As a devoted planespotter I always wanted to do “something” with the signal that most of the live ATC sites use for tracking and displaying planes. They use responses from a Secondary Surveillance Radar interrogating the aircraft within it’s beam, more precisely the mode S also known as ADS-B (not sure if there is a difference between the two, yet).
First thing I did was to tune my handheld radio scanner to 1090MHz, as this is the frequency for the replies, and sure enough, I could hear some squeaks. As the data rate is 1Mpbs, the headphones output was no use, and tapping the IF output on my I-com receiver proved to be too complicated and risky. So after some research I picked up the miniADSB kit from guys at http://miniadsb.web99.de/. I put it together like that:
And with the my scanner’s antenna I could even see some pulses on the scope that looked about right!
So I tried to come up with my own antenna, somewhat smaller (the one from scanner is some 0.7m high) and designed for the band.
Don’t laugh at my ignorance of RF design, but I gave the good old quarter wave a shot. It may not be a best choice for these frequencies, but I wanted something as isotropic as I could get. I figured if this doesn’t work out, I’ll try the co-linear ones, as they seem next in line for my skill and needs.
I bougt a BNC connector which looked perfect for the job: it had four mounting holes around the centre pin. I cut some thick copper wire to the adequate length, which was some 65.4 mm (I know, that’s not a hell of a wire, but that’s what you get at 1GHz), same for the radials and vertical radiator, allowing for the differences in total height/length caused by elements to which I was attaching them. The end result looked like this:
Doesn’t it remind you of Sputnik? After attaching it to the receiver, I was surprised to find that it would still pick up some signal! It’s difficult to tell what’s the actual gain or loss compared to what I had, I will have to try to gather data for some representative period of time with both designs and we’ll see.
I used my VGA Comparator circuit for a quick-and-dirty check if I could get some usefull digital output (AD8313 gives something like 0.7-1.0V on average for what I pick up with the antennas), but it didn’t have any RC filters so I quickly dropped it. Now I’m experimenting with what DL4MEA’s PICADSB9 is using as it’s analog frontend, with some mixed results. Will report in next post how it’s coming along.
For now the shielded miniASDB box with my antenna looks like this:
I also implemented a Verilog module that detects the ADS-B sync pattern (16 bits) and decodes the Manchester data that follows. It shifts the data out using a synchronous serial interface that feeds nicely as SPI into a microcontroller of choice (complete with active low Chip Select signal). And it fits into CPLD (xc9536xl, the smallest Xilinx’s). More on that will follow as well.