TITLE: 02/03 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY:
LVEA was Bifircated via a Laser Hazard areas around HAM 1& 2, while the rest of the LVEA is LASER SAFE unless at height. See M2600004 for details.

PM1 was mysteriously dancing & Saturating too much for Rahul's transfer function measurement all day long. While he was troubleshooting. Update!!! Rahul dragged the EE guys out and found a dead Coil Driver! Mystery solved!

JAC work continued all day starting with mode matching in HAM1, and starting a JM2 swap. JAC table now has all the mounts and Cables but no optics yet. 

HAM7 & ITMY ISI Watchdogs tripped at 23:33 UTC due to mysterious high frequency ground noise that was louder at HAM7 than ITMY.  Jim & Betsy went to go look for fallen wrenches or other watchdog tripping phenomena..... No explanation was ever found. 

OPS info: 
New conda ENV doesn't run watchdog untripping scripts unless you run Conda Kill first.  Tagging CDS. 

LOG:

Dave, Oli, Marc, Fil, Rahul

This morning I switched PM1 (Tip Tilt suspension) in HAM1 chamber from safe state to damped state and immediately the DAC output was saturating. I suspected purge air (which Travis turned it down twice during the day) and later Masayuki also covered PM1 with foil. However, nothing stopped the suspension from saturating.

Then Oli, Dave and I did a model restart and later I power cycled the Satamps for PM1 in the LVEA - nothing helped.

Next, Marc, Fil and I went to the LVEA (checked the satsamp, which was fine) and CER and found the Coil Drivers to be faulty.

New Coil driver - E2400048 s/n 22001180

Old Coil Driver - E2100430 s/n S1106046

SUS PM1 is now not saturating. I will perform the heath checks later on.

Just noting that this kind of thing can happen when platforms are in different states. Something caused large-ish high frequency ground motion in a short burst this afternoon and tripped HAM7 and ITMY, because those chambers were damped, but not any of the other chambers because they had high gain isolation loops running, or were already tripped/locked. What ever it happened was louder at HAM7, than the ITMs, as shown in the top left of the attached plot. Because the motion was high frequency, where the isolation loops had enough gain to keep the ISI seismometers from saturating, only HAM7 and ITMY tripped because their isolation loops were off, and damping only engaged. And the event was small enough to not saturate the actuators.

Oli, Rahul, Tony, Marc, Fil, Daniel, Dave.

At 15:41 we restarted the h1susham1 model as part of the PM1 noise investigation. This did not change anything.

Later (16:40) Marc and Fil found that the problem was a dead power rail in the PM1 coil driver.

Used 1W into JAC.

9 (26.8dBm into EOM), 45 (27dBm) and 118MHz (10.76dBm) -> Used a single bounce beam from ITMX and scanned OMC. Successful for 9 and 45, nothing visible for 118. Detailed will be posted later.

24MHz (IMC, 14.2dBm) -> Scanned IMC length and tried to find something in IMC transmission. Nothing visible.

43MHz (JAC, 12.2dBm) -> Scanned JAC, saw nothing in JAC transmission. Boosted RF  power with an amp to 29.67dBm -> m=0.092. Without the amp modulation index would be 0.012.

Jennie W

This morning I re-centred the beam on MC TRANS in-air camera on IOT2. This was using the IO_MCT_M5 mirror, see layout here.

I did this while the JAC was locked and the IMC was flashing through modes.

I only had to move the steering mirror in yaw to centre the mode flashes on the camera.

As noted on Friday, RLF QPD A segment 3 railed Nov 29th, and has been railed since then.  

Filiberto, Kar Meng, Marc and I went to the rack and swapped the two cables for WFS1 + WFS2 on D2000552.  The saturaed segment moved with the cable which could indicate the problem is the in vacuum QPD.  

The 105kHz segment went quiet at the same time as the problem on the DC segment.  This happened on a Saturday, before the chamber was vented.  It had been two days of not locking the squeezer or the IFO before this happened.  

Edited to add:  This is the QPD used for filter cavity length control.  We did lock the filter cavity after the segment broke, December 4th.  

The LVEA is now in a LASER Safe Bifricated configuration, which means that HAM1 is still in LASER HAZARD along with any work done at height.  There are barricades up designating the Bifricated area.

WP: https://services2.ligo-la.caltech.edu/LHO/workpermits/view.php?permit_id=13010
SOP: https://dcc.ligo.org/LIGO-M2600004

The rest of the LVEA is Laser Safe.  So if you are heading into the LVEA, have your laser goggles with you so if you are forced to walk over near HAM1 you'll have your goggle with you. 

