
some 100 MHz).

88
 Parameter Dependencies

Tracking bandwidth filter

A (preselector) Input signal

f
little
A suppression

Input signal
of 1st mixer

Fig. 4-35   
Suppression of input sufficient
suppression
signals outside the
displayed spectrum by
a tracking YIG filter
fStart fStop f

Displayed frequency range (span)

Unlike overdriving of the first mixer, distortion products caused by over-

driving of analog IF signal processing components do not appear in the
displayed spectrum. They are suppressed by the IF filter and the subse-
quent narrowband resolution filters (see Fig. 4-37).

89
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

1st IF amplifier 1st IF filter

IF signal
RF after 1st
IF filter
a b c
1st LO

Input signals
A Strong input signal outside
frequency range of interest

Signal
to be displayed

a
f
Frequency range
to be displayed

Input signals
converted to 1st IF
A
1st IF filter

Sweep Sweep

b
f1st IF f

IF signal after
A 1st IF filter

Sweep

c
f1st IF f

Fig. 4-36  Suppression of mixture products at the 1st IF by the first IF filter

The spectrum analyzer described here has overload detectors at the 2nd
and 3rd IF so that overdriving of the analog IF processing stages can be
indicated (44 and 45).

90
 Parameter Dependencies

1st IF amplifier 1st IF filter

2nd mixer IF signal
1st IF after 1st
f1st IF = 3476.4 MHz IF filter
a b c
fO = 3476.4 MHz 2nd LO
B = 200 MHz

Sinusoidal input signal

A converted to 1st IF

a
f1st IF f

Harmonics of IF signal due to

A nonlinearities in 1st IF filter

b
f1st IF 2f1st IF f

Harmonics suppressed
A 1st IF filter by 1st IF filter

c
f
f1st IF 2 f1st IF

Fig. 4-37  S
 uppression of distortion products produced in analog IF signal
processing components

91
 P r a ct i c a l R e a l i z a t i o n of an Analyzer Operating on the ­H e t e r o d y n e P r i n c i p l e

Settable IF amplifier and subsequent stages

As already mentioned, the IF gain depends from the setting of the ref-
erence level.
If a signal exceeds the reference level in the displayed spectrum, the
settable IF amplifier and subsequent signal processing stages will be
overdriven. Their response depends on the selected settings. Based on
the block diagram of the spectrum analyzer shown on the fold-out page,
the following cases are possible:

Use of analog IF filters

Exceeding the reference level causes overdriving of the log amplifier
(with logarithmic level display) or overdriving of the envelope detector
(with linear level display).
It is not possible to perform measurements on an input signal whose
level exceeds the reference level. Level measurements on weak sig-
nals in the immediate vicinity are however not affected by overdriving
(Fig. 4-38). As shown in the block diagram, the resolution filter is made
up of several individual circuits. The filter circuits ahead of the setta-
ble IF amplifier provide for suppression of strong input signals outside
the passband. Therefore there will be no distortion products that might
impair the displayed spectrum.

Use of digital IF filters or FFT filters

When using digital IF filters or FFT filters, the IF signal is sampled with
the aid of an A/D converter. If in the displayed spectrum a signal level
exceeds the reference level, the A/D converter may be overdriven. Unlike
analog filters, mixture products are produced which become visible in
the displayed spectrum (Fig. 4-39).

92
 Parameter Dependencies

a)

RBW 100 kHz Marker 1 [T1]

*VBW 300 Hz –56.07 dBm
Ref –30 dBm Att 10 dB SWT 420 ms 862.01000000 MHz
–30
A
–40
1 AP
CLRWR
–50 1

–60
PRN
–70

–80

–90

–100

–110

–120

–130
Start 860.5 MHz 500 kHz/ Span 5 MHz

b)

RBW 100 kHz Marker 1 [T1]

*VBW 300 Hz –56.06 dBm
Ref –30 dBm *Att 10 dB SWT 420 ms 862.01000000 MHz
–30
A
–40
1 AP
CLRWR
–50
1

–60
PRN
–70

–80

–90

–100

–110

–120

–130
Start 860.5 MHz 500 kHz/ Span 5 MHz

Fig. 4-38  L evel measurement on a weak input signal signal in the presence of
a very strong signal, with normal driving (a) and overdriving of the
settable IF amplifier (b). Overdriving has no effect on the measurement
result

93
 Performance Features of S p e ct r u m A n a l y z e r s

a)

*RBW 30 kHz Marker 1 [T1]

*VBW 300 Hz –89.18 dBm
Ref –40 dBm Att 10 dB SWT 1.15 s 862.00000000 MHz
–40
IFOVL A
1 AP –50
CLRWR
–60

–70
PRN
–80
1
–90

–100

–110

–120

–130

–140
Center 862 MHz 500 kHz/ Span 5 MHz

b)

*RBW 30 kHz Marker 1 [T1]

*VBW 300 Hz –69.93 dBm
Ref 0 dBm Att 30 dB SWT 1.15 s 862.00000000 MHz
0
A
1 AP –10
CLRWR
–20

–30
PRN
–40

–50

–60
1
–70

–80

–90

–100
Center 862 MHz 500 kHz/ Span 5 MHz

Fig. 4-39  M
 ixture products due to overdriving of A/D converter converter when
using digital IF filters or FFT filters (a); displayed spectrum in case of
correct driving (b)

94
 Inherent Noise

5 Performance Features of Spectrum Analyzers

5.1 Inherent noise

Inherent noise is understood as the thermal noise which characterizes

both receivers and spectrum analyzers. Due to inherent noise the signal-
to-noise ratio of an input signal is reduced. Therefore, inherent noise is
a measure of the sensitivity of the spectrum analyzer. It allows conclu-
sions to be drawn as to the minimum level required for the input signal
to be detectable.
The inherent noise of receivers can be specified in different ways,
usually it is expressed as the noise factor or noise figure.
The non-dimensional noise factor F of a two-port network is the ratio
between the signal-to-noise ratio at the input of the network and the sig-
nal-to-noise ratio at the output of the network. The following applies:
PS1 / PN1
F=  (Equation 5-1)
PS2 / PN2

where PS1/PN1 signal-to-noise ratio at the input of the network

PS2/PN2 signal-to-noise ratio at the output of the network

The noise figure NF is then obtained by

NF = 10 dB ⋅ lg F  (Equation 5-2)

F1 , G1 F2 , G2 Fn , Gn
Fig. 5-1   
Several cascaded
noisy networks Ftotal , Gtotal
 

95
 Performance Features of S p e ct r u m A n a l y z e r s

The total noise factor Ftotal of cascaded networks is determined by

F2- 1 F3- 1 F -1
Ftotal = F1 + + + … + n -n1  (Equation 5-3)
G1 G1⋅ G2
∏G i
i=1

where Fi noise factor of an individual stage

G i gain of an individual stage

The following holds true for passive lossy networks such as cables or
attenuator pads:
a
10 dB
F = 10 and NF = a (Equation 5-4)

where F and NF noise factor and noise figure of network

a attenuation of network

Equation 5-3 reveals that the noise factor of the first stage is fully taken
into account in the total noise factor of a cascaded circuit. The atten-
uator is located at the input of a spectrum analyzer – a passive stage
whose noise factor can be calculated by means of Equation 5-4.
The total noise factor of the analyzer therefore depends on the atten-
uator setting. Increasing the attenuation by 10 dB, for instance, results
in a 10 dB higher total noise figure. Maximum sensitivity is therefore
attained with an attenuator setting of 0 dB (see Fig. 5-2).

The sensitivity of spectrum analyzers is usually specified as the dis-

played average noise level (DANL), a parameter that can be directly read
from the display of the spectrum analyzer.
The noise produced in a receiver is thermal noise, which means that
it does not contain any discrete components. The probability of a noise
voltage occurring in a specific voltage range can be derived from the
Gaussian distribution so that the designation Gaussian noise is also
used.

96
 Inherent Noise

*RBW 300 kHz

*VBW 3 kHz
Ref –40 dBm *Att 20 dB SWT 30 ms
–40
*
A
1 AP –50
CLRWR
–60
2 AP
VIEW –70
PRN
–80
RF Att = 20 dB
3 AP
EXT
VIEW
–90
RF Att = 10 dB

–100
RF Att = 0 dB

–110

–120

–130

–140
Center 890 MHz 1 MHz/ Span 10 kHz

Fig. 5-2  D
 isplayed average noise level of spectrum analyzer as a function of RF
attenuation

The displayed noise corresponds to the noise voltage picked up at the

envelope detector. The corresponding noise power can be calculated by
integration of the noise density over the noise bandwidth of the receiver
which would be the noise bandwidth of all stages ahead of the detec-
tor. In the case of spectrum analyzers, this bandwidth is determined by
the noise bandwidth of the IF filter. Accordingly, the displayed noise
depends on the resolution bandwidth setting.

Since the spectral power density of thermal noise is constant within the
noise bandwidth, the displayed average noise level can be calculated as
follows provided the noise figure of the analyzer and the noise band-
width of the selected IF filter are known:

97
 Performance Features of S p e ct r u m A n a l y z e r s

 k ⋅ T ⋅ BN, IF  
LDAN(mW) = 10 dB ⋅ lg   + NFSA- 2.5 dB (Equation 5-5)
 1 ⋅ mW 

where LDAN displayed average noise level, relative to 1 mW and

1 Hz bandwidth
k Boltzmann’s constant, k = 1.38 · 10 –23 WS/K
T ambient temperature
BN, IF noise bandwidth of IF filter
NFSA noise figure of spectrum analyzer
-2.5 dB underweighting of noise by sample detector and
­averaging of logarithmic level values

For an ambient temperature of 290 K, the displayed average noise level

is determined by:
BN, IF
LDAN (mW) = - 174 dB + 10 dB ⋅ lg + NFSA- 2.5 dB  
Hz
 (Equation 5-6)

The value of -174 dBm (1 Hz) corresponds to the available thermal

noise power across an ohmic resistance in 1 Hz bandwidth at an ambi-
ent temperature of 290 K. This is the noise floor, or the absolute mini-
mum noise level at a specified temperature.
The sample detector usually used for noise measurements with spec-
trum analyzers (see chapter 4.4 Detectors) determines the arithmetic
mean of the noise. In the case of Gaussian noise this is 1.05 dB below
the RMS value (i. e. the effective noise power). Due to averaging of the
results on a logarithmic scale (e. g. by averaging over several traces) the
displayed noise is reduced by a further 1.45 dB. In calculating the dis-
played average noise level in accordance with Equation 5-6, this is taken
account of by the subtraction of 2.5 dB. This correction is only permissi-
ble for Gaussian noise, which can be assumed for thermal noise.

The following relationship can be derived from Equation 5-5 for the vari-
ation of the displayed noise as a function of varying the IF bandwidth
setting of BIF, 1 to BIF, 2:
BN, IF, 2
DLDAN = 10 dB ⋅ lg  (Equation 5-7)
BN, IF, 1
where BN, IF, 1 , BN, IF, 2 noise bandwidth of IF filter before and after
variation of setting

98
 Inherent Noise

DLDAN variation of displayed noise as a function of

varying the IF bandwidth

If both IF filters have the same relationship between 3 dB bandwidth

and noise bandwidth, the difference in the displayed noise can also be
calculated from the 3 dB bandwidths. The following then applies:
BIF, 2
DLDAN = 10 dB ⋅ lg  (Equation 5-8)
BIF, 1

where BIF, 1, BIF, 2 3 dB bandwidth of IF filter before and after

variation of setting

Fig. 5-3 shows the effects of different IF bandwidths on the displayed

noise. Due to the different practical realization of the IF filters of a spec-
trum analyzer, the noise figure of the analyzer may also depend on the
selected resolution bandwidth. The actual variation of the displayed
average noise level may therefore differ from the value worked out with
Equation 4-8.

*RBW 300 kHz

*VBW 300 Hz
Ref –60 dBm Att 10 dB SWT 840 ms
–60
*
A
1 AP –65
CLRWR
–70
2 AP
VIEW –75
PRN
–80
3 AP 3 MHz EXT
VIEW
–85
1 MHz
–90 300 kHz

–95

–100

–105

–110
Center 891.5 MHz 3 MHz/ Span 30 MHz

Fig. 5-3  Displayed average noise level at various resolution bandwidths

The sensitivity limit of the analyzer can also be determined from the dis-
played average noise level. This is understood as the minimum level of

99
 Performance Features of S p e ct r u m A n a l y z e r s

an input signal required to yield a noise increase of 3 dB in the display

of the analyzer, and is called the minimum detectable signal. Since on
the spectrum analyzer the sum of input signal and noise (PS + PN) is dis-
played, this condition is fulfilled with an input level that corresponds to
the effective noise level of the inherent thermal noise (PS = PN). In this
case the signal-to-noise ratio is determined by

PS + PN  P + PN 
= 2 and 10 dB ⋅ lg  S  = 3 dB  (Equation 5-9)
PN  PN 

PN corresponds to the displayed noise when using an RMS detector.

Displayed average noise level

(0 dB RF attenuation, RBW 10 Hz, VBW 1 Hz, 20 averages,
trace average, span 0 Hz, termination 50 W)
Frequency
9 kHz <–95 dBm
100 kHz <–100 dBm Fig. 5-4   
Typical specifications
1 MHz <–120 dBm, typ. –125 dBm for displayed noise of
a spectrum analyzer
10 MHz to 7 GHz <–138 dBm, typ. –143 dBm
(extract from data sheet)

Data sheet specifications for the displayed average noise level must
always include the resolution bandwidth and attenuator setting. Typi-
cal settings are 0 dB RF attenuation and the smallest resolution band-
width available.
For a stabilized noise display, appropriate averaging is required
which can be achieved with a narrow video bandwidth (1 Hz in the
above data sheet extract) and averaging over several traces (trace aver-
age). In our example 20 traces are averaged.

100
 Inherent Noise

  Maximum sensitivity
The maximum sensitivity of spectrum analyzers is obtained with
an attenuator setting of 0 dB. It is important that the first mixer
of the analyzer is not overdriven by a high-level signal – even out-
side the displayed frequency range.
To further reduce the displayed noise, the resolution band-
width is reduced. A trade-off has to be found between low dis-
played noise and high measurement speed. For the display of
input signals with a very low signal-to-noise ratio, it is useful
to reduce the video bandwidth as well as the resolution band-
width or to increase the sweep time when using the RMS detec-
tor. The trace is thus smoothed so that the input signal is clearly
displayed. The measured levels are thus stabilized – a prerequi-
site for accurate, reproducible result.
If the sensitivity of the spectrum analyzer is unsatisfactory, it
can be enhanced with aid of an external preamplifier. The total
noise factor of the preamplifier and spectrum analyzer arrange-
ment can be calculated from Equation 5-3. F1 and G1 correspond
to the noise factor and gain of the preamplifier, F2 to the noise
factor of the spectrum analyzer.
For level measurements it is essential to know the frequency-
dependent gain of the preamplifier. This gain must be subtracted
from the measured levels. Many spectrum analyzers offer the
possibility of taking into account the frequency-dependent gain
with the aid of transducer tables. Recorded spectra can thus auto-
matically be displayed with the correct levels.
High sensitivity of the spectrum analyzer is especially impor-
tant for applications in which the resolution bandwidth is pre-
scribed by standards. In these cases a reduction of the displayed
noise by decreasing the resolution bandwidth is not permitted.
The sensitivity is also important for fast measurement speeds.
To attain sufficiently low displayed noise, narrowband IF filters
are required with lower sensitivities, which in turn increases the
sweep time. Spectrum analyzers featuring a low noise figure
allow the use of greater resolution bandwidths and hence shorter
sweep times (chapter 4.6: Parameter dependencies).

101
 Performance Features of S p e ct r u m A n a l y z e r s

5.2 Nonlinearities

An ideal linear two-port network provides for distortion-free transfer of

signals from its input to its output. The voltage transfer function of such
a network is

() ()
v out t = Gv⋅ v in t  (Equation 5-10)

where vout(t ) voltage at output of network

vin(t ) voltage at input of network
G v voltage gain of network

Such ideal networks can only be realized with the aid of passive com-
ponents. Resistive attenuator pads, for instance, can be assumed to be
ideal.
Networks containing semiconductor components, for instance ampli-
fiers or mixers, exhibit nonlinearities. In this case the transfer function
can be approximated by a power series as follows:

∞
 
() () () ()
 v out t = ∑ a n⋅ v inn t = a1⋅ v in t + a 2⋅ v in2 t + a 3⋅ v in3 t + …()
n=1
 (Equation 5-11)

where vout(t ) voltage at output of network

vin(t ) voltage at input of network
a n coefficient of nonlinear element of voltage gain

In most cases it is sufficient to consider the square and cubic terms so

that the power series defined by Equation 5-11 is taken up to n = 3.
For many components, such as mixers or level detectors, the nonlin-
ear response is desirable. Spectrum analyzers however should feature
distortion-free display of the input signals. Accordingly, linearity is an
essential criterion for the assessment of a spectrum analyzer.
The effects of a network’s nonlinearities on its output spectrum
depend on its input signal.

Single-tone input
If an input to a network is a sinusoidal signal vin(t) of

102
 Nonlinearities

() (
v in t = Û in⋅ sin 2π fin,1⋅ t ) (Equation 5-12)

where Ûin peak value of vin(t)

fin,1 frequency of vin(t)

is present, this is referred to as a single-tone input. By substituting

Equation 5-12 in Equation 5-11 it can be seen that due to the nonlinear-
ities, harmonics of the input signal are produced with the frequencies
fn,H = n · f1 (see Fig. 5-5).

Nonlinear
Ain amplifier Aout

fin,1 f fin,1 2 fin,1 3 fin,1 f

Input signal Output signal

Fig. 5-5  Spectrum before and after nonlinear network

The levels of these harmonics depend on the coefficients an in Equation

5-11. They are also dependent on the order n of the respective harmonic
as well as on the input level. When the input level is increased, the levels
of harmonics increase with their order. That is, a variation of the input
level by DdB causes a variation of the harmonic level by n · DdB.
Data sheet specifications on this type of signal distortion usually
refer to the second harmonic only for which the level difference ak2 from
the fundamental at the output of the network is specified. The speci-
fications are valid for a certain input level Lin, which has always to be
stated. When comparing spectrum analyzers, it should therefore always
be checked whether the specifications of the various models refer to the
same mixer level.

Level-independent specifications that can be made with the aid of the

intercept point (known as the second harmonic intercept SHI) are much
more convenient for comparisons. The second harmonic intercept cor-
responds to the assumed input or output level at which the second har-
monic of the input signal at the output of the network attains the same
level as the fundamental (Fig. 5-6).

103
 Performance Features of S p e ct r u m A n a l y z e r s

L
SHIout SHI

Lout L2nd H

1 dB/dB 2 dB/dB Fig. 5-6   

Second harmonic
intercept
SHIin Lin

In practice, this point can never be attained since the network, as shown
in Fig. 5-6, already provides compression at lower input levels. The inter-
cept point can be referred both to the network’s input or output level
and is therefore designated as input or output intercept point (SHIin and
SHIout in our example).

Since the output intercept point depends on the gain of the network,
the input intercept point (with selected RF attenuation, usually 0 dB) is
always stated in the spectrum analyzer specifications.
With a specific input level Lin and harmonic level difference ak2 of the
second harmonic, input intercept point can be calculated as follows:

SHI in(mW) = a k2 + Lin(mW)  (Equation 5-13)

SHIout relative to the output is calculated as

SHI out(mW) = SHI in(mW) + g  (Equation 5-14)

where g power gain of network

Two-tone input
With a two-tone input, signal vin(t) consisting of two sinusoidal signals
of equal amplitude is applied to the input of the network. The input sig-
nal is determined by:

104
 Nonlinearities

( )
v in(t ) = Û in⋅ sin 2pfin,1⋅ t +Û in⋅ sin 2pfin, 2⋅ t  ( ) (Equation 5-15)

where Ûin peak value of the two sinusoidal signals

fin,1, fin,2 signal frequencies

By substituting Equation 5-15 in the nonlinear transfer function in

accordance with Equation 5-11, the mixture products listed in Table 5-1
are obtained among others at the output of the twoport. The angular fre-
quency w is always stated as w1= 2 · p · fin,1 and w2= 2 · p · fin, 2 .

DC component a2 · 0.5(Û in,1

2
+ Û in,2
2
)
Fundamentals a1  · Ûin,1 · sin(w1t)
a1  · Ûin,2 · sin(w2t)
2nd harmonics a2  · 0.5 · Û in,1
2
 · cos(2 · w1t)
a2 · 0.5 · Û in,2
2
 · cos(2 · w2t)
Intermodulation products a2 · Ûin,1 · Ûin,2 · cos(w1- w2)t
of 2nd order
a2 · Ûin,1 · Ûin,2 · cos(w1+ w2)t
3rd harmonics a3 · 0.25 · Û in,1
3
 · sin(3 · w1t)
a3 · 0.25 · Û in,2
3
 · cos(3 · w2t)
Intermodulation products a3 · Û in,1
2
 · Ûin,2 · 0.75 · sin(2w1+ w2)t
of 3rd order
a3 · Û in,1 · Û in,2
2
 · 0.75 · sin(2w2+ w1)t
a3 · Û in,1
2
 · Ûin,2 · 0.75 · sin(2w1- w2)t
a3 · Û in,1 · Û in,2
2
 · 0.75 · sin(2w2- w1)t

Table 5-1  Mixture products with two-tone input

Input signals
Products of 2nd order
A aIM3 ak2 aIM2
Products of 3rd order

9.54 dB

6 dB

0 fin,1 fin,2 2 fin,1 2 fin, 2 3 fin,1 3 fin, 2 f

fin, 2 – fin,1 2 fin,1 – fin,2 2 fin, 2 – fin,1 fin,1 + fin, 2 2fin,1 + fin,2 2 fin, 2 + fin,1

Fig. 5-7  O
 utput spectrum of nonlinear network with two-tone input (magnitude
spectrum)

105
 Performance Features of S p e ct r u m A n a l y z e r s

In addition to the harmonics, intermodulation products are produced,

also referred to as difference-frequency distortion. The order of inter-
modulation products corresponds to the sum of the ordinal numbers of
the components involved. For the product at 2 · fin, 1+ 1 ·  fin, 2 for instance
the order is 2  + 1 = 3. Table 5-1 provides products up to the 3rd order.
While even-numbered intermodulation products always occur far
from the two input signals in frequency, odd-numbered intermodulation
products of low order are always found in the close vicinity of the input
signal.

