[Sheila, Elenna]

Previous alogs (64350 and 64139) we noted some ring ups in PRC2 pitch during lock acquisition. I adjusted the PRC2 P loop shape to further supress 1 and 3 Hz oscillations that show up due to the plant resonances. I updated the FM8 filter to be a modified plant inversion that also has an ELP at 2.5 Hz. "Modified" plant inversion in that we still are suppressing the 1.1 Hz peak with a pair of complex zeros but the low pass takes care of the ~3 Hz peak. In the attachment, the blue trace shows the new filter design, while the red trace is the old filter design. The new filter reduces the phase at the UGF (0.1 Hz), by 4 degrees, but the phase margin was ~65 degrees, so that shouldn't cause a problem. See the PRC2 P measurement in userapps asc/h1/templates/PRC2 for reference.

****EDIT*****

This was not the right solution. After implementing the change in the guardian, we had three locklosses right after acquiring DRMI and turning on the DRMI ASC, all with 0.95 Hz oscillations. I reverted to FM7 and there was still a 1 Hz oscillation when the ASC offloaded, but I was able to avoid lockloss (I think) by dropping the PRC2 P gain to -1 again. (Or it would have resolved itself anyway. Who knows). I will think more about the best solution tomorrow.