Q:
I am seeing a small jump in errors at both 90 and 270 degrees. Is there anything I can do to improve this?
---------------------------------------------------------------------------------------------------------------------------------------------------------
A:
These "spikes" in the error pattern (at 90 and 270 degrees) might be due to small errors in the accelerometer that is closest to the axis of rotation. Switching to a two-axis computation eliminates that error source, along with the spikes at 90 and 270 degrees. The MSC_CTRL register provides an undocumented option for switching between a two-axis and three axis angle computation. At the time of this post, the ADIS16210 datasheet does not reflect this function. The following two figures reflect the present and future datasheet descriptions of the MSC_CTRL register. For now, set MSC_CTRL[9] = 1, by setting DIN = 0xB502, update the flash and re-run your angle measurements to see if this make a difference. I hope that helps!
Very best,
NevadaMark
CURRENT (NOT CORRECT)
FUTURE REVISION (CORRECT)
