Tektronix FUNDAMENTALS OF SPECTRUM ANALYSIS TCZL/

Tektronix FUNDAMENTALS OF SPECTRUM ANALYSIS TCZL/
TCZL/ SPECTRUM
I L l X ANALYZERS
26W-5360
FUNDAMENTALS OF
SPECTRUM ANALYSIS
Tektronix
COMMITTED TO EXCELLENCE
CONTENTS:
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nature of Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Types of Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
1
2
Primary Controls . . . . . . . . . . .
Amplitude . . . . . . . . . . . . . . . .
Scale Factor (Vertical Display)
Reference Level . . . . . . . . . .
................................
................................
...............................
................................
2
2
2
3
Typical Spectrum Analyzer Controls Photo . . . . . . . . . . . . . . . . . . . . . . .
Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Frequency Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Span Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Resolution Bandwidth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4
6
6
6
6
Secondary Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sweep T i m e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Video Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
7
8
Digital Storage . . .
Frequency Range
Phase Lock . . . . .
Preselector . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8
9
9
9
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Amplitude Modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Harmonic Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Intermodulation Distortion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tracking Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . : . . . . . . . . . .
Pulsed RF(Radar) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Noise M e a s u r e m e n t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Antenna Sweeps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9
9
10
11
11
12
14
15
Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
right © 1983, Tektronix, Inc. All rights reserved.
Fundamentals of Spectrum Analysis was written by Bill Benedict,
Engineering Operations Manager, Frequency Domain Instruments, Tektronix, Inc.
Preface
This handbook explains the fundamentals, or basics, of spectrum
analysis. It describes the essential
controls and how to use them, how
to make elementary measurements,
and how to interpret the display.
There are other articles available
from Tektronix, Inc. and others, that
describe in more detail the operation of an analyzer and the interpretation of the display. After reading
this handbook, an individual familiar
with basic electronics and primary
electronic communication theory will
be able to make basic measurements with an analyzer.
For the best results, use an analyzer
when reading the text, especially the
section on Primary Controls. The
material will be much more meaningful. Trying to duplicate the photos
is the most effective way to understand the function of each control.
A multi-function signal generator will
provide most of the signals used in
the photos. Gaining the basic knowledge of how to use a Spectrum
Analyzer will make it easier to
switch from one model analyzer to
another.
This text will not discuss all the controls of an analyzer as many of them
are for special functions and will vary
between analyzers and manufacturers. The operator's manual for a
particular analyzer should be consulted regarding the exact operation of all controls.
electrical signal during the measurement interval with respect to
time. Likewise, a Spectrum Analyzer
permits observation of the amplitudes
and frequencies of the various discrete sinusoidal signals during the
measurement interval. In both cases,
the results are displayed on a cathode-ray tube (crt) with the vertical
axis being the amplitude scale and
the horizontal axis being the time
scale for an oscilloscope or the frequency axis for a Spectrum Analyzer.
Figures 1, 2, and 3 show various
waveforms as displayed on both an
oscilloscope and a Spectrum Analyzer.
where the positive peak (vertical axis)
indicates the amplitude of signal and
the single signal (horizontal axis) indicates there is only one frequency or
sine wave present. [You will note the
presence of a zero hertz marker. It is
present due to system design of a
Spectrum Analyzer and is present
regardless of the input signal. All
signals to the left of the zero hertz
marker are not negative frequencies
as one might think; they are images
or reflections of those signals to the
right of the zero hertz marker].
The second example (Fig. 2) is a
modulated carrier where both the
modulation frequency and carrier
frequency can be determined. The
Spectrum Analyzer indicates the
carrier as the larger signal and the
modulation as the two smaller signals (upper and lower sidebands).
In the first example, a sine wave is
displayed. The oscilloscope displays
the peak-to-peak voltage (vertical axis)
of the signal with respect to time
(horizontal axis). The Spectrum Analyzer shows the same sine wave
Oscilloscope Waveform: 3 MHz Sine Wave
Spectrum Analyzer Waveform: 3 MHz
Sine Wave
Figure 1.
Introduction
Nature of Measurement
All electrical waveforms or signals
are composed of a combination of
sinusoidal signals of varying amplitudes and frequencies. The combination of sine waves can be observed
in the time domain with an oscilloscope, or in the frequency domain
with a Spectrum Analyzer. The oscilloscope enables observation of
the amplitude and shape of an
Oscilloscope Waveform: Modulated Carrier
at 1 MHz, 15 kHz Modulation
Spectrum Analyzer Waveform: Modulated
Carrier at 1 MHz, 15 kHz Modulation
Figure 2.
The third example (Fig. 3) shows the
signal appearing on the oscilloscope
as a square wave. The Spectrum
Analyzer displays a "fundamental"
sine wave at the same frequency as
the square wave and the other frequencies of diminishing amplitude
(as the frequency increases) that
make up a square wave. These
other frequencies are identified as
the 3rd, 5th, 7th, etc. (odd) harmonics of the fundamental frequency.
Oscilloscope Waveform: 100 kHz Square
Wave
Figure 3.
Types of Measurements
Composite voltage waveforms are
displayed by an oscilloscope. The
Spectrum Analyzer, as the name
implies, analyzes the composite waveform and displays the individual frequency components and the relative
power each component contributes
to the total waveform.
Since the Spectrum Analyzer has
this characteristic, it is well suited
for work that involves oscillators, RF
carriers, RF spectrum surveillance,
etc. With an analyzer, it is possible
to observe:
• an oscillator
• RF carrier
• amount and frequency of modulation
• unexpected modulation
• carrier suppression in single
sideband radio
• harmonic level of oscillators and
RF carriers
With a sweeping oscillator or "Tracking Generator", filter response,
amplifier frequency response, and
antenna standing wave ratio (SWR)
can all be checked, along with other
measurements described in the Applications section dealing with the
Tracking Generator.
Primary Controls
(Refer to front panel photo on pages
4 and 5 for typical Spectrum
Analyzer controls).
Spectrum Analyzer Waveform: 100 kHz
Square Wave
Amplitude
The Spectrum Analyzer has two
major amplitude controls. The first
controls the scale factor (volts/div or
dB/div) and the second determines
what input signal amplitude is necessary to produce a signal display
up to the top line on the crt, which
is called the Reference Level.
Scale Factor (Vertical Display)
Most oscilloscope graticules are
divided vertically into eight major
divisions. Each major division is further divided into five minor divisions.
Thus, a signal of one minor division
in amplitude can be accurately
measured and another signal of
eight divisions in amplitude can be
measured and compared to determine the larger one as being
8 div (5 minor div/div)
1 minor division
= 40 times greater than
the smaller signal.
To determine this ratio in dB, use
with the standard eight division
screen to display 80 dB of range.
As an example, with 80 dB of onscreen range, two signals can be
measured simultaneously; one of 1 W
( + 30 dBm) and the other of 0.01 /WV
(-50 dBm). That is a voltage ratio
of 10,000:1 , far greater than the 40:1
ratio possible with the oscilloscope.
Before going further, note the basic
equations that can be used to convert to dB, dBm, dBV, and dBmV.
Once you begin to use the Spectrum
Analyzer, you will find that most measurements will be in dB or dBm and
no conversion will be necessary. It is
not important that you conquer these
equations before going further.
Signal ratios are expressed in dB:
or1 0 log
* Power (1)
Power into a known load (50, 75,
600 ohms, etc.) is expressed in:
Power*
dBm = 10 log
1 mW
* (at specified impedance)
= 20 log ^ = 32dB.
Since many Spectrum Analyzers are
capable of displaying ratios of 80 dB
on screen, either a different scale
factor is required or a crt display
with 2,000 major vertical divisions is
required! The obvious solution is to
use a logarithmic scale of 10 dB/div
dBV = 20 log
dBmV = 20log
1V
vm
1 mV
* (volts are RMS volts)
The obvious problem with having a
scale factor that allows such a large
range of signals on screen simultaneously is that two signals appear-
ing close in amplitude may in reality
vary significantly in amplitude. As an
example, assume there is one signal
of 1 mW and another signal of 2 mW
power. Using the equations, it is
apparent they are
=10log2 =
1 mW
apart in amplitude, or 1.5 minor divisions with a scale factor of 10 dB/div.
To allow accurate measurements of
signals of close amplitudes, an analyzer typically has a Display Mode of
2 dB/div where, as in the previous
example of two signals being 3 dB
apart (2X), the display would indicate
1.5 major divisions of separation. A
third common display mode is linear
scale factor, where the RMS value
of the signal is displayed with a calibration of volts/div.
OPTIMUM INPUT
LEVEL -30 dBm
( + 13 dBm MAX)
+ 30 dBm MAX
STEP
ATTENUATOR
10 log
Reference Level
The Reference Level is one of the
three main controls of a Spectrum
Analyzer. The purpose of this control
is to obtain an adequate display of
signal amplitude on screen. This
control sets the level of the signal
necessary to produce a full-screen
deflection (i.e., the top of the screen
is the Reference Line). Thus, if the
Reference Level control is set for 0
dBm with a Vertical Display of 10
dB/div, a 0 dBm signal would rise
to the top crt graticule marking. A
-20 dBm signal would be 2 divisions down from the top [0 dBm
- 2 div (1 0 dB/div) = - 20 dBm].
Some analyzers separate the Reference Level control into two individual controls. Together they represent
the Reference Level, but separately
each controls an individual section
of the analyzer. The two independent
sections of the analyzer are the RF
Attenuator control and the IF Gain
control.
The RF Attenuator control selects
the amount of RF attenuation the
signal experiences just after it enters
the analyzer. For optimum analyzer
performance, the input signal must
be attenuated to a level specified by
CONTROL
Figure 4. Spectrum Analyzer Input Indicating Point of Optimum Input Level.
the manufacturer for the 1st mixer
(optimum input level, see Fig. 4). For
example, Tektronix 490 Series Spectrum Analyzers have an optimum
signal level for the 1st mixer of -30
dBm. Therefore, if the signal being
measured with the analyzer is -10
dBm in level, the RF attenuator
should be set for 20 dB of attenuation. The first mixer would then see:
-10 dBm (input) -20 dB (attenuation) = -30 dBm (1st mixer level).
without damage. Therefore, if you
are unsure of the level of the input
signal, select the largest RF attenuation available. Once the signal is
displayed on the screen, the attenuation can be removed one step at
a time to bring the largest signal to
the top of the screen. Typically, if
the input is less than 0 dBm, the analyzer will not be damaged regardless of how the Reference Level
controls are set.
The IF Gain control selects the proper amount of gain within an amplifier stage to keep the instrument
within amplitude calibration. This
control does not have any restrictions for proper operation.
Most RF power meters indicate the
total amount of power available at
the head of the power meter from
all signals present on the cable.
Thus, if there are many discrete
sinusoidal signals present on the
cable, the amplitude of any one
signal cannot be determined with
the power meter. The Spectrum Analyzer allows each signal to be viewed
separately for both amplitude and
frequency. However, the input (attenuator and 1st mixer) circuitry is
like the power meter in that it is exposed to all signals present. Therefore, the rules regarding maximum
input level apply to the sum of all
signals present on the input, regardless of whether they are all being displayed on the screen or not.
As an example, if two signals of +20
dBm and one of -50 dBm are present on a cable, the input circuitry
is actually being exposed to over
+ 23 dBm. Remember (from a previous example), if you double the
power, you have a signal level 3 dB
higher [(+ 20 dbm) + (+ 20 dBm) >
( + 23 dBm)]. With over +23 dBm on
Some analyzers, like the Tektronix
490 Series, contain a microprocessor
that selects the proper ratio of RF attenuation and IF gain, depending on
the Reference Level selected. This
eases operator responsibility, because the operator is only required
to keep the signal at or below the
top graticule line by selecting an
appropriate Reference Level.
All analyzers have a maximum input
level that must be observed. Typically,
this level is +30 dBm (1 W). It is extremely important to observe this
limit, because extensive and expensive damage may occur to the
input circuitry. Usually, both the RF
attenuator and the 1 st mixer have a
maximum input level, and quite often
they are not the same level. The RF
attenuator can handle a significantly
larger signal level than the 1st mixer
VERTICAL DISPLAY:
SCALE FACTOR FOR
VERTICAL DEFLECTION (i.e.
Y-AXIS SCALE FACTOR)
TRIGGERING:
SELECTS MODE OF
TRIGGERING SWEEP
TIME/DIV:
SELECTS RATE AT WHICH
FREQUENCY SPECTRUM IS
ANALYZED
CALIBRATOR:
PROVIDES ACCURATE
SIGNAL FOR AMPLITUDE
AND FREQUENCY
CALIBRATION
REFERENCE LEVEL:
DEFINES AMPLITUDE (LEVEL)
OF SIGNAL NECESSARY FOR
FULL SCREEN DEFLECTION
F
D
V\
0
FREQUENCY:
DEFINES FREQUENCY
WHICH REFERENCE DOT
ON SCREEN REPRESENTS
FREQUENCY RANGE:
SELECTS FREQUENCY BAND
OF ANALYSIS (INDICATED
ON LOWER SCREEN
READOUT)
FREQUENCY SPAN/DIV:
CONTROLS MAGNITUDE OF
FREQUENCY SPECTRUM BEING
ANALYZED (i.e., X-AXIS
SCALE FACTOR)
RESOLUTION BANDWIDTH:
DEFINES ABILITY OF
ANALYZER TO IDENTIFY
ADJACENT SIGNALS
DIGITAL STORAGE:
SELECTS MODES OF
ACQUIRING AND DISPLAYING
SIGNALS FOR PRESENTATION
ON SCREEN
the input and a 1 st mixer that works
best with -30 dBm, we need 53 dB
of attenuation for optimum operation
[ + 23 dBm (input) - 53 dB (attenuation) = -30 dBm (1st mixer signal
level)]. If the analyzer is tuned to
shift the larger signals off screen,
the RF attenuation still cannot be
removed to shift the -50 dBm signal up on screen for better viewing,
because the input circuitry is still
being exposed to the two larger
signals. However, IF gain may be
added to increase the displayed
level of the smaller signal.
Unlike an oscilloscope, a Spectrum
Analyzer is ordinarily susceptible to
damage from dc voltages. This is
extremely important to remember.
If a dc voltage can be applied to an
analyzer, it will usually be indicated
on the front panel near the input
connector. If dc voltage is a possibility, always use an external Blocking Capacitor. Suitable blocking
capacitors with good VSWR are
available from several vendors.
lyzer that defines the Span control
as just "span" will sweep that many
"Hz" across the screen.) As an example, a span of 1 MHz/div, would
sweep across a frequency spectrum
of 10 MHz. Just exactly which 10
MHz would depend on the Frequency control. If the Frequency
control was set for 100 MHz, then
the analyzer would sweep from 95
MHz to 105 MHz (see Fig. 5).
In Fig. 5, note that the large signal
is 100 MHz in frequency and has a
level of -17 dBm. The smaller signals are at 98 MHz and 102 MHz at
a level of -62 dBm. Since the smaller
signals are symmetrical about the
center signal, they could be the modulation of the carrier at 100 MHz. In
that case, the above example would
be referred to as a "signal" or "carrier" at 100 MHz with 2 MHz sidebands down 45 dB from the carrier
(or -45 dBc). (The term dBc means
below the carrier.)
Frequency
Frequency Control
The Frequency control is the second
of the three main controls. This control identifies the frequency of a particular point on the display. Customarily,
this is the center of the screen. In
some modes of operation, however,
it could be some other point on the
screen. On many analyzers, there is
a dot or other indication on the display that indicates the point on
screen that represents the specified frequency.
Span Control
The Span control or Span/div control
is the third of the three main controls.
With this control, the width of the frequency spectrum being analyzed
can be varied. When referred to as
Span/div, it indicates "X" Hz/division;
therefore, a 10 division screen would
be sweeping across a frequency
spectrum of 10x"X" Hz. (An ana-
Figure 5. With Frequency control set at 100
MHz and a Span of 1 MHz/div,
the displayed spectrum extends
from 95 MHz to 105 MHz.
The Span control has two settings
that are not calibrated in hertz. Turn
this control clockwise to eventually
reach a position of maximum (MAX)
span. In this position, the analyzer
sweeps across its maximum frequency spectrum for the band of
frequencies selected. In a "band"
that extends from 0 Hz to 1800 MHz,
the analyzer sweeps the frequency
spectrum from 0 Hz to 1800 MHz
to look for signals when in MAX
span. Although the analyzer is only
specified from 50 kHz to 1800 MHz,
a certain amount of oversweep is
common. Turn the Span control
counterclockwise, and the spans get
smaller and smaller in frequency until
the "zero" span position is reached.
In this position, the analyzer no longer
sweeps across a frequency spectrum,
but behaves like a superheterodyne
receiver. The analyzer now basically
works like a typical oscilloscope
where the display indicates the
modulation of any signal at the frequency selected by the Frequency
control.
Resolution Bandwidth (RBW)
Ideally, the display or graph of amplitude vs. frequency should be vertical
lines of minimum width to allow signals of very close frequency spacing
to be individually discernible as
shown in Fig. 6. Note the pair of
sidebands located very close to the
carrier. If a wide pen had been used
to draw the figure, as in Fig. 7, the
sidebands might have been overlooked as denoted by the slight
width change near the bottom of the
carrier. Resolution Bandwidth (RBW)
performs much the same function as
varying the width of the pen when
plotting the display on the screen.
As the frequency spectrum being
displayed on screen varies as a function of the span/div, the width of the
"pen" that is calibrated in hertz must
also change. If an extremely narrow
"pen" is used with an extremely
wide frequency span, signals will
appear very narrow and may be
overlooked.
Most modern Spectrum Analyzers
through the use of microprocessors
have the capability to select the optimum bandwidth (resolution bandwidth) depending on the span/div
and time/div selected. There will be
times, however, when manual control of this function will be desired.
Figure 6. Spectral graph drawn with fine tip
pen clearly showing closely
spaced signals.
Figure 7. Spectral graph drawn with broad
tip pen masking closely spaced
signals.
Resolution Bandwidth is a functional
control that selects one of several
bandpass filters physically located
in the instrument's Intermediate Frequency (IF) chain. It is defined in the
term Hertz (Hz) and is a measure of
the width of the filter either 3 dB or
6 dB down from its peak, depending
on analyzer manufacturer. The shape
of the signals being traced out on
screen are, in reality, a combination
of the shape of the Resolution Bandwidth filter, and the signal, not just
the shape of the signal being
analyzed.
The limitations imposed on an analyzer by the Resolution Bandwidth
filter are significant. Sweep speed
(the rate the analyzer sweeps through
the frequencies present) must be slow
enough to allow the filters to reach
peak amplitude or an inaccurate
signal amplitude will result. When
analyzing pulse type signals such
as radar, Resolution Bandwidth is
very important or erroneous results
will be obtained. This application will
be covered in more detail in the
Applications section.
Unless a special requirement dictates
a specific Resolution Bandwidth, the
Resolution Bandwidth selected should
be somewhat greater than 1/50 the
span/div. Figs. 8 and 9 show the two
extremes of useful Resolution Bandwidth for a particular span. In each
case, the signal being displayed is
the same, with only the Resolution
Bandwidth of the analyzer changing
between the two figures. Fig. 8 is
displayed with an extremely wide
RBW for the span/div selected. Fig. 9
has a more optimum RBW selected,
and we can now see sidebands on
the signal that were not visible in
Fig. 8. If the bandwidth continued
to narrow, the sweep speed of the
analyzer would have to slow down
to allow the signal to trace the correct amplitude through the filter, and
the display would be less viewable.
Figure 8. 1 MHz RBW. Wider Than
Optimum Bandwidth
Another characteristic not yet mentioned, which works in our favor, is
that as the RBW is decreased, the
noise floor of the analyzer goes down.
(The term noise floor refers to the
baseline or lowest horizontal part of
the trace. Because of its appearance,
this part of the signal is sometimes
referred to as the "grass".) For each
decade decrease in RBW (e.g. from
100 kHz to 10 kHz), the noise floor
of the analyzer decreases by 10 dB.
This is extremely important when
looking for very small signals.
Figure 10 is a composite of two
RBW's that show a signal which
was initially buried in the noise. The
only parameter changed is the RBW,
which in effect, pushed the noise of
the analyzer below the level of the
signal's sidebands.
Figure 10. Composite photograph illustrating
sidebands obscured by noise in
wider resolution bandwidth.
Secondary Controls
Sweep Time
Figure 9. 100 kHz RBW. Optimum
Bandwidth Showing Signals
Masked By Filter Skirts in
Figure 8.
Like Resolution Bandwidth, most
newer analyzers have an Auto position in which a microprocessor selects
the optimum sweep speed, depending on other parameters. When analyzing a frequency spectrum, this
control determines the rate at which
the analyzer sweeps through the determined spectrum. If the spectrum
is swept too fast, the RBW filters may
ring or fail to reach full amplitude. If
swept too slow, there are no disadvantages, unless the analyzer does
not have "digital storage" or some
form of waveform storage. Without
storage, by the time an extremely
slow sweep is complete, the operator could have forgotten the content
of the original spectrum.
When the analyzer Span/div control
is set for Zero Span, the Sweep control functions like an oscilloscope's
time control. As previously described,
the display is a time domain presentation of the modulation at the center
frequency selected when in Zero
Span.
Video Filter (sometimes referred
to as a Noise Averaging Filter)
This filter is used primarily as a
smoothing filter to remove or smooth
out the short duration noise spikes at
the bottom of the display. When the
analyzer is in Auto sweep speed,
note that the sweep rate decreases
when a Video Filter is turned on. In
most analyzers there are usually
several Video Filters to choose from.
Care must be taken, much like selecting the Resolution Bandwidth
filter.
When analyzing a signal such as
pulse radar or if the Resolution Bandwidth is very narrow for the span (i.e.,
narrow signals displayed on screen),
the Video Filter should not be selected, as this will not allow the
amplitude of the analyzed signals
to reach full amplitude due to its
video bandwidth limiting property
(i.e., a low-pass filter).
Digital Storage
In many older Spectrum Analyzers,
a storage oscilloscope was used as
the display. This was necessary
because of the slow sweep speeds
required to maintain amplitude calibration. With the advances in digital
hardware, it is now possible to divide
the screen into small horizontal segments and digitize the amplitude as
the analyzer sweeps through each
segment and store the data in RAM
(random access memory). This data
can then be accessed, converted to
analog signals, and sequentially displayed on the screen in the proper
horizontal sequence at slightly above
flicker rate.
This procedure of digitizing a signal
occurs after the signal has been processed by the Resolution Bandwidth
(RBW) circuitry, the Logarithmic (log)
circuitry, and Video Filter circuitry. It
usually occurs just prior to being
amplified for the crt.
Once the data is stored in RAM, we
usually have an option as to the
method of display. If we desire to
"SAVE" a particular waveform (e.g.,
"A" waveform), we can select the
"SAVE A" function and the "A
memory" within the analyzer will be
frozen and not updated. The B waveform in memory will continue to be
updated with each sweep of the
analyzer; thus we would view separate traces on the screen. If the
"SAVED" trace was not needed for
immediate viewing, the "VIEW A"
function could be disabled and the
"A memory" would not be displayed
on the screen; but, its data would
still be available for future reference.
If the "VIEW B" function is simultaneously disabled, the display portion
of Digital Storage is disengaged and
the sweeping signal of the analyzer
will be displayed, with the refresh
rate determined by the Time/Div
control.
The "B-SAVE A" control is used to
display the difference between two
waveforms. As the name implies, it
subtracts the SAVED A waveform
from the active B waveform. This
function is most useful when the
analyzer is used with the Tracking
Generator. (This application will be
discussed in the Applications section dealing with the Tracking
Generator.)
The "MAX HOLD" function is used
to capture the maximum Y deflection
(amplitude) for any X axis position
(frequency), regardless of how many
sweeps must be made to capture
these extremes. This is accomplished
by the digital storage digitizing new
amplitude data for a particular point
on screen, then checking the amp-
litude in memory for that specific
memory location and saving the
larger of the two. The usefulness of
"MAX HOLD" is in capturing a frequency spectrum where a signal
randomly appears, then disappears.
Once the signal has been analyzed
and stored, the digital storage will
continue to display the signal, regardless of whether or not it reappears on succeeding sweeps.
A different application might be to
monitor an FM'ing or drifting signal
and note the frequency excursions.
This can be accomplished by selecting the desired frequency carrier and enabling the "MAX HOLD"
function. On each succeeding sweep,
the analyzer will analyze the carrier
at its precise frequency at the moment of analysis and save this value
in memory. With repetitive sweeps,
the maximum excursions will be filled
and viewable for analysis. It is important to check the drift specifications of the analyzer to ensure the
analyzer is more stable than the signal to be checked.
