APM 2.6 compass calibration issue

[speedbird]

Hi, I've recently got an APM 2.6 with separate GPS and compass module, here's where I bought it from:



All has arrived that was listed, and today I proceeded to start calibrating the controller for an H quad build. However, I got stumped by pretty much the first thing, it all connected fine with the usb, and I selected the H quad no problem, but when it came to compass calibration I hit a snag. I hit the 'Live Calibration' button and moved the controller through all axis, but it just came back and said, "not enough data points received", so I tried again and again, both with and without the compass/GPS module attached, but it keeps on saying the first thing. I'm using APM planner 2.0.15, and I'm not sure if I should use a previous version.
Some help would really be appreciated as I've never used APM before, only KK2, so this is a new experience for me!

 

[makattack]

If it's a real APM 2.6, it doesn't have the onboard compass... If it's an APM 2.5 masquerading as a 2.6 board, there's a trace / solder pad that's cut to disable the I2C connection on the onboard compass. At any rate, if you have a 2.6, and that GPS antenna as the built in external compass, and you have two sets of cables going from that to the APM (the GPS serial lines, and the Compass I2C lines), and the software is setup with the "External Compass" as documented here:

http://copter.ardupilot.com/wiki/ac_compasssetupupadvanced/

Do you see that XYZ/3D graph with the colors dots as you rotate the aircraft around all it's axis?

You might try to see what the compass graph or the magnetometer values are. This is for mission planner, but I think it's pretty similar to what you'll find in APM Planner:
http://plane.ardupilot.com/wiki/common-diagnosing-problems-using-logs/#compass_interference

Lastly, you might need to go to the CLI interface, in APM Planner -- using the "Advanced Mode" you select from the "File" menu, and I believe the command is "test compass" if the above doesn't show anything.

Note that when trouble shooting a compass problem, the APM flight data display will make it "appear" as if the compass is working, but that's only because it's using the accelerometers to detect orientation change... it probably won't match real world compass headings.

 

[speedbird]

Thanks for the advice, I've downloaded mission planner, loaded the compass calibration with the 3 axis graph, and nothing happens when I move it around with the compass plugged in.

 

[speedbird]

I've just tried it again, and it seems to have worked for some reason.

 

[makattack]

I've had problems with some of the made in China APM 2.6 boards with bad I2C busses -- or so it seems to me. Different compass modules just don't work on them -- even ones that work on other arduino boards. Sorry to break the bad news. It could be a bad cable or just a bad compass, but unless you have a logic probe to test with, it's hard to say.

 

[makattack]

Oh great! Could it have been a loose connection? Those micro JST connectors/molex or whatever they're called are a PITA...