I can't get past FIND_IR consistently, the last attempt COMM was doing nothing so I found it by hand then it was stuck in an endless cycle of "found seomthing, sleeping for 5 seconds". Next attempt was successful but there were no flashes on DRMI, went to CHECK_MICH which offloaded, still no flashes. I brought PR3 back to where it was and still didn't get any flashes and we went back into CHECK_MICH which again finished, then we were able to lock PRMI but not DRMI, got PRMI again then lost lock. I'm going to send out some calls for help now. Keita said he's going to check it out.

TITLE: 11/05 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Lock Acquisition
INCOMING OPERATOR: Ryan C
SHIFT SUMMARY:

• Shift started with opportunistic PEM commissioning
• 23:32 Resumed observing
• 01:35 Dropped observing from SQZ ISS pump dropping out (tagging SQZ)
  • Guardians brought everything back automatically
• 01:37 Resumed observing
• 04:12 Lockloss - alog
• Struggling to relock; locklosses at various places
• 05:50 Starting initial alignment
• 06:23 IA finished, trying to lock again
• Still no luck; issues seem to be around TR_CARM steps. Still having the problem where ALS_DIFF is not stable when looking for IR resonance during FIND_IR, but I can usually get it good enough to progress past CHECK_IR
• Looking into possibility of bounce/roll modes being rung up, but they seem to be around this level at this point in the lock sequence
• Walking through steps of START_TR_CARM line-by-line (the problem state); lockloss happens soon after ALS-C_REFL_DC_BIAS_GAIN is set to 24
• Now struggling to get even up to locking DRMI; green in arms is having trouble staying locked, possibly due to higher microseism?

Tagging in Ryan C to continue troubleshooting.

LOG:

After our most recent LL from NLN we haven't been able to relock, microseism has increased quite a lot over the past 12 hours, yarms pretty unstable. The COMM beatnote isnt great -16, PR3 looks to have drifted a tiny bit in yaw? Not sure if its a real culprit. I think im going to try another IA and try to get the COMM beatnote better by following Jennes alog 73892 then call for help if I can't make any progress.

Lockloss @ 04:12 UTC - no obvious cause.

ETMX saturation right before; looks like LSC DARM saw first motion. Secondary microseism has been on the rise over the past 12 hours as well.

State of H1: Observing at 152Mpc

H1 has been locked for just over 24 hours. Quiet evening so far since PEM commissioning wrapped up.

FAMIS 26505

3-month trend for the SUS HWWDs attached. As previously seen, ETMX had the fewest switches of the bunch, the last one being on August 3rd, just beyond 3 months ago. This seems like a borderline case of weather action is required or not, but worth monitoring.

TITLE: 11/04 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Commissioning
OUTGOING OPERATOR: Austin
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 7mph Gusts, 4mph 5min avg
    Primary useism: 0.03 μm/s
    Secondary useism: 0.28 μm/s
QUICK SUMMARY: H1 has been locked for over 20 hours. We've just started opportunistic PEM commissioning while L1 is down (no more than 30 minutes).

• All systems look good
• Manual damping of IY05/06 and IY08 is on and working well; I'll keep an eye on it

TITLE: 11/04 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Observing at 153Mpc
INCOMING OPERATOR: Ryan S
SHIFT SUMMARY:

- EX saturation @ 17:40

- Robert is currently running PEM measurements until LLO comes back up

Quiet shift otherwise, nothing else to note.
LOG:

No log for this shift.

Quiet day so far, H1 has been locked for 17:19 hours and all systems appear to be stable. Range does appear to be on the lower end of the spectrum, hovering around 150 Mpc.

I've created a new ndscope trends web service which, for each ndscope yaml defined, generates 1 day, 1 week, 1 month and 3 months trend plots.

The URL is https://lhocds.ligo-wa.caltech.edu/ndshooter.html

Previously the yaml file's title line defined the lookback time, but this lead to many duplicate yamls. The new python code defines the lookbacks in a dictionary, which defaults to 1, 7, 30 and 90 days.

Please contact me if you would like to see more trends added.

Currently the update time is every 15 minutes, which includes a 4 minute acquisition time.

Sat Nov 04 10:09:53 2023 INFO: Fill completed in 9min 50secs

There have been no gps errors since the timing master was replaced 17 hours ago.

As was briefly mentioned in alogs 73853 and 73917, I injected into thermistor 2 (hot one which is directly mounted on the compression ring of the OM2 mirror) and thermistor 1 (cold one mounted on the back of the T-SAMS structure) independently, and measured the transfer fucntion from the injection to DCPD-SUM (and some other TFs).

All traces in the attached show the transfer function and coherence from OMC-TEST_EXC (calibrated to the voltage across the Thermistor) and DCPD_SUM (in mA) when H1 is in high power. Red is the injection into Thermistor 2 (hot one), green is into Thermistor 1 (cold one), both while the heater was ON. There are many peaks, most notably the one centered at around 277Hz (that's where CW people observed the most problematic comb structure) but there are others like 78, 93, 105, 144, 154, 166, 314 and 356Hz with good coherence.

The coupling is actually very small for broadband noise, it's only problematic for lines like we've experienced. For example, even if we take the peak at 277Hz (-75dB = 1.8E-4 mA/V) for Thermistor 2, in order for the voltage noise to be comparable to e.g. the shot noise for 40mA DC (~1.1E-7 mA/sqrtHz), the voltage noise should be ~600uV/sqrtHz (because 1.1E-7 [mA/sqrtHz] / 1.8E-4 [mA/V]) at 277Hz, which is a big number.

Thermistor 1 TF amplitude is smaller by more than 30dB than Thermistor 2 measured at around the 277Hz peak. This cannot be explained by the current difference due to lower resistance of the hotter thermistor (thermistor 1 is 7.41k, thermistor 2 is 4.08k, measured on the floor). It should be the location of the thermistor. FYI Thermistor 2 is mounted on the compression ring of the T-SAMS and is therefore much closer to the mirror than Thermistor 1 that is mounted on the back.

Within the resolution of this measurement, frequencies of Thermistor 1 peaks that are visible do match with Thermistor 2, but the resolution is not great (~2Hz at around 277Hz) so it's not clear if we're looking at the same resonance driven by different thermistors, or if this is actually two different resonances, each belonging to one thermistor but not the other.

Blue is the injection into Thermistor 2, but this time the measurement was done immediately after the heater was turned OFF. I only measured down to 220Hz or so, but anyway I see no real difference between the blue and the red. This means that the magnetic field formed by the heater loop around the mirror is NOT relevant for the coupling. Since our experience was that the comb was NOT there while the OM2 was cold, maybe the coupling is dependent on the tightness of the compression ring.

Injection amplitude was constant (A=8000 i.e. 16000cts pp sine wave for all frequencies, corresponding to ~4.9Vpp across the thermistor), and I was making huge lines in DARM (2nd attachment).

After the measurement, all breakout boards and temporary cables and handheld voltage reference and things were disconnected from the system (but were left on the work table by HAM6, I'll pick them up during the next maintenance). EXC cable was connected back to the DCPD whitening, and Beckhoff cable was connected back to the OM2 heater driver.