Open ahendrix opened 11 years ago
[schlicht] Software ticket was opened.
Tested R+W (-2) coupler but the low level oscillation increased by a factor of 3. This was somewhat expected.
Will try and test out using a motor with encoder instead of the separate encoder we use now. the idea is to both decrease cost and eliminate the low level osc.
The U-Penn Lsled was making a loud groaning noise when running at low speed. Testing here at WG we found a low level oscillation on the motor drive current. While the main fix was to square up the mechanical mounting bracket(s) to reduce the motor effort, it makes sense to eliminate the oscillation to prevent this from happening in the future. We tried reducing the controller loop gain from 8 to 4 and the amplitude of the oscillation dropped by 80%. But this gain reduction is probably not ideal at high speeds so a speed dependent gain adjustment makes sense where the gain starts out at 8 for the fastest speed and reduces to below 4 at the slowest speed. The exact gain range needs to be determined.
This ticket is associated to PR2SCDS ticket #1739.
trac data: