ELEC 332

In the Lab

Part 1: Spectrum Analyzer

The scope can tell us the amplitude and frequency of a single sinusoid (or square wave, or any other periodic function for that matter), but when more complex signals or multiple signals are present, the display becomes difficult to interpret. In such cases, looking at the frequency domain, rather than the time domain, may be more informative, and this is what the spectrum analyzer allows us to do.


Initial Configuration

Connect together the scope, signal generator, spectrum analyzer, and Wavetek function generator as shown in the diagram, using BNC Tees and coax cables. The purpose is to view the signal generator output on both the scope and the spectrum analyzer. The function generator is used to provide a modulating waveform (the builtin 400 Hz and 1 kHz modulating signals are two low in frequency to produce visible AM sidebands on the spectrum analyzer).

(Click to enlarge)

Set the function generator to produce a 10 kHz, 2.5 V sine wave.

Initializing:

When the spectrum analyzer is turned on, it returns to its previous configuration. Since the spectrum analyzer has a large number of settings, some of which can cause it to do distinctly odd things, it's a good idea to start by returning it to some sort of "normal" configuration before starting. Fortunately, most of the settings are displayed on the screen, so we can check the configuration quickly if we know what to look for. Here's what we want (starting at the upper left of the screen):
REF
Should be 0.0 dBm. If not: press the LEVEL button, enter 0 on the numeric keypad, and press the GHz +dBm dB key.
ATT
Should be 10dB. If not: press the LEVEL key, then use the ATT soft key to select AUTO.
A_wrt
If this says A_avg, A_blnk, or something else: press the TRACE key, then press the Write A soft key.
B_blnk
If something else is displayed in this position: press the TRACE key, the Trc Menu A/B key, and Blank B.

Verify that the bandwidth is in the automatic mode: Press the BW key, then use the soft keys to insure that both RBW and VBW have AUTO selected.

Observing a signal:

Set the signal generator to produce a 10 MHz, -10 dBm, unmodulated signal. Set the spectrum analyzer to a center frequency of 10 MHz and a span of 1 MHz. We would expect to see a single line in the center of the display, but what we actually get looks more like a tent. This is because the bandwidth chosen in the automatic mode is fairly wide, to provide for a fast sweep rate. Try changing the span size and notice that the bandwidth (labeled RBW (for Resolution Bandwidth) at the bottom of the screen) also changes, keeping the shape of the display roughly the same.

Changing the bandwidth:

We can get a more line-like display by reducing the bandwidth. To do this, press the BW key, then use the softkeys to set the RBW mode to MNL (manual). Use the knob or the arrow keys to decrease the bandwith and note how the display changes. The lowest available bandwidth is 1 kHz. This is the 3 dB bandwidth, so the width of the "spike" at its base will be significantly wider.

Using the marker:

Set the SPAN to 100 kHz and the BW to 1 kHz. Press PK SRCH key under MARKER to set the marker to the peak value. The frequency and amplitude of the marker are displayed in the upper right portion of the screen.

Comment on the accuracy of single frequency measurement with the spectrum analyzer.

Resolving signals:

Set the function generator as specified in the initial configuration (10 kHz, 2.5 Vpp sine wave). Set the signal generator modulation to extenal AC and turn the modulation on. You should see the two sidebands 10 kHz to either side of the carrier frequency. The additional sidebands are due to the fact that the sine wave output of the function generator is not very accurate, so there is significant harmonic distortion. To see lots of harmonics, briefly set the function generator to square wave.

Return the function generator output to sine wave, and change the frequency to 1 kHz. We would expect to see sidebands 1 kHz either size of 10 MHz, but because of the limited resolution of the spectrum analyzer, all we see is a mess. The moral: be careful in interpreting what you see on the spectrum analyzer.

Small signals:

Reduce the signal generator output in steps of 10 dBm while watching the signal on both the scope and the spectrum analyzer. Adjust the Volts/div of the scope and the LEVEL of the spectrum analyzer as necessary to maintain a satisfactory display.

Using averaging:

As you reduce the reference level, you will start to see a fuzzy band (sometimes called grass) at the bottom of the display. This is noise, both on the input signal, and internal to the spectrum analyzer. To see how much of the noise is internal, temporarily unplug the input connector.

Since the noise is random, its effect can be reduced by averaging. To turn on averaging, press the TRACE key, use the bottom soft key to select the second half of the trace menu, press the AVG A key and turn averaging ON. This should significantly reduce the amplitude of the noise, while leaving the input signal undisturbed.

Continue reducing the amplitude of the input until it is no longer possible to detect it. Comment on the magnitude of the minimum detectable signal for both the scope and the spectrum analyzer.

Part 2: Real World Signals

While looking at the signal generator output may be educational, it gets boring fairly quickly. Now that we are experts in using the spectrum analyzer, we can try looking for some more interesting signals.


Setting up:

Set the spectrum analyzer to the following configuration: reference level (REF) 0.0 dBm, center frequency 100 MHz, span 200 MHz, bandwidth automatic, averaging off. Remove the coax cable from the spectrum analyzer input and replace it with a banana plug to BNC adapter. Fully extend one of the whip antennas and plug it into the red terminal of the adapter. With a little care, friction will hold the antenna in the vertical direction (it helps to rotate the adapter so that the red terminal is on the bottom).

