jenne.driggers@LIGO.ORG - posted 14:13, Wednesday 03 October 2018 - last comment - 20:50, Wednesday 03 October 2018(44316)
Wed AM locking
- Running A2L helped sensitivity. Script seems like it reverted the A2L gains, and we lost the improvement after the script finished ETMY P2L. I trended the P2L value, and setting it back to the 2.2 it was during the time that the range was good (vs. the 2.8 it decided on) brought the range back up. This is a 5% range increase.
- Probably when the A2L value started far from the correct answer, we need to run it again. We should add a warning or summary of then/now a2l gains at the end of the a2l script.
- Happened again, after moving CSOFT P offset.
- 6:45am, starting to go up in power, very slowly.
- 7:03, arrive at 15W injected.
- Moving SOFT and PRC1 offsets:
- Dither CSOFT_P at 0.03 Hz. Can see with even amplitudes of 0.003 that we could improve the power buildup in the arms.
- Had to move PRC1_P offset to keep POP_B_18 high and POP_B_90 low. Alternated between CSOFT_P and PRC1_P offsets.
- DSOFT_P had almost no effect on circulating power.
- CSOFT_Y improving circulating power. No need (yet) to change PRC1 offsets. POP18 staying high, POP90 actually getting better.
- Since SRC ASC loops are open, Gabriele moved SRM to get POP 90 power low.
- Moved offsets from the loop error points to the QPDs themselves for TRX, TRY, POPA.
- Ending recycling gain = 49
- Dither CSOFT_P at 0.03 Hz. Can see with even amplitudes of 0.003 that we could improve the power buildup in the arms.
- Ran A2L
- Increasing power again at 8:45am to 20W injected.
- PRCL gain is too low (MICH, SRCL okay). PRCL UGF was 30 Hz with 26 deg phase margin with digital gain of 8. Increasing to 12.
- Increasing power again at 8:57am to 26W injected.
- Fast lockloss after ~5 min at 26W injected. Guardian lockloss time: 1222617719.9375. Actual lockloss is at about -2.88s from there.
- CHARD sees something at -3.9s
- LSC-MICH sees a small glitch at -3.9s
- LSC-POP_A_RF45_DEMOD_RFMON starts walking upward from -27 toward +8.5, starting at -5.2s, and hitting +8.5 just before the lockloss. Other DEMOD_RFMONs don't see this walk-away. POPAIR_B_RF90 doesn't start increasing until much closer to the lockloss.
- Re-exectued TCS CO2 guardians so that it will turn on preheating.
- Initial alignment.
- Xarm green ASC worked, but only after I hand aligned it to be quite close.
- Relocked to 15W with no problems, other than a long time to acquire both PRMI and DRMI.
- Gabriele has measurements of the HARD ASC loops, they seem much more marginal than they were in O2. Other than a sign flip in the filter bank gain (haven't checked if there is a matching / undo sign flip elsewhere yet), the gains and filters are the same as during O2. The DHARD gains are a little higher now, but all the filter modules have stayed the same.
- Lost lock reducing CHARD_P gain while trying some lownoise ASC tuning.
Images attached to this report
Comments related to this report
At 15 W, we could reduce CHARD_P gain by a factor 2 and engage the low pass filter.
We tried the same with CHARD_Y, and lost lock while reducing the gain by a factor of 2.