Department Of
Electrical And Electronics Engineering
Applied Microwave Engineering
Lab
Lab # 1
Frequency and Time Domain Representations
Prepared by
Sana Khan & Hassan Sajjad
Laboratory #1
LAB PROCEDURES
Note: Do not use the very short SMA cables for this lab such as for the connection between T-Junction and SA input. It is recommended that longer SMA or BNC cables be used for this lab. DO NOT FORGET TO USE DC BLOCK
WITH SPECTRUM ANALYZER (SA).
Part I: SA Measurement of two signals close in frequency
(What you should see: The ability to clearly identify two signals that are close in frequency as separate signals depends
strongly on the measurement settings on the spectrum analyzer. You will explore the effect of the span and resolution
bandwidth settings on this type of measurement. The resolution bandwidth is really the key in this case, but the two settings are linked within the SA.)
We will use two SIGNAL GENERATORS in this lab and combine the signals using a combiner/splitter provided to
you as shown in Fig. 1. The combined signal will be then observed in the SA to determine the frequency contents of the
signal and to visualize the resolution bandwidth of two signals close in frequency.
Figure 1: SA Measurement of two signals close in frequency
1. Setup the Signal Generator at FREQ 300 MHz and power level -5dBm.
2. Setup the 2nd Signal Generator provided to you by the TA and set the FREQ to 301MHz and power level -5 dBm.
3. Using the two long SMA cables connect the output of the two SIGNAL GENERATORS to PORT 1 and Port 2 of
the COMBINER/SPLITTER (SBCTJ-1W 1 to 750MHz) respectively.
4. Connect the Ouput Port of the Combiner to the SA with a DC BLOCK.
5. Set-up the SA for a 300MHz center frequency and a 20 MHz span. Push FREQUENCY 300 MHz and SPAN
20MHz. At this point, TURN RF ON for the SIGNAL GENERATOR 1. You should see a signal at 300 MHz.
6. Now Turn RF ON for the second SIGNAL GENERATOR and you should be able to see another signal at 301 MHz.
7. This will establish two signals at spectrum analyzer input that differ by 1MHz. With the specified SA settings
you should be able to see two signal peaks on the SA. Record the frequency and amplitude of the two peaks
corresponding to the two signal sources in Table 1.
8. Using the arrow keys, increase the SPAN while keeping the center frequency fixed at 300MHz . Note that the
resolution bandwidth automatically changes with the SPAN setting to keep the sweep time approximately constant.
This is what we want to happen for this part. Note that setting the span more than 100MHz will cause the SA to
display frequencies below 0Hz. Hence, the maximum span should not be extended beyond these values.
2
Laboratory #1
Table 1: Observed signal peaks with calibration reference signal and full-span frequency display (pre-set).
Freq. of Peak
Amplitude of Peak
Comments
OBSERVATION: For what span setting can you no longer distinguish between the two signals?
What is the corresponding (automatic) setting for the resolution bandwidth? (You can find this easily
by pushing BW after you have the SPAN set).
9. Set the SPAN back to 20MHz. Increase the resolution bandwidth, while keeping the span and center frequency
fixed. This is done by pressing BW then using the arrow keys or knob to vary the value of the resolution bandwidth.
Observation:For what the value of resolution bandwidth can you first no longer distinguish that there
are two frequencies present in the display (if two frequencies are already not distinguishable report that
number)?
10. Now decrease the resolution bandwidth using the arrow or knob functions.
Observation: For what the value of resolution bandwidth can you first clearly distinguish the two
signals?
Part II: SA Measurement of SIGNAL GENERATOR signal at varied frequency and amplitude
(What you should see: this part will give you experience, or practice, with setting up various amplitude and frequencies using the SIGNAL GENERATOR as a signal source, and then finding and measuring the resulting output signal using
the SA. You should see correlation between the settings of the SIGNAL GENERATOR and the measurements on the SA.)
Figure 2 shows the test set-up for this part of the experiment.
1. Connect the SIGNAL GENERATOR test signal directly to the Spectrum Analyzer using a DC Block.
2. Push PRE-SET on the spectrum analyzer. Next set 20 dB internal attenuation of the Spectrum Analyzer to help
ensure that the SA remains in its linear operating region. This is done by pushing AMPLITUDE ATTENUATION
20 dB . Also, set the REF LEVEL to 10 dBm by pushing AMPLITUDE REF LEVLEL 10 dBm.
3. Using analogous procedures as in the previous part, set the SIGNAL GENERATOR test frequency to a value of 25
MHz. Use a power level setting of +0 dBm, and make sure the RF is ON.
4. On the SA set the center FREQUENCY to coincide with the SIGNAL GENERATOR test frequency and set the
SPAN to correspond to a 10% value of the center frequency. For instance, if the center frequency set to 300 kHz,
the span should be 30 kHz. Use the PEAK SEARCH and/or MARKER keys to identify the observed frequency and
amplitude peaks corresponding to the 25 MHz or SIGNAL GENERATOR test signal.
