High Internal Compass offsets

Hi All,

I am after some help with a compass issue.

I have a Pixhawk running Copter 3.3.3 and after performing a compass calibration I always wind up with offsets around: x:150 y:-700 z-260.
The calibration has been performed in the frame, out of the frame, with and without the external compass connected, and in each case I get the same result. At this point the Pixhawk has been completely re-flashed and the compass calibration re done (nothing attached except the USB cable and in a different physical location) with the same result.
External compass when connected calibrates fine.

Based on this is there any fix or do I have a faulty Pixhawk?

Hope to hear an answer - I have the same issue.
Not as high (480) but offset number is yellow.

Might want to give this a go and see what happens with yours.

What I tried was to pass a fairly strong magnet all over the outer case of the Pixhawk and then try the calibration again.
After doing this a few times the offsets are now coming up green more often than not when the calibration is performed.
Not say it is fixed, but it certainly showed a change,
On a side note, I now have 3.4rc5 install with the latest version of MP installed and the compass calibrates after only doing 2 axis. Will change back to 3.3 and see what happens.

Meh, I take mine to the field for compass calibration.
No power lines, at least 50 feet away from the car and at least 15 feet away from the laptop with volume on full waiting for “Compass Calibration Complete”.

I have the DJI450 clone frame.

I did note this morning when I opened Mission Planner there was a new version.
I never read release notes so I am not sure if the offset might have been linked?
Only 2 ways to find out - read or do a compass calibration.

So far the magnet trick has seemed to solve it for me, now I can calibrate it in any location with both the latest and previous versions of MP and is seems OK. I had tried calibrating in some wide open spaces where there was little external interference and had the same problem. (Also had another Pixhawk on hand which was calibrated in the same location as a reference). Sure looks like the odd Pixhawk had become magnetised in some way.