Signals in the lab:

You should see a bump just to the left of the center of the screen. This is the FM broadcast band. These are the only signals with the proper combination of power and wavelength to be able to penetrate deep into the interior of the building where the lab is located.

However, there are plenty of signals originating inside the building which can be picked up, mostly originating from the large number of switching power supplies in the lab and in surrounding rooms. This cacophony usually blends together into a random looking noise, but by touching the antenna or moving it to a different orientation, you may be able to pick up some individual signals (recognizable by their regularly spaced harmonics).

There's another signal of much higher frequency that we can pick up. Set the spectrum analyzer center frequency to 2.4 GHz, and the span to 500 MHz. Return the whip antenna to the vertical position. Turn on the microwave oven.

Signals outside the lab:

Fortunately we have some antennas outside the building which can pick up a better selection of interesting signals. Return the spectrum analyzer settings to 100 MHz center frequency, 200 MHz span. Remove the whip antenna and banana plug adapter. Plug in the outside antenna cable (marked with a piece of blue tape).

You should now see a number of bumps. The FM band is still there, with significantly higher amplitude. The hump at the far left is the AM broadcast band and the international shortwave broadcast bands. The pair of spikes around 60 MHz is Channel 2. Channel 8 and Channel 13 are near the right side of the display, and you may be able to see some pager and two-way radio traffic around 150 MHz.

Try using the FREQUENCY, SPAN, and BW, controls to zoom in on some of these signals.

Part 3: Mixers

If we wanted to build our own radio system, we could use the signal generator as a transmitter, and the spectrum analyzer as a receiver (there's an earphone jack on the back, so we can listen to the received signals as well as looking at them). Although this would give us considerable flexibility in our choice of frequency, it would be very expensive and not very portable. Fortunately, inexpensive components are available from which we can build a considerably less expensive system.

The mixer we will use is the SA602, which is based on a Gilbert Cell multiplier. This is a fully differential circuit, but there are various ways of using it with single ended signals, including simply ignoring one of the inputs or outputs.

Here's our circuit:


(Click to enlarge)

As usual, we've built a number of options into our module. These include

  1. A choice of external or on board local oscillator.
  2. A choice of traditional or surpressed carrier AM.
  3. A choice of direct, buffered, or transformer coupled output.

We will be using two different modules, each with a different configuration. The first has an external oscillator input, no offset on the baseband input, and a transformer coupled output. Since there is no offset added to the input, the output will be of the form
 x(t) = A s(t)cos(ωct)
where ωc is the frequency of the external oscillator.


(Click to enlarge)

The second has an on board 13.56 MHz crystal oscillator, adds an offset to the modulating input to produce AM, and has an emitter follower output buffer. Because of the offset, the output is
 x(t) = A (s(t)+1)cos(ωct)


(Click to enlarge)


Connect the components:

We'll be using the same equipment as in Part 1, but connected differently. Set the function generator to produce a 1 kHz, 50 mV p-p sine wave (you will probably need to use the 20 dB attenuator). Set the signal generator to produce a 13.56 MHz, unmodulated output, also with a 50 mV p-p amplitude.

Select the mixer module with the external local oscillator input (the one with 3 SMB connectors) and plug it into the breadboard. Connect the function generator output to the IN connector of the mixer(the one on the left side of the board), and to channel 1 of the scope. Set the scope to trigger on channel 1. Connect the signal generator output to the LO IN connector (upper right). Connect the OUT connector (lower right) to the scope (channel 2) and spectrum analyzer.

Observe signals:

Turn everything on and adjust the scope and spectrum analyzer to produce a satisfactory display. Since there is no DC component in the input, the specturm should contain no component at the carrier frequency.

Adjust levels:

Increase the amplitude of the function generator and signal generator outputs until the modulated waveform begins to show signs of distortion. Where does the distortion first become visible, on the scope or the spectrum analyzer? Reduce each amplitude slightly.

Adjust frequencies;

Change the frequency of the function generator and signal generator and verify that the modulated signal behaves as it should. Switch the function generator output to square and triangle waveforms and verify that the spectrum is correct.

Change mixers:

Turn off the power and restore the instruments to their original settings. Disconnect the signal generator and turn off its output to avoid interference. Replace the mixer with the module having the on board oscillator. This module also adds a DC offset to the input (via R27) so that the carrier frequency will appear in the output.

Observe signals and adjust levels:

As before, observe the input and output signals in the time and frequency domains. Adjust the function generator output to product the largest possible output without significant distortion. Vary the function generator frequency and waveform and compare the behavior to that of the previous mixer.

AM transmitter:

Attach the whip antenna module to the breadboard.

(Click to enlarge)
Use all four screws to ensure good support. Disconnect the output of the mixer from the scope and spectrum analyzer and connect it to the antenna. Tune the Kaito radio to 13.56 MHz and extend its antenna.

(Click to enlarge)
Since this mixer is producing a conventional AM signal, you should be able to hear it on the radio. Vary the function generator frequency to insure that you are in fact listening to your signal. What is the maximum distance at which you can receive the signal?