3
Laboratory #1
Figure 2: Test configuration for measurement of SIGNAL GENERATOR as a single frequency source
5. Repeat steps 3 and 4 for varied test signals in order to make the measurements indicated in Table 2 shown below. To
do this you will be changing the frequency, and in some cases power level on the SIGNAL GENERATOR control
window, including the center frequency and span settings on the SA.
Table 2: Observed frequency and amplitudes for varied SIGNAL GENERATOR CW test signals
Source
test
Frequency
(MHz)
25
25
300
300
600
900
Source Power
Level Setting
(dBm)
0
-3
0
-3
0
0
SA Measured
Peak Frequency
(MHz)
SA Measured
Amplitude
(dBm)
Comments
OBSERVATIONS:
By how much does the SA measured signal amplitude drop at 25 MHz and 300 MHz when the SIGNAL
GENERATOR power level setting is reduced by 3 dB?
Other comments and observations:
Part III: Oscilloscope Measurement of the SIGNAL GENERATOR signal at varied frequency and amplitude
(What you should see: you will measure the same signals in the time domain on an Oscilloscope that you measured
above on the SA. You should see agreement between the O-Scope and SA frequency measurements at low frequency, but
will see differences at higher frequencies. Amplitude correlation between the two instruments requires you to properly
understand and calculate powers in dBm from voltages across a 50 ohm load and vice-versa. The 1 MΩ vs 50 Ω coupling
setting is critical to getting this right.)
The test configuration for this part is shown below. You should read the reference guide for the oscilloscope provided in
handout material before completing this part.
1. Connect a cable directly between the SIGNAL GENERATOR output and the Ch1 input connector of the oscilloscope.
2. Set-up the SIGNAL GENERATOR for a test frequency of 25 MHz, with a power level setting of +0 dBm.
3. Press the CH1 Push AUTOSET on the OSCOPE, then adjust the vertical and horizontal position and scale as needed
to see a sinusoidal signal with a few signal periods in the display. Press the VERT MENU button and select DC
coupling.
4
Laboratory #1
Figure 3: Configuration
for time domain measurement of SIGNAL GENERATOR test signal using oscilloscope
4. Use the vertical line cursors to determine the period T of the displayed signal and calculate the frequency as f=
1/T.
OBSERVATIONS: T=
s
f=
Hz
5. Use the horizontal line cursors to determine the peak-to-peak voltage levels for both coupling impedance options:
1MΩ and 50 Ω. Pushing the VERTICAL MENU button and then selecting the coupling impedance soft key allows
you to change the coupling impedance until the desired coupling impedance is highlighted.
OBSERVATIONS: Vpp (1MΩ coupling) =
Vpp (50Ω coupling) =
mV
mV
6. Use the automatic measurement features to take the same measurements made in parts 4 and 5. After pressing the
“MEASURE” key, the measurements you want to select are: Frequency, and Pk-Pk. Tabulate your measurements,
along with the results of the next step, in the table (Table 3) shown below.
7. Change the frequency and amplitude settings as indicated in the table shown below and use the automatic measurement features to complete the table.Use DC coupling with a coupling impedance of 50 ohms. You will need to hit
AUTOSET on the OSCOPE each time frequency is changed.
When making the frequency measurement you will find that the readout changes as you look at it, making it difficult to specify a number. To estimate the precision of the measurement try to “eyeball” a nominal value and an
uncertainty. For example if the 300MHz measurement is changing between 300.9 and 298.9, state the measured
frequency as 299.9 +/- 1.0.
You should notice more difficulties in making the measurements at 600MHz and 900MHz than at lower frequencies.
Comment on this in the space provided in and/or below the table.
Table 3: Summary of Oscilloscope Amplitude and Frequency Measurements
Signal
Test
quency
(MHz)
25
25
300
300
600
900
gen.
Fre-
Signal gen.
Power Level
(dBm)
OSCOPE Meas.
Frequency
(MHz) fo +/-∆
0
-3
0
-3
0
0
5
OSCOPE
Measured Vpp
Z: 1MΩ
OSCOPE
Measured Vpp
Z: 50Ω
Comments
Laboratory #1
By what factor does the peak voltage Vpk (= Vpp /2) change when the power drops by 3 dB?
Additional comments and observations:
Part IV: Oscilloscope Measurement of Composite Signal (2 Sinusoids)
(What you should see: Prior to this lab session you should have become familiar with a Matlab or Mathcad file that
plots the sum of two sinusoidal signals in the time domain. The goal of this part is to get a good visual correlation
between the Matlab/Mathcad representation of the result of 2 signals and a measurement of the same thing. You will see
that the composite waveform is very sensitive to the relative amplitude and phases between the two signals, even after you
set the frequencies to be the same as those in the measurement setup.)
This section can be completed using MATLAB.
1. Setup the equipment configuration shown in Fig. 4. Also pull-up on the PC the Matlab file you examined previously
(in Lab #0 or Pre-lab assignment) that plots the result of two sinusoids in the time domain.
