Fri Mar 13 10:15:08 2026 INFO: Fill completed in 15min 4secs

Tony, Filiberto, Sheila, Jennie Wright, Oli

This morning we tried to do an OMC scan to check the modulation depths.  The OMC high voltage was off, Filiberto found that the supply was powered on but the output disabled, so he enabled it.  

We used 10W input power single bounce off ITMX.  I compared this to a reference from 65828, it seems as thought the 9MHz is too small. 

Jennie Wright checked modulation depths in 89456, at that time the EOM driver was set to 27dBm for 9MHz, on 10 am on March 10th it was set to 23.4dBm.  This was in SDF as 23.4dBm, but this channel is not monitored and this table is not in SDF revert. ISC_LOCK prep for locking sets these values because they get changed in low noise, I've changed it in lsc params now.  

TITLE: 03/13 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    SEI_ENV state: EARTHQUAKE
    Wind: 5mph Gusts, 3mph 3min avg
    Primary useism: 1.02 μm/s
    Secondary useism: 0.21 μm/s 
QUICK SUMMARY:

Site is wet, wind is low.
When I walked in there was an active 6.3M earthquake rolling through from Chile.
I'll give the earthquake about 10 more minutes to roll on through then start locking green arms and PRX for the comissioners.
 

TITLE: 03/12 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: Corey
SHIFT SUMMARY:

Jenne/Jennies locked pseudo-PRMI earlier in the shift.  Ran trend plots of sus for comparison between today and Nov2025 (see Jenne's alog).
LOG:

• 2346 Grabbing laptop (sophie)
• 0002 Jenne had PRMI locked for about a minute!
• 0318 Closed the HAM6 shutter due to OMC Trans increasing on nuc31

It would be helpful to have a comparison of the top mass osem readbacks for optics, comparing this time in the attachment from Nov 15 2025, to tonight.  In both of these times PRMI is locked with arms held off resonance with ALS (ISC_LOCK state 50). 

Past time for comparison: 15 Nov 2025 00:26:00 UTC, plus or minus 1 minute (but check that ISC_LOCK_STATE_N was 50, and we want closer to the end time of that state 50 time). 

Now time to compare: 2026-03-13 00:40:16.264360 UTC, plus or minus 10 mins. 

Optics to trend:

• IMs 1-4
• RMs 1-2
• PM1
• MC1-3
• PRM, PR2, PR3
• BS
• ITMs, ETMs, TMSs
• others optional, like SR2, SR3, OMs1-3. 

[Sheila, Jennie, Jenne, RyanS, Corey]

We have been able to get PRMI locked (with arms held off resonance using ALS).  To do this we:

• Increased POP18 analog whitening gain, and redid its dark offsets
• Phased POP90 by -180 deg.  RyanS noticed that POP90 was flashing negative, when it should have been positive.  We realized that POP18 was fine because the RF9s didn't need rephasing with the new EOM, but the RF45s needed to be phased by 90 degrees, so the RF90 would need a 180 deg phase.
• Increased REFLAIR_A_RF9 analog whitening gain. No dark offset change needed
• Increased REFLAIR_A_RF45 analog whitening gain and adjusted dark offsets.
• Took the ratio of REFLAIR RF peak heights now vs a time with the same IFO config on Nov 16th 2025, and increased the MICH and PRCL LSC input matrix values by those ratios.
• Jennie calculated what the updated MICH LSC trigger thresholds should be, but with the whitening gain up on POP18 we ended up being okay with the default from guardian.

PRMI locked without too much fuss.  Since we've effectively also increased the noise on the error signals, it's a bit delicate, and I lost lock twice while touching up alignment. The third lock I was able to go a bit slower when touching up BS and PRM alignment, and we've now been locked for more than 20 mins (aka infinity).  Jennie is taking PRCL and MICH OLGs.

The attached image shows all the things that I had changed. 

• POP18 analog gain to 45dB
• POP18 dark offsets
• REFLAIR_RF9 analog gain to 36 dB
• REFLAIR_RF45 analog gain to 30 dB
• REFLAIR_45_Q to MICH input matrix element to 22.7 (5.6x times default of 4.0)
• REFLAIR_9_I to PRCL input matrix element to 18.6 (10.6x times default of 1.75)

Nothing is in guardian (except maybe some trigger thresholds which should be reverted), and nothing is saved in SDF. 