Depending on the application, products of both even-numbered and odd-

numbered order may cause interference. For measurements on CATV
systems, where a frequency range covering more than one octave is to
be examined, harmonics as well as intermodulation products of even-
numbered order fall within the range of interest. For this application the
requirements placed on the SHI of the spectrum analyzer are therefore
very stringent, particularly as in such systems usually a large number of
signals with very high level occurs.

106
 Nonlinearities

Lout / dBm
IP 2out IP 2
50

IP 3out 30 IP3

Lout LIM3 LIM2

10
g

–10

2 dB/dB
1 dB/dB

–30

–50
–50 –30 –10 0 10 IP3in 30 IP2in 50
Lin / dBm

3 dB/dB

Fig. 5-8  Intercept point of 2nd and 3rd order

Like for harmonics of higher order, a level variation of the two sinusoi-
dal carriers at the input by D dB causes a level variation of the respec-
tive intermodulation product by n  · DdB. The level differences between
intermodulation products and the fundamentals of the sinusoidal car-
riers must therefore always be specified together with the input level
since otherwise no conclusion can be drawn as to the linearity. It is
therefore of advantage to calculate the intercept point of nth order too.
The intercept point of nth order relative to the input is determined by:

107
 Performance Features of S p e ct r u m A n a l y z e r s

a IMn
IPn in(mW) = + Lin(mW)  (Equation 5-16)
n-1

where IPnin input intercept point of nth order, relative to 1 mW

aIMn level difference between intermodulation ­product of
nth order and fundamental of the input signal
Lin level of one of the two input signals, relative to 1 mW

In most cases the intercept points of 2nd and 3rd order are specified
(see also Fig. 5-8). They are designated as IP2 or SOI (second order inter-
cept) and IP3 or TOI (third order intercept) The input intercept points of
2nd and 3rd order are determined by

IP 2in(mW) = a IM2 + Lin(mW)  (Equation 5-17)

and
a IM3
IP 3in(mW) = + Lin(mW)  (Equation 5-18)
2

The output intercept points can be calculated from the input intercept
points by adding the gain g of the network (in dB). In spectrum analyzer
specifications the intercept points are referenced to the input.

The 2nd order intermodulation products with two-tone input as well

as the 2nd harmonics with single-tone input are produced due to the
square term of the nonlinear transfer function. There is a fixed relation-
ship between IP2 and SHI (see [5-1]):

SHI (mW) = IP 2 (mW) + 6 dB  (Equation 5-19)

Therefore only IP 2 or SHI is usually specified in data sheets, but rarely

both values are specified. Intercept points are usually stated in dBm.
The higher the specified intercept point, the more linear is the spectrum
analyzer, which is an essential prerequisite for a large dynamic range
(chapter 5.5: Dynamic range).

108
 Nonlinearities

Intermodulation
3rd order intermodulation
Intermodulation-free dynamic range
level 2 · –30 dBm, Df >5 · RBW or 10 kHz, wichever is
the greater value
Frequency
20 MHz to 200 MHz >70 dBc, TOI >5 dBm
200 MHz to 3 GHz >74 dBc, TOI >7 dBm
3 GHz to 7 GHz >80 dBc, TOI >10 dBm
Fig. 5-9    2nd harmonic intercept point (SHI)
Typical specifications
<50 MHz 25 dBm
for intermodulation
characteristics of a 50 MHz to 3 GHz 35 dBm
spectrum analyzer
3 GHz to 7 GHz 45 dBm
(extract from data sheet)

Intercept points specified in data sheets are only valid for a specific
attenuator setting, usually 0 dB. As will be explained later, the intercept
point increases with increasing RF attenuation.

Examples: Comparison of data sheet specifications of two spectrum

analyzers

1. Analyzer 1:
With two-tone input of -30 dBm each, the 3rd order intermodula-
tion products are at least 70 dB below the input signal.
Analyzer 2:
With two-tone input of -40 dBm each, the 3rd order inter-
modulation products are at least 100 dB below the input signal.

Solution

70 dB
Analyzer 1: IP3in(mW) =
2
( )
+ -30 dB = + 5 dB

100 dB
Analyzer 2: IP3in(mW) =
2
( )
+ -40 dB = + 10 dB

109
 Performance Features of S p e ct r u m A n a l y z e r s

2. Analyzer 1
With two-tone input of -30 dBm each the 2nd order inter-
modulation products are at least 65 dB below the input signal.

Analyzer 2
An SHI of +35 dBm is specified in the data sheet.

Solution

Analyzer 1: (
IP2in(mW) = 65 dB + -30 dB = + 35 dB )
Analyzer 2:
IP 2in(mW) = IPk 2 (mW) - 6 dB = 35 dB - 6 dB = + 29 dB

Often the intermodulation-free dynamic range is specified in the data

sheets. This is understood as the level difference between IM products
and input signals. Unless otherwise stated, these data solely refer to
3rd order intermodulation products (products occurring in the vicinity
of the input signals). An essential parameter is the signal level at the
input of the first mixer which always has to be specified, too.
For the 3rd order intercept point specified in Fig. 5-9 (for the input
frequency range 200 MHz to 3 GHz in our example), the intermodulation-
free range can be calculated from IP3in by using Equation 5-18:

( ) (
 a IM3 = 2 ⋅ IP 3in(mW) - Lin(mW) = 2 ⋅ 7 dB - -30 dB = 74 dB ( ))
 
 (Equation 5-20)

Attenuator pad or amplifiers ahead of first mixer

If a preamplifier or attenuator pad is connected ahead of the first mixer
of the spectrum analyzer, the total input intercept point of the arrange-
ment is affected. The following applies to two cascaded stages [5-2]:

110
 Nonlinearities

  g1 + IP 3in, 1 (mW )  IP 3in, 2 (mW )  

    

 10 dB 

+ 10 
10 dB
 IP 3in, total(mW) = IP 3in, 1(mW) + IP 3in, 2(mW) - 10 dB ⋅ lg 10  


 
 (Equation 5-21)

where IP3in,total 3rd order input intercept point of cascaded

stages, relative to 1 mW
IP3in,1, IP3in,2 3rd order input intercept points of individual
stages, relative to 1 mW
g1 gain factor of first stage

Assuming an ideal linear attenuator pad – a prerequisite that is realistically

implemented using resistive, mechanically switched attenuators – almost
any value can be inserted into Equation 5-21 for IP3in, 1. An increase of the
RF attenuation, for instance, from 0 dB to 10 dB (g  = -10 dB in Equa-
tion 5-21) causes an increase of the intercept point by the same amount
(10 dB in our case). At the same time, the noise figure of the analyzer is
degraded to the same degree. Therefore, an increase of the RF attenua-
tion does not increase the dynamic range (chapter 5.5: Dynamic range).

If a preamplifier is connected ahead of the analyzer, this will cause a

degrading of the total intercept point.

Example:
An input intercept point of +7 dBm is specified for the spectrum ana-
lyzer. To increase the sensitivity, a preamplifier with a gain factor of
20 dB and an input intercept point of -10 dBm is to be connected. The
total 3rd order input intercept point is then determined by

  20 dB - 10 dB   7 dB  
 
   
IP 3in, total(mW) = -10 dB + 7 dB - 10 dB ⋅ lg 10  + 10 10 dB   = - 14.8 dB
 10 dB   

 
 

111
 Performance Features of S p e ct r u m A n a l y z e r s

  Identification of intermodulation products

A typical application of the spectrum analyzer is distortion mea-
surement on devices under test such as amplifiers or mixers. Dis-
tortion in the form of higher-order harmonics or intermodulation
products is not only produced in the DUT but also in the spec-
trum analyzer. Especially with high signal levels at the input of
the first mixer this may lead to incorrect measurements since the
harmonics or intermodulation products generated in the spec-
trum analyzer are added to those produced in the DUT. Linear-
ity measurement would indicate poorer characteristics than it is
actually the case.
The spectrum analyzer's mixer and IF amplifier primar-
ily determine the linearity of a spectrum analyzer, whereas the
input RF attenuator has practically no effect. If the RF attenu-
ator is used to vary the mixer level, the levels of intermodula-
tion products generated in the spectrum analyzer are determined
by order. The level of intermodulation products generated in the
DUT remains constant.
With the aid of the RF attenuator, it can thus be determined
where the intermodulation products displayed on the spectrum
analyzer are generated. The measurement yields correct results
if the relative levels of harmonics or intermodulation prod-
ucts remain constant despite an increase of the RF attenuation
(Fig. 5-10a). If the relative level of the harmonics displayed on
the spectrum analyzer varies however, the measurement result
is incorrect.

112
 Nonlinearities

a)
*RBW 300 kHz
*VBW 3 kHz
Ref –10 dBm *Att 20 dB SWT 15 ms
–10
*
A
1 AP –20
CLRWR
–30
2 AP
VIEW –40
PRN
–50
EXT
–60
RF Att = 30 dB
–70

–80
RF Att = 20 dB
–90

–100

–110
Center 2.02004 GHz 500 kHz/ Span 5 MHz

b)
*RBW 100 kHz
*VBW 3 kHz
Ref –30 dBm *Att 10 dB SWT 45 ms
–30
*
A
1 AP –40
CLRWR
–50
2 AP
VIEW –60
PRN
–70
EXT
–80
RF Att = 10 dB
–90

–100
RF Att = 0 dB
–110

–120

–130
Center 2.02004 GHz 500 kHz/ Span 5 MHz

Fig. 5-10  I dentification of intermodulation products  

a) intermodulation products of DUT (measurement is correct)  
b) intermodulation products of analyzer (incorrect measurement)

113
 Performance Features of S p e ct r u m A n a l y z e r s

5.3 Phase noise (spectral purity)

Phase noise is a measure for the short-time stability of oscillators, like

the ones used also in a spectrum analyzer for the conversion of input
signals into IF. Phase noise is caused by variations of phase or frequency
and amplitude of an oscillator output signal, although the amplitude
effect is negligible in most cases. These variations have a modulating
effect on the oscillator signal.
The phase noise is usually specified as single-sideband phase noise
referenced to the carrier power and as a function of the carrier offset.
The values are stated as relative noise levels at a noise bandwidth of
1 Hz, where the suffix c designates the reference to the carrier. The mag-
nitude symbol is L(Pc), and the unit is dB. The unit dBc (Hz) is used as a
short form. Since the phase noise level is lower than the carrier power,
negative numeric values will be specified.
The effects of phase noise are shown in Fig. 5-11. Assuming a suf-
ficiently high resolution, one would expect a single spectral line for a
purely sinusoidal signal in the frequency domain. In fact, the spectrum
of a signal generated by a real oscillator is wider than a single line.

An oscillator signal exhibits phase noise that may be more or less dis-
tinct. By choosing appropriate circuit configurations, phase noise can be
minimized to a certain degree but never be fully eliminated. In modern
spectrum analyzers the local oscillators are implemented as synthesizers
locked to a high-precision reference (such as 10 MHz) via phase-locked
loops (PLL) as described in [5-3]. The phase noise characteristic will be
influenced by the PLL bandwidth of the frequency locking circuitry. The
spectrum is divided into the following subranges (see Fig. 5-11):

Close to carrier (offset approximately up to 1 kHz):

In this range the phase noise corresponds to the phase noise of the ref-
erence signal relative to the output signal of the local oscillator. Due to
the multiplying effect in the PLL this phase noise is higher than that of
the reference oscillator.

Range extending to the upper limit of the PLL bandwidth

(offset greater than 1 kHz)
Within the PLL bandwidth, the phase noise corresponds to the additive
noise of several PLL components, such as divider, phase detector and of
the multiplied reference signal. The upper limit of this range depends on

114
 P h a s e N o i s e (S p e ct r a l P u r i t y )

the spectrum analyzer, or more precisely on the type of oscillator used. It

is typically in the range between 100 kHz and 300 kHz.

Range outside the PLL bandwidth

Outside the PLL bandwidth, the phase noise is practically exclusively
determined by the phase noise of the oscillator in non-synchronized
mode. In this range it decreases by 20 dB per decade.

L
Phase noise of Free-running
reference oscillator oscillator (VCO)
referred to output
frequency f0 (OCXO)

VCO locked to OCXO:

narrow PLL bandwidth

medium PLL bandwidth

wide PLL bandwidth

1 2 3 f
f0

Fig. 5-11  P
 hase noise of OCXO, VCO and of VCO locked to OCXO at different PLL
bandwidths

Fig. 5-11 shows the phase noise at different PLL bandwidths. It is of par-
ticular interest to compare the phase noise of a free-running oscillator
with the phase noise of an oscillator locked to a reference oscillator at
different PLL bandwidths. The following cases have to be considered:

Wide PLL bandwidth

The loop gain of the PLL is so high that the oscillator noise is reduced to
the reference oscillator noise. Due to the phase rotation of the filter cir-
cuit, the phase noise is increased far from the carrier.

Medium PLL bandwidth

The loop gain is not sufficient to attain the reference oscillator noise
close to the carrier. The increase of the phase noise far from the carrier
is, however, much smaller than with a wide PLL bandwidth.

115
 Performance Features of S p e ct r u m A n a l y z e r s

Narrow PLL bandwidth

The phase noise far from the carrier is not degraded compared to the
free-running oscillator. Close to the carrier it is, however, considerably
higher than that with medium and wide PLL bandwidth.

To optimize the phase noise for the specific application, the PLL band-
width should be variable.

Input signal
A A
IF signal
Input IF

LO

fin,1 fin, 2 f fIF f

LO signal

fLO f

Fig. 5-12  Internal phase noise transferred onto input signal by reciprocal mixing

The LO phase noise of a spectrum analyzer is transferred onto the input

signal by reciprocal mixing in the conversion stages (Fig. 5-12). This
means that even with an ideal sinusoidal input signal, the displayed
spectrum will reflect the combined phase noise of all LOs. If the input
signal also exhibits phase noise (which in practice is always the case),
the trace displayed on the spectrum analyzer corresponds to the com-
bined phase noise of the input signal and LOs.
The displayed phase noise is always relative to the carrier of the
input signal irrespective of the input signal level. This means that for
phase noise measurements on input signals (chapter 6.1: Phase noise
measurements) the dynamic range for close-to-carrier measurements
cannot be maximized by increasing the input signal level – which is in
sharp contrast to the effect of thermal noise.
In particular with measurements close to the carrier, the phase noise
of the spectrum analyzers therefore marks the limit of the measure-
ment range.

116
 P h a s e N o i s e (S p e ct r a l P u r i t y )

Apart from this restriction, the resolution and the dynamic range of the
analyzer will also be limited by the phase noise. Signals with little offset
from the carrier of a signal with much higher levels may not be detect-
able (Fig. 5-12).

Example:
A sinusoidal signal with a level of -10 dBm is present at the input of a
spectrum analyzer. At a carrier offset of 100 kHz the phase noise of the
spectrum analyzer is assumed to be -100 dBc (1 Hz).
What level must a second signal at the same offset of 100 kHz have
to be detectable at a resolution bandwidth of 1 kHz (in our example the
resolution bandwidth is assumed to correspond to the noise bandwidth
of the filter)?

Solution:
Due to the resolution bandwidth of 1 kHz the phase noise produces a
displayed noise level LN of
1 kHz 
LN (Pc ) = -100 dB + 10 dB ⋅ lg   = -70 dB 
 1 Hz 
This means that the input signal must have at least a level LS, min of
( )
L S,min mW = -10 dB - 70 dB = -80 dB in order to be detectable fol-
lowing a noise increase by 3 dB.
This limited resolution has also an adverse effect on adjacent-chan-
nel power measurements (chapter 6.3: Channel and adjacent-channel
power measurement) since dynamic range is reduced by phase noise
occurring in the adjacent channel.

Data sheets often specify residual frequency modulation (residual FM).

From the carrier-offset-dependent phase noise, the RMS value of the
residual FM can be calculated by integration as follows:

117
 Performance Features of S p e ct r u m A n a l y z e r s

2
 L P ,f 
∞  (20cdBoff ) 
DFRMS = 2 ⋅ ∫ 10 ⋅ foff  dfoff
   (Equation 5-22)
foff = 0  

where DFRMS RMS value of residual FM

foff frequency offset from carrier
L(Pc , foff) phase noise level as a function of carrier offset,
relative to carrier power Pc and 1 Hz bandwidth
(dBc (1 Hz))

Similarly, the RMS value of the residual phase modulation (residual jM)
can be calculated from the phase noise:

∞ (
L Pc, foff )
Dj RMS / rad = 2 ⋅ ∫ 10 10 dB
dfoff  (Equation 5-23)
foff = 0

∞ (
L Pc, foff )
180°
Dj RMS = 2 ⋅ ∫ 10 10 dB
dfoff  (Equation 5-24)
π foff = 0

where DjRMS RMS value of residual jM

foff frequency offset from carrier
L(Pc,foff) phase noise level as a function of carrier offset,
­relative to carrier power Pc and 1 Hz bandwidth
(dBc (1 Hz))

High residual FM of the LO signal may produce a smearing effect of the

displayed spectrum. This leads to a reduction of the resolution and thus
determines the lower limit for the useful smallest resolution bandwidth.
Since in modern spectrum analyzers the local oscillators are imple-
mented as synthesizers as described above, this effect is practically of
no relevance.

118
 P h a s e N o i s e (S p e ct r a l P u r i t y )

Spectral purity (dBc (1 Hz))

SSB phase noise, f = 500 MHz
Carrier offset
100 Hz <–90, typ. –94
1 kHz <–100, typ. –108
10 kHz <–106, typ. –113
100 kHz1) <–110, typ. –113
1 MHz1) <–120, typ. –125
10 MHz typ. –145
Residual FM
(f = 500 MHz, BN = 1 kHz typ. 3 Hz
sweep time 100 ms)

Typical values for SSB phase noise

Carrier fin = 500 MHz fin = 3 GHz fin = 7 GHz
offset

Fig. 5-13    100 Hz 94 dBc (1 Hz) 90 dBc (1 Hz) 84 dBc (1 Hz)

Typical specifications 1 kHz 105 dBc (1 Hz) 100 dBc (1 Hz) 94 dBc (1 Hz)
for phase noise and 10 kHz 113 dBc (1 Hz) 108 dBc (1 Hz) 104 dBc (1 Hz)
residual FM of a 100 kHz 113 dBc (1 Hz) 108 dBc (1 Hz) 106 dBc (1 Hz)
spectrum analyzer
1 MHz 125 dBc (1 Hz) 118 dBc (1 Hz) 118 dBc (1 Hz)
(extract from data sheet)
1)  Valid for span >100 kHz.

In view of the above restrictions, phase noise is an essential criterion for

evaluating a spectrum analyzer. Depending on the application, phase
noise may be important for the user both in the case of small offsets
(such as measurements on radar systems) and large offsets (such as
measurements on mobile radio equipment). Data sheets therefore
always provide specifications at different offsets, usually in decade
steps (see Fig. 5-13).
As shown above, phase noise is largely influenced by the PLL band-
width. In spectrum analyzers the PLL bandwidth is usually variable to
allow adaptation to the specific measurement task. Switchover is often
implicit, as is illustrated with the analyzer described in our example. The
PLL bandwidth is coupled to the frequency range to be displayed or to
the selected resolution bandwidth. Especially if large frequency ranges
are to be displayed (span >100 kHz, in our example, see footnote 1 in
the extract from data sheet), the minimum phase noise far from the car-
rier is usually of interest. Therefore a narrow PLL bandwidth is automat-
ically selected for this setting.

119
 Performance Features of S p e ct r u m A n a l y z e r s

To better define parameter dependencies, resolution bandwidths and

span settings are often stated in data sheets in addition to the phase
noise at the various carrier offsets. Settings other than those specified
may result in poorer phase noise values.

For analyzing input signals of very high frequency, the LO signal must
be multiplied (chapter 4.1). Just as it is found in a frequency-modulated
signal, the frequency deviation is multiplied, thus causing degrading of
the phase noise. The following applies:

Lmult  (Pc, foff ) = L (Pc, foff ) + 20 dB ⋅ lg (n )  (Equation 5-25)

where Lmult(Pc, foff) phase noise level after multiplication as a

function of carrier offset, relative to carrier
power Pc and 1 Hz bandwidth (dBc (1 Hz))
L(Pc,foff) phase noise level of original signal as a func-
tion of carrier offset, relative to carrier power
Pc and 1 Hz bandwidth (dBc (1 Hz))
n multiplying factor

Due to this degradation, phase noise specifications are always refer-

enced to a specific signal frequency. Typical phase noise curves are often
provided for several signal frequencies, allowing an estimation of the
expected phase noise in the frequency range of interest.

