aLIGO LHO Logbook

H1 SQZ

ryan.crouch@LIGO.ORG - posted 21:12, Saturday 02 November 2024 - last comment - 21:59, Saturday 02 November 2024(81027)

SQZ not locking

The FC keeps losing IR lock, it says "LO SERVO RAILED" pretty much as soon as the LO_LR grd startings working after the FC locks IR.

Comments related to this report

ryan.crouch@LIGO.ORG - 21:55, Saturday 02 November 2024 (81029)

Link

RyanC, Vicky

I tried bringing SQZ_MANAGER to FDS_READY then I tried to set FC{1,2} back to where they were the last good SQZ lock ~5 hours ago according to top mass OSEMs... same issue. Trending the ZMs showed some drifting over the past 5 hours, I was thinking about moving these back just as with FC2 then Vicky suggested to try Clearing ASC history. I tried this following Camillas alog80519, but the issue persisted as clearing ASC also cleared FC ASC which misaligned the FC. Vicky then logged on and starting trying to manually lock everything, successfully.

To summarize

SQZ AS42 ASC got misaligned between locks
clearing ASC also cleared FC ASC, so then FC was misaligned too
Vicky manually walked ZM6 P/Y, then when was some sign on OMC RF3 TRANS, locked the LO loop, then manually enabled SQZ AS42 ASC
then SQZ_MANAGER to FREQ_DEP_SQZ worked
then manually lowered FC ASC trigger threshold, so get FC_ASC to turn on

victoriaa.xu@LIGO.ORG - 21:59, Saturday 02 November 2024 (81030)

Link

Ryan C, Vicky

What I think was the issue and what I did:

SQZ AS42 ASC got misaligned (see how the range ran away at the start of lock)
- This is likely because squeezer ASC ran away, which I think is related to SQZ angle ADF servo running away. When I found it, the SQZ angle (as controlled by H1:SQZ-CLF_RF6_PHASE_PHASEDEG) was railed at 268 deg, whereas normally the RF6 demod phase is around 195 degrees. Because the known SQZ angle to ASC coupling, the ADF Angle servo running away could likely cause ASC to run away, which causes LO loop not to lock, and then FC-IR can't lock.
Ryan cleared ASC as he described above
- TO-DO for us next week: We should clean up what the SQZ ASC clearing does.
First, I manually walked ZM6 P/Y to get some signal on OMC RF3 TRANS (manually walked the RF3_MON up from -22 to -10 or so). Then I was able to lock the LO loop (LO loop was railing when SQZ ASC had it totally misaligned). Then I manually enabled SQZ AS42 ASC, and it worked to align ZM4/6 to maximize RF3_MON.
I then adjusted OPO temperature to optimize RF6_ABS, see screenshot. Here I just wanted to make sure the OPO temperature being wrong did not cause the SQZ angle to be wrong, which in turn could've caused ASC to run away.
After locking the LO loop, then SQZ_MANAGER to FREQ_DEP_SQZ worked, and SQZ_FC locked successfully.
But FC ASC did not turn on, because FC was misaligned (the issue described above). I manually lowered FC ASC trigger threshold (FC_ASC screen > TRIG: I lowered H1:SQZ-FC_ASC_TRIGGER_THRESH_ON from nominal 3 to 1.5), in order to force FC_ASC to turn on and align FC1 and FC2 for us.
Then I reverted the SDF's on FC_INJ_ANG_{GAIN,TRAMP}, which (I didn't realize) then applied the FC_ASC offsets stored in the filter bank histories to ZM3. This was good actually. Restoring this ZM3 alignment restored the FC_REFL power.
Then all SDF's were clear, and Ryan could go to observe.

Images attached to this comment