TITLE: 03/12 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: Corey
SHIFT SUMMARY: More alignment work today to recover H1; see other alogs for more detailed accounts of the day, some of which will be posted after this one). This afternoon we are attempting to lock PRMI using the main locking steps in ISC_LOCK, which will give us a better sense of our overall IFO alignment. And it happened for just over a minute!
LOG:

TITLE: 03/12 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 8mph Gusts, 4mph 3min avg
    Primary useism: 0.03 μm/s
    Secondary useism: 0.25 μm/s 
QUICK SUMMARY:

Crew continues recovery work as I walked in:  they were working on a pseudo-PRMI and then moved on to CHECK MICH FRINGES.  And allof this was after work out on the floor at ISCT1 (see 89485).  Much calmer breezes.

Operational Note: 

What we can do alignment-wise with ALIGN IFO:  Green Arms, PRX_LOCKED (for PRM touch ups).  But basically have  issues for any IR cavities alignment-wise that need WFS at this point.  (Some MICH & SRC states also work, but we don't need them at this point)

[Tom R, Ibrahim]

We glued 6x of the QOSEM flags, attaching the lens (D2500277) to the aluminium body (D2500253). This was done using a jig (D2600054) to ensure the lenses were aligned to the body of the flag. We used EP30.

Only a tiny drop of glue is required, and was applied directly to the lip of the body using a thin wire. The flags will be left to dry overnight before another batch of 6 will be glued.

Oli, Sheila, Jennie Wright, Ryan Short, Jenne Driggers

Today we aligned the ISCT1 refl and pop paths.  First, we remembered that both refl and pop are now on the same top periscope mirror: 84558.  Then we realized that the RF signal from PRMI flashes on POPAIR A was unrelated to light on the diode, it didn't go away when we blocked the diode, closed the beam diverter or shutter, or increase when we did get beam on the diode.  

We had a bit of difficulty aligning onto the pop diodes because PRMI flashes are infrequent.  With 2W PRMI flashes right now, we have about 1/10th of the POPAIR B LF light we expect, and roughly similar low RF18 flashes.  

[Commissioners]

As yet another thing to try alignment-wise, I moved the IMC's uncontrolled alignment degree of freedom.  

In the attached plot, the y-axis references are set from our Nov 16th, 2025 15:32:00 UTC reference time.  During this time today PRX was locked, so the absolute values on the ASC POP QPDs aren't really comparable to the reference time, but since there was no PRX ASC, there should still be changes on the POP QPDs. 

I moved MC1 (top left traces in attachment) temporarily to the top mass osem position from Nov 16th while IMC ASC DoFs 1,2,3 were all closed (with a little higher than normal gain) and indeed see that MC3 responds in the expected way for the uncontrolled DoF.  For a reminder on the uncontrolled degree of freedom, I'm looking at Jeff's G1301131 page 23. I expect the sign of the yaw on MC3 to be the same as the sign on MC1 (and it is), and I expect the sign of the pitch motion on MC3 to be opposite of that of MC1 (and it is).  So, I think that indeed I've moved the uncontrolled DoF (called DOF4 in that slide deck).  

When the uncontrolled degree of freedom is moved, I'd expect the pointing of the beam coming out of the IMC to change (I have not calculated by how much).  In the attachment, I see that IM4 trans indeed looks like it has moved by about 0.04 normalized counts.  I don't see much change on POP A or POP B QPDs - if there's any change, it's extremely small.  

Anyhow, I'm not sure there's anything deep here, but perhaps we could go back to this MC1 position from Nov 2025 and then move IM3 to get back to the reference position on IM4 trans and then move IM4 to get back to the reference position on POP QPDs, and see if that changes anything.  We've moved more than this on the IM4 and POP QPDs already over the last few days, so I'm not sure that doing so will magically fix things.

Thu Mar 12 10:14:28 2026 INFO: Fill completed in 14min 24secs

We had a couple of power glitches last night, reported by the GC UPS briefly going onto battery power (no reports from CDS UPS)

Both were very brief, 4 or 5 cycles in duration, impacting all 3 phases.

