It seems current release of Mission Planner (perhaps some earlier too) has a bug in the compass calibration routines - offsets almost always are calculated as 0 0 0.
Compass works fine in the terminal (shows correct heading etc) but after calibration the compass tries to get back to North.
Yet I calibrated compass in the APM Planner 2 without any problems, calculated offsets are resonable and the compass behaviour is normal after calibration.
Anyone else with similar problem?
can you post a tlog?
The one thing I notice recently is a lot more compass "bad health". Not sure why. I asked my compass and he said he is feeling fine. :-)
At the moment I'm not quite sure which log is which, but it's pretty possible that tlogs 25 and 26 are after the calibration in the MP 1.2.9 and tlogs 41 and 42 are after the calibration in the APM Planner 2. I've tried to record new logs, but perhaps the board only logs anything when it's armed? And currently it's out of the frame.
Instead I also recorded two movies - one after the compass has been 'messed up' during the calibration in the current version of the Mission Planner (offsets again 'defaulted' to zeros after the compass calibration) and then I connected the board to the APM Planner 2 and performer another calibration.
Freshly after the compass calibration in the Mission Planner 184.108.40.206:
Compass calibration in the APM Planner 2 and after the calibration...
Let me know what do you think - I can perform more tests etc tomorrow (it's 4 am here and I have to take some sleep ;) )
PS: I have recently replaced the 3.3v regulator (same low noise Texas Instruments regulator as the original) and the schottky protection diode in this board, but I don't think it's the cause for this issue as the tests performed in the terminal are looking fine and the calibration procedure works fine in the APM Planner as well (offsets are pretty high, but I'm performing this calibration inside, and some big metal objects are in the 1 m radius from the board).
PPS: Of course the calibration procedure has been perfomed just like it's described here: http://copter.ardupilot.com/wiki/initial-setup/configuring-hardware... or here: http://ardupilot.com/forum/viewtopic.php?t=6726&p=13164
the tlogs are recorded in MP while you are doing the calibration.
You have given me logs. there should be tlogs that should have a date on them that im after. This way I can replicate 100% whats happening.
Yes, my bad.
I performed the calibration in the mission planner again. Again 0 0 0 offsets and strange 'rubbery' compass behaviour after the calibration. tlog attached.
PS: And again - everything back to normal after the calibration in the APM Planner 2 (but APM Planner 2 doesn't collect any logs, right?).
please try the latest beta mp. it should be fixed again.
Yes, it's fixed now. Thank you, you are the best.
BTW - Polish MP translation is very poor (and full of mistakes), I can give you a hand with that.