The Peak/Average cursor is used to
determine data processing prior to
loading the digitized information in
RAM for a particular horizontal point
on the screen. For each horizontal
point on the screen (of which there
are 1000) the digitizer may digitize
from 2 to 10,000 samples (depending on the sweep speed) to represent the Y value to be stored for a
particular horizontal point. If the
amplitude of the signal at this horizontal point is above the cursor, the
storage will select the maximum
value digitized and load this number in memory; thus, the term "Peak
Detect". When the amplitude of the
signal at this horizontal point is located below the cursor, the digital
storage will take the mathematical
average of the digitized numbers
and load this number in memory;
thus, the term "Average Detect".
The necessity for having a "Peak/
Average" function is to ensure that
the maximum value of a narrow
pulse can be stored to represent the
maximum amplitude of that pulse,
and "noise" or "grass" can be
averaged before storing in RAM to
offer the maximum possible signalto-noise ratio.
Frequency Range
This function operates much like a
"Band Select" switch on a shortwave receiver. Each succeeding
selection of either the "up" or "down"
control will place the instrument in a
higher or lower frequency band of
operation.
Phase Lock
An analyzer usually has two or more
internal oscillators, one or more of
which will be swept (or moved) as
the analyzer is sweeping through a
frequency spectrum. When in wide
spans, such as 100 kHz/div or greater,
a slight amount of drift in one of the
internal oscillators is usually not noticeable. However, as the span is
reduced to several kHz/div or less,
the instability of the internal oscillators
becomes apparent. The screen indication is of an apparently drifting
signal, when the real problem is a
drifting oscillator within the analyzer.
Therefore, when an analyzer is operating in the narrower spans, the
oscillator causing the drift problem
is typically phase locked to a stable
reference to prevent the drift. In
wider spans/div, this oscillator is
typically being swept; therefore, it
cannot be locked at all times. When
the phase lock circuitry is operating,
a front-panel indicator will typically
inform the operator. This indication
requires no action on the user's part
and will usually not affect the measurement in an adverse way.
Preselector
A Preselector is a filter located just
slightly behind the input connector.
The function of the filter is to select or
allow only a narrow band of frequen-
cies to pass into the analyzer. It is a
sweeping filter that tracks the frequency the analyzer is tuned to at
any particular point in time. The function it performs is to inhibit harmonic
mixing within the first converter. By
eliminating the harmonic conversions,
unwanted mixing products do not
appear as signals in the spectrum.
In addition, any large signals (up to
+ 30 dBM) present on the input but
out of the frequency range being
analyzed are prohibited from reaching the input mixer, thus eliminating
the need to use attenuation to protect the input mixer from burnout.
The preselector is almost transparent
to the user, except that it needs to
be "peaked" occasionally. This is
usually accomplished with a frontpanel control. The "Peaking Control"
allows the user to offset the tracking
filter slightly forward or backward
with respect to the frequency the
analyzer is tuned (i.e., to be centered around the tuned frequency).
If the filter is mis-peaked and is
completely offset from the tuned
frequency, the analyzer will indicate
a complete lack of signals in the
preselected bands. Preselection
occurs only in bands 2 through 5
(1.7 — 21 GHz) in the 490 Series of
Spectrum Analyzers.
From this display, the frequency
of the carrier (fc) and the frequency
of the modulation (fm) can also be
determined. The percent of modulation can also be determined by noting the difference in amplitude (12
dB) between fc and fm and using
the table in Fig. 13.
Figure 11. AM Modulation (50%) in time
domain.
•REOUENCV
SPAWDIV
Figure 12. AM Modulation in frequency
domain.
Applications
Amplitude Modulation
(AM) Notes
i Modulations 200 x 10
- (dB down from Carrier)
20
An Amplitude Modulated signal,
when viewed in the time domain
(as with an oscilloscope), might appear as in Fig. 11. From this photo,
we can determine the frequency of
the carrier (fc) and the frequency of
the modulation (fm). In addition, the
percent of modulation can be calculated from the equation:
%Mod =
x100.
Figure 12 represents the same signal being displayed in the frequency
domain on a spectrum analyzer.
Figure 13. Chart of dB vs. °/o of Modulation.
Figures 11 and 12 were prepared
under controlled test conditions. In
normal operation, the modulation
will not be a pure sine wave, but
will be a composite of multiple sine
waves, and their frequencies cannot
be determined in the time domain.
However, the Spectrum Analyzer will
accurately display all frequencies
present.
A suppressed carrier system would
be displayed on the analyzer as in
Fig. 14. The typical measurements
to be made in this system would be
carrier suppression. The measurement is the difference in carrier amplitude between when the carrier is
turned on and when it is turned off.
Fig. 14 indicates the carrier is suppressed by 40 dB.
VARIABLE
FREQUENCY
OSCILLATOR
Figure 15. Test setup for sweeping audio flatness.
system described, and Fig. 16 is a
photo of such a sweep.
From Fig. 16 we can see the system
flatness is 1.3 dB (which in reality may
be a type of emphasis placed on
the audio), and the system 3 dB
bandwidth is in excess of 8 kHz.
Both lower and upper sideband envelopes should be symmetrical. If
the transmitter were seriously mistuned or was working into a poor
antenna match, the Spectrum Analyzer would show how each sideband was individually affected.
modulator should be driven to a
specified percent of modulation
and the Spectrum Analyzer should
be checked for the presence of only
the signal that is put into the modulator. If harmonic distortion is occurring, additional products will appear
on the screen at multiples of the
modulating frequency. Fig. 17
shows the result of a modulator being driven with a 5 kHz test signal.
Harmonic distortion will show up as
signals at 10 kHz, 15 kHz, 20 kHz,
etc. from the carrier. The Total Harmonic Distortion (THD) can be determined by noting the amplitude
difference between the modulation
signal and its harmonic products.
-igure 14. Suppressed Carrier System
Similarly, if the lower sideband was
suppressed as well, we could determine the amount of this suppression
by noting the difference in amplitude between the upper and lower
sideband.
Another type of measurement that
could be made on an AM system
would be to check system flatness
by sweeping the audio input with an
audio generator of known or verified
flatness. The RF carrier could then
be monitored in a narrow span/div
and a deflection (scale) factor of 2
dB/div. By using the MAX HOLD
function, we could construct a waveform to indicate the flatness of the
total system. This waveform would
also indicate any "emphasis" placed
on the audio. Figure 15 shows the
Figure 16. Audio flatness of AM system as
measured at RF frequency.
Distortion
Distortion is the result of electronic
circuits operating in a non-linear
mode. Two of the most common
methods of checking for distortion
involve driving the equipment with
known signals and monitoring the
equipment output for signals other
than those present at the input.
(Harmonic Distortion) A typical
Harmonic Distortion measurement
would be set up as in Fig. 15 with
the variable frequency oscillator set
at some specified frequency. The
10
Figure 17. Harmonic Distortion at 10 kHz,
15 kHz, and 20 kHz from Carrier.
The sum of all the harmonic products
must be used to determine the percent of harmonic distortion. The Total
Harmonic Distortion (THD) can be
determined by noting at what level
below the fundamental each harmonic lies, and determining the
percent ratio for each harmonic
from Fig. 18 and substituting in the
following equation. This equation is
only accurate if the upper and lower
harmonic pairs are within one or two
dB of each other.
THD(%)= V(2nd Harmonic %)2
+ (3rd Harmonic %)2
+ (4th Harmonic %)2
+ etc.
From Fig. 17, the THD is
V0.03252 + 0.0122 + 0.00182
= 0.035 = 3.5%
When making this measurement, it
is important to be sure the modulating signal from the audio oscillator is
free from any harmonics. To do this,
check the signal source with a
Spectrum Analyzer.
10
20
30
40
50
60
dB BELOW FUNDAMENTAL
Spectrum Analyzer should be used
to check the output of the directional
bridge or combiner for any signals
other than those applied prior to
modulating the transmitter. The frequency of the modulating signals
depend on the type of test to be
performed and the type of equipment being checked. Our example
uses a 4 kHz (f-|) and 5 kHz (f2) signal. There are multiple IM products
created, of which the first one is
called the second order IM product,
which will occur around the carrier
at f 1 +f 2 , f-i -f 2 and/or f 2 -f 1 (9 kHz
and 1 kHz from the carrier). The third
order IM products will occur at
2f-| +f 2 , 2f-| -f 2 , 2f2 + f 1 , and/or
2f 2 -f 1 (13 kHz, 3 kHz, 14 kHz, and
6 kHz, from the carrier.) Fig. 19
shows a typical response and identifies the various 2nd and 3rd order
products.
