3DR Antenna Tracker Groundstation (WIP)

[Static65]

In my quest to stay at legal power levels and still get a good telemetry and FPV links I';ve started work on a servo-driven Pixhawk-based 3DR antenna tracker. I was going to run it with an old APM2.6 that';s gathering dust but it seems that even for AT purposes the Pixhawk is going to be the future hardware of choice. Plus the Pixhawk can handle multiple telemetry links with ease whereas the APM2';s require some fettling. I had a Pixhawk lying around that doesn';t work with power modules any more (keep getting ';check board voltage'; warnings when powered via module) that seemed perfect for the job.

There are some off-the-shelf antenna trackers available but they are almost always massive, available only in the US and out of stock, e.g. http://www.readymaderc.com/store/index.php?main_page=product_info&cPath=11_27&products_id=96. I wanted something a bit less bulky that I could drive with standard servos and hopefully build on the cheap.

My next issue was the size and weight of most directional 433MHz antenna designs. I can see why some of the off-the-shelf offerings are so bulky - it';s to accommodate the giant patches required for the 70cm wavelength. A bit of googling later and I had found the Moxon antenna design: http://www.moxonantennaproject.com/

This proved to be very lightweight and only about 250mm x 100mm in size for 433MHz. It seems these antennas need to be made with relatively high care to be sensitive on the correct frequency range, so I made a 3D printed mount that I could press-fit some 2mm wire into. Time will tell if it actually works well! I know a bit about RF but I';m no expert so it';s possible I';ve got the polarisation wrong - might have to fit a different antenna to the aircraft than the standard omni.

With the antenna as a starting point I then began to design the rest of the tracker around it. I found 2 metal gear servos on a set of RCTimer retracts that had a little unexpected meeting with the ground a while back, tested and thankfully they were both working and hadn';t stripped. I';ve built the pitch mount that holds the Pixhawk and 433MHz/5.8GHz antennas, and also the yaw mount which holds the GPS antenna on a 200mm length of 4mm CF rod. So far I';ve got the Pixhawk sending test signals to the servos but have yet to get them tracking an aircraft (real or simulated). I need to work out how to drive the HITL simulator!



As you can see I keep my workshop super tidy at all times

I';ve gone for direct drive rather than adding a gear ratio to the servos, mostly because I have yet to research how to calculate gear ratios and such. Seems to work but I';m getting some ';shaking'; from the yaw servo due to the mass of the tracker on top of it. If I stick my finger on it for a moment it stops, if I tap it then it starts again. Not sure if I need a beefier servo or whether some servo PID tuning will calm it down.

Here it is with the antennas mounted:



Things still to do:
Design base to allow mounting on a tripod, attaching battery, possibly mounting an FPV monitor too.
Add Raspberry Pi as a wifi access point to communicate via MAVProxy (design RPi mount into base)

Any engineering suggestions are welcome, this is my first foray into antenna tracking

[Static65]

Progress! I redesigned it slightly so that the pitch and yaw servos are inside the parts they are moving, which makes the whole thing a bit more compact. I also redesigned the moxon antenna mount with a lot less material to make it lighter.

Got the base designed and printed and my Rapberry Pi installed. I can now connect to the tracker over WiFi via MAVProxy which is pretty cool





Unfortunately I still have shakey servos! Does anybody know servos well? They are powered from a 5V 5A SBEC and the Pixhawk and RPi are powered from a separate 5V 3A reg, but everything shares the same battery. There';s a fair bit of play in the servos, much more so in the yaw servo. Once it gets to the edge of that ';play'; it tries to move back quickly and ends up simply oscillating. The pitch servo buzzes a bit because the weight isn';t quite central but I can also see small dips every couple of seconds.

Do I just have rubbish servos?