18:16:15 Wed11Mar2026 PDT

21:26:26 Wed11Mar2026 PDT

No sign of any impact on CDS, also HEPI pump controllers look good.

S. Muusse

Calibrating sensors

I have been working on calibrating the sensors on the cheta tables. When calibrating the power at the pick off  that will be used to monitor CHETA power on ITM using beckhoff I was seeing voltage jumps up to saturation at 2.5V as shown in attached plot. This was determined to be due to the auto range on the thermal power head being on. This setting was turned off ( as outlined in T2600081 ) and a range of 0-22mV was selected as max incident power was ≈ 11mV.  After which a linear relationship was observed and shown in the attached plot. The laser power vs laser current and the measured V vs laser current have been takem but I havent calculated the responsivity value for the beckhoff screen yet. 

Profiling CHETA beam 

The beam profiler has been shipped back from LLO to check the alignment against the model and profile the beam size as the second lens is moved by the translation stage. This is how we aim to control the CHETA beam size.
On the Y-table the beamsize was measured over a 60mm distance and the profile was fit with a non linear fit for w0 and z0 setting M2 =1. This is shown in the attached plot.

--- Horizontal (A1) Fit ---

Fit converged: w₀=896.3+/-37.1 µm, z₀=-528.0+/-47.8 mm

Model: w₀=1020.9 µm, z₀=-704 mm

--- Vertical (A2) Fit ---

Fit converged: w₀=977.3+/-40.1 µm, z₀=-449.0+/-40.1 mm

Model: w₀=999.36 µm, z₀=-459 mm

The vertical fit agrees for both w0 and z0 within the fit uncertainties and the horizontal profile was just outside the uncertainty bounds. 
Subsequently, the beamsize has been measured along the same positions with L2 at its minimum and maximum displacements on the translaiton stage (±50mm) but this data hasnt been processed yet at will be added in a comment. 

TITLE: 03/11 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: Ryan S
CURRENT ENVIRONMENT:
    SEI_ENV state: CALM
    Wind: 41mph Gusts, 30mph 3min avg
    Primary useism: 0.10 μm/s
    Secondary useism: 0.31 μm/s 
QUICK SUMMARY:

Got an overall summary on the status of H1 recovery from RyanS.  

Fil, Patrick, Jonathan, Erik, Dave:

Fil connected the IOT1 JAC digital camera to the CER POE switch (sw-lvea-aux 0/4).

Erik added the new camera to DNS as h1cam38 (10.106.0.61).

Patrick added this camera as the first cam on h1digivideo6, using the multicast address 239.192.106.67

I added the camera to camera.yaml and updated the camera overview MEDM (see attached).

The camera is currently in a dark location, so no image.

Jennie W, Keita K, Rahul K, Ryan S,

This morning Keita turned the power up to 1 W and the purge air down in order to scan the OMC and measure the modulation index.

We had some work to do to lock the JAC, the cavity wouldn't scan properly when managed by the guardian code.

Ryan and I have looked back at it and think there may be something wrong with the feedback to the PZT. It can't be the PZT itself though, because Keita and I were able to lock on the TRANS PD signal with the dither locking done manually by tuning the PZT offset to find the resonance and then turning on the feedback.

I have enclosed a picture to aid problem solving next week.

We then had problems keeping the IMC locked so Keita had to change the threshold used to check if the cavity is locked down to 2. This was changed at line 133 in ISC_library.py at userapps/h1/isc/guardian/.

trans_pd_lock_threshold = 2

We then managed to do three scans,

one with 45 MHz off, one with both on, and one with 9 MHz off. Detailed analysis in progress.

After turning the power back down Keita and Rahul swapped the mirror we currently have in front of the TRANS PD, with the uncoated laser window that is meant to be there (JACT_BS1).

They had to top gun and swab it to remove two dust particles first.

This replacement means the power to the PD is lower and so we needed to power up to 1W to check this alignment.

I have left the mirror we took out on the table on the -Y side labelled as 'HR mirror taken out of JACT-BS1'. Need to check which exact mirror this is with Masayuki on Monday.

Keita locked the cavity manually with the REFL PD and optimised the amount of light falling onto the TRANS PD.

I powered down to 100mW and de-energised the rotation stage and closed the light pipe at the end of the day. We also had the rotation stage de-energised and locked out when we were working in chamber.