POP_A_LF and AS_AIR RF90 aren't looking good. I while I was wondering what's going on I noticed differences in SDF and ASC INP and PRC1 pitch and yaw gains were 0 (setpoint=1). Was this a leftover from a2l script error? I reverted it, accepted the changes and hoped for the best. No luck though. IFO just lost lock at NOMINAL LOW NOISE. CSOFT was running away.
Seeing how a2l script left us with problems tonight I WILL NOT RUN IT until I know that it's safe to run.
Lost lock at ENGAGE ASC PART3 twice in a row. I've attached the log and the lockloss plots of both locklosses (and the ASC Striptool from the recent lockloss). Sorry about the terrible screenshots of the log (only showing the DOWN state).
I called Kiwamu to make sure the ASC loop gains were correct. Turned out INP and PRC1 (pitch and yaw) gains were 0 when they're supposed to be 1. I put in the gains by hand before realized that PRC1 had no ramp time at all. There was a spike from PRC1 yaw and the ifo lost lock shortly after. So I gave PRC1 ramp time of 10 s.
Nutsinee, Sheila, Kiwamu,
Nutsinee had a number (5-ish times) of lock losses in ENGAGE_ASC_PART3. We blame TMSY for these lock losses.
Nutsinee went through the initial alignment and it corrected TMSY in pitch by 2 urad. See the attached screenshot showing 2 day trend of TMS angles. A large step shown in the middle of the screenshot must be the temporary correction that Corey introduced in the last night (alog 22535). You can see that the TMSY angle is now back to where it was 2 days ago. Tonight, the symptom was that when we engaged the SOFT loops, something started running away and decreasing the cavity power everywhere. We think that TMSY was pulling the SOFT loops to some kind of bad alignment point due to the different TMSY angle. It was nothing to do with the 20 dB boost in the SOFT loops (for example, alog 21587) which we suspected at the very beginning of the investigation. After the initial alignment, everything seems to be going as smooth as usual and we made it back to low noise (alog 22579). We are happy.
In additon, there was another issue with INP1 and PRC1 where they were not engaged in ENGAGE_ASC_PART1 and PART2. This was fixed by reloading the ISC_LOCK guardian since it had been edited in the last evening but was not reloaded since then.
