(Elenna C [in spirit], Ibrahim A, Corey G, Tony S, Jim W)

Summary:  H1 survived a magnitude 6.0 EQ in Alaska via new ASC High Gain button/state.  This preserved a 24+hr lock.  This probably saved us (at minimum 2.5hrs of downtime!):  (1) ~90 min for the earthquake seismic noise to die down & (2) for 1-2hrs of locking.  L1 went DOWN due to this EQ about 20min after we saw it rolling through LHO!

Had my first experience getting to ENGAGE the new RED "ASC Hi Gn" button (see Elenna's alog) for an earthquake!  This tool is still fairly new and "manual" (& cool!) for operators.  Main items for engaging at this point (for me, as a newbie) was waiting for SEI_ENV to go to EARTHQUAKE mode and also watched for any colors on the Picket Fence.  While we were dealing with this earthquake, Jim walked in to also observe and answer questions for us; Ibrahim also gave advice since he's engaged this EQ mode twice.  All the details for what happened are logged below.

How Everything Went "Down"!

2102 VerbalAlarm:  "EQ alert for M6.0 in Aleutians"  (it said "Incoming earthquake from Canada")

• 2111 Seismic:  EQ Mode Activated! (due to R-Wave arrival)
• 2112 Out of Observing & clicked Earthquake on Observatory Mode
• 2113 Hit the "ASC Hi Gn" (red) button (on this bright green pop-up terminal, I typed in "yes".....but had to open another one, because it only needs "y").  Within a minute or two, the script completed the transition to "ride" out the earthquake.
  • Also around this time, started seeing yellow on picket fences, followed by some oranges.  This EQ started to show up as a big spike on the earthquake band on the wall (see 1st attachment).
    • Also saw this EQ in secondary microseism.
  • 2114 ASC oscillations appear
  • 2116 ALL Picket Fences are yellow or orange!
  • 2117 Peakmon reaches max just a hair under 2500 nm/s on Peakmon
  • 2117, 2121:  EX saturations on VerbalAlarms
  • (2134 L1 goes DOWN due to EQ....2nd attachment for the earthquake when it took L1 down)
  • 2154 Seismic back to CALM
    • SEI_ENV took about a minute or so to make the transition to CALM (all guardian nodes on ISI_CONFIG eventually go green).  Once everything was GREEN....
    • On ISI_CONFIG, clicked the blue "ASC Low Noise" button & this pops up another bright green terminal, and I entered a "y" and this transitioned H1 back to NLN with our usual High Sensitivity.
  • 2157 H1 Back to OBSERVING!
  • (2254 L1 back to OBSERVING)

Using Vibration Sensors To Gauge Health Of HVAC Fans Site Wide FAMIS 26592

H0:VAC-EX_FAN2_570_1 & 2 both seem to have gotten significantly more noisy for about 27 hours from 1800 UTC on Sep 1st to 2100 UTC on Sep 2nd

Sheila and I repeated her 86683 yesterday's FC injections

• 10,000 Gain: refs 0-3 (compare to yesterday's refs 23-25)
• 3,000 Gain: refs 4-7 (compare to yesterday's refs 20-22
• Quiet time: refs 8-11 (compare to yesterday's refs 14-16)

Interestingly the shape of the injection has changed,  meaning the FC loop shape is changing, see quite time today vs yesterday on FC2_M3 plot. The coupling into DARM is also significantly less today. This isn't;t a surprise as we see the FC causing noise into DARM only intermittently 86608, but we don't know the reason why yet. Sheila checked the FC OLG and it looked fine.

Addressed TCS Chillers (Wed [Sept3] 957-1040am local) & CLOSED FAMIS #27823:

Had a little bit of a rigmarole since I hadn't done this for a while.  In my quick procedure scan, I didn't see how to read the "floating red ball" in the chillers.  I had remembered using the bottom of the ball, but wasn't sure, so I consulted Camilla & TJ.  I ended up using the bottom for my initial measurements, but later learned we measure the top of the ball!  Either way, they were a little low, but I overfilled them since I used the bottom of the ball for readings.  Also not clear what exactly is the "Max level" (reason for overfill).  Either way, chillers had received water since June, so adding water an ok thing!  :)  Attached is a photo of what I'm talking about, and the red ball is about 4mm in diameter so I updated readings I initially entered when filling by adding 4mm to all water level readings.

• TCSx Chiller:  Had level of 29.5 cm.  Added 550mL for 30.9cm.
• TCSy Chiller:  Had level of 9.8 cm.  Added 425mL for 10.7cm.  
• There was no water in the small cup under the slow leak.

Wed Sep 03 10:15:46 2025 INFO: Fill completed in 15min 42secs

Gerardo confirmed a good fill curbside.

With the Monday holiday and no calibration run last Sat, ran a calibration this morning to start of a Wed Commissioning session.

