georgia.mansell@LIGO.ORG - posted 22:33, Saturday 05 January 2019 - last comment - 06:11, Monday 07 January 2019(46248)
Locking today
Magnitude 6.8 earthquake in Brazil hit before I arrived. Nothing tripped, seismic FOM 0.03-0.1Hz peaked at 3.5 um/s, decided to wait it out without hitting the "very large EQ" button.
While powering up saw huge 240 Hz peak, and larger-than-expected 60 Hz line. Unplugged SR785 from PSL racks (was in the IFO common mode servo board) just in case, remembering the ground loop issues from earlier this week, peaks went away and did not return.
Wanted to turn down ASC dither line frequencies (and amplitudes). Found the dither line implementation in guardian confusing. Currently the dither lines (without servos), i.e. YAW1 YAW2 YAW8 YAW9, are turned on in prep_asc, and the dither servos are turned on in engage_asc. Moved the setting of the dither servo LO frequencies to prep_asc.
Note to self: current assignment of ADS lines (this info is already in the log in a few places)
Frequency (Hz)
LO actuator
Demod actuator
Yaw1
16.131
SRM
-
Yaw2
17.37
PR2
-
Yaw8
23.25
CHARD
-
Yaw9
21.11
CSOFT
-
Yaw3
18.37
PRM
PRM
Yaw4
22.347
EX
IX, EX
Yaw5
21.9
EY
IY, EY
Pit3
19.653
PRM
PRM
Pit4
20.131
EX
IX,EX
Pit5
20.789
EY
IY,EY
Naively tried to get away with just halving these dither frequencies putting the dither lines around 10 Hz to make the noise at 20 Hz cleaner.
I tried this, got to NLN (22:43:22 UTC), but lost lock 6 mins later due to a familiar 0.45 Hz oscillation in the soft loops (pitch). I don''t think i was saturating anything with the dither lines but they were very large in DARM.
The broadband noise at 20 Hz looked bad anyway, so I did not put the dither frequency change into the guardian.
Reached NLN again at 02:35:02 UTC with the nominal dither line frequencies.
Reduced amplitudes of lines for spot position servos, added lower amplitudes to guardian in lownoise_asc state, as yet untested in lock acquisition.
Lost lock due to overzealous TMS shaking
In an intermediate locking attempt another earthquake from Japan hit as we were engaging ASC for full IFO. Hard loops oscillated, all TM PUMs tripped, violin modes got rung up. Had a problem with ETMY violins - mode 1 and mode 6 are very close together in frequency, and one of them was getting rung up with the guardian damping settings. Turned off the damping feedback to mode 6, reduced the gain on the feedback to mode1, this did the trick.
After Dave restarted all the EY models I untripped the watchdogs. I was able to get up to FIND_IR again at one point, but ALS was being flakey so I gave up locking for the night.
Comments related to this report
peter.fritschel@LIGO.ORG - 06:11, Monday 07 January 2019 (46252)
The L1 ASC dither frequencies are all in the range of 6-9 Hz. Why were the H1 dither frequencies set so much higher in the first place?