70
Figure 18. dB below Fundamental to '
Distortion.
(Intermodulation Distortion) An
additional measurement common
to amplifiers or transmitters is the
Two Tone Intermodulation Distortion
(IM) test. This test is similar to the
Harmonic Distortion check, except
it requires an additional audio signal generator. The two audio generators are combined, and the result
is applied to the modulator. The
method of combining the two signals is very important, as mixing the
two sources with each other can
create unwanted products. Combining should occur in a directional
bridge. A "T" connection or combiner can be used, provided each
generator is sufficiently padded. A
Figure 19. Distortion showing input signals
and both Harmonic Distortion
(2f,, 2f2, etc.) and IM products
(second order: f2-'1, \\ + f2),
(3rd order: 2f 1 -f 2 ,2f 2 -f l ,
2f, + f2, 2f2 + f1), plus higher
order products.
For more information on these and
other tests on AM systems, see
Tektronix, Inc. Application Notes
AX-3266, "AM BROADCAST MEASUREMENTS USING THE SPECTRUM ANALYZER", and 26W-4889,
"NO LOOSE ENDS — REVISED.
THE TEKTRONIX PROOF OF PERFORMANCE PROGRAM FOR
CATV."
Tracking Generator — TG
(with Spectrum Analyzer
for Swept Measurements)
A Tracking Generator (TG), when
used in conjunction with a Spectrum
Analyzer (SA), allows such items as
filters, amplifiers, couplers, etc. to be
observed with respect to frequency
(i.e., Frequency Response). This is
performed by connecting the output
of the TG (TG output frequency is
synchronized to frequency being
analyzed by Analyzer at any point
in time i.e., "Tracking Generator") to
the input of the device being tested,
and monitoring the output of the
device with the SA (as shown in
Fig. 20). This type of measurement
is known as an S12 magnitude only
measurement, since the phase shift
of the signal through the device is
not displayed.
The response displayed on the
screen of the analyzer will be a
combination of the unflatness of the
TG and the response of the device
being tested. The unflatness of the
SPECTRUM ANALYZER
INTERFACE
CABLES
TRACKING
GENERATOR
o
DEVICE
UNDER TEST
Figure 20. Spectrum Analyzer (SA) and Tracking Generator (TG) Test Setup.
11
TG/SA can be removed by using the
"B-Save A" function of the SA. First,
connect the TG to the SA and save
the flatness (or unfiatness) of the
TG/SA in the A memory by using
the SA "Save" function, and using
the "Vert Display" mode that will be
used in the measurement. Then,
connect the TG to the device being
tested and monitor the device with
the analyzer. Once a sweep has
been made, the analyzer display
will indicate the system response.
By activating B-Save A, the saved unfiatness of the TG will be subtracted
from the response of the system,
and the corrected display will indicate the corrected frequency response of the device being tested.
only capable of 7 dB of ultimate due
to a re-entrant mode.
For more information on the subject
of SA/TG measurements, see
Tektronix, Inc. Application Note
26W-5121, "THE TRACKING GENERATOR/SPECTRUM ANALYZER
SYSTEM."
Figure 22. Filter response at wide sweep
indicating re-entrant mode.
Crystal
Fig. 23 shows the response of a
The photographs that follow indicate
the typical responses of the systems
shown.
crystal. The series resonance (fs)
and parallel resonance (fp) frequencies are identified on the photo. Also
Filter
note the crystal spurs located between 300 kHz and 400 kHz above
the crystal resonance.
Figure 21 shows a 9 MHz filter being swept with a SA/TG system. We
can determine the filter loss as be-
ing approximately 8 dB by noting
the difference in amplitude between
the TG response and the filter response. The filter is approximately
400 kHz wide 3 dB down from the
peak. Note the unsymmetrical shape
near the base of the filter. The filter
ultimate is better than 68 dB.
monitoring the output level for 1 dB
ncreases to determine the 1 dB
compression point (approaching
saturation where output does not
follow input with linear changes).
For more information on crystal testing, see Tektronix, Inc. Application
Note AX-3525, "CRYSTAL DEVICE
MEASUREMENTS USING THE
SPECTRUM ANALYZER."
Figure 24. Amplifier response to SA/TG
system.
Pulsed RF (Radar)
Pulsed waveforms when viewed in
the time domain will appear as in
Fig. 25. Different types of modulators will generate different types of
pulses, but the commonality is a
carrier that is turned on for a period
of time, then off for a specified period The period of time the pulse is
on will be referred to as (tpw) and
the pulse repetition rate will be f r .
RESOLUTION
Figure 23. Crystal response to SA/TG
system.
Amplifier
Figure 21. Filter response of filter using
SA/TG.
Fig. 22 shows the same filter as
shown in Fig. 21 when being swept
over a wider frequency range. The
filter is being tested from 0-900 MHz.
At 350 MHz we can see the filter is
In Fig. 24, an amplifier is being
tested. The input is at -40 dBm
and the output is at - 10 dBm, thus,
a gam of 30 dB is realized. The 3 dB
roll-off is m excess of 1100 MHz. The
flatness up to 1100 MHz is less than
3 dB peak-to-peak. Further tests
might include increasing the level
of the input signal in 1 dB steps and
12
Figure 25. Time domain display of pulse
waveform.
The same pulse waveform when
displayed in the frequency domain
would appear as shown in Fig. 26.
Note that the pulse width (tpw) and
the repetition rate (fr) can be determined from the spectral display.
Figure 26. Frequency domain display of
pulse waveform.
sweep time/div 10 ms
* r= #pulses/div =H'0~=1 ms
1______1
W - lobe width ~ 200 kHz ~ 5 ^
In the introduction, we learned that
all waveforms can be described as
a combination of various sinusoidal
waveforms of differing amplitudes.
The pulses in Fig. 25 are likewise
composed of an infinite number of
discrete sinusoidal frequencies of
differing amplitudes. Since there are
an infinite number of signals, we are
primarily interested in the envelope
of the amplitude of the signals. In
our example, this is described by a
sin x
x
display shown in Fig. 27. We can
see that the amplitudes in Fig. 26
lie within the area described by Fig.
27. The big question is "Why do we
see discrete signals in Fig. 26 if the
waveform is composed of an infinite series of frequencies?"
Figure 27. Envelope of pulsed signal
/ sin x
\
I —— envelope I
The answer lies in the fact that swept
frequency analyzers only analyze a
specific frequency at a specific time
as the beam traces across the screen.
Each time a pulse is generated, the
analyzer will analyze the amplitude
of the frequency component at the
frequency being analyzed at that
instant. If the pulse repetition period
(tr) of the pulse was 1 ms and the
analyzer was sweeping through the
frequency spectrum at 1 ms/div, we
would see one spectral line/div. If
we slowed the sweep speed to 100
ms/div, we would obtain 100 spectral
lines/div, which would clearly show
the envelope display of Fig. 27. We
need to remember that although we
are varying the sweep speed, we
are not changing the span of frequencies being analyzed, just the
rate at which we are analyzing them.
To compute the repetition rate from
Fig. 26, determine the number of
divisions/spectral line and multiply
by the sweep speed/division.
To display an optimum waveform of
pulsed RF, the Resolution Bandwidth
(RBW) should be selected narrow
enough to display each spectral
line. As the RBW is narrowed, the
amount of energy from the pulse
reaching the detector within the analyzer is reduced and the display will
indicate a lower level signal than is
actually present. The optimum
Resolution Bandwidth (RBW) is approximately 0.1/pulse width (tpw) or
t p w xRBW equal-to-or-less-than 0.1.
Figure 28 shows the optimum RBW
as a function of pulse width, and Fig.
29 shows the approximate sensitivity
loss or signal amplitude loss as a
function of the product of tpw x RBW
(pulsewidth x Resolution Bandwidth).
Note that the type of Resolution
Bandwidth Filter in the analyzer
will vary the amount of loss between
Pulsed RF and a CW signal of equal
amplitude. 490 Series filters are of
the rectangular response. Since there
is a signal loss through an analyzer
due to RBW limitations, it is important
to remember that the front end of the
13
analyzer is being driven harder than
the signals on the screen indicate.
Care should be taken not to overdrive the input mixer.
