Another Compass Offset problem!

Any advice on how to lower those compass offset numbers? any kind of shield to avoid electrical interference? I’m currently using the ArduCopter 3.4 and is working pretty good, the hexa is flying better and better but with some toilet bowl effect, very slow but it has it. any advice will be welcome. external?



test it with just usb power with the same results, bet that if i pull it out of the frame will be the same. Any idea? burning hours on this project, none stop since last Wednesday, but i like it!!


Hello there!

Due to a petty inconsistency in APM source code, the issue cannot be solved in a clear way. Now APM uses HMC5883L’s hardcoded compass offsets to perform pre-arm checks. Since we use another compass (namely, AK8963), the offsets are different. Fortunately, the offsets don’t play a significant role in the position sensing. That’s why you can safely ignore it for now. (disable the failing pre-arm check in Parameters section in your GCS of choice).

Hopefully, this’ll be fixed in the next release. There’s already a PR pending that addresses this exact problem.

@porelmundo The offsets values that you are seeing during the calibration are different from what you would expect from a Pixhawk or another APM board. It does not mean that they are too high, they are just different units. We are now working to push changes to APM that would allow all compasses to report offsets and values in the same units. That should make direct comparison easier.

1 Like

Ok Thank you, i will continue testing and let everybody know, for now is working ok and i didn’t disable the pre-arm.
Thank you for your support