5.4 1 dB compression point and maximum input level

The 1 dB compression point of a network marks the point in its dynamic
range at which the gain is reduced by 1 dB due to saturation (see
Fig. 5-14). Similar to the intercept point, the 1 dB compression point can
be referenced to the input or to the output level. For power amplifiers,
the output level at which the 1 dB compression occurs is usually speci-
fied, and the input level is specified for spectrum analyzers.

120
 1 dB C o m p r e s s i o n P o i n t and Maximum Input Level

Ideal
Lout / dBm Output level as a network
function of input level

1 dB
L1 dB, out

real
network

Fig. 5-14   
1 dB compression point
Lin / dBm
L1 dB, in

Maximum input level

RF attenuation 0 dB
DC voltage 50 V
CW RF power 20 dBm (= 0.3 W)
Pulse spectral density 97 dBµV (Hz)
RF attenuation 10 dB
CW RF power 30 dBm (= 1 W)
Fig. 5-15    Max. pulse voltage 150 V
Typical specifications
Max. pulse energy (10 µs) 1mWs
for 1 dB compression
point and maximum 1 dB compressionn of output level
input level (extract
0 dB RF attenuation, f >200 MHz 0 dBm nominal
from data sheet)

The 1 dB compression point is mainly determined by the first mixer and
usually specified at an attenuator setting of 0 dB. The specified input
level is also referred to as mixer level. By increasing the RF attenuation
the 1 dB compression point is increased to the same degree.

To avoid unwanted products due to distortion, the maximum input level

(reference level) to be displayed should be kept clearly below the 1 dB
compression point.
Due to the coupling of reference level and attenuator setting (chap-
ter 4.6: Parameter dependencies), the maximum reference level is lim-
ited (in our example to -10 dBm) with an RF attenuation of 0 dB. The

121
 Performance Features of S p e ct r u m A n a l y z e r s

1 dB compression point cannot be measured directly. Nevertheless, it is

an important criterion in many measurements.
For example, when making phase noise measurements a single sinu-
soidal signal is applied to the spectrum analyzer input. Even when driv-
ing the spectrum analyzer close to its 1 dB compression point, no inter-
modulation products will be generated that would appear in the vicin-
ity of the input signal. Due to the high level of the input signal, only har-
monics of it will be produced in the spectrum analyzer, which usually do
not degrade phase noise measurements. The wide dynamic range makes
for a large signal-to-noise ratio so that the maximum dynamic range at
large carrier offsets is obtained for measurements.

In contrast to the 1 dB compression point, which provides information

about the overdrive capacity of the spectrum analyzer, the maximum
input level denotes the upper limit for damage-free operation. To avoid
damage to the analyzer, this value should not be exceeded.
The limit value is usually determined by the first critical component
in the signal processing chain. Accordingly, the selected RF attenuation
has always to be taken into account:

❙ RF attenuation 0 dB
The attenuator is not loaded in this case and the input signal is not
attenuated. Therefore the first mixer is usually decisive for the maxi-
mum input level. Due to the generally high load capacity of diplexer
and tracking bandpass filter, the same holds true for the high-frequency
input section (above 3 GHz in our example).

❙ RF attenuation > 0 dB (≥ 10 dB in our example)

The input signal is attenuated by the attenuator so that the subsequent
stages can usually be ignored. The specified value reflects the load
capacity of the attenuator.

Both of the above cases are important for practical operation, so they
are specified separately in data sheets.
Further distinction is made as to the type of the input signal
(Fig. 5-15):

122
 1 dB C o m p r e s s i o n P o i n t and Maximum Input Level

DC voltage
For DC-coupled spectrum analyzers this value corresponds to the max-
imum mixer-compatible DC voltage. Usually a value of 0 V is specified
irrespective of the RF attenuation.
For AC-coupled spectrum analyzers the specified value corresponds
to the dielectric strength of the coupling capacitor at the spectrum ana-
lyzer input. In the above data sheet extract a value of 50 V is specified.

Continuous wave (CW) RF power

This value specifies the maximum total power of all input signals that is
permissible without any time limit. It is assumed that all input signals
are stationary.

Pulse spectral density

Pulsed signals feature a very wide spectrum with many spectral compo-
nents whose sum power should not exceed a specific value. For pulse
spectra, a spectral density is usually specified as the voltage level rel-
ative to a specific bandwidth, typically 1 MHz. In the above data sheet
extract (Fig. 5-15), 97 dBµV (Hz) is specified as the limit value.

Maximum pulse energy and maximum pulse voltage

With extremely short pulses, the pulse power may exceed the value spec-
ified for the CW RF power for long intervals of time. There is a limit-
ing threshold set by the maximum pulse energy, in our example speci-
fied in mWs for a certain pulse period, as well as by the maximum pulse
­voltage.
Often the maximum pulse power is specified which can be calculated
from the pulse energy and pulse period as follows:
EP
PP = and E P = PP⋅ t P  (Equation 5-26)
tP
where PP pulse power
EP pulse energy
tP pulse duration

With the values specified in the data sheet extract in Fig. 5-15
(EP = 1 mWs, tP = 10 µs) a maximum pulse power of 100 W is
obtained.

123
 performanCe feaTures of speCTrum analyzers

PP

PP,1

EP,1 EP,1 = EP, 2

PP,2

EP, 2
Fig. 5-16 
Pulses with different 
pulse width but equal 
pulse energy
tP,1 tP, 2 tP

For constant pulse energy EP, the permissible pulse power is even higher
for a short pulse period in accordance with Equation 5-26 (Fig. 5-16).
By decreasing the pulse period, the pulse power may not be increased
arbitrarily since the maximum permissible pulse voltage must not be
exceeded. In the cited specifications, a value of 150 V is stipulated. For
the rectangular pulse as shown in Fig. 5-16, the permissible peak voltage
yields a maximum pulse power of

(150 V)
2
Û P2
PP = = = 450 W (Equation 5-27)
R 50 W

where ÛP pulse peak voltage

R input impedance of spectrum analyzer

This maximum pulse power as well as the maximum permissible pulse

energy (1 mWs in our example) should not be exceeded under any cir-
cumstances. Equation 5-26 can be used to calculate the maximum pulse
period for a pulse of maximum pulse power, which in our example is
2.2 µs.
For longer pulse periods and constant pulse energy, the pulse power
has to be reduced. The relationship shown in Fig. 5-17 is then obtained
(valid for the specifications provided in Fig. 5-15).

124
 Dynamic Range

1000
Max. pulse power Pp / W

100

10

1
0.1 1 10 100 1000
Pulse period tp / µs

Fig. 5-17  M
 aximum pulse power as a function of pulse period (for max. pulse
voltage of 150 V and max. pulse energy of 1 mWs)

5.5 Dynamic range

The dynamic range provides information about the analyzer’s capability

to simultaneously process signals with very different levels. The limits
of the dynamic range depend on the measurement to be performed. The
lower limit is determined by the inherent noise or phase noise. The upper
limit is set either by the 1 dB compression point or by distortion prod-
ucts occurring in the analyzer in the case of overdriving. The dynamic
range can be defined in different ways. It should not be confused with
the display range.

125
 Performance Features of S p e ct r u m A n a l y z e r s

Max.
input level
1 dB compression
of 1st mixer

range / max. harmonic supression

Optimum
mixer level

intermodulation-free
Max. dynamic range
Level display range

Max.

Fig. 5-18   
Comparison of level display range,
Displayed overrange noise level maximum dynamic range and max.
with selected resolution bandwidth
intermodulation-free range

Level display range

The dynamic range does not correspond to the level display range which
is also specified in data sheets. The display range is the range from the
displayed average noise level to the maximum input level (Fig. 5-18). For
displaying a signal with a level corresponding to the maximum input
level of the analyzer, an RF attenuation greater than 0 dB is usually
selected, which means that the minimum displayed noise (the lower
limit of the specified level display range) will not be attained.

Maximum dynamic range

A maximum dynamic range is often specified which is limited by the
displayed noise (usually at the smallest resolution bandwidth) and the
1 dB compression point (Fig. 5-18). If, however, an input signal at the first
mixer reaches the mixer’s 1 dB compression point, there will be very
high levels of the distortion products generated due to nonlinearities of
the mixer. Using sufficiently small bandwidth, the distortion products
can become visible in the displayed spectrum (they are not covered up
by the inherent noise). The spectrum display is then no longer unambig-
uous.
In modern spectrum analyzers, the reference level selectable for a
specific RF attenuation is therefore limited even with manual decoupled
attenuator setting. Signals attaining the 1 dB compression point at the
mixer input considerably exceed the reference level.

126
 Dynamic Range

The value stated for the maximum dynamic range is therefore of limited
significance and only relevant for certain applications, such as phase
noise measurements far from the carrier.

Maximum intermodulation-free range for maximum harmonic

suppression
As explained in chapter 4.6.2, Reference level and RF attenuation, a com-
promise has to be found in the selection of the mixer level. If the RF
attenuation is high, so that the mixer level is low, the levels of the distor-
tion and intermodulation products generated in the analyzer are also low,
but at the same time the signal-to-noise ratio of the input signal is small.
In this case the dynamic range is limited at the lower end by the inherent
noise. If, on the other hand, the mixer level is high, then distortion and
intermodulation products are generated whose levels exceed the inherent
noise level and therefore become visible (Fig. 4-31). In practice, it is impor-
tant to have a level display range in which the displayed spectrum is free
from such products. Depending on whether intermodulation products or
harmonics of higher order limit this range, one speaks of an intermodu-
lation-free range or maximum harmonic suppression. Both parameters
depend on the mixer level and the selected resolution bandwidth. A max-
imum is attained if the levels of the intermodulation products or har-
monics of higher order are equal to the noise level. The ideal mixer level
required for this purpose can either be calculated or graphically deter-
mined. For elucidation the graphical method is described first.

–40 40
LIM3,rel Lk2, rel
–50 50
Intermodulation-free or distortion-free
dynamic range in dB
LN, rel, Lk2,rel, LIM3, rel in dB

–60 60

–70 70

–80 80
LN, rel (BR = 1 kHz)
–90 90

–100 100
LN,rel (BN = 1 Hz) LN, rel (BN = 10 Hz)
–110 110

–120 120
–70 –60 –50 –40 –30 –20 –10 0
Lmix,opt Mixer level in dBm
Fig. 5-19  I ntermodulation-free range and maximum harmonic suppression as a
function of mixer level (NF = 24.5 dB, IP3in= 7 dBm, SHIin= 40 dBm)

127
 Performance Features of S p e ct r u m A n a l y z e r s

For a specific noise bandwidth of the IF filter and noise figure of the
spectrum analyzer, the noise power level is to be calculated relative to
the mixer level using the following:

 BN, IF 
LN, rel(Pmix ) = LN(mW) - Lmix(mW) = -174 dB + 10 dB ⋅ lg   + NF - Lmix(mW)  
 Hz 
 
 (Equation 5-28)

where LN, rel relative noise level referenced to mixer power Pmix
Lmix mixer level, relative to 1 mW
LN noise level, relative to 1 mW
BN, IF noise bandwidth of resolution filter
NF noise figure of spectrum analyzer

When using a sample detector and averaging by a narrowband video fil-

ter, a further 2.5 dB is to be subtracted from the calculated value LN, rel
due to underweighting of the noise. The relative noise level is plotted in
Fig. 5-19 for different resolution filters versus the mixer level. A noise fig-
ure of 24.5 dB is assumed in this example. It is shown that the relative
noise level decreases with increasing mixer level.

The following relationship can be derived from Equation 5-16 for the
relative level of nth order intermodulation products referenced to the
mixer level:

( )
LIMn, rel Pmix = - (n - 1) ⋅ (IPn in(mW) - Lmix(mW))  (Equation 5-29)

where LIMn, rel relative level of nth order intermodulation ­products

referenced to mixer power Pmix
IPnin input intercept point of nth order of spectrum analyzer
(RF attenuation 0 dB), relative to 1 mW
Lmix mixer level, relative to 1 mW

Usually the 3rd order intermodulation products are the most disturbing
ones in practice since they occur in the immediate vicinity of the input
signals. The relative level of such products is determined by:

( )
LIM3, rel Pmix = -2 ⋅ (IP 3in(mW) - Lmix(mW))  (Equation 5-30)

128
 Dynamic Range

Likewise, the relative level of 2nd order distortion products (2nd order
harmonics) can be derived from Equation 5-13:

( )
Lk2, rel Pmix = -(SHI in(mW) - Lmix(mW))  (Equation 5-31)

where Lk2, rel relative level of 2nd order distortion products, refer-
enced to mixer power Pmix
SHIin input second harmonic intercept of spectrum analyzer,
relative to 1 mW

Since the distortion and intermodulation products are always calculated

from the mixer level, the results are independent of the RF attenuation.
Therefore the intercept points referenced to the input of the first mixer
must always be substituted for IP3in and SHIin. The values correspond
to the intercept points of the analyzer with a RF attenuation of 0 dB.
The relative level of the 3rd order intermodulation products as well
as of the 2nd harmonics are shown in Fig. 5-19 as a function of the mixer
level. For IP3in , a value of 7 dBm has been assumed. SHIin is 40 dBm.

Depending on the specific measurement application, 3rd order intermod-

ulation products or harmonics of higher order may limit the dynamic
range. The optimum mixer level for the specific application and hence
the maximum attainable dynamic range is obtained from the point of
intersection of the noise level line and the line of the 3rd order inter-
modulation products or 2nd order harmonics. At this point of intersec-
tion, the level of the intermodulation or distortion products is equal to
the noise level, and the representation is still unambiguous.
In Fig. 5-19, a maximum intermodulation-free range of about 98 dB
can be found at the noise bandwidth of 10 Hz. A mixer level of -42 dBm is
required for this range. If two sinusoidal signals each having a level of
-42 dBm are applied to the spectrum analyzer (RF attenuation 0 dB), the
3rd intermodulation products will attain a level of -42 dBm - 98 dB =
-140 dBm

The optimum mixer level and the attainable dynamic range can also be
calculated.

With optimum mixer level, the noise level corresponds to the level of the
intermodulation products. Equation 5-28 and Equation 5-29 have to be
equated and solved for Lmix . This yields

129
 Performance Features of S p e ct r u m A n a l y z e r s

(n - 1) ⋅ IPnin(mW) - 1774 dB + 10 dB ⋅ lg 

 BN, IF 
 + NF
 
(n - 1) ⋅ IPnin(mW) + LN, rel(mW)  Hz 
 L (mW ) =
mix, opt
n
=
n
 
 (Equation 5-32)

where Lmix, opt optimum mixer level, relative to 1 mW

IPnin input intercept point of nth order of spectrum analyzer
(RF attenuation 0 dB), relative to 1 mW
n order of intermodulation or distortion products by
which the dynamic range is limited
BN, IF noise bandwidth of resolution filter
NF noise figure of spectrum analyzer

For n = 3, that is the limitation of the intermodulation-free range by 3rd

order intermodulation products, the following is obtained:
 BN, IF 
2 ⋅ IP 3in(mW) - 174 dB + 10 dB ⋅ lg   + NF  
2 ⋅ IP 3in(mW ) + LN, rel(Pc )  Hz 
 Lmix, opt(mW) = 3
=
3
 (Equation 5-33)

The optimum mixer level for maximum harmonic suppression is deter-

mined by
B 
SHI in(mW ) - 174 dB + 10 dB ⋅ lg  N, IF  + NF
SHI in(mW) + LN, rel(Pc )  Hz 
Lmix, opt(mW) = =  
2 2
 (Equation 5-34)

With optimum mixer level, the dynamic range corresponds to the level
difference between mixer level and nth order intermodulation products
or noise level. The following applies:

n -1 n - 1   BN, IF  
 DRmax = ⋅ (IPn in(mW) - LN(Pc )) = ⋅ IPn in(mW) + 174 dB - 10 dB ⋅ lg   - NF 
n n   Hz  
 
 (Equation 5-35)

where DRmax maximum dynamic range

IPnin input intercept point of nth order of spectrum analyzer
(RF attenuation 0 dB), relative to 1 mW
n order of intermodulation or distortion products by
which the dynamic range is limited

130
 Dynamic Range

BN,IF noise bandwidth of resolution filter

NF noise figure of spectrum analyzer

For n = 3, a maximum intermodulation-free range of

2 2   BN, IF  
DRmax = ⋅ (IP 3in(mW) - LN(Pc )) = ⋅ IP 3in(mW) + 174 dB - 10 dB ⋅ lg   - NF 
3 3   Hz    
 
 (Equation 5-36)

or (for n = 2) a maximum harmonic suppression of

1 1   BN, IF  
DRmax = ⋅ (SHI in(mW) - LN(Pc )) = ⋅ SHI in(mW) + 174 dB - 10 dB ⋅ lg   - NF   
2 2   Hz  
 
 (Equation 5-37)

can be derived.

Equation 5-35 reveals that both a high intercept point and a low noise
figure are required to obtain a high intermodulation-free range. For fast
assessment of the dynamic range of a spectrum analyzer a figure of
merit (FOM) can be used as follows:

FOM = IP 3in(mW) - NF  (Equation 5-38)

The intermodulation-free dynamic range becomes wider with increas-

ing FOM. Modern high-end analyzers attain a figure of merit of 0 dBm
with a typical 3rd order intercept point of 15 dBm and typical noise fig-
ure of 15 dB.
The above discussions regarding the dynamic range were with refer-
ence to the signal level at the input of the first mixer. If the signal level
at the spectrum analyzer input is higher than the optimum mixer level,
it must be reduced by an appropriate RF attenuation. The required RF
attenuation can be calculated as follows:

a RF = Lin(mW) - Lmix(mW)  (Equation 5-39)

where aRF RF attenuation

Lin signal level at spectrum analyzer input, relative to
1 mW
Lmix mixer level to be set, relative to 1 mW

131
 Performance Features of S p e ct r u m A n a l y z e r s

When setting the mixer level, the attenuator steps are important:
If in the above example the input level is -17 dBm and the atten-
uator steps are 10 dB, the signal level can only be reduced to a mixer
level of -37 dBm (with 20 dB RF attenuation) or -47 dBm (with 30 dB
RF ­attenuation). Accordingly, the intermodulation-free range is then
92 dB (Lmix= -47 dBm) or 88 dB (Lmix= -37 dBm). To utilize the maxi-
mum intermodulation-free range, the level can be reduced to -22 dBm
by means of an external 5 dB attenuator. By applying an RF attenuation
of 20 dB, the optimum mixer level of -42 dBm and an intermodulation-
free range of 98 dB are obtained again.
For some spectrum analyzers, an attenuator with 1 dB steps is avail-
able. It is then not necessary to use external attenuator pads or an exter-
nal RF attenuator.

Effects of phase noise on dynamic range

As described in chapter 5.3, the phase noise of the local oscillators of
the spectrum analyzer is transferred onto the input signals by recipro-
cal mixing. The dynamic range for phase noise measurements on input
signals is therefore limited by the spectrum analyzer phase noise – in
particular at small carrier offsets. The phase noise of the DUT must be
higher than that of the measuring instrument for accurate measure-
ments (chapter 6.1: Phase noise measurements).
If weak signals in the immediate vicinity of very strong input sig-
nals are to be displayed (such as for measuring the 3rd intercept point
of a DUT), the phase noise of the analyzer needs to be as low as possible.
Otherwise, the weak input signal may be covered up by the phase noise
transferred onto the strong neighboring signal (Fig. 5-12). The phase
noise should be accounted for in such cases in the calculation of the
dynamic range.
Since the phase noise transferred onto the input signal depends on
the signal carrier level, varying the carrier level cannot influence the
phase noise effect. In Fig. 5-20 the phase noise contribution is repre-
sented accordingly as a horizontal line. If the signal frequency is greater
than the carrier offset at which measurements on weak signals are to
be performed, harmonics and 2nd order intermodulation products are
insignificant. In Fig. 5-20 only 3rd order intermodulation products are
therefore taken into account. The phase noise level has to be calculated
for the specified resolution bandwidth. The following applies:

132
 Dynamic Range

 BN, IF 
LPN (Pc, foff ) = L (Pc, foff ) + 10 dB ⋅ lg    (Equation 5-40)
 Hz 

where LPN phase noise as a function of carrier offset within band-

width BN, IF, relative to carrier power Pc
L(foff) phase noise as a function of carrier offset, relative to
carrier power Pc and 1 Hz bandwidth
BN, IF noise bandwidth of IF filter
foff carrier offset

The effects of thermal noise, intermodulation products and phase noise

have to be added linearly.
The sum trace (Lsum) represented in Fig. 5-20 holds true for a phase
noise of -122 dBc (1 Hz), a 3rd order intercept point of 7 dBm and a noise
figure of 24.5 dB. The selected resolution bandwidth of 10 kHz should
correspond to the noise bandwidth. Maximum dynamic range is attained
at a mixer level at which the sum trace is at its minimum.
–40 40

–50 50
LN, rel, LIM3, rel, LPN,RBW in dB

Dynamic range in dB

–60 60
LSum
–70 70

–80 80
LPN, RBW
–90 90
LIM3,rel LN, rel
–100 100

–110 110

–120 120
–70 –60 –50 –40 –30 –20 –10 0
Lmix, opt
Mixer level in dB
Fig. 5-20  D
 ynamic range taking into account thermal noise, phase noise and 3rd
order intermodulation products (NF = 24.5 dB, IP3in = 7 dBm,  
L(foff ) = -122 dBc (1 Hz), BN, IF = 10 kHz)

Similarly, the dynamic range for adjacent-channel power measurements

is limited by phase noise. Further details on the dynamic range for this
type of measurement can be found in chapter 6.3: Channel and adjacent-
channel power measurements.

