Sheila, RyanC
I ran the full (all 6 dof) undamped TFs and the OLG templates for P and Y for PR2. For P and Y the magnitude of the TF was off. On the OLG for Y the magnitude is low, and we're not fully sure why. We doubled the gain of the P and Y damping loops from -0.5 to -1.0, this reduced the oscillation seen when moving sliders by ~half, see before and after... Also noting that PR2_Y sees the oscillation both in M3 and M1 whereas PR2_P only really sees it in M3. This is likely not the final solution but it helps for now.
I have accepted these two gain changes in the PR2 SDF safe file.
After Talking with Rahul, we're thinking it could be a coil driver or sat amp issue since the OSEMs themselves are working as expected alog91527. We should try power cycling the coil drivers and or sat amps then swapping them if the power cycle doesn't fix it. Looking at the other HSTS suspensions M1 DAMP gains, they're all -1.0, execpt for the recycyling cavity optics which are all -0.5.
Some more investigation into this revealed that the magnitude difference in the undamped transfer functions is from differing M1_TEST_{P,Y}_GAINs, the previous TFs were taken when said gains were lower, P and Y were both reduced to 1.0 for about a week before being reverted. So that makes sense why the undamped TFs magnitude was high than the reference.
The OLG magnitudes being lower than the references was also due to a gain difference as compared the reference time, P and Y's DAMP gains were set to -1.0 during the reference time whereas the current value was at -0.5, that measurement was taken before alog68903 where the HSTS gains were adjusted.
The differences between the traces magnitudes on the undamped TFs was ~equal to the differences in DAMP gains, ~0.8 and ~1.7 for P and Y which use the TEST_EXC channels which care about the TEST_GAINs.
PR2 seems fine now, the increased gains have reduced the oscillations.