Figure 4: Measurement of 2 sinusoidal signals combined on the O-Scope
2. Set the SIGNAL GENERATOR to 350 MHz at 0 dBm power level, so with the 20 dB pad attached the signal will
be at approximately –20 dBm at the BNC T-Jcn combiner.
3. Display the combination of two sinusoidal signals on the Oscilloscope. IMPORTANT: Use averaging (Acquire –
Mode - Average) with about 30-40 averages to smooth out the display discontinuities caused by the two signal
generators not being tied to the same time base.
4. Note you have represented on the scope a signal that could mathematically be represented by
s0 (t) = A0 .cos(2.π.f0 .t + θ0 )
s1 (t) = A1 .cos(2.π.f1 .t + θ1 )
s(t) = s0 (t) + s1 (t)
Where s0 (t) and s1 (t) are the two signals. Try determining A1 and A0 separately by first disconnecting one, then
the other from the T-junction.
6
Laboratory #1
Observations:
What values do you estimate for A0 and A1 for your measurement configuration?
A1 =
A0 =
Enter these values into the Matlab file, and compare to the O-Scope display.
5. One aspect that is harder to quantify and remains an ambiguity in the comparison (Matlab vs. measurement) is the
phase relationship between the two signals. Vary the phase of one of the two Matlab signals until you get a response
that looks similar to what you have on the O-Scope display.
Observations: Printout the Matlab file with plots that are closest to what you can measure for the two
signals on the O-scope.
(Optional) See if you can capture the digital waveform from the scope and obtain a plot to compare
to your Matlab file.
Hint: You may have to vary the time scale (ns/division) on the scope and the total time span of the Matlab file in
looking for similar waveform representations. By increasing the total time span of the screen you should definitely
see a repetitive “beat signal” pattern that repeats on both the scope and on the Matlab representations. By varying
the phase you should be able to get the O-scope screen view to look similar to the Matlab view.
Part V: Simulations and additional analysis
(What you should see: if done properly you should see reasonable agreement between SA and oscilloscope measurements of frequency and power – i.e. powers calculated from voltage for OSCOPE – for lower frequencies. You will see
problems at higher frequencies with this agreement.)
Convert your peak-to-peak voltage values from part IV for the 50 ohm coupling case to power in dBm and compare
to the results obtained with the spectrum analyzer in Table 4 below. Also summarize the frequency results from the SA
and OSCOPE. The following equation may be used to convert the 50 ohm voltages to power in dBm.
(Vpp )2
400 × 10−3
where Vpp is the peak-to-peak voltage in volts measured on the OSCOPE. Use of Matlab or a spreadsheet facilitates these
calculations.
P (dBm) = 10 log
Table 4: : Comparison of Spectrum Analyzer and Oscilloscope Measurements
Signal gen.
Test
Freq.
(MHz)
25
25
300
300
600
900
Signal gen.
Power Level
(dBm)
0
-3
0
-3
0
0
OSCOPE Meas.
Freq. (MHz) fo +/
-∆
7
OSCOPE
Measured Vpp
Z: 1MΩ
OSCOPE
Measured Vpp
Z: 50Ω
Comments
Laboratory #1
OBSERVATIONS:
How well do the SA and OSCOPE frequency measurements compare?
How well to the SA and OSCOPE amplitude measurements compare?
Other comments:
Part VII (Optional) : Oscilloscope Measurement of Noisy Signal.
(This part is required for graduate students, but optional for undergraduates.)
(What you should see: you will measure a very low level signal sent through a “wireless link” and observe the different measurement requirements that such “noisy” signals may require. You will be receiving your signal as well some
other signals.)
In this portion of the lab you will construct a simple wireless link and note the effect of link losses and noise on timedomain and frequency-domain measurements. The test configuration is shown below.
Figure 5: Configuration to observe the effect of link losses and noise on Time-Domain and Frequency - Domain measurements using OSCOPE and SA respectively.
1. Set the SIGNAL GENERATOR to a frequency of 300 MHz plus twice your roll numbers last digit (in MHz); e.g.,
if last digit of your roll number is #4, use 308 MHz. Make sure the SIGNAL GENERATOR RF power is off.
2. With the power supply off, build the system, and connect the output of the ZFL-1000LN amplifier to the SA. If you
have any questions, especially in connecting the power supply to the amplifier, be sure to request assistance from
the TA. You can leave the antenna on the bench top
3. Call for the TA. He/ She will check out your configuration, set the SIGNAL GENERATOR power to +7 dBm, and
“power-up” the amplifier. Ask the TA or an instructor if you have any questions about the setup (e.g. selection of
antenna).
4. Cycle the SIGNAL GENERATOR RF power on and off to verify that you are viewing your own signal. Do you see
signals from any of the other benches? If so, what are their powers?
5. Carefully and slowly move the antenna around (a foot or two at most), and move your hand over it. Note the effect
on the signal level in both cases. How much variation do you observe (in both dB and percent)?
6. Now connect the ZFL-1000LN/ (Any Amplifier provided to you by TA) output to the OSCOPE. Explore the effects
of changing the Acquisition mode under the Acquire menu. Try each mode in sequence and notice the effects on
the signal display. Record your observations below.
8