wiki/ news/ 2005-03-09 - Fun with a spectrum analyzer: communication team field day at Maxim Semiconductor

Introduction

Jeromy works at Maxim Semiconductor in Hillsboro. He obtained permission for us to use their testing facilities, specifically their HP 8561E spectrum analyzer. Thanks Jeromy!

Our mission: try and find out why the ATV system seems to be interfering with the GPS, 2m and 802.11b systems.

comm_test_2005_03_09_20.thumb.jpg
The testers

Test 1: Examine ATV signal

The first tests we wanted to do is to determine if the ATV system was putting out spurs (unwanted frequency components outside of intended band). We used the ATV mini-vertical into a F to BNC to N to SMA (!) connector to the spectrum analyzer.

comm_test_2005_03_09_12.thumb.jpg
The setup

Here's what we saw:

comm_test_2005_03_09_01.thumb.jpg comm_test_2005_03_09_02.thumb.jpg
"Nothing" on. The spur on the right turns out to be from the 2m receiver on the recovery node which was, in fact, on because of it's backup battery. Whoops.
comm_test_2005_03_09_04.thumb.jpg
ATV Exciter (transmitter) on, no power amp and no input signal so it's a pure carrier at 1.253 GHz. Spur on the right is from the 2m system which is still on at this point.

We zoomed in to capture the ATV signal in more detail:

comm_test_2005_03_09_05.thumb.jpg comm_test_2005_03_09_07.thumb.jpg comm_test_2005_03_09_09.thumb.jpg comm_test_2005_03_09_10.thumb.jpg
ATV system: 1) Carrier with no amplication or modulation (exciter only), 2) with modulation, 3) with 8V power amp on, and 4) with 12V power amp on.

Test 2: Look for ATV signals in Other Frequency Bands

Next we looked for spurs from the ATV system in the GPS, 802.11b and 2m bands... but we didn't find anything. We did find Andrew's laptop searching desperately for wifi access points, and we again saw spikes from the local oscillator (LO) on the 2m receiver on the recovery node although we didn't realize it at the time.

comm_test_2005_03_09_14.thumb.jpg comm_test_2005_03_09_15.thumb.jpg comm_test_2005_03_09_16.thumb.jpg
1) GPS band (1.57542 GHz), 2) 802.11b band (2.412 GHz), and 3) 2m band (146.050 MHz)

Note that we thought we saw the noise in the GPS floor rise by about 3 dB when we turned on the ATV power amp, but we couldn't reproduce it later.

Here's a shot with Andrew's handheld 2m radio transmitting:

comm_test_2005_03_09_17.thumb.jpg
Andrew's 2m handheld on low power

And we also checked the whole 1.3 GHz frequency range just in case:

comm_test_2005_03_09_18.thumb.jpg
1.2 to 1.6 GHz

Test 3: Try 2m receiver with ATV PA on/off

As we expected, turning the ATV power amp on and off degraded the 2m receiver's ability to receive DTMF tones. It got some, but not many. On a lark, we swamped out antennas for the 2m receiver and plugged in Andrew's ht's "rubber ducky" vertical. Suddenly, even with the power amp on, it worked like a charm. Aha! Somehow, using the "airframe" antenna couples the ATV power amp to the 2m and jams the 2m system.

In the process, we also found all sorts of crazy spurs from the 2m receiver ! Somehow the local oscillator (LO) on the 2m receiver was being nicely radiated out. And it showed up in the GPS antenna:

comm_test_2005_03_09_19.thumb.jpg
"2m receiver LO being picked up on the GPS antenna.

Conclusion

We don't think the ATV is sending out of band signals to the other receivers. The 2m receiver is, surprisingly, although their bands shouldn't be affecting the various receivers since they're so narrow.

That said, we think that the ATV is desensitizing the Jupiter GPS receiver because it may have no filter on the front end. We'll have to try it with the MG5001 GPS which has a SAW (Surface Acoustic Wave) filter built in.

We don't know how the ATV PA is affecting the 2m system, but it clearly is through the 2m antenna.

Here's what's next:

  1. Make a GPS SAW filter (a 5 MHz bandpass centered at 1.57542 GHz) to go between the GPS LNA and the GPS antenna.
  2. Quelch 2m receiver LO "transmissions".
  3. Try putting a "rubber ducky" vertical inside the recovery module and do a range test on it.

After fixing problems 1 - 3, we'll then move on to the interference between the ATV and 802.11b radios.