Just setting up a 500Trex with APM 2.5 and firmware 3.1, and the compass calibration did not go well. At least the “Live Calibration”. I did it three times with these offsets detected:
+66 -27 -28
+22 -78 +10
+99 +71 -97
All done at the same spot outside, away from magnetic interference, with roughly the same movements of airframe. Incredible! Naturally, none of these calibrations had the resulting yaw anywhere near accurate. Anywhere from 40 to 80 degrees off!
So… I entered values for X and Y axes manually, until I got N-E-S-W accurate, aligning the airframe manually with known direction and a good quality hand-held compass. Offsets necessary were
X-axis N/S -48
Y-axis E/W -40
I didn’t set anything for the Z-axis. Considering these offsets are not very large, I assume then that interference was not the cause of the poor results of the “Live Calibrations”.
Since Live Calibration does not appear to work properly, I will have to go ahead with test flights using my manually calculated offsets. But if I enable “Learn Offsets”, will my manually-set offsets become more accurate?
Anyone else experience such problems with compass calibration?
Wow, I’ve experience some variability in the calibration, but not that much! Can you show how the mag is mounted?
You must have some ferrous metal nearby.
In the past, my best mag calibration was to do a manual calibration, to get something reasonable. Then take it for a nice manual sport flight. All orientations and directions. Then use the “Log calibration” procedure. That usually gets me something really good.
After further testing today I am going to assume it is the proximity of the motor, which deflects my hand-held compass quite a lot at a similar distance. Since joly had mounted his board on the 450 not all that far from the motor, I had hoped my position would be ok: on an aluminum tray underneath the frame, at the rear.
So… will install an external mag out on the boom where I can see little or practically no compass deflection. I’ll report back with results in a couple of days. When you say “Then take it for a nice manual sport flight,” I assume you mean with the learning parameter enabled?
No, I never use inflight learning. I have gotten bad results from that in the past. I do the offline log calibration, that way I can sanity check it. I find it does a better job.
But typically, I have zero issues with compass on helis when they’re back on the tail.
However, mounting it where you have it now, it might work, in flight. Once that field is rotating, everything changes. I’ve never done it. But I respect Joly and if he says it worked for him, I believe it. Maybe ask him for more info.
Installed the external one on the boom today - seems much better but I couldn’t take it outside and Live Calibrate - snowing here! While off its shelf and the case apart, I could also put a light-proof shield over the baro and re-wire ch7 for my camera control. I’ll have to experiment more with the gimbal control, and take a look at the forum for that, So far it has proved impenetrable for me, BUT, if at first you don’t succeed, read the instructions!