Measurement NOTES:

• Following instructions (link)
• H1 at NLN for for ~18hrs (had recent SQZ issue however)
• Calibration_Monitor medm screenshot is attached
• 1500 utc Out of Observing for CALIBRATION
•  Broadband measurement started at 1500 utc ended 1505utc
• Measurement at:  /ligo/groups/cal/H1/measurements/PCALY2DARM_BB/PCALY2DARM_BB_20250903T150035Z.xml
•  Simullines measurement
  • gps START time: 1440947220 (1506utc)
  • gps STOP time: 1440948616 (1529utc)
  •  File written out to: /ligo/groups/cal/H1/measurements/DARMOLG_SS/DARMOLG_SS_20250903T150641Z.hdf5
  • File written out to: /ligo/groups/cal/H1/measurements/PCALY2DARM_SS/PCALY2DARM_SS_20250903T150641Z.hdf5
  • File written out to: /ligo/groups/cal/H1/measurements/SUSETMX_L1_SS/SUSETMX_L1_SS_20250903T150641Z.hdf5
  • File written out to: /ligo/groups/cal/H1/measurements/SUSETMX_L2_SS/SUSETMX_L2_SS_20250903T150641Z.hdf5
  • File written out to: /ligo/groups/cal/H1/measurements/SUSETMX_L3_SS/SUSETMX_L3_SS_20250903T150641Z.hdf5
•   1530utc ISC_LOCK taken back to NLN & started commissioning

TITLE: 09/03 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 149Mpc
OUTGOING OPERATOR: Tony
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 3mph Gusts, 1mph 3min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.11 μm/s
QUICK SUMMARY:

H1 had issues last night with the SQZ (as noted by Tony in his OWL alogs), but h1 has been locked/thermalized for just over 17hrs37min.  Bring up thermalization, because the plan is for a calibration suite  of measurements to be run at 8am.  Also looks like there will be commissioning right after this measurement TODAY (830-1130amPDT).

On nuc23, the DARM .5-2kHz bands have been recovering from the SQZ issue starting 2hrs ago and look like they are at normal levels for the last hr or so (see attached).

@ 2:43 am (Local time)  H1 called for assistance.
I noticed that it was a SQZ issue, specifically the SHG "PZT was out of range" error on the SQZ_SHG Guardian.
It was bouncing from Locking to Locked then scanning. etc.
I tried Init-ing SQZ_SHG manager..... SQZ_Manager.
Then I took all SQZ GRD Nodes to down  except SQZ_PMC & SQZ_SHG and tried to troubleshoot the SQZ_SHG directly.
Looking at the SQZ troubleshooting guide: There wasn't a section for SQZ _SHG troubleshooting. So I searched the ALOG with no hits for "PZT Out of Range".
I tried to adjust the OPO temp to maximize H1:SQZ-CLF_REFL_RF6_ABS_OUTPUT. But changing the OPO temp had no impact.
This is when I started to realize it may take me awhile to find out what is going on, since it's too late to call a SQZ expert, I should just take us to NO-SQZ.

So I ran the noconda python switch_nom_sqz_states.py without script.

Sadly I did accept the SDFs and these too, because I thought they were causing an SDF issue. But it may have been SQZ_ANG_ADJUST needed to be at ADJUST_SQZ_AND_ADF  instead of DOWN for the SQZr to be considered ready for OBSERVING. 
Made it back to OBSERVING at 11:30:44 UTC
Now that we have made it back to OBSERVING albeit without SQZing I'm using a finer toofed comb on my troubleshooting.

I went looking for Sitemap>SQZ>SQZT0>SHG That last SHG Button was a bit difficult to find when your eyes are still crossed..

I then took a screenshot of the SHG screen before any changes.
I then locked the SQZ_PMC and the SQZ_SHG. The SQZ_SHG was cycling between locked and Unlocked. Dropped from Observing at 11:57:53 UTC.
Then while trying to maximize H1:SQZ-SHG_GR_DC_POWERMON I changed H1:SQZ-SHG_TEC_SETTEMP from 35.89 to 35.61.
After this I  ran the noconda python switch_nom_sqz_states.py with script to try and get us the SQZ SQuoZe.
SQZ_MAN was having an issue with the SQZ_FC losing lock at Transition_IR_LOCKIN, I then re-touched up the OPO temp.

Success! The SQZr Is SQUOZE!!
Now I need to accept all the SDF's again (and these too ) cause I didn't follow the directions when i ran the initial no sqzing script.
Observing reached again at 12:51:32 UTC.



 

TITLE: 09/03 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Observing at 152Mpc
INCOMING OPERATOR: Tony
SHIFT SUMMARY: Very quiet shift with H1 locked and observing throughout; lock stretch is currently up to 8 hours. A M5.3 EQ out of Japan rolled through about an hour ago but didn't even put us into EQ mode.

Bi-weekly Histograms of Locking statew times.

