I have recently been carrying out mapping flights in somewhat more turbulent and windy conditions and it has brought to light a possible issue with noise in the accelerometer data from IMU 2. Specifically poor position holding, with a lot of drift and occasional loss of 1 or 2m altitude.
The issue only came up once I reduced vibrations to the FC to a reasonable level using hobbyking orange foam to suspend the pixhawk between the top and bottom plates of the frame.
As you can see from the attached figures the signal to noise ratio on all axes of the IMU2 accelerometers is much worse than the information from IMU1. Y axis is the least bad, Z axis the worst.
As you can see the range of the data and the averages are almost the same, but the (random?) noise in the IMU2 ACC data seems to reduce the signal to noise ratio in that data to almost nothing, surely making it much less useful than the IMU1 data.
Vibe charts also attached - they seem well within limits.
I decided to disable IMU2 and have noticed a significant improvement in the performance in pos. hold. and auto mode
Can anyone offer any insight into the issue?
Thanks very much.
Replies