600
400
*-_
OJ
^
(£> CO/-
CVJ
T
(D COO
O
O
PULSE WIDTH (!,„)- pi"
Figure 28. Resolution bandwidth setting
for pulsed RF computed from
[t pw ]xRBW = 0.1
.001
.01
.1
1
PULSEWIDTH-BANDWIDTH PRODUCT (t^xRBW)
Figure 29. Sensitivity loss of pulsed
signals vs CW.
For best results when analyzing
pulsed RF, Digital Storage should
be disabled until the optimum combination of sweep speed, span/div,
RBW and Reference Level have
been achieved. Once the desired
waveform has been acquired, the
storage can be activated with the
Peak/Average cursor placed at the
bottom of the screen. The Auto
Sweep speed (time/div) and Auto
RBW should not be used, as the
algorithm used to compute the optimum setting is not valid for pulsed
RF.
Typical observations of an RF spectrum would be the following:
1. For a rectangular pulse, the 1st
sidelobe should be approximately
13.3 dB below the main lobe
(see Fig. 26).
2. If the nulls are not well defined,
the pulsewidth (tpw) is FM'ing
(see Fig. 27).
3. Poor carrier on/off ratio shows
up as a response buried under
the main lobe (see Fig. 30).
4. If the carrier is FM'ing, the lobes
could be unsymmetrical (see
Fig. 31).
Figure 30. Note void in main lobe and
pulse extension on top of main
lobe caused by poor carrier
on/off ratio.
For further information on this subject, see Tektronix, Inc. Application
Notes AX-4217, "PULSED RF
SPECTRUM ANALYSIS" and
AX-3259, "NOISE MEASUREMENTS
USING THE SPECTRUM ANALYZER
— PART TWO; IMPULSE NOISE."
Noise Measurements
Noise measurements are often
made as carrier-to-noise (C/N) measurements, oscillator spectral purity,
white noise level, etc. The noise referred to is the level of the baseline
signal or "grass" of a spectrum display. The unit of measure when dealing with random noise is usually
dBm/Hz or Watts/Hz. The noise bandwidth must always be specified, because each decade of change in
noise bandwidth will vary the measurement by 10 dB. Random Noise
implies the noise is being analyzed
through an idealized square-shaped
filter. Since most filters are not of the
idealized square shape, a correction
factor may have to be generated to
convert from the Spectrum Analyzer's Resolution Bandwidth (RBW)
to the effective Noise Bandwidth of
each filter. This correction is ex-
plained in Tektronix, Inc. Application
Note AX-3260 "NOISE MEASUREMENTS USING THE SPECTRUM
ANALYZER - PART ONE: RANDOM
NOISE." If this correction is not
made for the RBW used, errors of
up to 2 dB can occur in the measurement.
Another source of error when making noise measurements occurs in
the detector and logarithmic circuitry.
These two errors cause the measured
noise to appear lower in level than
the actual noise by the following
factors:
LIN display mode: 1.13 dB
LOG display mode: 2.5 dB
An additional source of error involves
dealing with very low level signals or
in this case, system noise located
close to the noise floor of the analyzer. To test for the analyzer's noise
floor, disconnect the input and note
the amplitude of the noise. When a
signal or system noise is located
within 10 dB of the analyzer noise
floor, the amplitude of the measured
signal or system noise will be indicated as being higher than it really is
by a factor as determined in Fig. 32.
INDICATED SIGNAL
AMPLITUDE ABOVE
ACTUAL SIGNAL
AMPLITUDE
Figure 31. FM'ing carrier.
WARNING
Radar applications require relatively
large amounts of power for proper
operation. Signal access points on
radar systems often have large signal levels that can be lethal to both
people and Spectrum Analyzers.
The input circuitry of Spectrum
Analyzers is fragile. Use caution
and plenty of external attenuators
when observing unknown signals.
2dB
4dB
6dB
8dB
10 dB
12 dB
DIFFERENCE BETWEEN ANALYZER NOISE FLOOR
AND LOW AMPLITUDE SIGNALS (OR SYSTEM NOISE) IN dB
Figure 32. Amplitude correction for signals located within 10 dB (low level signals) of analyzer
noise floor.
14
In our example of Fig. 33, the difference of analyzer noise floor to
system noise floor is 5 dB. From
Fig. 32, a correction factor of 1.7
dB must be subtracted from the indicated system noise amplitude to
obtain the true noise level. (Remember that two signals of the same amplitude will indicate 3 dB more
power than the amplitude of either
of the signals. Thus, a signal measured 3 dB above the noise is actually
at the same amplitude as the noise.)
Antenna Sweeps (SWR)
ror and RBW/Noise Bandwidth correction factor and signal noise floor/
analyzer noise floor correction factor.
Each analyzer's RBW/Noise bandwidth correction factor must be
compiled per the previously mentioned Tektronix, Inc. Application
Note AX-3260. Let us assume a 1
dB error.
C/N at 4 MHz noise bandwidth = C/N
at 4 MHz RBW + signal noise floor/
analyzer floor correction - RBW/noise
bandwidth correction factor - Log
Error
Antenna sweeps are performed on
antenna systems to determine if the
antenna is "tuned" for the frequency
at which it will transmit or receive.
An improperly tuned transmitting
antenna can cause much of the
energy created by a transmitter to
be reflected back into the transmitter causing Intermodulation Distortion thus causing a loss of effective
power being radiated. A properly
tuned antenna will have its characteristic impedance at the frequency
of intended use. The measurement
of a system standing wave ratio
(SWR) can be made using a Spectrum Analyzer, Return Loss Bridge,
and a Tracking Generator or Sweeper
capable of operating at the frequency of antenna operation. From
the SWR, you can determine the system impedance at any frequency
over which the SWR was measured.
For our example, then
C/N at 4 MHz noise bandwidth =
49 dB at 4 MHz RBW +1.7 dB -1
dB RBW/noise bandwidth-2.5 dB
(C/N of example = 47.2 dB at 4 MHz
noise bandwidth)
For more information on the subject
of Noise, see Tektronix, Inc. Application Notes AX-3260 "NOISE
Figure 33. Carrier-to-noise ratio measurement including correction for
low amplitude system noise.
A system will quite often have a
noise specification of a noise bandwidth in other than a common Spectrum Analyzer RBW. To get from
one bandwidth to another bandwidth,
the following formula can be used.
C/N at Specified Bandwidth (dB) =
C/N at Measured Bandwidth (dB)
_1f).
Specified Bandwidth (Hz)
9
Measured Bandwidth (Hz)
MEASUREMENTS USING THE
SPECTRUM ANALYZER — PART
ONE: RANDOM NOISE", AX-3259,
"NOISE MEASUREMENTS USING
THE SPECTRUM ANALYZER —
PART TWO: IMPULSE NOISE",
and 26W-4889, "NO LOOSE ENDS
- REVISED: THE TEKTRONIX
PROOF OF PERFORMANCE PROGRAM FOR CAW."
SWR measurements are made using
the mentioned equipment and connected as shown in Fig. 34. A Return
Loss Bridge designed for the characteristic impedance of the antenna
must be used.
SPECTRUM ANALYZER
INTERFACE
CABLES
TRACKING
Using Fig. 33 as an example, we
measured the carrier-to-noise ratio
of a system as being 65 dB in a 100
kHz RBW. The system specification
requires the result to be a 4 MHz
noise bandwidth.
GENERATOR
Q
RETURN LOSS
BRIDGE
C/N at 4 MHz = 65 dB at 100 kHz
SOURCE
minn 4 MHz
- 10 '° g 100kHz
= 65 dB-16 dB = 49 dB at 4 MHz
RBW
REFLECTED
LOAD
The actual C/N at 4 MHz Noise
Bandwidth is then determined by
accounting for the analyzer's log erFigure 34. SWR test setup.
15
The system operates by the signal
source (Tracking Generator in this
case) launching a signal at a specific
frequency to the Return Loss Bridge
(Bridge) The Bridge routes the signal to the antenna or system under
test, but not to the analyzer. If the
termination at the end of the line
looks like the system characteristic
impedance, all the energy is absorbed and nothing is reflected. If
the termination is not at the characteristic impedance, a portion of the
energy will be reflected back to the
Bridge where it will be routed to the
Spectrum Analyzer and displayed
on screen. As the sweeper or tracking generator sweeps across the
frequency band selected, the analyzer will plot a graph of Return
Level or Return Loss (in dB) vs.
frequency.
