The local ham club I am a member of was handed several decommissioned or defective K-band police radar heads similar to the one shown on the right. We knew these contain a gunnplexer and can potentially carry Wideband FM (WBFM) radio contacts. This would allow our club to add a band to the ARRL June VHF QSO Party. At the last club meeting, I raised my hand when it came to who would try to put these 24 GHz police radar heads on the air. So here is my report on how things are progressing.
The first thing I did was to try to find some technical data on the web on these radar units. Very little available. Not surprising, considering the nature of the beast and the application... All I found is that their nominal output power is 10mW and that the horn beamwidth is 12 degrees. So I disassembled the first unit and found that it had indeed an M/A-Com K-band Gunnplexer, with a separate mixer diode for the IF detection (doppler) outside of the gunnplexer assembly. Also noted is the fact that there is no varactor-tuning diode. This means that the modulation will have to be injected into the main Gunn diode supply. I then searched for the gunnplexer specsheet and found what I wanted. The Gunn diode inside works on +5V, with +5.5V as an absolute max. A +5V regulator will be perfect here.
Looking at the waveguide assembly and at the horn, it is clear that this is circularly polarized, with opposite sense of rotation on Receive vs. Transmit. This was expected considering that a hard wave reflection on a vehicle inverts the sense of rotation. This is not good for a gunnplexer-gunnplexer communication link though.
There is also a PCB inside the unit with lots of custom electronics, mainly the various DC supplies, the doppler frequency detection and the data transmission to the controller. Not useful to us, so I took all that out.
I powered up one gunnplexer and immediately detected RF at ~ 24.130 GHz, using my HP 5350B frequency counter and a quarter-wavelength probe made of a connectorized piece of semi-rigid cable. Looking good. That beast drifted a lot at power up and warm up, but then stabilized to withing a MHz or so afterwards.
Next thing to try was to inject modulation. I used a function generator and sent some audio tone through a capacitor right on the Gunn diode bias. That worked too. I could hear the tone by sending the IF output of the frequency counter into my HP 8901A modulation analyzer and by listening to its modulation output. Looking even better.
The one hurdle left to solve was the opposite circular polarization. And I have done that too. But that will come on my next update of this entry.
OK, I have some good development to report. First the opposite sense of rotation to resolve. The fact to recognize is that a received signal having the same sense of rotation as the transmitted signal will obviously return to the Gunnplexer. So this is where you want to detect the received signal.
I noticed that the Gunnplexer housing had holes to populate mixer diodes but none was present on our police radars. Clearly, the same housing is used for various models of Gunnplexers. Some units have the IF port Pin 3 present (see picture) but the radar detectors do not, as the mixing occurs at the opposite end in the waveguide assembly, outside of the Gunnplexer. By the way, there are even models that have two mixer diodes populated for stereophonic detection of movement. But I had a couple of defective Gunnplexers on hand, and those had two mixer diodes each. These diodes are only held in place mechanically. So I simply transferred a diode in each of the two radar Gunnplexers, and Voilà. The same polarization received signal will head back into the Gunnplexer assembly and get mixed with the LO (the Gunn diode TX source) and produce the IF.
Now onto testing. I am happy to report that the first test was conclusive. With a spacing of 100 feet (30m), and a frequency difference of 88 MHz between the two Gunnplexers, I could establish a solid link (no background noise) between the two units. All I used is an FM radio not electrically connected to the IF port, just picking up the IF signal over the air.
So my next step is to try a much longer path, and with better IF radios at both ends. I plan to try 1800 feet, as there is a quiet straight road near where I live that offers that capability.
I will report back. Stay tuned!
I finally received the two cheap FM radios I had purchased on eBay. I then modified their tuning range to cover from 20MHz to 40MHz and I added a low pass filter to their antenna input to reduce interference from the FM broadcast bands. I also re-tuned the two Gunnplexers for 24.125 GHz and 24.155 GHz respectively, yielding an IF frequency of 30 MHz. So I was ready for an outdoors test. This time I went for a > 600-foot spacing. It worked surprisingly well, with a lot of margin left on the signal strength. The only issue was significant drift due to the direct sun and the high wind constantly varying the Gunnplexer temperature, which shifts its frequency.
I put together a chart package for my local ham club. It shows a lot of details on the electronics and has many pictures showing the internal guts of my modified units. It is on my website. See http://ve2zaz.net/Presentations/Presentations.htm to download the package (.PDF form)
Great website and blog, I enjoyed your excellent article on modifying a Jabra 210 bluetooth system.
You might be interested to see new product for wireless internet that will soon be sharing the 24ghz band. (I am not connected with this company). They claim it will pass data at over 1 gigabit and up to 13 km with no licence.