133
 Performance Features of S p e ct r u m A n a l y z e r s

For the simple determination of the dynamic range as a function of

noise figure, 3rd order intercept point and phase noise of the spectrum
analyzer, a spreadsheet in MS Excel™ is available (file DYN_CALC.XLS,
Fig. 5-21) which can be obtained via the Rohde&Schwarz website
(www.rohde-schwarz.com). In the spreadsheet 2nd order harmonics are
taken into account, so that the dynamic range can easily be calculated
for practically any application. The spreadsheet consists of two sheets:

❙ Input & Diagram
Input of noise bandwidth, noise figure, IP3, SHI and phase noise of the
spectrum analyzer (Fig. 5-21a, yellow highlighted fields top left). Graphi-
cal output of relative noise level with respect to input signal level, phase
noise level as well as relative levels of 2nd harmonics and 3rd order inter-
modulation products. In addition, the sum of the contributions of ther-
mal noise, phase noise and 3rd order intermodulation products is output.

❙ Num. Results
Numeric output of results, which are graphically output on the
Input&Diagram sheet.

The phase noise value to be entered depends on the frequency offset

from the strong signal at which a weak signal is to be represented.

a)

134
 Immunity to Interference

b)

Fig. 5-21  S
 preadsheet DYN_CALC.XLS (a) Input & Diagram sheet: input fields
for noise bandwidth, noise figure, TOI, SHI and phase noise as well as
graphical representation of resulting dynamic range (b) Num. Results
sheet: numeric output of results

5.6 Immunity to interference

The signal at the spectrum analyzer input may give rise to unwanted
components which spectrally do not show any relationship to the input
signal. There are different causes for such unwanted components which
are to be dealt with in the following section. Unlike harmonics or inter-
modulation products generated in the spectrum analyzer due to nonlin-
earities, immunity to interference cannot as a rule be improved by opti-
mizing the mixer level as it is mostly independent of the selected RF
attenuation.

Immunity to interference
Image frequency >70 dB
Intermediale frequency >70 dB
Spurious responses (f >1 MHz, without
Fig. 5-22   <–103 dBm
input signal, 0 dB attenuation)
Typical data sheet speci-
Other surious with input signal, mixer
fications for immunity to <–70 dBc
interference level <–10 dBm, Df >100 kHz)

135
 Performance Features of S p e ct r u m A n a l y z e r s

Image frequency
As described in chapter 4, the conversion of a signal by mixing is not
unambiguous. At a specific LO frequency, there is always an image fre-
quency in addition to the wanted signal frequency. From Equation 4-4
and Equation 4-5 the following relationship between image frequency
and input frequency can be derived:

fim = fin + 2 ⋅ f1stIF  (Equation 5-41)

Input signals at the image frequency are suppressed by suitable filters

depending on the configuration of the front end as described in chapter
4.1. Due to the limited isolation of technically feasible filters, the achiev-
able suppression has its limits. In the above data sheet extract, a value
of > 70 dB is specified.
Applied to the analyzer described in chapter 4, this means that an
input signal with a frequency of 7100 MHz and level of -10 dBm will
cause in the displayed spectrum a response at 147.2 MHz with a maxi-
mum level of (-10 dBm- 70 dB) = -80 dBm.

As shown in chapter 4.1, an image frequency also occurs in the second con-
version and has to be suppressed accordingly. For the analyzer described
in our example, the following relationship can be derived for the frequency
the input signal must have to be converted to the image frequency of the
second mixer and to become visible at the frequency fin:

fim,2nd IF = fin + 2 ⋅ f2nd IF  (Equation 5-42)

where fim, 2ndIF spurious response due to image frequency at 2nd IF

fin frequency at which spurious response becomes visible
in displayed spectrum
f2ndIF second intermediate frequency

IF feedthrough or reception at intermediate frequency

Due to the limited isolation between the RF input and IF output of the
first mixer, input signals may be fed through directly (without conver-
sion) to the IF (chapter 4.1), which is known as the IF feedthrough. If the
frequency of an input signal corresponds to the first IF, the signal will be
reproduced in the frequency range of the displayed spectrum irrespec-
tive of the LO frequency. Signals with a frequency corresponding to the
first IF therefore must be suppressed before the first mixer by appro-

136
 Immunity to Interference

priate filters which are required for image frequency rejection. The ana-
lyzer described here uses an input lowpass filter (3) for this purpose in
the RF input section for up to 3 GHz and a tracking bandpass filter (20)
for the frequency range above 3 GHz. The signals to be suppressed are
at 3476.4 MHz and 404.4 MHz.
In the above data sheet extract, a value of > 70 dB is specified for the
immunity to interference at the IF. This means that for an input signal
of 3476.4 MHz and -10 dBm, a maximum value of -80 dBm will be dis-
played in the frequency range 9 kHz to 3 GHz.

Spurious responses
Inherent spurious responses
Inherent spurious responses are signals in the displayed spectrum
that are generated in the spectrum analyzer itself. They are caused,
for instance, by clock signals of microprocessors, which may be prop-
agated via supply voltage lines and coupled into the analog signal pro-
cessing circuitry. Distinction has to be made whether inherent spurious
responses are permanently present or only occur if a signal is present at
the input of the spectrum analyzer. Spurious of local oscillators belong
to the latter group, being present only if an input signal is present. Data
sheet specifications for inherent spurious responses produced by the
input signal are therefore related to the carrier level of the input signal
(in dBc). In the data sheet extract shown in Fig. 5-22, a value of -70 dBc
is specified, and for inherent spurious responses that are independent of
the input signal -103 dBm is ­specified.

Spurious responses
Harmonics of the input signal are produced among others in the first
mixer of the spectrum analyzer. If the input level is sufficiently high,
the harmonics will be displayed. Harmonics of the input signal are con-
verted to the first intermediate frequency by means of the fundamen-
tal and the harmonics of the LO signal in accordance with Equation 4-1.
For input frequencies fin, N , for which Equation 4-1 is fulfilled with m ≥ 1
and n > 1 at a specified IF and LO frequency range, spurious responses
are generated.

Example:
A spectrum analyzer for the frequency range of 10 MHz to 5 GHz con-
verts the input signal to a high first IF of 5.8 GHz with the aid of a LO sig-

137
 Performance Features of S p e ct r u m A n a l y z e r s

nal tunable from 5.81 GHz to 10.8 GHz. A signal of 3.87 GHz is applied to

the analyzer input and displayed at a frequency of 3.87 GHz.
Simultaneously, higher-order harmonics of the signal are produced
in the first mixer of the analyzer. The 3rd harmonic, for instance, is at
11.61 GHz. If the analyzer is tuned to an input frequency of 10 MHz, the
LO frequency is 5.81 GHz. The 3rd harmonic of the input signal is thus
converted to the IF:

fIF = 3 ⋅ fin- fLO = 3 ⋅ 3.87 GHz - 5.81 GHz = 5.80 GHz

The input signal at 3.87 GHz also causes a spurious response at 10 MHz

in the displayed spectrum.

Such spurious responses are inherent in the concept. To avoid the gen-
erated spurious from disturbing the displayed spectrum, particularly
stringent requirements have to be observed for the first mixer of the
spectrum analyzer regarding linearity specifically intercept point. At the
same time, the mixer level should not be too high, which is a require-
ment that can be fulfilled by appropriate setting of the RF attenuation.

5.7 LO feedthrough

In passive mixers as used in spectrum analyzers for the first conversion

of the input signal, the LO signal is coupled into the IF path due to the
limited isolation. The block diagram of the analyzer described here (see
fold-out page) shows that if very low-frequency input signals are con-
verted (such as 9 kHz), the frequency of the LO signal (3476.409 MHz
in our example) practically corresponds to the first IF. Especially with
large resolution bandwidths (0.5  · BIF > fin) the LO signal coupled into
the IF path is therefore not suppressed or only slightly by the IF filter.
The LO signal is then sent to the detector and displayed (Fig. 5-23); this
effect is referred to as LO feedthrough. Due to the phase noise of the
LO signal, the displayed average noise level close to the minimum start
frequency is increased and as a result the sensitivity in this frequency
range decreased. The LO feedthrough is usually not stated explicitly in
data sheets. It can, however, be recognized from the displayed noise
level specified for the frequency range close to zero.

138
 F i l t e r C h a r a ct e r i s t i c s

RBW 20 kHz RF ATT 10 dB

VBW 2 kHz
Ref Lvl –20 dBm SWT 6 ms Unit dBm
–20
*
A
–30

–40

–50
1AVG
2VIEW RBW = 20 kHz 1SA
–60 2SA
3VIEW RBW = 10 kHz
4VIEW 3SA
–70 RBW = 5 kHz 4SA
RBW = 1 kHz
–80

–90

–100

–110

–120
Start 9 kHz 9.1 kHz/ Stop 100 kHz

Fig. 5-23  LO feedthrough as a function of selected resolution bandwidth

The LO feedthrough can be reduced by reducing the resolution band-

width as shown in Fig. 5-23.
For spectrum analyzers featuring a very low input frequency limit,
such as 20 Hz, this is possible to a limited extent only. Due to the very nar-
row resolution bandwidths required to reduce the LO feedthrough, the
sweep time is drastically increased. Therefore complex circuitry is often
implemented in such analyzers in order to reduce the LO feedthrough.
The LO signal can, for instance, be coupled into the IF path with opposite
phase, thus causing partial cancellation and LO suppression.

5.8 Filter characteristics

The main characteristics and different methods of implementing resolu-

tion filters have been described in chapter 4.2. In addition to the shape
factor, which determines the selectivity characteristics, the minimum
and maximum resolution bandwidths of a spectrum analyzer play an
important role. For measurements requiring high sensitivity, very nar-
row bandwidths are needed (chapter 5.1), whereas for pulse measure-
ments and measurements in the time domain (chapters 6.2 and 6.3), very
large resolution bandwidths are necessary.

139
 Performance Features of S p e ct r u m A n a l y z e r s

To allow shorter sweep times, FFT filters are advantageous for narrow
resolution bandwidths. However, it is essential that there is also a choice
of analog or digital filters as it may not be possible, for instance, to carry
out pulse measurements with FFT filters (chapter 3.1).

The accuracy of the bandwidth is important for applications where a

measured signal level is referenced to the measurement bandwidth. The
accuracy is usually stated as a percentage. The method of calculating
the measurement accuracy is described in detail in chapter 5.10.

5.9 Frequency accuracy

The local oscillators in modern spectrum analyzers are synchronized to

a stable reference oscillator via phase-locked loops. The frequency accu-
racy of the spectrum analyzer thus corresponds to the accuracy of this
reference and is also influenced by the temperature and long-term sta-
bility of the reference.
References, usually at a frequency of 10 MHz, are implemented as
temperature-compensated crystal oscillators (TCXO) or oven-controlled
crystal oscillators (OCXO). The generated reference frequency depends
on the ambient temperature and varies due to aging during operation.
To ensure a high absolute frequency accuracy of the spectrum analyzer,
the reference frequency has to be adjusted at regular intervals. With
modern spectrum analyzers, the user can make this adjustment with a
D/A converter provided that a frequency counter or signal of known fre-
quency is available.

Internal reference frequency (nominal)

Aging per year 1) 1 · 10 –6
Temperature drift (+5 °C to 45 °C) 1 · 10 –6
with optional OCXO
Fig. 5-24   
Aging per year 1) 1 · 10 –7 Typical specifications of
frequency accuracy of
Temperature drift (+5 °C to 45 °C) 1 · 10 –8
spectrum analyzers
1) After 30 days of operation.

Fig. 5-24 shows a data sheet extract with frequency accuracy specifica-
tions for a spectrum analyzer. A distinction is made between the accu-
racy of the basic unit and the accuracy with built-in optional OCXO. It

140
 L e v e l M e a s u r e m e n t A cc u r a c y

can be seen that with an OCXO, a much higher temperature stability

and a much smaller temperature drift are achieved. The total frequency
error is made up of temperature drift and long-term stability. The long-
term stability is only effective, however, if the instrument remains perma-
nently switched on. If the instrument (or the OCXO) is switched off and
on again, a retrace takes place [5-4] whereby the oscillator frequency
assumes another value.

5.10 Level measurement accuracy

The measurement of signal levels always involves some uncertainty. In

the case of level measurements using spectrum analyzers, several com-
ponents contribute to this uncertainty. Spectrum analyzers are, there-
fore, factory-calibrated prior to their delivery by recording the individual
measurement errors and storing them in the analyzer as correction val-
ues. These measurement errors are considered in the displayed level so
that the accuracy is increased.
Since the analyzer characteristics are also subject to temperature
drift and aging, most analyzers feature an internal, temperature-stabi-
lized signal source (43) as well as self-adjustment functions allowing
critical measurement errors to be determined during operation and
appropriate corrections to be taken.

To ensure minimum measurement uncertainty, calibration at regular

intervals (usually at the manufacturer’s factory) is nevertheless required
since even the signal source in the analyzer used for the self-adjustment
is subject to certain aging, however small, and parameters such as fre-
quency response can only be checked with the aid of external measure-
ment equipment. A calibration interval for factory calibration is there-
fore recommended in spectrum analyzer data sheets.
The calibration by the manufacturer is also subject to measurement
uncertainties that are entered into the calibration results. Spectrum
analyzer data sheets specify maximum measurement errors. The indi-
vidual uncertainty components are explained in the following. System-
atic measurement errors due to insufficient signal-to-noise ratio are not
taken into account. These will be discussed separately and in detail at
the end of this chapter.

141
 Performance Features of S p e ct r u m A n a l y z e r s

5.10.1 Uncertainty components

Max. uncertainty of level measurement

at 128 MHz, - ­ 30 dBm  <0.2 dB (s = 0.07 dB)
(RF attenuation 10 dB, RBW
10 kHz, ref. level -20 dBm)
Frequency response
≤50 kHz <+0.5 / -1.0 dB
50 kHz to 3 GHz <0.5 dB (s = 0.17 dB)
3 GHz to 7 GHz <2.0 dB (s = 0.7 dB)
Frequency response with electronic attenuator
switched on
10 MHz to 3 GHz <1 dB (s = 0.33 dB)
3 GHz to 7 GHz <2.0 dB (s = 0.7 dB)
Attenuator <0.2 dB (s = 0.07 dB)
Reference level switching <0.2 dB (s = 0.07 dB)
Display nonlinearity LOG /LIN (S/N >16 dB)
RBW ≤100 kHz
0 dB to –70 dB <0.2 dB (s = 0.07 dB)
–70 dB to –90 dB <0.5 dB (s = 0.17 dB)
RBW ≥300 kHz
0 dB to –50 dB <0.2 dB (s = 0.07 dB)
–50 dB to –70 dB <0.5 dB (s = 0.17 dB)
Bandwidths switching uncertainty (relative to
RBW = 10 kHz) Fig. 5-25   
10 Hz to 100 kHz <0.1 dB (s = 0.03 dB) Data sheet specifications
for the level measure-
300 kHz to 10 MHz <0.2 dB (s = 0.07 dB)
ment accuracy of a
1 Hz to 3 kHz FFT <0.2 dB (s = 0.03 dB) spectrum analyzer

Uncertainty after self-adjustment at the reference frequency

The total gain of the analog signal processing of a spectrum analyzer
may vary due to temperature drift or aging. To correct the resulting level
error, a signal can be applied to the input of the analyzer (i. e. ahead of
the RF attenuator) as shown in the block diagram on the fold-out page.
If the level of this signal is known, the actual transfer constant of the
analog stages can be determined and the level error due to temperature
drift or aging can be compensated. As a prerequisite, the signal level
must be constant throughout the temperature range of the analyzer. The
stability of the built-in signal source used for the self-adjustment is a
determining factor for the absolute accuracy of the analyzer.

To ensure precise level measurements throughout the temperature

range, the self-adjustment function is called up after the warmup time

142
 L e v e l M e a s u r e m e n t A cc u r a c y

specified in the data sheet (e. g. 30 minutes). This function allows the
error to be determined and corrected during measurements.

The frequency of the signal used for the self-adjustment is usually con-
stant, i. e. the transfer constant of the signal processing stages includ-
ing the first mixer can only be determined at one frequency. The cor-
rection therefore is only valid at this frequency (128 MHz in our exam-
ple). The measurement uncertainty is increased by the magnitude of
the frequency response if the measurement is carried out at another
­frequency.
Since parameters such as IF bandwidth, RF attenuation, IF gain (set
via reference level) and linearity of the log amplifier also influence the
measurement uncertainty, they are included in the specification.
The above data sheet specifications are valid specifically at a level of
-30 dBm (corresponding to the level of the calibration source), 10 dB RF
attenuation, -20 dBm reference level and 10 kHz resolution bandwidth.

Frequency response
Measurement errors due to the frequency response are entered into the
total measurement uncertainty if level measurements are not carried
out at the frequency of the signal source used for self-adjustment (i. e.
128 MHz in our example).

For frequency ranges for which a YIG filter is switched in the signal path
of the analyzer (above 3 GHz, for example) additional conditions have
often to be met to attain the specified values. Due to their magnetic cir-
cuit, YIG filters exhibit hysteresis as well as temperature drift of the cen-
ter frequency. Therefore, it is not always possible to measure exactly at
the same point of the transfer function, i. e. the insertion loss of the filter
varies in the passband.
Spectrum analyzers therefore often use a peaking function. If a sig-
nal is applied to the analyzer input, this peaking function can be used
for fine adjustment of the center frequency to the maximum signal level
which results in higher level accuracy.
During this fine adjustment, the YIG filter is tuned in a very small fre-
quency range with relatively low speed. Due to the dynamic response of
YIG filters, measurements are again not carried out exactly at the point
determined by fine adjustment, particularly at very high sweep speeds.
At very short sweep times (<10 ms/GHz in our example), an additional
measurement error is therefore produced.

143
 Performance Features of S p e ct r u m A n a l y z e r s

Display nonlinearity
In the ideal case, a variation of the input level by n dB causes a variation
of n dB in the displayed level. The display nonlinearity specifies the max-
imum deviation from the expected variation in the displayed level.
With logarithmic level display the log amplifier primarily determines
this nonlinearity. Often the maximum nonlinearity is specified for a spe-
cific level range relative to the reference level. (Fig. 5-25, e. g. <0.2 dB for
displayed levels that are maximally 70 dB below the reference level with
resolution bandwidths ≤100 kHz). If the level varies within this range,
the displayed level differs from the actual value within the ­specified
error.
It is customary to specify the maximum total nonlinearity as a func-
tion of the displayed level relative to the reference level.
For example, the maximum nonlinearity for the level range 0 dB
to 70 dB below the reference level is 0.3 dB + 0.01 ⋅aR (aR = offset
from reference level). Accordingly, the maximum display nonlinearity
for a signal with a display level 70 dB (aR) below the reference level is
0.3 dB + 0.01 ⋅ 70 dB = 1 dB.
The nonlinearity specified in this way is especially relevant for abso-
lute level measurements. For relative level measurements, the devia-
tion of the displayed level variation from the expected level variation
is of interest and has to be specified as an arithmetic sum. It is usually
stated as maximum display nonlinearity for a specific level variation –
for example, 0.4 dB/4 dB (0.4 dB deviation for a level variation of 4 dB).
For linear display mode, the nonlinearity is stated as a percentage
relative to the reference level.

Attenuator uncertainty
Attenuator settings are subject to uncertainties. In modern instruments,
these uncertainties are determined during the self-adjustment procedure
and used as a correction value in the displayed level. The value specified
for the attenuator uncertainty is regarded as a residual deviation due to
long-term effects such as drift due to temperature v­ ariations.

IF gain uncertainty or uncertainty of reference level setting

Similar to the attenuator setting, the IF gain setting is also subject to
uncertainties. Since the IF gain can be set only indirectly via the refer-
ence level, the uncertainty is also referred to as the uncertainty of the
reference level setting. In addition to the specification of the maximum

144
 L e v e l M e a s u r e m e n t A cc u r a c y

uncertainty, as for instance in Fig. 5-25, the uncertainty is often specified

as a function of the set reference level.

For example, the maximum reference level deviation for a reference

level of -20 dBm is 0.3 dB. For other reference levels the deviation is
0.3 dB + 0.01 ⋅ aR (aR = deviation from a reference level of -20 dBm). If,
for instance, the reference level is set to +10 dBm, the maximum refer-
ence level deviation is 0.3 dB + 0.01 ⋅  (+10 dBm - (-20 dBm)) = 0.6 dB.

Bandwidth switching uncertainty

When switching between different resolution bandwidths, level devia-
tions occur that have to be taken into account. This deviation can be
determined during the self-adjustment procedure and compensated by
a correction factor. The specified uncertainty corresponds to the resid-
ual uncertainty due to long-term effects such as temperature drift.

