Motor interference compensation in AC 2.9.2

A bench-test demonstration of the impact motors can have on a multicopter's heading and the improved results after the new AC 2.9.2 compass-motor compensation is enabled.

0:51 ~ 2:01 : interference of 40 degrees caused by motor
2:38 ~ 3:33 : motor compensation set-up method
4:25 ~ 5:37 : interference reduced to only 6 degrees

[rmackay]

damo642,
No, it's quite different. We're measuring and then trying to counter-act the interference that increasing the throttle has on the compass. So we measure the change in the compass mag field as we increase the throttle. Then as we're flying we look at the throttle and dynamically subtract the expected interference from the field before calculating the heading.

[rmackay]

It's quite different because we're measuring (and trying to counter-act) the interference that increasing the throttle has on the compass. During normal compass calibration we measure the offsets and scaling of the compass without interference.

[rmackay]

DesperatePilot,
Possibly. I've also heard of people talking about using MuMetal to try and shield the APM from magnetic interference but I've never tried those solutions. I think many people suffer from the interference in the video above and it results in "toilet bowl"-ing, curved paths, copter flying off in the wrong direction, etc. I'm slightly amazed at how bad this copter is compared to my others. I'm sure it's the power distribution wires.
thanks for the suggestion

[rmackay]

We call it "motor" interference but I think it's actually the power distribution wires that are in the body of the hexa (the APM sits on top). There's about 2cm between the wires and the APM.

[DESPERATEPILOT]

Randy, can't the wires be shielded? Perhaps by wrapping them in foil?

[kamashi]

this isn't similar to compass calibration?
what differs?