System calibration requires terminating the antenna end of the cable
with an "open" or "short" to reflect
all the energy back, and adjusting
the analyzer with a display at the top
of the screen. Then, by terminating
the antenna end of the cable into
the characteristic impedance, the
operator can determine the display
level representing the characteristic
impedance. Fig. 35 shows a typical
display of an antenna trimmed or
tuned for operation at 135 MHz.
The Return Loss Bridge specification should be checked for power
handling capability
Figure 35 demonstrates a narrow
band antenna being swept from 35
MHz to 235 MHz. The antenna is
showing a 40 dB "Return Loss" at
135 MHz. From Fig. 36 we can determine the antenna's SWR as being
1.02:1. At 110 MHz, the SWR = 2.0:1
WARNING
Extreme caution must be practiced when
operating an analyzer near high power RF
equipment. Excessive power applied to
the input will damage a Spectrum Analyzer.
One of the limitations and problems
associated with the setup of Fig. 34
is: Most signal sources are only
capable of generating between 1
mW and 1 W of power. Therefore,
the analyzer will be set with very little
RF attenuation. If another nearby
transmitter broadcasts during the
period the test is being conducted,
excessive power could be received
by the antenna being tested and
damage the Spectrum Analyzer. If
an amplifier is available to place
between the tracking generator and
the Bridge to boost the power (amplifier system should be checked for
flatness), then external attenuators
can be placed between the Bridge
and Spectrum Analyzer to reduce
the signal and protect the analyzer.
RETURN
LOSS
dB
J1
40
B
"--
20
H-4
Other Application Notes of interest:
"EMI MEASUREMENTS USING A
SPECTRUM ANALYZER" 26W-4971.
"EMI APPLICATIONS USING THE
SPECTRUM ANALYZER" AX-3406-1.
"DIGITAL RADIO MEASUREMENTS
USING THE SPECTRUM ANALYZER" AX-4457.
"FM BROADCAST MEASUREMENTS USING THE SPECTRUM
ANALYZER" 26AX-3582-3.
REFLECTION COEFFICIENT
.02
N
30
For more information on SWR
Measurements, see Tektronix, Inc.
Application Notes 26W-5121, "THE
TRACKING GENERATOR/SPECTRUM ANALYZER SYSTEM" and
AX-3842, "TROUBLESHOOTING
TWO-WAY RADIOS WITH THE
SPECTRUM ANALYZER."
C
.05
r
i1!;
.1
.2
SWR, P ETURN LOS S, R EFL EC Tl ^^
CH AR T.
s5S
rm
ai: 5
1
.7
(;OEFFICIEN J
L.
*s
.
•^*s
—
EH & ~:
::
?S4, S
T—— •
_..
4i
~
S
:"
*S«
TS
10
.5
rr:—
•s-. S.
E:
jj
••»'•' o_>* 1
™
:X
::::;:
S. ^
r—
^
-^
...
"~~^.
'_
--—,
•~-- — .
-4-
1 02
1.05
1.1
1.2
1.4
1.6
1.8 2.0
SWR
' /Return Loss dB
~
"
1 - 1 0 " /Return Loss dti \
FREO
Figure 35
\
RESOLUTION
Return Loss of Narrow Band
Antenna with 40 dB Return
20 "" "
/
Figure 36. SWR, Return Loss, and Reflection Coefficient Chart.
Loss
16
3.0
5.0
7.0 9.0
Glossary
NOISE: Unwanted disturbances superimposed upon a useful signal that tend to
AMPLITUDE MODULATION (AM): The process, or the result of the process,
whereby the amplitude of one electrical quantity (carrier frequency) is varied in
accordance with some selected characteristic of a second quantity (modulating
frequency)
B-SAVE-A: A mode of display whereby a waveform which is stored in a digital
memory is subtracted from a waveform stored in a second memory with the result
being displayed on screen.
BASELINE CLIPPER: A means of blanking the signal at the baseline portion of
the display.
CALIBRATOR: A signal generator whose output is used for purposes of calibration, normally either amplitude or frequency or both.
CARRIER: The wave (frequency) to which modulation is applied.
CENTER FREQUENCY: That frequency which corresponds to the center of a frequency span, (Hz).
COMB GENERATOR: A signal source which produces a frequency and multiple
harmonics of the fundamental frequency. Signals are equally spaced at the frequency of the fundamental.
DEGAUSS: To neutralize the residual magnetic polarity of an electronic device by
electric means.
DELTA F: A difference in frequency. A mode of operation of an analyzer where the
NOISE SIDEBAND: Undesired response caused by noise internal to the spectrum
analyzer appearing on the display around a desired response.
OPTIMUM INPUT LEVEL: Design parameter of first mixer which allows for maximum dynamic range (largest carrier to noise ratio) and minimum distortion.
OSCILLOSCOPE: An instrument primarily for making visible the instantaneous
value of one or more rapidly varying electrical quantities as a function of time or of
another electrical or mechanical quantity.
PEAK/AVERAGE CURSOR: A manually controllable function which allows the
user an option to the type of signal processing of data prior to storage in a digital
storage system.
PEAKING: The adjusting of a circuit for maximum amplitude of a signal by aligning
internal filters.
PHASE LOCK: The control of an oscillator or periodic generator so as to operate
at a constant phase angle relative to a stable reference signal source. Primary use
in analyzers is for frequency stability of oscillators.
obscure its information content.
PRESELECTOR: A device placed ahead of a frequency converter or other
device, that passes signals of desired frequencies and reduces others.
PRODUCTS: The resultant frequencies produced through mixing of two or more
signals.
PULSE STRETCHER: A pulse shaper that produces an output pulse whose duration is greater than that of the input pulse and whose amplitude is proportional to
difference in frequency of two signals can be read out directly.
DIGITAL STORAGE: A means of storing the display in modern spectrum analyzers. Allows for flicker-free displays that may be held in memory. Also includes
capabilities such as digital averaging and storing maximum signal excursions.
that of the peak amplitude of the input pulse.
REFERENCE LEVEL: A selected level or amplitude associated with the top
graticule of the CRT. Any signal displayed whose amplitude reaches the top
graticule is said to have an amplitude equal to the Reference Level quantity.
REFRESH RATE: The rate or frequency at which a swept CRT display is refreshed (updated). This rate is typically greater than 50 Hz to avoid flicker.
DIPLEXER: A device capable of simultaneously directing one signal out and
receiving another signal on the same port. The received signal is then routed out in
a separate port.
DISTORTION: An undesired change in waveform caused by signal processing in
a non-linear device or system.
DYNAMIC RANGE: The maximum ratio of two signals simultaneously present at
RESOLUTION BANDWIDTH (RBW): The bandwidth of the most selective
amplifier/filter
RF ATTENUATOR: A device which reduces the amplitude of an input signal to a
level required by the input mixer. The term RF implies linear operation into the high
frequencies.
RF INPUT: The input connector or circuitry directly behind the input connector.
RING: An overshooting condition where the signal will exceed its steady state condition momentarily before stabilizing after a perturbation.
SAVE A: A mode of display whereby a waveform which is stored in digital memory
is not modified by succeeding sweeps (i.e., the waveform is frozen).
2ND LO OUTPUT: A port on a spectrum analyzer where the 2nd local oscillator
frequency is made available for use outside the analyzer
SENSITIVITY: Measure of a spectrum analyzer's ability to display minimum level
signals, at a given IF bandwidth, display mode, and any other influencing factors
and expressed in decibels (e.g., - 120 dBm)
the input which can be measured to a specified accuracy.
EXTERNAL MIXER: A device used to mix the 1st local oscillator of a spectrum
analyzer with RF frequencies. This mixer is external to the analyzer. Typically the
mixing is occurring within a waveguide.
FILTER: A circuit for separating signals on the basis of their frequency.
1ST LO OUTPUT: A port on a spectrum analyzer where the 1st local oscillator frequency is made available for use outside the analyzer.
FLATNESS: The unwanted variation of the displayed amplitude over a specified
frequency span, expressed in decibels
FREQUENCY BAND: A range of frequencies that can be covered without switching (in units of Hz).
FREQUENCY MODULATION (FM): The process, or the result of the process,
whereby the frequency of one electrical quantity (carrier frequency) is varied in
SHAPE FACTOR (Skirt selectivity): A measure of the asymptotic shape of the
resolution bandwidth response curve of a spectrum analyzer. The ratio between
the frequency difference between two widely spaced points on the response
curve, such as the 6 decibels and 60 decibels down points.