Effect of bandwidth uncertainties

Bandwidth uncertainties are the deviation of the actual from the set
resolution bandwidth. A specified bandwidth uncertainty of 5 % means
that with a set resolution bandwidth of 10 kHz, for example, the actual
bandwidth may be anywhere between 9.5 kHz and 10.5 kHz. This uncer-
tainty is only significant for applications in which the measured power
has to be referenced to the measurement bandwidth or the measure-
ment bandwidth must be known for further calculations. This is the case
with (phase) noise measurements or channel power measurements (see
chapter 6.3).
The resulting level uncertainty can in turn be calculated from the
bandwidth uncertainty in percent. The following holds true for noise or
noise-like signals:
 DBN 
DLBN = 10 dB ⋅ lg 1 -   (Equation 5-43)
 100 % 

where DLBN level uncertainty due to bandwidth uncertainty

DBN bandwidth uncertainty

Measurement uncertainty due to mismatch

An ideal spectrum analyzer with an input reflection coefficient of zero
would absorb the applied input power completely irrespective of the out-
put impedance of the signal source.

145
 Performance Features of S p e ct r u m A n a l y z e r s

However, the reflection coefficient at the input of a real spectrum ana-

lyzers is > 0, so that there is a mismatch. The measurement result, there-
fore, also depends on the output reflection coefficient of the source
which again is typically > 0. The measurement uncertainty MU due to
mismatch is determined by:

( )
2
M U = 1 + rs ⋅ rl - 1  (Equation 5-44)

where MU measurement uncertainty

rs magnitude of source reflection coefficient
rl magnitude of spectrum analyzer reflection c­ oefficient

The following approximation applies:

M U ≈ 2 ⋅ rs ⋅ rl  (Equation 5-45)

For spectrum analyzers, level uncertainties are stated in dB. Equation

5-44 can be reformulated as follows:

DLr = 20 dB ⋅ lg (1 - rs ⋅ rl )  (Equation 5-46)

where DLr level measurement uncertainty due to mismatch

The input matching of an analyzer or output matching of a DUT is often

stated as voltage standing wave ratio (VSWR) or return loss. Based on
such data, the corresponding reflection coefficients can be calculated
as follows:

s -1
r =  (Equation 5-47)
s +1

where r reflection coefficient

s VSWR

and
ar
-
20 dB
r = 10

where ar return loss

146
 L e v e l M e a s u r e m e n t A cc u r a c y

Substituting Equation 5-47 in Equation 5-46 yields

 s - 1 s l - 1 
DLr = 20 dB ⋅ lg 1 - s ⋅   (Equation 5-48)
 s s + 1 s l + 1

  Improvement of input matching

The RF attenuation of a spectrum analyzer should always be set
to at least 10 dB provided that the sensitivity is sufficiently high.
In this way the first mixer is protected against damage by too
high input signals and the input matching is improved. For exam-
ple, if an ideal attenuator pad with an attenuation a = 6 dB is
connected ahead of a twoport with a return loss of ar = 10 dB at
the input, the total return loss ar, total is ar + 2 · a or 22 dB. Fig. 5-26
shows the spectrum analyzer with the attenuator pad.
ar, total ar
Fig. 5-26   
Improvement of spectrum
analyzer input matching by
a
connecting an attenuator
pad ahead of analyzer Attenuator pad Spectrum
analyzer

The return loss of real attenuator pads, including the attenuator

of a spectrum analyzer, is limited, so the theoretical values can-
not be attained under certain conditions. The input matching of
an attenuator is usually much better than that of the broadband
first mixer. Especially for measurements on DUTs with poor out-
put matching, the level measurement accuracy can be dramati-
cally increased with an attenuator setting of ≥10 dB.
In spectrum analyzers, the RF attenuation can usually be cou-
pled to the reference level. In this coupled mode, the minimum RF
attenuation, 10 dB, is therefore set for the above reason even for
very low reference levels.

147
 Performance Features of S p e ct r u m A n a l y z e r s

5.10.2 Calculation of total measurement uncertainty

The factors that affect total measurement uncertainty depend on the

type of measurement. In the following sections, uncertainty sources
encountered in frequent measurement applications are described.

Measurement of absolute level

If the absolute level of a sinusoidal signal is to be measured, the follow-
ing factors usually contribute to the total measurement uncertainty:
❙ Reference frequency uncertainty
❙ Frequency response
(only if the signal frequency distinctly differs from the frequency of
the internal calibration source)
❙ Attenuator
(only if the attenuator setting deviates from that specified in the data
sheet)
❙ IF gain
(only if the set reference level deviatess from that specified in the
data sheet)
❙ Linearity
The display nonlinearity to be taken into account depends on the
spacing of the input signal from the reference level.
❙ Bandwidth switching
(only if the set bandwidth deviates from that specified in the data
sheet)

An additional bandwidth uncertainty has to be taken into account in

noise or channel power measurements.

Relative level measurement

The following uncertainty components have to be taken into account
when measuring the level difference between two sinusoidal signals:

148
 L e v e l M e a s u r e m e n t A cc u r a c y

❙ Frequency response
(only if the signal frequency strongly varies between the individual
measurements)
❙ Attenuator
If the attenuator setting is not varied during measurement, this com-
ponent can be ignored.
❙ IF gain
If the reference level is not varied during measurement, this compo-
nent can be ignored.
❙ Linearity
❙ Bandwidth switching
If the bandwidth is not varied during measurement, this component
can be ignored.

An additional bandwidth uncertainty has to be taken into account in

noise or channel power measurements if the resolution bandwidth is
varied between the measurements.
Resolution bandwidth, attenuator setting (RF attenuation) and refer-
ence level should not be varied during the measurement to minimize the
uncertainty of relative level measurement. Only the nonlinearity uncer-
tainties and, if applicable, frequency response will then contribute to the
total measurement uncertainty.

Table 5-2 shows the uncertainty components that have to be consid-

ered in typical measurements. The maximum deviation (worst case)
can be calculated from the individual components simply by adding
the relevant components. The calculated maximum uncertainty has a
confidence level of 100 %, i. e. the actual measurement uncertainty never
exceeds the calculated error limits.

149
 Measurement Absolute Harmonic 3rd order inter- 3rd order Channel Adjacent­ Power versus Phase noise, Phase
level distortion modulation intercept power channel time (e. g. far off carrier, noise,
of CW products (close power for TDMA with variation close to
signal to carrier) ratio signals), of RF attenua- carrier
Uncertainty ­ relative tion and refer-
component ence level
Reference frequency • • •
uncertainty
Frequency response • • • •

Attenuator uncertainty • • • •
Performance Features

IF gain uncertainty • • • •

150
of

Nonlinearity • • • • • • • • •

Bandwidth switching • • •
uncertainty

Bandwidth uncertainty • • • •
S p e ct r u m A n a l y z e r s

Uncertainty due to limi- • •

ted number of samples

Mismatch • • • •

Table 5-2  Uncertainty components in typical measurements using a spectrum analyzer

 L e v e l M e a s u r e m e n t A cc u r a c y

In practice, however, the maximum uncertainty does not occur. If the

total uncertainty is the sum of many individual uncertainties stemming
from completely independent sources, it is statistically a very rare event
for all individual uncertainties to occur in a measurement at the same
time with their maximum value and same sign.

It is more realistic to calculate the total uncertainty with a certain confi-

dence level, which is typically 95 % or 99 %.
The extent to which the individual uncertainty components affect
the calculation depends on their probability distribution. The following
discussions are based on [5-5].

For random uncertainties, that is for all uncertainties listed above with
the exception of mismatch, a rectangular distribution is assumed. The
variance s2 of the individual uncertainties is determined by:

a2
s2 =  (Equation 5-49)
3
where s2 variance
a max. measurement uncertainty (error limit); unit: dB

If the data sheet specification for the level measurement uncertainty is

not defined as worst case but with a specific confidence level, the vari-
ance has to be calculated from this value first. The following applies:
2
a 

s =  CL  
2
(Equation 5-50)
 k 

where s2 variance
aCL specified measurement uncertainty with defined
­confidence level (unit: dB)
k coverage factor

The value of k depends on the confidence level of the value specified in

the data sheet. The following applies:
 CL 
  
k = 2 ⋅ erfinv  (Equation 5-51)
 100% 

where erfinv inverse error function

CL confidence level

151
 Performance Features of S p e ct r u m A n a l y z e r s

Fig. 5-27 shows k as a function of the confidence level. For a confidence

level of 95 %, k assumes a value of 1.96, and for 99 %, a value of 2.58.
4 4

3.5
Factor k

Factor k
3
3
2 k = 2.58
2.5
Confidence level 68% k = 1.96
1 2

0 1.5
0 20 40 60 80 100 90 92 94 96 98 100
Confidence level in % Confidence level in %
a) b)
Fig. 5-27  Coverage factor k as a function of confidence level

In some cases the standard deviation s is specified in addition to the

level error limit. This makes the calculation defined by equation 5-50
unnecessary. The variance can be calculated from the specified standard
deviation by squaring it.

Bandwidth uncertainties are usually specified as a percentage. The fol-

lowing applies:
2
 
10 dB ⋅ lg 1 + DBN 

2

  100 % 
s =  (Equation 5-52)
3
where s2 variance
DBN bandwidth uncertainty

Uncertainties due to mismatch have a U-shape distribution. The vari-

ance s2 is determined by:
2
 
20 dB ⋅ lg 1 - s s - 1 ⋅ s l - 1 
   

2
(20 dB ⋅ lg (1 - rs ⋅ rl ))  s s + 1 s l + 1
 s 2 = =
2 2
 
 (Equation 5-53)

where s2 variance
rs magnitude of source reflection coefficient
rl magnitude of spectrum analyzer reflection c­ oefficient

152
 L e v e l M e a s u r e m e n t A cc u r a c y

The reflection coefficient can be calculated from Equation 5-47.

Error limits Calculation of variance

(max. uncertainties)
Max. level uncertainty at
the reference frequency
Frequency response
 a 
2 Equation 5-49
Attenuator uncertainty a 2 and s 2 =  CL 
s2 =    and
IF gain uncertainty
 3  k  Equation 5-50
Nonlinearity
Bandwidth switching
uncertainty
Bandwidth 2
 
10 dB ⋅ lg 1 + DBN 
   
 2  100 %  Equation 5-52
s =
3
Mismatch
2
2 (20 dB ⋅ lg (1 - rs ⋅ rl ))  Equation 5-53
s =
2

Table 5-3  Calculating the variances of the individual uncertainty components

The total standard deviation stot can be calculated from the variances si2
of the individual uncertainty components as follows:

stot = s12 + s22 + … + sn2  (Equation 5-54)

The measurement uncertainty has a confidence level of 68% (Fig. 5-27a).

To obtain the measurement uncertainty at some other confidence level,
stot has to be multiplied by the coverage factor k which can be derived
from Fig. 5.27. For a confidence level of 95 %, k = 1.96 and for 99%
k = 2.58 is obtained again.

Example:
For the absolute level measurement of a sinusoidal input signal of 1 GHz
(VSWR of signal source 1.2 : 1), the total measurement uncertainty is to
be determined with a confidence level of 95 %. The resolution bandwidth
set on the spectrum analyzer is 30 kHz, the RF attenuation is 20 dB and
the reference level 0 dBm. The signal level is about 20 dB below the ref-
erence level.

153
 Performance Features of S p e ct r u m A n a l y z e r s

Which components contribute to the total measurement uncertainty?

❙ Max. level uncertainty at the reference frequency

❙ Frequency response
❙ Attenuator uncertainty
❙ IF gain uncertainty
❙ Nonlinearity
❙ Bandwidth switching uncertainty

Since the input is a sinusoidal signal, the bandwidth uncertainty does

not affect the total measurement uncertainty.

The required data is taken from the spectrum analyzer data sheet:

Specified max. Variance

measurement (si/dB)2
uncertainty
(error limit)

Max. level uncertainty at the reference frequency 0.2 dB 0.0133

Frequency response 0.5 dB 0.0833

Attenuator uncertainty 0.2 dB 0.0133

IF gain uncertainty 0.2 dB 0.0133

Nonlinearity 0.2 dB 0.0133

Bandwidth switching uncertainty 0.1 dB 0.0033

Mismatch

VSWR at spectrum analyzer input 1.5

VSWR at signal source output 1.2 0.0127

The total standard deviation can be calculated from the variances s 2i

with the aid of Equation 5-54 to yield s tot = 0.39. The total measurement
uncertainty of 0.76 dB is obtained at a confidence level of 95 % by multi-
plying the standard deviation by a factor of 1.96.

To simplify calculations of measurement uncertainty, a spreadsheet in

MS Excel® is available (file FSP_ERR.XLS, see Fig. 5-28) which can be
obtained from the Rohde &Schwarz website (www.rohde-schwarz.com).

154
 L e v e l M e a s u r e m e n t A cc u r a c y

❶ ❷ ❸ ❹

❻ ❼

❽
Fig. 5-28  Spreadsheet FSP_ERR.XLS

Legend for spreadsheet FSP_ERR.XLS

All entry fields in the table are highlighted in yellow. The fields for inter-
mediate results as well as for the resulting total level error uncertainties
are highlighted in light and dark blue.

❶ You can choose whether the values entered in ❷ are error limits
(worst case) or a standard uncertainty.
❷ Input fields for specified values.
❸ Output of the variances calculated from the input values.
❹ You can choose whether the value entered under ❷ is to be taken
into account in calculating the total uncertainty. Uncertainty compo-
nents can thus very easily be ignored without having to set the value
entered under ❷ to zero.
❺ Output of the calculated total uncertainty with a confidence level of
95% or 99%. Uncertainties due to mismatch are not yet considered in
this result.
❻ You can choose whether the mismatch of the DUT or spectrum ana-
lyzer is entered as VSWR (v) or as return loss (a).
❼ Input field for specified maximum mismatch of the DUT or spectrum
analyzer.
❽ Output of the calculated total measurement uncertainty with a con-
fidence level of 95  % or 99 %. All uncertainty components are consid-
ered in the result.

155
 Performance Features of S p e ct r u m A n a l y z e r s

5.10.3 Measurement error due to low signal-to-noise ratio

The signals displayed on a spectrum analyzer are the sum of the input
signal (S) and the superimposed thermal noise (N). If the signal level
is high relative to the noise, this has no adverse effect since the result-
ing level variation due to the superimposed noise relative to the mea-
surement level is low. Level measurements on signals with low signal-
to-noise ratio, however, produce measurement errors that are not neg-
ligible.
The measurement error can be corrected if the ratio between total
signal plus noise power and inherent noise (PS + PN)/PN is known. The
thermal noise power without input signal is determined first at the mea-
surement frequency. With the same spectrum analyzer settings, the level
of the input signal including the superimposed noise is then measured
and the ratio (PS + PN)/PN calculated. As described in the following, a
correction factor can be found and subtracted from the measured total
power level (LS + N) to obtain the true signal level (LS). Both the type of
input signal and detector used have to be taken into account.
To increase the measurement accuracy, it is necessary to smooth the
trace by employing appropriate methods.

RMS detector
When using the RMS detector, the effective value of power is measured
in both measurements of the thermal noise power and the input signal
level with superimposed noise. As already explained in connection with
the sensitivity limit (chapter 5.1), an input signal with a level correspond-
ing to the thermal noise level causes a noise increase by 3 dB. Accord-
ingly, the correction value is 3 dB. The following applies:

 PS + PN 
 
 PN   PS + PN 
cN = 10 dB ⋅ lg   = 10 dB ⋅ lg  
 SP + P   PS 
 P
N
- 1 
N

 (Equation 5-55)

where cN correction factor

(PS + PN)/PN measured ratio between total signal power and
noise power

156
 L e v e l M e a s u r e m e n t A cc u r a c y

10
Correction factor cN / dB

0
0 2 4 6 8 10 12 14 16
Total power / inherent noise / dB
Fig. 5-29  C
 orrection factor cN as a function of total power/inherent noise
(PS + PN )/PN for measurement of noise or noise-like signals using RMS
detector

Example:
The channel power of a digitally modulated signal at a low level is to
be measured and the result corrected. A special measurement function
of the spectrum analyzer is used to determine the channel power. The
channel bandwidth is 4.096 MHz.
Because of the noise-like signal, the RMS detector is used for the
measurement.

157
 Performance Features of S p e ct r u m A n a l y z e r s

a)

*RBW 30 kHz Marker1 [T1]

*VBW 300 kHz –102.79 dBm
Ref –80 dBm *Att 0 dB *SWT 10 s 2.20000000 GHz
–80
–90 1
–100 A
–110
1RM*
–120
CLRWR
–130
–140
–150
–160 C0
–170 C0 PRN
–180
Center 2.2 GHz 1 MHz/ Span 10 MHz
Tx Channel
Power –81.95 dBm Bandwidth 4.096 MHz
Reference 0.00 dBm

b)

*RBW 30 kHz Marker1 [T1]

*VBW 300 kHz –107.36 dBm
Ref –80 dBm *Att 0 dB *SWT 10 s 2.20000000 GHz
–80
–90
–100 1 A
–110
1RM*
–120
CLRWR
–130
–140
–150
–160 C0
–170 C0 PRN
–180
Center 2.2 GHz 1 MHz/ Span 10 MHz
Tx Channel
Power –86.08 dBm Bandwidth 4.096 MHz
Reference 0.00 dBm

Fig. 5-30  C
 hannel power measurement. Measurement of total power (a) and
inherent noise (b)

Step 1: measurement of total power level, relative to 1 mW

LS+N(mW) = - 81.95 dB  (Fig. 5-30 a)

Step 2: measurement of inherent noise level in specified channel,

­relative to 1 mW

LN(mW) = - 86.08 dB  (Fig. 5-30 b)

158
 Sweep Time and Update Rate

Step 3: calculation of difference between total power level and inher-

ent noise level

LS + N(mW) - LN(mW) = - 81.95 dB - (-86.08 dB) = 4.13 dB

Step 4: determination of correction factor (from Equation 5-55 or Fig. 5-29)

c N = 2.12 dB

Step 5: calculation of signal level from total power level

LS(mW) = LS+ N(mW) - c N = - 81.95 dB - 2.12 dB = - 84.07 dBm

Sample detector
As already explained in chapter 4.4 Detectors, the displayed level of
noise and noise-like signals is too low if the sample detector is used and
averaging over a logarithmic level scale. However, sinusoidal signals are
not affected. The correction factor to be applied therefore depends on
the type of the input signal.
If measurements are carried out on noise signals, the displayed level
of both the input signal and the inherent noise is 2.5 dB too low. The
resulting ratio between total power and inherent noise corresponds to
the result that would be obtained with the RMS detector. The correction
factor can be calculated in accordance with Equation 5-55 or derived
from Fig. 5-29.
If the input signal is a discrete spectral line, for instance a sinusoi-
dal signal, the measured level is not falsified by the sample detector and
averaging over a logarithmic level scale. Since a lower level is displayed
for the inherent noise, the ratio between total power and inherent noise
is greater than when using the RMS detector. Correction factors calcu-
lated in accordance with Equation 5-55 are not valid. It is recommended
that the RMS detector is used in such cases.

5.11 Sweep time and update rate

The minimum sweep time, that is the minimum time required for record-
ing a certain frequency span, is determined by various factors.
❙ Resolution and video bandwidths
❙ Setting time of LO

159
 Performance Features of S p e ct r u m A n a l y z e r s

❙ Data processing
❙ Sampling rate of A/D converter
❙ Maximum sweep speed of YIG filter, if applicable

The dependency of the sweep time on the resolution and video band-
widths as well as on the span was described in chapter 4.6: Parame-
ter dependencies. As described, the required minimum sweep time
increases with decreasing resolution bandwidth so that for such cases
the use of FFT filters is recommended provided their use is permitted by
the specific measurement application.
But even at very large resolution and video bandwidths, the sweep
time cannot be reduced without any limitation. For setting the local oscil-
lator and collecting measured data, a certain minimum time is always
required which in turn depends on the set span, so that there is a limit
(2.5 ms in our example) that cannot be reduced even under the most
favorable circumstances.
In the case of large spans, the minimum sweep time is addition-
ally influenced by the permissible tuning speed of the local oscillator.
For the analyzer described here, a sweep time of 5 ms, for instance, is
required for a span of 1 GHz. In concepts using tracking YIG filters for
image frequency rejection (above 3 GHz for the analyzer described here),
the sweep speed is further reduced by the ‘inertia’ of the tunable mag-
netic circuit. Sweep times of less than 6 ms at 1 GHz span can hardly be
attained in such cases.
Data sheets usually specify the minimum sweep time that is achiev-
able under the most favorable conditions, such as large resolution and
video bandwidth and small span in a frequency range for which the
tracking YIG filter is not required. In our example, 2.5 ms sweep time can
be achieved under these circumstances.
If the spectrum analyzer remains tuned to a fixed frequency during
the measurements, which is referred to as zero span, the minimum mea-
surement time only depends on the data acquisition of the analyzer. The
minimum measurement times achievable in this mode are very short
(1 µs, in our example).
An essential criterion in this mode is the time resolution. That is, the
minimum time between two samples (125 ns in our example). The sam-
pling rate of the A/D converter is the limiting parameter in this case.

The update rate (number of sweeps per unit time), is again important
both for manual and remote-controlled operation of the spectrum ana-

160
 Sweep Time and Update Rate

lyzer. Additional time is required for data processing, display and, where
applicable, data transfer via the IEEE bus or other interfaces so that the
maximum update rate is considerably lower than the reciprocal value of
the achievable minimum sweep time (Fig. 5-31).
If FFT filters are used, the difference is especially high due to the
very complex calculations.

Manual operation

Sweep time Data processing Display update

Remote-controlled operation
Display on
Display update and
Sweep time Data processing data output via IEEE bus

Display off
Data output
Sweep time Data processing via IEEE bus

Fig. 5-31  Sequence of a measurement

A high screen update rate is desirable in the manual mode, such as for
tuning operations so that changes in the recorded spectrum are dis-
played almost immediately. Update rates of about 20 measurements per
second are sufficient for such applications as they yield almost flicker-
free display. For automated measurements, such as in production, where
control commands and measurement results are transferred via inter-
faces like the IEEE bus, the update rate can never be high enough to
achieve minimum test times and high production throughput.
As shown in Fig. 5-31, a certain time for displaying the results on
the screen is required in remote-controlled mode. To achieve maximum
update rates it is therefore advisable to deactivate the screen display.

161
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

6 Frequent Measurements and Enhanced Functionality

6.1 Phase noise measurements

As described in chapter 5.3, the phase noise of an oscillator is a measure

of the oscillator’s short-term stability and hence an essential quality cri-
terion. Therefore, special phase noise measurement equipment or, when
requirements on the dynamic range are not stringent, spectrum analyz-
ers are used for phase noise measurement.

Phase noise measurement with a spectrum analyzer is referred to as a

direct measurement. As a prerequisite, the DUT must have a small fre-
quency drift relative to the sweep time of the spectrum analyzer or else
the measurable frequency variation of the oscillator would be too large
and invalidate the measurement results. Spectrum analyzers are there-
fore mainly suitable for measurements on synthesized signal sources
locked to a stable reference rather than for measurements on free-run-
ning oscillators.

6.1.1 Measurement procedure

For oscillators, the SSB phase noise is usually specified at a certain car-
rier offset relative to the carrier level within 1 Hz bandwidth (Fig. 6-1).
Accordingly, the unit is dBc (1 Hz).

L Oscillator signal
LT

1 Hz
LPN
Fig. 6-1  Definition of
SSB phase noise
fT f
foff

162
 Phase Noise Measurements

Phase noise measurement with a spectrum analyzer requires two

steps:
❙ Measurement of carrier level LC
❙ Measurement of phase noise level LPN at carrier offset foff

For evaluation, the phase noise measured at a carrier offset foff and res-
olution bandwidth BIF is first referenced to a 1 Hz bandwidth. The fol-
lowing applies when using the RMS detector:
 BN, IF 
LPN, foff (mW) = LPNM, foff (mW) - 10 dB ⋅ lg    (Equation 6-1)
 Hz 

where LPN, f phase noise level at carrier offset foff and noise
off
bandwidth BN, IF , relative to 1 mW and 1 Hz band-
width
LPNM, f phase noise level measured with RMS detector at
off
carrier offset foff and noise bandwidth BN, IF , rela-
tive to 1 mW
BN, IF noise bandwidth of resolution filter

Depending on the filter implementation, the noise bandwidth of the res-

olution filter can be calculated from the 3 dB filter bandwidth with the
aid of the conversion factors provided in Table 4-1. If the sample detec-
tor is used instead of the RMS detector and the trace averaged over a
narrow video bandwidth or over several measurements, the noise will
be underweighted as described in chapter 4.4: Detectors. The following
then applies:
 BN, IF 
LPN, foff (mW) = LPNM, foff (mW) - 10 dB ⋅ lg   + 2.5 dB  
   Hz 
 (Equation 6-2)

where LPNM, f averaged phase noise level measured with sample

off
detector at carrier offset foff and noise bandwidth
BN, IF , relative to 1 mW

The phase noise level within 1 Hz bandwidth must now be referenced to
the carrier level:

163
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

off
( )
Lf Pc = Lc(mW) - LPN,f (mW) 
off
(Equation 6-3)

where Lf relative phase noise level within 1 Hz bandwidth at

off
carrier offset foff , relative to carrier power Pc
LPN,f phase noise level within 1 Hz bandwidth at carrier off-
off
set foff , relative to 1 mW
Lc carrier level, relative to 1 mW

To simplify phase noise measurements, most spectrum analyzers fea-

ture marker functions which allow direct readout of the phase noise at
a specific carrier offset. Usually, noise bandwidth and correction factors,
which are necessary due to the underweighting of noise signals when
using the sample detector, are already taken into account.

Fig. 6-2  Marker function for easy phase noise measurement

With such marker functions the phase noise can only be determined at
a certain carrier offset. However, the phase noise is often of interest in a
wider range (such as 1 kHz to 1 MHz carrier offset). To make these mea-
surements simple, application software is available for some spectrum
analyzers. Fig. 6-3 shows the result of a phase noise measurement using
this kind of software.

164
 Phase Noise Measurements

Fig. 6-3  Phase noise measurement over wide offset range

6.1.2 Selection of resolution bandwidth

When measuring the phase noise at a certain carrier offset, care should
be taken that the selected resolution bandwidth be appropriately small.
If the resolution bandwidth is too large, the carrier at the offset foff will
not be sufficiently suppressed by the IF filter (Fig. 6-4a). The level of the
residual carrier at the input of the envelope detector or A/D converter
will be greater than the phase noise and thus the measurement result
will be falsified. The phase noise will have an apparently higher value
than it should (Fig. 6-4b). The maximum permissible resolution band-
width depends on the carrier offset and skirt selectivity (the shape factor
of the IF filter). A generally valid relationship principle, therefore, can-
not be established.

165
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

a)

*RBW 30 kHz Delta 2 [T1 PHN]

VBW 300 kHz –112.76 dBc/Hz
Ref –10 dBm *Att 0 dB *SWT 5 s 100.00000000 kHz
1
–10 PHN –12.292 dBm
Marker 1 [T1 FXD]
–12.25 dBm A
–20
1 RM* 497.00000000 MHz
CLRWR
–30

–40
PRN
–50

–60

–70
2
–80

–90

–100
PHN
–110
Center 497.06 MHz 20 kHz/ Span 200 kHz

b)

*RBW 30 kHz Delta 2 [T1 PHN]

VBW 30 kHz –116.29 dBc/Hz
Ref –10 dBm *Att 0 dB *SWT 5 s 100.00000000 kHz
1
–10
PHN –12.292 dBm
Marker 1 [T1 FXD]
–12.25 dBm A
–20
1 RM* 497.00000000 MHz
CLRWR
–30

–40
PRN
–50

–60

–70

–80

–90 2

–100
PHN
–110
Center 497.06 MHz 20 kHz/ Span 200 kHz

Fig. 6-4  C
 hoosing the right resolution bandwidth: (a) resolution bandwidth too
large, carrier suppression is insufficient; (b) resolution bandwidth is
small enough

166
 Phase Noise Measurements

Due to the high minimum sweep times at narrow IF bandwidths, high

resolution bandwidths are desirable in practice. With a wideband IF fil-
ter, the resolution bandwidth should be reduced in steps until the mea-
sured phase noise values no longer decrease.

6.1.3 Dynamic range

The phase noise of the local oscillators is transferred to the converted

input signal by reciprocal mixing in the converting stages of the spec-
trum analyzer (chapter 5.3: Phase noise). If the input signal is of suffi-
ciently high level so that the effects of thermal noise of the spectrum
analyzer are negligible, the achievable dynamic range at small carrier
offsets is solely determined by the LO phase noise of the analyzer. The
limitation imposed by the system-inherent phase noise is independent
of the input signal level.
Since it is always the sum of the DUT phase noise and LO phase
noise of the spectrum analyzer that is measured, such measurement can
only be carried out on DUTs with relatively high phase noise.

As shown in Fig. 5-11, the system-inherent phase noise of spectrum ana-

lyzers decreases with increasing carrier offset. At large carrier offsets
the dynamic range is limited to an increasing extent by the thermal
noise of the spectrum analyzer. There is no clear-cut difference between
limitation due to system-inherent phase noise and thermal noise.
To minimize the limitation caused by thermal noise, a high signal-
to-noise ratio is required via a high signal level at the input of the first
mixer. A high 1 dB compression point is also important to achieve a wide
dynamic range far off the carrier.

Due to the high signal level, harmonics of the input signal are produced
in the first mixer. If the maximum offset up to which phase noise is to
be examined is smaller than the input frequency, the harmonics are out-
side the frequency range of interest and do not cause any disturbance.
If the input signal level is greater than the dynamic range of the
spectrum analyzer, it has to be reduced by setting a suitable RF attenua-
tion. Due to the attenuator step sizes, the maximum dynamic range may
not be fully utilized.

167
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

Example:
The 1 dB compression point of the spectrum analyzer is assumed to be
+10 dBm (mixer level). To avoid measurement errors, the signal level at
the first mixer should not exceed +5 dBm. The RF attenuation can be
set in 10 dB steps.
A signal level of +17 dBm is applied to the analyzer input, so an RF
attenuation of at least 20 dB is required. The signal level at the first
mixer is then -3 dBm. The dynamic range for measurements at large
carrier offsets is thus 8 dB lower than the maximum achievable dynamic
range.
To utilize the maximum dynamic range, the signal level in our exam-
ple would have to be attenuated to +15 dBm using an external 2 dB
attenuator pad. With 10 dB RF attenuation, a mixer level of +5 dBm is
then obtained.
To avoid distortion products due to overdriving, the maximum ref-
erence level settable at an RF attenuation of 0 dB is clearly below the
upper limit of the analyzer’s dynamic range (+5 dBm in the above exam-
ple). With the analyzer driven to its maximum input, measurement of the
carrier level as a subsequent reference for the phase noise would not
be possible. Likewise, measurement of phase noise close to the carrier
would not be possible. Phase noise is therefore measured in two steps:

1.  Measurement of carrier level and phase noise close to carrier

The RF attenuation of the spectrum analyzer is increased until the refer-
ence level is equal to the signal level (Fig. 6-5). The carrier level can then
be readily measured with the aid of markers, since the input signal does
not exceed the reference level. To be able to refer the measured phase
noise to the carrier level, the latter is stored. The carrier level is usually
stored automatically when activating the marker function for the phase
noise measurement.

168
 Phase Noise Measurements

*RBW 1 kHz Delta 2 [T1 PHN]

VBW 10 kHz –109.36 dBc/Hz
Ref 0 dBm Att 30 dB 1 *SWT 10 s 20.00000000 kHz
0 PHN –0.305 dBm
Marker 1 [T1 FXD]
–0.22 dBm A
–10
1 RM* 497.00020000 MHz
CLRWR
–20

–30
PRN
–40

–50

–60

–70
2
–80

–90
PHN
–100
Start 496.98 MHz 5 kHz/ Stop 497.03 MHz

Fig. 6-5  S
 etting RF attenuation and reference level for measurement of carrier
level and phase noise close to carrier

2.  Measurement of phase noise far off carrier

The RF attenuation is reduced until the maximum dynamic range limit
is attained with the signal applied to the input of the first mixer. The fol-
lowing applies:

a RF,min = Lin(mW) - Lmax(mW)  (Equation 6-4)

where aRF, min required minimum RF attenuation

Lin signal level at spectrum analyzer input,
relative to 1 mW
Lmax maximum dynamic range limit of analyzer,
relative to 1 mW

169
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

An overload detector before the first mixer is useful for determining

the minimum RF attenuation. The RF attenuation can then be increased
until overdriving is no longer detected.
If digital filters are used, deliberately exceeding the reference level
may result in overdriving of the A/D converter, which in turn produces
unwanted products in the displayed spectrum (chapter 4.6.3: Overdriv-
ing). This constraint is not significant, since measurements at large car-
rier offsets use relatively wideband analog resolution filters.
If digital filters are used and the input signal exceeds the reference
level, it has to be ensured that the carrier frequency is far from the
displayed spectrum. The carrier is then suppressed by the analog anti-
aliasing filter ahead of the A/D converter so that it cannot cause over­
driving.

It is very easy to determine whether the noise displayed is the phase

noise or the thermal noise of the analyzer. To perform this test, the
recorded trace has to be stored and a second measurement carried out
with the same settings, but without input signal at the analyzer. If at the
carrier offset of interest there is a clear level difference between the two
traces as shown in Fig. 6-6a, the measurement result is barely affected
by the thermal noise of the analyzer.
In the measurement shown in Fig. 6-6b, the dynamic range is limited
by the thermal noise and the result falsified.

The displayed phase noise is always the sum of the phase noise of the
DUT and the spectrum analyzer and the spectrum analyzer’s thermal
noise. If measurements are carried out very close to the dynamic range
limit, errors may be produced due to insufficient spacing between the
measured phase noise and the system-inherent noise. If the system-
inherent noise is known, a correction can be made in accordance with
Equation 5-55 (chapter 5.10.3).

170
 Phase Noise Measurements

a)

*RBW 100 kHz Delta 2 [T1 PHN]

VBW 1 MHz –144.47 dBc/Hz
Ref –10 dBm *Att 0 dB *SWT 10 s 3.00000000 kHz
–10
PHN –1.286 dBm
Marker 1 [T1 FXD]
–95.21 dBm A
–20
1RM* 496.99400000 MHz
CLRWR
–30

2RM*
–40
VIEW
PRN
–50

–60

–70

–80

–90 1 2

–100
PHN
–110
Start 496 MHz 700 kHz/ Stop 503 MHz

b)
*RBW 100 kHz Delta 2 [T1 PHN]
VBW 1 MHz –130.27 dBc/Hz
Ref –10 dBm *Att 20 dB *SWT 10 s 3.00000000 MHz
–10
PHN –1.286 dBm
Marker 1 [T1 FXD]
–81.01 dBm A
–20
1RM* 496.99400000 MHz
CLRWR
–30

2RM*
–40
VIEW
PRN
–50

–60

–70
1 2
–80

–90

–100
PHN
–110
Start 496 MHz 700 kHz/ Stop 503 MHz

Fig. 6-6  V
 erification of phase noise measurements. (a) Measurement is not not
affected by thermal noise of the analyzer. (b) Dynamic range is limited
by thermal noise of the spectrum analyzer

171
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

6.2 Measurements on pulsed signals

(Dipl.-Ing. Volker Janssen)

Communications systems for information transfer, which for a long time

have mainly been implemented in analog form, are increasingly being
replaced by digital components and systems. The latter often use pulse-
modulated signals, for instance in television, radar and mobile radio.
Due to the spectral distribution of such signals, a spectrum analyzer
used to measure the signals has to fulfil special requirements. The same
applies to other types of signals, namely high-frequency, broadband
interfering signals occurring in switching operations or in the clock gen-
eration of microprocessors. Almost any electronic circuit does not only
produce the wanted signals but also unwanted spurious emissions that
impair the function of the circuit itself or of other electronic compo-
nents. Inherent spurious emissions of electronic communications equip-
ment degrade, for instance, specifications like the signal-to-noise ratio
(S/N) or bit error rate (BER). The situation can be even worse if con-
ducted or radiated interference affects other devices and impairs their
performance or causes malfunction. The prevention of spurious emis-
sions and the provision of high immunity to interference are the domain
of electromagnetic compatibility (EMC). Worldwide uniform EMC stan-
dards and guidelines provide for reproducibility of interference mea-
surements and form the basis of international regulations with respect
to interference limit values to be complied with to ensure proper func-
tioning of electronic units, modules, devices and systems.
Theoretically, the energy of pulse-modulated signals is distributed
over the whole spectrum. The measured energy strongly depends on the
resolution bandwidth and on the point of measurement in the spectrum.
If the si function is measured close to a null in the envelope, overdriv-
ing of the input stage may be caused as a result of incorrect setting. The
total energy spectrum is applied to the input stage if no preselection fil-
ters are used. This reduces the spectrum and applies the spectrum to the
mixer of the first conversion stage in ‘slices’. Modern spectrum analyz-
ers feature low nonlinearities and high overload capacity (high intercept
points of 2nd and 3rd order and high 1 dB compression). Moreover, they
are equipped with internal overload detectors used for automatic cor-
rection of the analyzer settings in order to optimize the dynamic range
and shift it into a non-critical level range with the aid of automatic RF
attenuation settings (auto range function). This ensures ease-of-opera-
tion for the user and reliable measurement.

172
 Measurements on Pulsed Signals

6.2.1 Fundamentals

The description of pulse signals is based on an ideal, periodic rectangu-

lar pulse sequence. The general real Fourier representation yields the
following for the time-dependent voltage characteristic v(t):

 
  2npt 
t ∞
1  2n pt  
()
v t = Û 1 + 2∑

T

⋅ sin


cos n w1t + 1 - cos
 T 
 
 sin n w1t 



n =1 2n pt  T 

 T 
 
 (Equation 6-5)

where Û amplitude
v(t) time function
t pulse duration (width)
T period
w1 angular frequency
n order of harmonic oscillation

The coefficients of the Fourier series describe the spectral amplitudes as

npt
sin  (Equation 6-6)
t T
v n = 2Û
T npt
T

1
––
T

Fig. 6-7   
General spectral display
(si function) after Fourier
analysis with modulated
carrier frequency f0
f0 f
3 2 1 1 2 3
f0 – –– τ f0 – ––
τ f0 – –– τ f0 + ––
τ f0 + ––
τ f0 + ––
τ

While the Fourier representation furnishes contributions from -∞ to

+∞ and the coefficients may also have a negative sign (Fig. 6-1), the
spectrum analyzer only represents positive frequencies by their mag-
nitude; the two pulse sequence characteristics according to Fig. 6-8 are
obtained:

173
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

v(t) v(t)

τ τ

Û Û

t t
T T

|V (nf1)|
|V (nf1)|

1
—
f
∆f = f1

f f
1
f1 = ––
T
fn = nf1 f1 = –1T fn = nf1

1 1
––
τ ––
τ
si si
f1 f2
Fig. 6-8  L ine spectra of two rectangular voltage characteristics with different
mark-to-space ratio, shown in time domain and in frequency domain.
The envelope of the spectral line is an si function decaying proportional
to 1/f

The smallest frequency f1 is the fundamental, corresponding to the recip-

rocal value of the period T:

f1 = 1 T  (Equation 6-7)

The amplitude values of the harmonics according to Equation 6-6 occur

at intervals of Df = f1 = 1/T.

The first null of the si function occurs at the reciprocal value of the pulse
duration:

fsi1 = 1 t  (Equation 6-8)

Further nulls follow at fn = n · fsi1· intervals.

174
 Measurements on Pulsed Signals

T
v(t)
τ

0 Fig. 6-9   
t
Finite pulse time
constants with real
pulse signal

The nulls in the pulse spectra measured in practice are not always dis-
tinct, because they are somewhat blurred. The reason lies in the asym-
metries of real signals that cannot be avoided, since in contrast to the
theoretical ideal rectangular pulses, the finite exponential rise and fall
times of the real pulses have to be taken into account.

Before dealing with the different terms and the dependencies of the dis-
played spectrum on the measurement bandwidth, let us examine some
other pulse shapes, as well.

*RBW 1 kHz
VBW 10 kHz
Ref 0 dBm Att 30 dB SWT 50 ms
0
A
–10
1PK
MAXH
–20

–30
PRN
–40
EXT
–50

–60

–70

–80

–90

–100
Center 900.024 MHz 5 kHz/ Span 50 kHz

Fig. 6-10  R
 eal representation of envelope spectrum with a spectrum analyzer
(100 µs pulse duration, 1 ms pulse period, 900 MHz carrier frequency,
1 kHz measurement bandwidth)

175
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

*RBW 100 Hz
VBW 1 kHz
Ref 0 dBm Att 30 dB SWT 5 s
0
A
–10
1PK
MAXH
–20

–30
PRN
–40
EXT
–50

–60

–70

–80

–90

–100
Center 900.024 MHz 5 kHz/ Span 50 kHz

Fig. 6-11  R
 eal representation of line spectrum with a spectrum analyzer, same
settings as in Fig. 6-10, but 100 Hz measurement bandwidth

Triangular and trapezoidal pulses

The spectrum of a triangular pulse with equal rise and fall time exhib-
its an envelope that corresponds to a si2 function. The trapezoidal pulse
can be obtained from a combination of rectangular and triangular pulse.
The effect of additional time constants is noticeable in the differently
decaying amplitudes of the log-log density spectrum. While with the
trapezoidal pulse at 1/pt the envelope of the amplitude density spec-
trum decreases 20 dB per decade, it decreases 40 dB per decade in the
case of equal rise and fall time. If the time constants are different, there
is a decrease of 20 dB per decade at the first (smaller) corner frequency
and of another 20 dB at the second (larger) corner frequency, similar to
the characteristic of the trapezoidal pulse.

176
 Measurements on Pulsed Signals

v (f ) / dB
Fig. 6-12   
Amplitude density
spectrum for rectangular, 2ûτ
trapezoidal and trian-
gular pulses (diagram
is shown for trise= tfall )
1 1 lg f
fv = –– fv = ––
πτ πτr

Considering t → 0 clearly reveals that this corner frequency (Fig. 6-12)
is shifted towards infinitely high frequencies. Examination of the bound-
ary case that the period T → ∞ (Df = 1/T → 0), yields a single pulse
with infinitely high amplitude (Dirac function).
The Fourier series only allows the representation of periodic time
domain functions. With the aid of the boundary conditions, T → ∞, and
Df → 0, non-periodic functions can also be described. This is possible
with the aid of the Fourier transform and Fourier integral.
In practice, non-periodic events occur more frequently, such as switch-
ing operations, lightning strokes or electrostatic discharges.