Histogram of time spent at NLN or higher:
Length of data 22
max Duration 2613 Min
Average 770.5 Min
% below 5 minutes : 0.0
Between 2025-08-20 00:27:34  and  2025-09-03 00:27:34
 

PowerUp Histogram
Length of data 30
max Duration 39 Min
Average 24.966666666666665 Min
% above 30 minutes 73.33333333333333
Between 2025-08-20 00:45:04  and  2025-09-03 00:45:04
 

DRMI-Histogram
Length of data 51
max Duration 49 Min
Average 11.647058823529411 Min
% above 5 minutes 54.90196078431373
Between 2025-08-20 00:46:03  and  2025-09-03 00:46:03
 

ALS Histogram
Length of data 53
max Duration 30 Min
Average 5.471698113207547 Min
% above 5 minutes 32.075471698113205
Between 2025-08-20 00:46:40  and  2025-09-03 00:46:40

TITLE: 09/02 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 149Mpc
INCOMING OPERATOR: Ryan S
SHIFT SUMMARY: A fair amount of work done even though it was looking to be a slower maintenance day. We managed to stay locked for a few hours into it, while staying in light maintenance mode. We relocked automatically with an initial alignment, then lost lock 39 seconds later from an ETMX glitch. On the way up the second time, Elenna opportunistically tuned up PRMI ASC since it has shown some signs of non-convergence (alog86694). We have now been locked for 2.5 hours while a small dropout when the SQZr unlocked.
LOG:

• 1627 Lock loss from maintenance activities
• 1734 ISI BS WD tripped
• 1846 GRB short - Fermi E596650
• 1941 Back to NLN
• 1942 Lock loss
• 2109 Observing
• 2204-2207 Briefly out of observing while the squeezer unlocked and relocked

TITLE: 09/02 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Observing at 150Mpc
OUTGOING OPERATOR: TJ
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 10mph Gusts, 5mph 3min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.11 μm/s
QUICK SUMMARY: H1 has been locked for 2 hours.

• Dust monitors for the Optics Lab and Diode Room both report counts have not changed
• Wind and ground motion low
• All other systems nominal

FAMIS 31101

The fire alarm yesterday briefly shut off air handlers, so we see some temperature changes in several places that appear to have all settled back to nominal. Also, Jason's PMC, RefCav, and ISS adjustments this morning are seen clearly. No other major events of note in the past week.

FAMIS 26684

pH of PSL chiller water was measured to be between 10.0 and 10.5 according to the color of the test strip.

TonyS and RickS

Today, we measured the transmissivity of one of the SPI HR mirrors and the AR surface reflectivity of one of the SPI beamsplitters, as requested by Jeff Kissel.  The reflectivities of the beamsplitters were measured earlier, see this aLog entry.

HR mirror reflectivity:  We used a set of Gentec power meters procured for the Pcal effort.  An SPI HR mirror (see attached photo for details).  Aligned at 45 deg. AOI with p-polarized light, the power transmissivity (P_trans/P_inc) was measured to be 23 e-6 W/ 567 e-3 W or 4.1e-5 (0.004%). 

BS AR surface reflectivity: The reflectivity of the AR surface of a 50/50 beamsplitter was measured in two ways, both after mounting the BS with the AR surface closest to the input beam.

1. Using the PS5 Pcal power sensor and blocking the beam reflected from the 50 % reflectivity surface near the beamsplitter, the output voltage (after zeroing the no-laser-light level) was - 0.387 mV.  With both the main reflected beam (~ 50 % of the incident light) and the beam reflected from the AR surface, the voltage was - 1.3699 V.  Doubling the voltage for what would have been expected with the full incident power, the estimated AR reflectivity is  0.387 e-3 / (2 x 1.3699) = 1.4 e-4 or  0.014 %.

2. Using power meters to measure the power in the beam reflected from the AR surface and the power incident on the BS, we measure 52.6 e-6 W / 564 mW = 9 e-5 or about 0.01 %.

Test for Thursday morning at 7.45 am, assuming we are thermalised.

• In NLN.
• Turn off OMC ASC:
  • sitemap -> OMC -> OMC Control -> turn MASTER GAIN slider to 0 (its blue and on the bottom centre of screen).
• Run DARM offset step:
  • Go to /ligo/gitcommon/labtutils/darm_offset_step
  • run:

conda activate labutils

python auto_darm_offset_step.py

• Wait until program has finished ~20 mins.
• Turn OMC ASC back on by putting master gain slider back to 0.020.
• Heat up OM2 heater (~ 3 hours):
  • IFO out -> OM2 -> change H1:AWC-OM2_TSAMS_POWER_SET value to 4.6.
• After power has leveled out on H1:AWC-OM2_TSAMS_THERMISTOR_1_TEMPERATURE and H1:AWC-OM2_TSAMS_THERMISTOR_2_TEMPERATURE, turn off OMC ASC again by putting MASTER GAIN slider back to 0 on OMC Control screen.
• Run auto_darm_offset_step.py again.
• Put OM2 heater H1:AWC-OM2_TSAMS_POWER_SET back to 0.
• Test done, wait for thermistors to level out again.

Since the HAM1 vent, I have done a few different measurements of the ASC that provide information on how to calibrate WFS signals from counts to microradians. Here is a summary:

CHARD, INP1 and PRC2 results come from this alog

DHARD results come from this alog

SRM results come from this alog (if you are comparing values, I made a power normalization error in the linked alog)

BS results were taken but never alogged (shame on me)

All of these measurements were taken by notching all ASC loops at 8.125 Hz and injecting an 8.125 Hz line in the desired DoF. The osem wits provide the urad reference.

Unless otherwise specified, the witness channels are the bottom stage osems