Hello, community, I hope, that somebody has a solution for this.
I've faced a situation in which sometimes randomly z axis of the IAM-20680 gyro axis becomes stuck.
As on the picture below (sticked axe in rest.jpg) you can see that in rest all of the axis show pretty the same behaviour in terms of range of noise.
But when I'm starting to rotate the IMU around all of 3 axis, one of them shows the same values as it would be in rest (you can see it on the sticked axe in moving.jpg - line wz).
As it's randomly occurs, it could randomly dissolve - after shutting down the power after few hours the problem could gone away.
By now, we have an opinion that there could be two possible causes: vibration and overheating.
Regarding vibration:
we assume that one of the reasons for the failure of one of the IMU axes is external vibration effects at the self frequency of the gyroscope axis oscillation, which leads to resonance and sticking of the axis.
Sticking - because the axis can subsequently "peel off" and start working again.
Judging by the datasheet on the IAM-20680, the natural frequencies of the gyroscopes are 25-29 kHz.
By now, we are attempting to measure spectrum of noise on tram.
Overheating: we conducted research in which we had heated several IMU's for 1 day at 175 degrees Celcius (way out of the temperature range of the IMU!). And 3 of 7 as a result obtained a IMU with exactly z axis becomed freezed. Thats the exact behaviour as occurs on trams.
Does anybody has a work around - know possible causes or ways to deal with gyro z axis becomes stuck?