6.2.2 Line and envelope spectrum

The energy of the periodic pulse occurs at the discrete frequencies n  · f1
(Equation 6-7), or equivalently at n  · 1/T.
The envelope si function has nulls at the integer multiples as a func-
tion of the mark-to-space ratio t/T. If the pulse signal is used for modu-
lation of a carrier, the spectrum will be symmetrically distributed above
and below the carrier frequency. Depending on the measurement or res-
olution bandwidth, the following three cases are possible when using a
frequency-selective spectrum analyzer or test receiver for the spectrum
measurement:

1. If the measurement bandwidth is small relative to the offset of the

frequency lines (defined by 1/T = Df ), the individual spectral lines
can be resolved so that a line spectrum is obtained.

B <1 T  (Equation 6-9)

177
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

 A further reduction of the bandwidth yields equal amplitude values,

reduces the noise and thus improves the signal-to-noise ratio with
the bandwidth ratio 10 · lg(B1/B2).

2. The bandwidth B is greater than the spacing Df of the spectral lines,

but smaller than the spacing 1/t of the first null of the envelope
si function from the carrier frequency.
 The spectral lines cannot be resolved and the amplitude height of
the envelope depends on the bandwidth. This makes sense as the
amplitude depends on the number of spectral lines collected within
the measurement bandwidth.

1 t >B >1 T  (Equation 6-10)

The above condition is described as an envelope display. The envelope

amplitude increases with increasing bandwidth by 20 · lg(B2 /B1).

*RBW 100 kHz Marker 1 [T1]

VBW 1 kHz –22.35 dBm
Ref 0 dBm Att 30 dB SWT 1 s 900.02408000 MHz
0
Delta 2 [T1]
–10 –0.15 dB A
1PK 1.00000000 kHz
MAXH 1 2
–20

–30
PRN
–40
EXT
–50

–60

–70

–80

–90

–100
Center 900.02408 MHz 1 kHz/ Span 10 kHz

Fig. 6-13  L ine spectrum of pulsed signal (measurement bandwidth  

B = 100 Hz < 1/T = 1 kHz)

178
 Measurements on Pulsed Signals

*RBW 3 kHz Marker 1 [T1]

VBW 30 kHz –9.15 dBm
Ref 0 dBm Att 30 dB SWT 15 ms 900.02408000 MHz
0
1 Delta 2 [T1]
–17.48 dB A
–10
1PK 10.00000000 kHz
MAXH
–20
2
–30
PRN
–40
EXT
–50

–60

–70

–80

–90

–100
Center 900.02408 MHz 10 kHz/ Span 100 kHz

Fig. 6-14  E
 nvelope spectrum of pulsed signals  
(1/t = 10 kHz > B = 3 kHz > 1/T = 1 kHz)

3. The bandwidth B is greater than the null spacings of the envelope,

selectivity is no longer effective and the amplitude distribution in
the spectrum cannot be recognized any more. With increasing band-
width the impulse response of the filter approaches the time func-
tion of the pulse-modulated carrier.

B >1 t  (Equation 6-11)

179
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

*RBW 300 kHz

VBW 3 MHz
Ref 0 dBm Att 30 dB SWT 2.5 ms
0
A
–10
1AP SGL
CLRWR
–20

–30
PRN
–40
EXT
–50

–60

–70

–80

–90

–100
Center 900.02408 MHz 100 kHz/ Span 1 MHz

Fig. 6-15  T ransition to display in time domain. Pulse duration of 100 µs and
period of 1 ms can be clearly recognized

To put it simply:
❙ In the case of the line spectrum the number of lines does not vary
as a function of the bandwidth or frequency span, the amplitude
remains constant.
❙ In the case of the envelope spectrum the number of lines varies as a
function of the bandwidth and not as function of the frequency offset.
The displayed amplitude increases with the resolution bandwidth due
to the larger energy component within the measurement bandwidth.

With pulse modulation, the displayed amplitude decreases with decreas-

ing bandwidth, this effect being referred to as pulse desensitization. The
relationship can be expressed by the determination of the pulse desen-
sitization factor (PDF):

PDFline = 20 dB ⋅ lg (t T )  (Equation 6-12)

for amplitude values in the line spectrum and

PDFenvelope = 20 dB ⋅ lg (tKB )  (Equation 6-13)

180
 Measurements on Pulsed Signals

for amplitude values in the envelope spectrum. The shape factor K

depends on the type of the resolution filter used and is described in
detail in the following section. Typical examples are K = 1 for Gauss-
ian filters and K = 1.5 for rectangular filters. For pulse signal measure-
ments, a compromise has to be found since with small resolution band-
widths the displayed amplitude may become too small, whereas with
large resolution bandwidth the displayed amplitude will be larger but
the resolution degraded to an increasing extent. In practice, the follow-
ing value has been empirically determined:

t ⋅ B = 0.1  (Equation 6-14)

0
A in dB

–20
Rectangular
filter
–40
Fig. 6-16   
Gaussian
Amplitude loss as a
filter
function of time/band- –60
width product t · B
0.001 0.01 0.1 1
tP · B

Examples:
A pulse of the duration t = 2 ms and pulse repetition frequency of 5 kHz
(= 1/T ), corresponding to a period T = 200 ms, is measured with a
­Gaussian filter (K = 1) of bandwidth B = 1 kHz.

The condition defined by Equation 6-9 applies (B < 1/T ), so we have a

line spectrum. Equation 6-12 then yield:

PDFline = 20 dB ⋅ lg (2 µs/200 µs) = -40 dB

Accordingly, the displayed amplitude value of the unmodulated carrier

would be 40 dB higher.

181
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

*RBW 100 kHz Marker 1 [T1]

VBW 1 MHz –40.39 dBm
Ref 0 dBm Att 30 dB SWT 2.5 ms 900.02428000 MHz
0
–10 *
–20 A
1AP –30
1
VIEW –40
–50
–60
–70
–80
–90 PRN
–100 EXT
Center 900.02408 MHz 200 kHz/ Span 2 MHz

RBW 100 kHz Marker 1 [T1]

VBW 300 kHz –13.17 dBm
Ref 0 dBm Att 30 dB SWT 2.5 ms 900.02000000 MHz
0
1
–10
–20 B
1PK –30
MAXH –40
–50
–60
–70
–80
–90
–100
Center 900 MHz 500 kHz/ Span 5 MHz

Fig. 6-17  P
 ulse spectra measured with different bandwidths. The markers display
the desensitization factor. The unmodulated carrier level is 0 dBm

The same measurement is repeated with the same parameters, but with
a measurement bandwidth B of 100 kHz. The relationship defined by
Equation 6-10 applies (1/t > B > 1/T ). Equation 6-13 then yield:

PDFenvelope = 20 dB ⋅ lg (2 ⋅ 10-6 ⋅ 1 ⋅ 100 ⋅ 103 ) = 20 dB ⋅ lg (2 ⋅ 10-1 ) = -14 dB

The maximum amplitude of the spectrum is 14 dB lower than that of the
unmodulated carrier.

6.2.3 Resolution filters for pulse measurements

The spectral lines of broadband pulse signals are correlated, so the dis-
played level doubles when the measurement bandwidth is doubled. To
determine the actual pulse bandwidth, the displayed level with the use
of a real filter is compared with the displayed level from an ideal rect-

182
 Measurements on Pulsed Signals

angular filter. For Gaussian filters, which are mostly used due to their
favorable transient response, the following relationship is obtained:

B I = 1.506 ⋅ B3 dB  (Equation 6-15)

where BI pulse bandwidth

The pulse bandwidth of Gaussian or Gaussian-like filters corresponds

approximately to the 6 dB bandwidth. For spectrum analyzers, the 3 dB
bandwidths are usually specified, whereas for EMI measurements,
where often pulses are measured, 6 dB bandwidths are stated.

The relationship between 3 dB, 6 dB, noise and pulse bandwidths for dif-
ferent filters was described in chapter 4. The conversion factors can be
directly taken from the table below.

Initial value is 3 dB b­ andwidth 4 filter circuits 5 filter circuits Gaussian filter
(analog) (analog) (digital)

6 dB bandwidth (B6 dB) 1.480 · B3 dB 1.464 · B3 dB 1.415 · B3 dB

Pulse bandwidth (BI) 1.806 · B3 dB 1.727 · B3 dB 1.506 · B3 dB

Initial value is 6 dB bandwidth

3 dB bandwidth (B3 dB) 0.676 · B6 dB 0.683 · B6 dB 0.707 · B6 dB

Pulse bandwidth (BI) 1.220 · B6 dB 1.179 · B6 dB 1.065 · B6 dB

It should be noted that the corner frequencies determined by the pulse

duration t and the period T, or the pulse spectrum, must not be affected
by a video filter. Modern measuring instruments have selectable cou-
pling factors between measurement or resolution bandwidth and video
bandwidth as well as between measurement and settling times. Depend-
ing on the conditions involved, a preset coupling factor or a user-select-
able factor may be defined for pulse measurements so that the video
bandwidth may exceed the measurement bandwidth by a factor of 10.

Implementation of the resolution bandwidth using digital filters has no

effect on the weighting of pulse signals. Digital filters are suited just as
well as analog filters and, in addition to temperature and long-term sta-
bility, they feature the advantage of being highly selective, so filters with
lower shape factor can be realized.

183
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

If spectrum analyzers exclusively operate on the basis of fast Fourier

transform (FFT), they are not suitable for pulse measurements. With FFT,
the spectrum is calculated from a section of the time domain signal. As
explained in chapter 3.1, the measurement results depend on the choice
of this section, so that the FFT analysis is unsuitable for pulse signal
analysis. It is therefore important that analyzers provide both analog /
digital filters and FFT analysis.

6.2.4 Analyzer parameters

The above explanations have clearly shown that measurement and

assessment of pulse signals is more complex with many more details
involved than with sinusoidal signals. In spectrum analyzer or test
receiver data sheets, the maximum input level for sinusoidal (CW) signals
is specified. For pulse measurements, parameters such as pulse spec-
tral density, maximum pulse energy or pulse voltage are important to
avoid erroneous measurement or damage to the measuring instrument.
It makes sense to define a parameter such as the pulse spectral density
and to relate a pulse voltage to a reference bandwidth (chapter 5.4: 1 dB
compression point and maximum input level). A reference bandwidth of
1 MHz has been defined for this purpose. The parameters are therefore
specified in the units µV/1 MHz and dBµV (MHz).

The pulse spectral density can easily be calculated from the measured
level using the following correction factor:

K 1 = 20 dB ⋅ lg (B1 1 MHz)  (Equation 6-16)

Example:
In a spectrum analyzer the measurement bandwidth is determined by
a Gaussian filter with a 3 dB bandwidth of 10 kHz. A pulse signal mea-
surement yields a displayed level of -67 dBm. What is the pulse spec-
tral density?

First, the measured value has to be converted into the unit dBµV. A

level of 0 dBm corresponds to 107 dBµV. That is, -67 dBm corresponds
to +40 dBµV. The pulse bandwidth is calculated for Gaussian filters in
accordance with Equation 6-15 at 1.506  · B3 dB to yield BI = 15 kHz.

184
 Measurements on Pulsed Signals

The correction factor of -36.5 dB is obtained from Equation 6-16:

K 1 = 20 dB ⋅ lg (B1 1 MHz) = 20 dB ⋅ lg (15 ⋅ 103 1 ⋅ 106 ) = - 36.5 dB

The measured value of +40 dBµV corresponds to a pulse spectral den-

sity of 76.5 dBµV (MHz).

A detailed description of maximum pulse energy and voltage, which are

further important parameters for the assessment of spectrum analyzers
and test receivers, can be found in chapter 5.4.

Certain instruments can even have two separate RF inputs, one of them
being pulse-protected for handling high maximum pulse voltage peaks.
Pulse energy is applied to the subsequent attenuator which is designed
to dissipate the resulting heat. In the case of inadequate heat dissipa-
tion, components may be damaged.

6.2.5 Pulse weighting in spurious signal measurements

So far we have considered wanted signals that serve for transmitting

information. As mentioned at the beginning of chapter 6.2, the measure-
ment and weighting of unwanted spurious signals is another important
aspect in the analysis of pulse signals.
The subject of electromagnetic compatibility is very complex because
almost every electronic device not only produces wanted but also
unwanted signals or couples the wanted signals into the signal path at
some point. The mechanism is made even more complex by the fact that
spurious emissions may be propagated on radiated or conducted paths.

Reproducibility of the spurious signal measurement is ensured by stan-

dards and guidelines. The relevant EMC standards define product-spe-
cific limit lines taking into account the DUT’s field of application (domes-
tic, industrial or military environment). For commercial measurements (in
contrast to military standards), a specially designed and calibrated detec-
tor known as quasi-peak detector is used for weighting continuous pulse
signals on the basis of their pulse repetition frequency. The weighted lev-
els can be compared to the defined limit lines. If the determined level val-
ues are below these limit lines, interference-free operation of the DUT
within the context of electromagnetic compatibility can be assumed.

185
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

With the matched circuitry, conducted pulses are not so hazardous pro-
vided that their energy does not exceed a certain limit value. It is more
often the case that the interference pulse triggers some kind of oscillator.

Examples:
According to the above and Fig. 6-12 it is shown that the bandwidth
occupied by an interference pulse is inversely proportional to its dura-
tion. This means that if an interference pulse with a pulse width of 1 µs
has its first null at 1 MHz in the frequency spectrum, a decay of the
spurious amplitudes becomes significant from about 300 kHz. A 100 ns
interference pulse shows this decay at 3 MHz.
Furthermore, a pulse amplitude of 1 V is assumed. Accordingly, a
pulse of 1 µs width has an energy (product of voltage and time) of 1 µVs.
A 100 ns interference pulse of 10 V amplitude also has an energy of 1 µVs.
With a pulse bandwidth setting of 10 kHz, the spectrum analyzer indi-
cates for the two pulses, both for the 1 µs and for the 100 ns pulse, a
voltage of 10 mV on its display relative to the RMS value of a sinewave
voltage. This means that the spectrum analyzer cannot differentiate
between the pulse amplitudes. From the observed voltage value, no con-
clusions as to overdriving can be made either since the same value will
be displayed for a 10 ns pulse of 100 V in amplitude.

6.2.5.1 Detectors, time constants

The peak detectors described in chapter 4.2, such as max peak, min peak,
auto peak and sample detectors, are standard in most spectrum analyz-
ers. RMS (root mean square) and AV (average) detectors are also imple-
mented in state-of-the-art instruments.
A special detector for interference pulse measurements referred to
as a quasi-peak detector (QP) is frequently available as an option. It
places high demands on the dynamic range and linearity of the input as
well as of the IF stage, which cannot be satisfied by a large number of
instruments available on the market. The requirements are the result of
the weighting characteristic for pulse sequences (prescribed by CISPR
16-1 standard) which, due to underweighting of pulse sequences at low
pulse repetition frequency (up to 40 dB for single pulses), calls for a
dynamic range that is wider than non-state-of-the-art-instruments by a
factor of 100.

186
 Measurements on Pulsed Signals

34
30 to 1000 MHz (band C/D)
32
Relative input level for constant display in dB

28
0.15 to 30 MHz (band B)
24
20 9 to 150 kHz (band A)
16
12

43.5 dB
8
4
0
–4
–8
–12
Single pulse 1 10 100 1 kHz
Pulse repetition frequency
Fig. 6-18  Pulse weighting to CISPR 16 for different pulse repetition frequencies

In accordance with the frequency bands specified by CISPR, the quasi-

peak detector is assigned defined charge and discharge time constants
and bandwidths. It is thus ensured that in the different CISPR bands, the
measured values are always collected with the same time constants and
same bandwidth (usually the pulse bandwidth) to provide for reproduc-
ibility and comparable limit lines. This is referred to as a weighted dis-
play of the QP detector and pulse weighting curve, which also contains
the time constant resulting from the inertia of mechanical meters.

The CISPR bands are defined as follows:

CISPR A 9 kHz to 150 kHz

CISPR B 150 kHz to 30 MHz
CISPR C 30 MHz to 300 MHz
CISPR D 300 MHz to 1000 MHz

187
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

Frequency band 6 dB mea- Charge Discharge Time

surement time of QP time of QP constant of
bandwidth ­detector detector meter
CISPR A 200 Hz 45 ms 500 ms 160 ms
CISPR B 9 kHz 1 ms 160 ms 160 ms
CISPR C 120 kHz 1 ms 550 ms 100 ms
CISPR D 120 kHz 1 ms 550 ms 100 ms

The weighting of different pulse repetition frequencies shows that the

greatest difference between the displayed levels is at low pulse repeti-
tion frequencies. With increasing pulse repetition frequency (PRF), for
instance PRF > 10 kHz, the levels displayed by all detectors (AV, RMS
and QP) approach the value of the peak detector.
0 PK
–10
dB OP
–20
–30 RMS
AV
–40
–50
–60
CISPR band B (150 kHz to 30 MHz)
–70
–80
10 0 2 3 5 10 1 2 3 5 10 2 2 3 5 10 3 2 3 5 10 4 2 3 5 10 5
Pulse repetition frequency / Hz

Fig. 6-19  L evels displayed with different detectors and pulse repetition frequen-
cies relative to peak display

Special standard pulse generators are available for calibration of the

QP display of spectrum analyzers and test receivers. Exacting require-
ments are placed on such standard pulse generators. For the calibration
of test receivers up to 1 GHz, pulses with a width of much less than 1 ns
are required, in practice usually 200 or 250 ps. In addition to the short
pulse duration, extremely short pulse rise and fall times can also be real-
ized. The pulse frequency of the generator must be variable for simulat-
ing the CISPR weighting curve.

188
 Measurements on Pulsed Signals

Peak detector
Another way of specifying the level in EMI measurements is the peak
value relative to a 1 MHz bandwidth. In this display mode the pulse spec-
tral density of the input signal is measured. The peak value at the output
of the envelope demodulator within the selected measurement time is
relative to 1 MHz. It should be noted that due to the measurement band-
width, the displayed peak value is increased by 20 dB · lg (1 MHz/BI).

Broadband and narrowband interference

Different pulse weighting methods are used in EMC. These are based on
the different limit values defined for broadband and narrowband inter-
ference. Broadband interference is relatively evenly distributed over the
spectrum, but due to this flat spectral distribution, the disturbance is
lower than that of a sinusoidal spurious signal (narrowband interferer)
of a very high level. The permissible spurious emission limit values are
10 dB higher than the narrowband limit values (depending on the rele-
vant standard). This means that a narrowband interferer should be sup-
pressed in a circuit (attenuated by at least 10 dB) by suitable RFI rejec-
tion or shielding.

The detector method and bandwidth detuning method are suitable to

differentiate between broadband and narrowband interference.
The detector method is based on the assumption that narrowband
interference weighted once by a peak detector and once by an aver-
age detector yields approximately the same result (difference < 6 dB).
If there is a greater difference, the interference is of broadband nature
and the limit values for broadband interference have to be applied. The
detector method can employ both a PK / AV and a PK / RMS detector
comparison, depending on the standard used.
The bandwidth detuning method assumes that a signal level is dis-
played with the selected peak detector, and used as a reference value.
The measurement is repeated with the same settings, but with the cen-
ter frequency shifted by ±(measurement bandwidth B ). If the two new
values are less than the critical threshold (6 dB) of the reference level,
the signal is considered to be a narrowband interferer. An interferer
that is not identified as being narrowband is considered to be broad-
band. The detuning method can also be carried out with the center fre-
quency shifted by ± 2 B. Both methods are permitted in the relevant
standards.

189
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

6.2.5.2 Measurement bandwidths

The measurement bandwidths specified in the standards for pulse mea-

surements within spurious emission measurements are to be under-
stood as pulse bandwidths. For commercial standards, these are the
bandwidths of

200 Hz, 9 kHz, 120 kHz (civil specifications, such as EN, VDE, FCC, VCCI,
etc),

whereas for military standards the following decade steps apply:

10 Hz, 100 Hz, 1 kHz, 10 kHz, 100 kHz, 1 MHz.

These bandwidths (designated as pulse bandwidths) as well as the QP

detector are additionally implemented in modern spectrum analyzers
which, due to their performance features regarding overload capability
and dynamic range, are suitable for spurious emission measurements.

6.3 Channel and adjacent-channel power measurement

(Dipl.-Ing. Roland Minihold)

6.3.1 Introduction

Advanced 3rd generation mobile radio systems operating on the CDMA

principle (code division multiple access) have a frequency multiplex
component like that of the 2nd generation TMDA systems (time domain
multiple access systems, such as GSM or IS-136) or the traditional 1st
generation analog FDMA systems (frequency domain multiple access,
such as AMPS).

190
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

a)

Amplitude

Frequency

n+2
n+1
Time n
n–1
Channels

b)
Amplitude

Frequency

n–1
n
Time
n+1
n+2
Time slots

c)
Amplitude

Frequency

Time

Fig. 6-20  V
 arious methods of channel generation in (mobile) radio systems by
signal multiplexing: FDMA (a), TDMA (b) and CDMA (c)

191
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

This means that in all these systems there are several adjacent radio
channels in the frequency band providing multiple access. The main dif-
ference between the various systems lies in the fact that compared to
the traditional analog radio systems, the radio channels occupy a larger
bandwidth. In the traditional analog radio system such as the American
AMPS system, each user is allocated a separate transmit and receive
channel, which are both occupied over the whole duration of active
radiocommunication. In TDMA systems, several users either share
transmit and receive channels in the time domain (frequency duplex
as in the GSM systems), or transmit and receive channel are identical
(time duplex as in DECT systems). In mobile radio systems operating on
the CDMA principle, many users (often approximately 128) share suffi-
ciently wide transmit and receive channels. The two channels are used
over the whole duration and the individual users are separated using
despreading codes.
To ensure undisturbed reception for a large number of users, it is
absolutely necessary to avoid interference with adjacent transmission
channels in the frequency band. An important criterion is a sufficiently
low adjacent-channel power specified either as absolute value (in dBm)
or relative value referred to the channel power in the transmit channel
(in dBc).
For cdmaOne systems (IS-95, 1.25 MHz channel bandwidth), addi-
tional limit values have been prescribed for signals emitted in neighbor-
ing analog radio channels of the AMPS systems (30 kHz channel band-
width).
In TDMA systems (such as IS-136 or GSM), the transmitter power,
and hence the unwanted power radiated in the adjacent channels, is
only applied in certain timeslots, so that special measures such as gat-
ing (measurement only during the active timeslot) are required. A dis-
tinction is usually made as to whether the spurious emissions in the
adjacent channels are caused by the modulated stationary transmit-
ter signal (spectrum due to modulation) or by the on / off switching of
the transmitter signal (spectrum due to switching). A spectrum analyzer
intended for performing measurements on TDMA systems should there-
fore feature suitable functions for adjacent-channel power measure-
ment as well as gating and trigger functions.