SINGLE SWEEP: Operating mode for a triggered sweep instrument in which the
sweep must be reset for each operation, thus preventing unwanted displays.
accordance with some selected characteristic of a second quantity (modulating
frequency).
FREQUENCY RANGE: That range of frequencies over which the instrument performance is specified (Hz to Hz). May refer to the range of frequencies available in
a particular band.
FREQUENCY SPAN: The magnitude of the frequency band displayed, expressed
in hertz or hertz per division.
HARMONIC: A sinusoidal component of a periodic wave or quantity having a fre-
SPECTRUM ANALYZER: A device which is generally used to display the power
distribution of an incoming signal as a function of frequency.
SPURIOUS RESPONSE: A characteristic of a spectrum analyzer wherein the
displayed frequency does not conform to the input frequency.
STABILITY: The property of retaining defined electrical characteristics for a pre-
quency that is an integral multiple of the fundamental frequency.
HARMONIC (N) MIXING: The product of one signal combining with harmonics of
a second signal. This method of mixing is used in spectrum analyzers to obtain
coverage in higher frequency bands than would otherwise be possible with fundamental conversions.
IDENTIFY CONTROL: A function which enables the user of an analyzer to
scribed time and environment (such as frequency stability or amplitude stability).
SWR (Standing Wave Ratio): The ratio of the maximum amplitude to the minimum
amplitude of a signal in a system caused by reflections at the termination of the system. The impedence mismatch causes reflections of the forward signal to combine
with the forward signal both in and out of phase to produce the peaks and nulls.
TIME/DIV: The sweep rate control which defines the rate at which the analyzer
sweeps through a defined frequency spectrum.
TRACKING GENERATOR: Signal source whose output frequency tracks in synchronism with the input frequency of a receiver, such as the spectrum analyzer.
TRIGGER: A pulse used to initiate a triggered sweep or delay ramp.
ULTIMATE: The ability of a filter to reject or suppress a frequency other than which
it was designed to pass.
VERTICAL DISPLAY FACTOR: The Y-axis scale factor for display on a CRT
VIDEO FILTER: A post detection low pass filter
VIEW A, VIEW B: Controls which allow two memories to be enabled for viewing or
disabled independently of each other
VSWR (Voltage Standing Wave Ratio): The ratio of the magnitude of the
transverse electric field in a plane of maximum strength to the magnitude at the
equivalent point in an adjacent plane of minimum field strength
ZERO SPAN: A mode of operation in which the frequency span is reduced to zero
determine if a signal being displayed represents a signal at the frequency indicated or is an undesired mixing product of the first mixer.
IF (Intermediate Frequency): A frequency at which the input signal is shifted
internally for processing.
INTERCEPT POINT: The theoretical points at which the fundamental (driving)
signals and the distortion products have equal amplitudes.
LINEAR DISPLAY: A display in which the vertical scale divisions are a linear function of the input signal voltage.
LOG DISPLAY: A display in which the vertical scale divisions are a logarithmic
function of the input signal power.
MAX HOLD: A mode of acquisition for a digital storage system where the maximum
amplitude achieved at every frequency being analyzed is retained and continuously
displayed for successive sweeps.
MAX SPAN: A mode of operation in which the spectrum analyzer scans an entire
frequency band.
MAXIMUM INPUT LEVEL: Maximum amount of power capable of being handled
by input circuitry without damage.
17
For further information, contact:
U.S.A., Asia, Australia, Central
& South America, Japan
Tektronix, Inc.
P.O. Box 1700
Beaverton, Oregon 97075
For additional literature, or the
address and phone number of the
Tektronix Sales Office nearest
you, contact:
Phone: 800/547-1512
Oregon only 800/452-1877
TWX: 910-467-8708
TLX: 15-1754
Cable: TEKTRONIX
Europe, Africa,
Middle East
Tektronix Europe B.V.
European Headquarters
Postbox 827
1180 AV Amstelveen
The Netherlands
Phone: (20)471146
Telex: 18312- 18328
Canada
Tektronix Canada Inc.
P.O. Box 6500
Barrie, Ontario L4M 4V3
Phone: 705/737-2700
Tektronix sales and service
offices around the world:
Albania, Algeria, Angola,
Argentina, Australia, Austria,
Bangladesh, Belgium, Bolivia,
Brazil, Canada, Peoples Republic
of China, Chile, Colombia, Costa
Rica, Czechoslovakia, Denmark,
East Africa, Ecuador, Egypt,
Federal Republic of Germany,
Finland, France, Greece, Hong
Kong, Hungary, Iceland, India,
Indonesia, Ireland, Israel, Italy,
Ivory Coast, Japan, Jordan,
Korea, Kuwait, Lebanon,
Malaysia, Mexico, Morocco, The
Netherlands, New Zealand,
Nigeria, Norway, Pakistan,
Panama, Peru, Philippines,
Poland, Portugal, Qatar, Republic
of South Africa, Romania, Saudi
Arabia, Singapore, Spain, Sri
Lanka, Sudan, Sweden,
Radar pulse width, pulse shape, repetition rate
and carrier frequency monitored on one display.
Switzerland, Syria, Taiwan,
Thailand, Turkey, Tunisia, United
Kingdom, Uruguay, USSR,
Venezuela, Yugoslavia, Zambia,
Zimbabwe.
Performance
worth the
name
Tektronix
COMMITTED TO EXCELLENCE
Copyright © 1983, Tektronix, Inc. All
rights reserved. Printed in U.S.A.
Tektronix products are covered by U.S.
and foreign patents, issued and
pending. Information in this publication
supersedes that in all previously
published material. Specification and
price change privileges reserved.
TEKTRONIX, TEK, SCOPE-MOBILE,
TELEQUIPMENT, and & are
registered trademarks. For further
information, contact: Tektronix, Inc.,
P.O. Box 500, Beaverton, OR 97077.
Phone: (503) 627-7111; TWX 910-467-
8708; TLX: 15-1754; Cable:
TEKTRONIX. Subsidiaries and
distributors worldwide.
5/83
26W-5360
For further information, contact:
U.S.A., Asia, Australia, Central
& South America, Japan
Tektronix, Inc.
P.O. Box 1700
Beavertpn, Oregon 97075
For additional literature, or the
address and phone number of the
Tektronix Sales Office nearest
you, contact:
Phone: 800/547-1512
Oregon only 800/452-1877
TWX: 910-467-8708
TLX: 15-1754
Cable: TEKTRONIX
Europe, Africa,
Middle East
Tektronix Europe B.V.
European Headquarters
Postbox 827
1180 AV Amstelveen
The Netherlands
Phone: (20)471146
Telex: 18312-18328
Canada
Tektronix Canada Inc.
P.O. Box 6500
Barrie, Ontario L4M 4V3
Phone: 705/737-2700
Tektronix sales and service
offices around the world:
Albania, Algeria, Angola,
Argentina, Austrajia, Austria,
Bangladesh, Belgium, Bolivia,
Brazil, Canada, Peoples Republic
of China, Chile, Colombia, Costa
Rica, Czechoslovakia, Denmark,
East Africa, Ecuador, Egypt,
Federal Republic of Germany,
Finland, France, Greece, Hong
Kong, Hungary, Iceland, India,
Indonesia, Ireland, Israel, Italy,
Ivory Coast, Japan, Jordan,
Korea, Kuwait, Lebanon,
Malaysia, Mexico, Morocco, The
Netherlands, New Zealand,
Nigeria, Norway, Pakistan,
Panama, Peru, Philippines,
Poland, Portugal, Qatar, Republic
of South Africa, Romania, Saudi
Arabia, Singapore, Spain, Sri
Lanka, Sudan, Sweden,
Switzerland, Syria, Taiwan,
Thailand, Turkey, Tunisia, United
Kingdom, Uruguay, USSR,
Venezuela, Yugoslavia, Zambia,
Zimbabwe.
Copyright © 1983, Tektronix, Inc. All
rights reserved. Printed in U S A .
Tektronix products are covered by U.S.
and foreign patents, issued and
pending. Information in this publication
supersedes that in all previously
published material. Specification and
price change privileges reserved
TEKTRONIX, TEK, SCOPE-MOBILE,
TELEQUIPMENT. and £j are
registered trademarks. For further
information, contact: Tektronix, Inc.,
P.O. Box 500, Beaverton, OR 97077.
Phone: (503) 627-7111; TWX 910-4678708; TLX: 15-1754; Cable:
TEKTRONIX. Subsidiaries and
distributors worldwide.
26W5360
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement