[Sheila, Ryan, Masayuki]

We got DRMI to lock for several minutes. By the end of the session it was re-locking every one to two minutes, though each stretch of lock lasted less than five minutes. 

After locking PRMI and aligning the BS and the PRM, we requested ACQUIRE_DRMI_1F and tried several combinations of the PRCL/SRCL/MICH gains. Eventually we found the combination which can grab the DRMI lock, and that was the original gains except for PRCL. The PRCL gain had been doubled yesterday (input matrix 1.75 -> 3); we reduced it to 2.5, and with that we could lock several times.

In the short stretches of lock we measured the OLG of each loop. PRCL and MICH both have a strange phase rotation around 10 Hz. The MICH loop gain was also on the low side, so I increased it and tried to lock again.

While waiting for the next lock in that configuration, I moved the SRM more or less at random by ~20 urad in pitch and yaw, and found that the better alignment made POP90 more stable (see attached plot). In that stretch I re-measured the MICH OLG, and its UGF was close to what we had for PRMI. So this looks more like an alignment issue than a loop gain or loop shape issue.

I then found that the MICH loop gain was by now too high to engage the boost (FM1 of MICH1), so I halved it. With that gain the guardian could engage all the filters in the DRMI_LOCKED state.

I measured the SRCL gain again and it came out lower than in our first measurement, though I am not sure that measurement is valid. I tried a higher SRCL gain and it seemed unstable, but that may have been the alignment degrading rather than the gain itself.

I stopped here. The guardian is left in the configuration that felt like it gave the most frequent locks.

The attached TFs are the first MICH/PRCL/SRCL gain measurements, and the SRCL gain measured just before I finished (second left hand side).