About compass heading errors

Hello everyone.
I’ve always used BlueROV2heavy.
I have a question about the compass.

The problem occurred after we changed the ROV’s Raspberry Pi from version 3 to version 4.
This time, the ROV was manually controlled using MAVROS to face the north direction.
For a few minutes after the ROV was started, the compass was correctly pointing in the north direction (0/360°). However, after 10 minutes, the north indicated by the compass gradually changed its heading to counterclockwise. (Finally, the compass stabilized at 283°.)
The ROV operating environment was a large outdoor area with a very large plastic case filled with water.

I would like to know about the magnetic orientation of the ROV.
Is there a significant change in magnetic heading due to heat with the BlueROV2 Pixhawk? Also, does the ROV need to be stabilized for several minutes to stabilize the magnetic heading?

Please lend me your wisdom.
Thank you in advance for your help.

Translated with DeepL Translate: The world's most accurate translator (free version)

Hi. All the sensors are on the Pixhawk, so the Raspberry Pi 4 shouldn’t matter.
I believe what you describe is the same issue here: Sub: Yaw drift issue · Issue #18229 · ArduPilot/ardupilot · GitHub

It usually goes away with a reboot. If you are booting in land (not a boar) you can set INS_GYR_CAL to 1 (calibrate gyros at startup).

1 Like

Hi. @williangalvani
Thanks for the advice.
I’ll give it a try today and see if there’s any yaw drift.