192
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

6.3.2 Key parameters for adjacent-channel power measurement

In addition to the channel bandwidth of the user channel and the adja-
cent channels, the channel spacings are important parameters for adja-
cent-channel power measurements. Channel spacing is understood as
the difference between the center frequency of the user channel and
that of the adjacent channel.
The number of adjacent channels in which the channel power is
measured is also important. The table below shows the channels to be
measured depending on the number of channels set:

No. of channels Channel power measurement

0 User channel only
1 User channel and upper/lower adjacent channel
2 User channel and adjacent channels + 1st alternate channels
3 User channel and adjacent channels + 1st alternate channels +
2nd alternate channels

As shown in Fig. 6-21, the adjacent channels have different designations

depending on their position relative to the user channel. In our example,
two channels will be set.

*RBW 30 kHz
*VBW 300 kHz
Ref –10 dBm Att 10 dB *SWT 100 ms
–10
–20
–30 A
1RM *–40 c12
CLRW –50 c12
c11
–60 c11
cu1
–70 cu1
–80 cu2
cu2
–90 c0
–100 c0 PRN
–110
Center 2.1 GHz 2.546574363 MHz/ Span 25.46574363 MHz

Tx Channel
Power –7.34 dBm Bandwidth 3.84 MHz

Adjacent Channel
Lower –58.75 dBc Bandwidth 3.84 MHz
Upper –59.24 dBc Spacing 5 MHz

Alternate Channel
Lower –59.32 dBc Bandwidth 3.84 MHz
Upper –58.84 dBc Spacing 10 MHz

Fig. 6-21  P
 osition of user channel and adjacent channels in frequency domain,
here for a WCDMA signal

193
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

6.3.3 Dynamic range in adjacent-channel power measurements

(see also chapter 5.5: Dynamic range)

The dynamic range achievable in adjacent-channel power measurements

using a spectrum analyzer is influenced by three factors, assuming suffi-
cient filter selectivity for user channel suppression and ideal signal:
❙ Thermal inherent noise of analyzer
In this case it is the signal-to-noise ratio achieved with the specific
device setting (measurement level on analyzer, RF attenuation, refer-
ence level).
❙ Phase noise of the analyzer
❙ Intermodulation products (spectral regrowth)
Intermodulation products falling within the adjacent channels are a
crucial factor especially in measurements on wideband CDMA sys-
tems.

The adjacent-channel power is obtained by linear addition of the above

individual contributions. The contributions made by thermal noise and
intermodulation products depend on the input level of the first mixer
of the spectrum analyzer. While the effect of thermal noise decreases
inversely proportionally with the mixer level, the intermodulation prod-
ucts increase. The sum of all power contributions gives the asymmetri-
cal ‘bathtub’ characteristic shown in Fig. 6-22. The maximum achievable
dynamic range can be determined for each mixer level.

–40 40

–50 50
LN, rel , LIM3, rel , LPN, RBW in dB

Dynamic range in dB

–60 60
LSum
–70 70

–80 80
LPN,RBW
–90 90
LIM3,rel LN,rel
–100 100

–110 110

–120 120
–70 –60 –50 –40 –30 –20 –10 0
Lmix,opt Mixer level in dBm
Fig. 6-22  D
 ynamic range taking into account thermal noise, phase noise and 3rd
order intermodulation products

194
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

6.3.4 Methods for adjacent-channel power measurement using a

spectrum analyzer

6.3.4.1 Integrated bandwidth method

The IF filters of spectrum analyzers are usually implemented in a rela-

tively coarse raster of 1, 3 or 1, 2, 3, 5 steps. Moreover, their selectivity
characteristics do not satisfy the requirements placed on channel filters.
Analog IF filters are usually implemented as synchronously tuned four- or
five-stage filters featuring optimized transient response to achieve min-
imum sweep times. The selectivity characteristic of filters with a shape
factor of approximately 12 for four-stage filters and approximately 9.5
for five-stage filters is rather poor and usually inadequate for sufficient
suppression of the signal in the user channel for adjacent channel mea-
surements. Digital resolution filters of modern spectrum analyzers that
are usually implemented as Gaussian filters are not suitable as channel
filters despite their better selectivity (shape factor of 4.6).
Therefore, spectrum analyzers usually offer power integration fea-
tures in the frequency domain for adjacent-channel power measurement.
Compared to the channel bandwidth, a very small resolution bandwidth
of typically 1 % to 3 % of the channel bandwidth is set to ensure appropri-
ate selectivity. The spectrum analyzer sweeps over the frequency range
of interest from the start of the lower adjacent channel to the end of the
upper adjacent channel depending on the number of measured adjacent
channels (Fig. 6-21).

195
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

*RBW 30 kHz Delta 2 [T1]

VBW 300 kHz 2.15 dB
Ref 10 dBm Att 10 dB *SWT 100 ms 1.23000000 MHz
10
Marker 1 [T1]
–14.11 dBm A
0
999.38300000 MHz
–10 2
1

–20
PRN
–30

–40

–50

–60

–70

–80

–90
Center 1 GHz 150 kHz/ Span 1.5 MHz

Fig. 6-23  Channel power measurement using integrated bandwidth method

The measured values corresponding to the levels of the displayed pixels

are integrated within the selected channel bandwidth on a linear scale.
The resulting adjacent-channel power is stated in dBc referred to the
power in the user channel.

The following individual steps are carried out:

❙ For all levels measured within a channel, the power is determined.
The following applies:

Pi
= 10( i
L( mW ) 10 dB)
 (Equation 6-17)
W
where Pi power of a measured value represented by pixel i
Li level of a measured value represented by pixel i,
­relative to 1 mW

❙ The power values of all trace points within a channel are added
together and divided by the number of trace points in the channel.
❙ The result for each channel is multiplied by the quotient from the
selected channel bandwidth and the noise bandwidth of the resolu-
tion filter.

196
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

From the above steps, the following relationship is obtained for the
absolute channel power:
B 1 n2
P 
LCh(mW) = 10 dB ⋅ lg  Ch ⋅ ⋅ ∑ i   (Equation 6-18)
 BN, IF 1 + n2- n1 n1 W 

where LCh channel power level, relative to 1 mW

BCh channel bandwidth
BN, IF noise bandwidth of IF filter
n1, n2 indexes of measured values to be added together
Pi power of a measured value represented by pixel i

Selection of resolution bandwidth (BN)

The selected resolution bandwidth BN should be small relative to the
channel bandwidth for the channel bandwidth to be measured accu-
rately. If the resolution bandwidth is too large, the selectivity of the sim-
ulated channel filter is insufficient and part of the main channel power
will be measured when measuring the adjacent channels, so the final
result will be incorrect. The well-chosen resolution bandwidth is typi-
cally 1 % to 3 % of the channel bandwidth. If the resolution bandwidth is
too small, the required sweep time becomes unduly long, and the mea-
surement time will be considerably increased.

Selection of detector
For power measurements within the channel bandwidth, the sample
detector and the RMS detector are suitable since only these two detec-
tors furnish results that allow power calculation. The peak detectors
(max peak, min peak, auto peak) are not suitable for measuring noise or
noise-like signals since a correlation between the detected video voltage
and input signal power cannot be established.
When using the sample detector, the measured value represented by
a pixel is derived from a sample of the IF envelope voltage. If the dis-
played spectrum is large relative to the resolution bandwidth (such as
the span/RBW  >500), discrete signal components (sinusoidal signals)
may get lost due to the limited number of pixels of the analyzer screen
(approximately 501), and the channel or adjacent-channel power mea-
surement will therefore be incorrect (chapter 4.4: Detectors).
Since digitally modulated signals are noise-like signals, the trace
obtained with a sample detector is subject to large variations. To obtain

197
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

stable results, averaging is necessary, in which case the displayed signal

will be underweighted and falsified (chapter 4.5: Detectors).
When choosing the RMS detector, the power represented by a pixel is
calculated from several measured values to obtain stable results. More-
over, the measurement time can be increased to allow averaging of the
trace. The power of discrete spurious signals contained in the channel is
also correctly determined. The RMS detector is therefore a better choice
than the sample detector for channel power measurements.

The RMS value is calculated from the samples of the video voltage as
follows:

1 N
VRMS = ⋅ ∑ v i2  (Equation 6-19)
N i =1

where VRMS RMS value of voltage

N number of samples allocated to the pixel c­ oncerned
vi samples of envelope

The reference resistance R can be used to calculate the power:

VRMS
2

P =  (Equation 6-20)
R

Some TDMA mobile radio standards (such as PDC) prescribe a peak

detector for measuring the adjacent-channel power (relative measure-
ment) to ensure better detection of the power transients.

Selection of video bandwidth / trace averaging

When using an RMS or sample detector, the video bandwidth must be
at least three times the resolution bandwidth to avoid averaging of the
video voltage, as this would lead to underweighting of noise-like signals,
resulting in channel power that would be too low. For this reason, trace
averaging over several traces should also be avoided.

6.3.4.2 Spectral power weighting with modulation filter

(IS-136, TETRA, WCDMA)

For determining the power in the main and adjacent channels of some
mobile radio systems, such as IS-136 (NADC), TETRA and WCDMA, it is

198
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

necessary to use a channel filter that corresponds to the modulation

filter of the respective system (typically root-raised cosine filter). This
leads to a more realistic weighting of the effect of the power emitted in
the adjacent channel since interference is mainly caused by signal com-
ponents in the center of an adjacent channel. Signal components close
to the channel boundaries are suppressed by the matched filter of the
receiver so that these cause less interference.

When using a spectrum analyzer for adjacent-channel power mea-

surements, the individual trace values in respective channels must be
weighted with a standard-specific modulation filter before each chan-
nel power is determined by integration of the measured values. Mod-
ern spectrum analyzers provide measurement functions with automatic
weighting.
0

–3
Attenuation / dB

–6

–9

–12

–15

–18

–21

–24

–27

–30
–18 –16 –14 –12 –10 –8 –6 –4 –2 fc 2 4 6 8 10 12 14 16 18
Frequency / kHz

Fig. 6-24  Channel filter for IS-136 (NADC) systems

Use of a weighting filter in channel power measurements with a spec-

trum analyzer can very easily be verified by using the following test:

With the channel power measurement activated, a sinusoidal signal with

a frequency corresponding to the channel center frequency is applied to
the spectrum analyzer input. The measured channel power is used as a
reference.
The frequency of the sinusoidal signal is then varied in steps towards
the channel boundaries (or alternatively by varying the channel center

199
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

frequency on the spectrum analyzer in case of a fixed-frequency sinu-

soidal signal) and the measured channel power observed. If the channel
power varies under these conditions, a channel filter is obviously being
used for weighting.

This test can also be carried out in the adjacent channels. It is recom-
mended to set the spectrum analyzer for absolute adjacent-channel
power measurement.

6.3.4.3 Channel power measurement in time domain

As explained in chapter 6.3.4.1, very narrowband resolution filters are

required for channel power measurements. Since these filters exhibit
low integration times, their use inevitably leads to relatively long sweep
times. If measurements are carried out over several adjacent channels,
frequency ranges between the individual channels will be included that
contain no information of interest, but contribute to the total sweep
time. All these drawbacks can be avoided by measuring the channel
power in the time domain.

With the aid of digital signal processing, practically any type of chan-
nel filter can be digitally implemented for modern spectrum analyzers,
such as root-raised cosine filters or near ideal rectangular bandpass fil-
ters as well as filters with very large bandwidth (such as 4 MHz). These
filters allow channel power measurement in the time domain with the
spectrum analyzer being tuned like a receiver to the center frequency of
the channel. In this way it is possible to avoid the limitation of the mini-
mum sweep time given by the transient time resulting from the narrow
resolution bandwidths that are required for measurements in the fre-
quency domain. In the time domain, a much better reproducibility of the
measurement results can be achieved with the same measurement time
as in the frequency domain, and the measurement time can be consider-
ably reduced compared to the conventional integration method.
If the power is to be measured in several adjacent channels, the spec-
trum analyzer is automatically successively tuned to the respective chan-
nel center frequencies. Frequency ranges between the various channels
of interest are skipped so that there is the further advantage in mea-
surement time compared to measurements in the frequency domain.
Another benefit of time domain measurements is the correct detec-
tion of transient signals caused by switching operations.

200
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

Ref 15.7 dBm *Att 30 dB SWT 200 ms

10
0
–10 A
1RM *
–20
CLRWR–30
–40
–50
–60
–70 PRN
–80
–90 EXT
Center 1 GHz 20 ms/
Tx Channel
Power 9.64 dBm Bandwidth 25 kHz

Adjacent Channel
Lower –34.96 dBc Bandwidth 25 kHz
Upper –35.12 dBc Spacing 30 kHz

Fig. 6-25  C
 hannel and adjacent-channel power measurement in time domain,
here with IS-136 signal

6.3.4.4  Spectral measurements on TDMA systems

For measuring the adjacent-channel power on switched signals in TDMA

systems, some special aspects have to be considered.
If the adjacent-channel power is to be determined from the mod-
ulation and phase noise of the transmitter, the detection of transient
signals caused by on / off switching operations must be avoided. Mea-
sured values should therefore only be collected within the active times-
lot (burst). This is possible by using gating features.
Derived from an external trigger signal or from a broadband level
detector within the spectrum analyzer (RF trigger), a corresponding time
window, or gate, is set during which measured values are collected. No
measured values are recorded outside this gate, during which the fre-
quency sweep is stopped.
With correct setting, the effective sweep time required for this mea-
surement to examine a certain frequency range is longer than a normal
sweep, namely by the reciprocal value of the on / off ratio ton/toff .

Many analyzers can be triggered by a video signal. This trigger source,

however, is not suitable for spectral measurements on TDMA systems
since the selectivity of the selected resolution bandwidths prevents gat-
ing from being activated. In this case, the sweep would not be t­ riggered.

201
 Frequent Measurements and E n h a n c e d F u n ct i o n a l i t y

Transient adjacent-channel power (power components in the adjacent

channels produced by switching operations) cannot be correctly detected
by integration in the frequency domain. The reason is that the necessary
filter is too narrow to be compared to the channel bandwidth (1 % to 3 %
of channel bandwidth), and cannot reach steady state for t­ ransients.

*RBW 1 kHz
Mix –10 dB *VBW 10 kHz
Ref –10 dBm Att 10 dB *SWT 500 ms
–10
–20
–30 A
1RM *–40 c12
CLRW –50 c12
c11
–60 c11
cu1
–70 cu1
–80 cu2
cu2
–90 c0
–100 c0 PRN
–110
Center 1 GHz 15.2805 kHz/ Span 152.805 kHz

Tx Channel
Power –6.65 dBm Bandwidth 25 kHz

Adjacent Channel
Lower –39.60 dBc Bandwidth 25 kHz
Upper –35.12 dBc Spacing 30 kHz

Alternate Channel
Lower –69.65 dBc Bandwidth 25 kHz
Upper –68.87 dBc Spacing 60 kHz

Fig. 6-26  A
 djacent-channel power measurement without gating, here with IS-136
signal active in one slot only

202
 Channel and A d j a c e n t -C h a n n e l P o w e r M e a s u r e m e n t

*RBW 1 kHz
Mix –10 dB *VBW 10 kHz
Ref –10 dBm Att 10 dB *SWT 1 s
–10
–20
–30 A
1RM *–40
c0 GAT
CLRW –50 c0
c12 TRG
–60 c12
–70 c11
c11
–80 cu1
–90 cu1
cu2
–100 cu2 PRN
–110
Center 1 GHz 15.2805 kHz/ Span 152.805 kHz

Tx Channel
Power –2.79 dBm Bandwidth 25 kHz

Adjacent Channel
Lower –33.51 dBc Bandwidth 25 kHz
Upper –32.32 dBc Spacing 30 kHz

Alternate Channel
Lower –67.10 dBc Bandwidth 25 kHz
Upper –67.48 dBc Spacing 60 kHz

Fig. 6-27  A
 djacent-channel power measurement on IS-136 signal with correctly
set gating (spectrum due to modulation)

RBW 1 kHz
Mix –10 dBm *VBW 10 kHz
Ref –10 dBm Att 10 dB SWT 10 ms
–10
A
–20
1RM *
CLRWR TRG
–30

–40
PRN
–50

–60

–70

–80

–90
GD GL
–100

–110
Center 1 GHz 1 ms/

Fig. 6-28  Gate setting in time domain

203
 Fundamentals of S p e ct r u m A n a l y s i s

References

[1-1] Engelson, M.: Modern Spectrum Analyzer Theory and Applica-

tions. Artech House, Dedham (USA), 1984.

[1-2] Engelson, M.: Modern Spectrum Analyzer Measurements. IMS,

Portland (USA), 1991.

[2-1] Brigham, E. O.: The Fast Fourier Transform and its Application.
Prentice Hall, 1988.

[3-1] Parks, T. W., Burrus, C. S.: Digital Filter Design. John Wiley & Sons,
Inc., New York, Chichester, Brisbane, Toronto, Singapore, 1987.

[3-2] Oppenheim, A.V., Schafer, R.W.: Discrete-Time Signal Processing.

Prentice Hall, 1989.

[4-1] Helszajn, J.: YIG Resonators and Filters. John Wiley & Sons Ltd.,
Chichester, New York, Brisbane, Toronto, Singapore, 1985.

[4-2] Rauscher, C.: Frequency Range Extension of Spectrum Ana-

lyzers with Harmonic Mixers. Application Note*) 1EF43,
Rohde & Schwarz.

[4-3] Volder, J. E.: The CORDIC Trigonometric Computing Technique.

IRE Transaction on Electronic Computers, vol. EC-8, p. 330–334,
Sept 1959.

[4-4] CISPR 16-1 (1997-07): Specification for radio disturbance and

immunity measuring apparatus and methods, Part 1: Radio dis-
turbance and immunity measuring apparatus.

[5-1] Winder, S.: Single Tone Intermodulation Testing. RF Design,

Dezember 1993, p. 34.

[5-2] Freed, J. G.: Equations Provide Accurate Third-Order IMD Analy-

sis. Microwaves & RF, August 1992, p. 75.

204
 References

[5-3] Rohde, U. L.: Microwave and Wireless Synthesizers Theory and

Design. John Wiley & Sons, Inc., New York, Chichester, Weinheim,
Brisbane, Singapore, Toronto, 1997.

[5-4] Beck, B.: Understand the Effects of Retrace and Aging in OCXOs.
Microwaves & RF, November 1998, p. 72.

[5-5] Wolf, J.: Level Error Calculation for Spectrum Analyzers. Appli-
cation Note* 1EF36, Rohde & Schwarz.

* Rohde & Schwarz Application Notes are available on the Internet under

www.rohde-schwarz.com.

205
Block diagram of spectrum analyzer described in this book

tunable
RF frontend bandpass filter 1st mixer IF amplifier
3 GHz to 7 GHz 404.4 MHz
3 GHz to 7 GHz
21
20 23
Diplexer 3.4 to
6.6 GHz N
1 22
19 39
Connection 20.4 MHz
IF output
for 7 GHz model
IF filter 3rd mixer IF amplifier
Attenuator 1st mixer IF amplifier IF filter 2nd mixer IF amplifier 14 15
RF input 3476.4
8 404.4 MHz
20.4 MHz To IF signal
9 kHz
9 kHz to 3 GHz processing
(9 kHz to 7 GHz) to 3 GHz MHz
4 fO = 3476.4 MHz
9 13 f0 = 404.4 MHz 18
1 2 Connection 3 7 12 3rd LO
for 3 GHz model 16 Input for
2nd LO Overload fLO = 384 MHz
44 27 external reference
fLO = 3072 MHz detector
1st LO
5 10
10 MHz
fLO = 3.4 to 6.6 GHz
43
Overload display
17 PLL
Signal source PLL 11 Output
fCal = 128 MHz 28 reference signal
RF frontend 6 PLL 10 MHz
9 kHz to 3 GHz 1st IF 2nd IF 3rd IF
24 25
26
fRef = 10 MHz
Reference oscillator

Max Peak
36
IF filter
Bandwidth setting Min Peak 40
Log amplifier 37
lin
32 Envelope A/D lin 41 42
30 detector converter Video filter log
From log
Video A Sample µP
RF frontend 20.4 MHz
signal D
38 Trace Display
f0 = 20.4 MHz IF amplifier f0 = 20.4 MHz 33 34 35 RMS evaluation

45 29 31
Overload AV
detector 39
IF signal
Overload display processing Video signal processing Detectors and trace output