I finally got around to setting up moving baseline support in my ArduRover (latest 4.5.6), and everything seems to be working correctly in terms of my setup— i.e. the GPS seems to be sending me very accurate yaw, except for one issue… The heading is approximately 90 degrees off. I double and triple checked my offset measurements for both the GPS master and slave antennas so I don’t think that’s the issue. But I did find something interesting.
Due to mounting constraints, my autopilot is mounted and configured for a Yaw270 orientation. On a hunch, I wondered what would happen if I modified the AHRS_ORIENTATION to None/0, and as soon as I did this, the yaw was perfect.
I would expect it to work out of the box with different AHRS_ORIENTATION values, but it does make sense if you look at the orientation of the GPS antennas with respect to the AP unit— the antennas are offset 90 degrees to the AP unit, even though they are properly positioned in the vehicle.
Questions:
- The documentation says the GPS antenna offsets should be measured with respect to the center and orientation of the vehicle (at least my understanding of them), but should the documentation be updated to reflect that the offsets should be measured with respect to the AP hardware?
- Or is the documentation correct, and this is actually a bug that should be fixed?
- If the offsets are meant to be measured against the AP, what if the AP is not mounted in the exact center of the vehicle? Should the offsets be measured in distance from the center of the vehicle, but in XYZ orientation of the AP? Or should the offsets be measured in both distance and XYZ orientation of the AP?
Thanks in advance!