Mon Feb 02 10:11:57 2026 INFO: Fill completed in 11min 53secs

PSL Status Report - Weekly Famis 39749

Laser Status:
    NPRO output power is 1.845W
    AMP1 output power is 70.43W
    AMP2 output power is 138.7W
    NPRO watchdog is GREEN
    AMP1 watchdog is GREEN
    AMP2 watchdog is GREEN
    PDWD watchdog is GREEN

PMC:
    It has been locked 4 days, 15 hr 14 minutes
    Reflected power = 27.01W
    Transmitted power = 103.8W
    PowerSum = 130.8W

FSS:
    It has been locked for 2 days 15 hr and 50 min
    TPD[V] = 0.4976V

ISS:
    The diffracted power is around 4.0%
    Last saturation event was 0 days 0 hours and 0 minutes ago

Possible Issues:
    PMC reflected power is high

TITLE: 02/02 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
OUTGOING OPERATOR: None
CURRENT ENVIRONMENT:
    SEI_ENV state: MAINTENANCE
    Wind: 8mph Gusts, 5mph 3min avg
    Primary useism: 0.05 μm/s
    Secondary useism: 0.53 μm/s 
QUICK SUMMARY:
Dust alarms for the VAC Prep lab were going off when I came into the control room. Likely due to the elevated winds we had yesterday.
Untripped ITMY ISI  Stage  2 WD. 

More HAM1 & 7 work is expected today Doors may start going on later this week. 
LVEA is bifurcated with LASER hazard work being done near HAM1. Please bring your LASER Saftey Goggles with you into the LVEA.






 

Summary

Based on the post-EOM IMC scan results showing significant mode mismatch, we performed a detailed mode-matching analysis using the as-built HAM1 layout. Optics positions were identified from chamber photos, beam traces were reconstructed, and the expected IMC mode-matching and higher-order mode fractions were calculated. The analysis shows that the dominant source of the mode-mismatch degradation is a longitudinal shift of JM2 introduced by the EOM installation, and that moving JM2 can effectively recover good mode matching.

Procedure

• The positions of the HAM1 optics were identified from chamber photos.
  • The identified optics locations are marked with purple circles in the attached layout figure.
  • Since the JM2 position differs before and after the EOM installation, two circles are shown for JM2 (pre- and post-EOM positions).

• Using these as-built positions, we reconstructed the beam trace and calculated the mode matching into the IMC and the expected 2nd-order mode fraction.

  • This prediction is shown in the first plot.

  • The lensing effect of the EOM was included using the test results from the actually installed crystal (see p12 of slides).

• The measured 2nd-order mode fractions obtained from the IMC scan (reported in the previous post) are overlaid on the same plot for direct comparison.

Notes

• From the reconstructed layout, JM2 is shifted by approximately 2 inches between the pre- and post-EOM configurations.
  • Due to the reflection geometry, this corresponds to roughly 4 inches of effective beam path length change.
  • This shift is identified as the primary cause of the observed degradation in mode matching after the EOM installation.
  • The calculated 2nd-order mode fraction agrees well with the independently measured values from the IMC scan, validating the beam-trace model.
• The model indicates that shortening the total path length by approximately 6.3 inches optimizes the mode matching for 100 W input power as shown in the second plot.
  • This corresponds to moving JM2 by about 3 inches, which is feasible within the current layout.
  • Around the optimum, a JM2 placement error of ±1 inch degrades the mode matching by only ~1%, suggesting that achieving <1% mismatch is realistic from an installation-accuracy standpoint.
• The third plot shows that, when JM2 is placed at the optimal position, the expected change in mode matching over 0–100 W input power variation is at the ~0.1% level, indicating low sensitivity to power-dependent effects.

Result

• The dominant contributor to the post-EOM mode mismatch is the longitudinal shift of JM2 associated with the EOM installation.
• Moving JM2 is expected to be highly effective in recovering good mode matching to the IMC.
• If JM3 is replaced with a tip-tilt mount in the future, keeping its longitudinal position within ±0.5 inch would be desirable to avoid introducing additional mode mismatch.

Proposal (Practical Implementation)

1. Install one iris in the reflected beam immediately after JM3, and place another iris on the IOT2L table.
2. Change the longitudinal position of JM2.
   • The target position will be documented in a follow-up figure.
   • After moving JM2 and roughly passing the beam through the irises, fix the mount using dog clamps.
   • Replace the JM2 mirror at this stage.
3. Perform alignment using the iris just after JM3 and the iris after the periscope.
   • Use JM2 to center the beam on the first iris and JM3 to center on the output iris.
   • This procedure is expected to provide sufficient alignment accuracy; if needed, the IOT2 table can be used for further refinement.
4. Repeat the IMC scan measurement to confirm the improvement in mode matching.

LIGO git repo: https://git.ligo.org/masayuki.nakano/lho_jac_installation/-/tree/2bd2518c1e08635017066b16d4b7c5f752ed0ea5/

Summary

We performed IMC scans to evaluate mis-polarization and spatial mode content associated with the JAC installation. First, reference measurements were taken before the EOM installation using intentional MC2 yaw/pitch misalignment to robustly identify higher-order modes. After the EOM installation, the same type of IMC scan was repeated. The post-EOM measurement shows that the mis-polarization signature observed before the EOM is strongly reduced, while a significantly larger higher-order mode content is present, indicating substantial mode mismatch in the current post-installation layout.

Procedure

• The HAM1 output beam was aligned to the IMC as well as possible using HAM1 mirrors.
• The FSS autolocker was stopped, and a laser frequency scan was applied using the laser PZT:
  • Triangle ramp: 0.1 Hz, ±8 V.
• IMC alignment was slightly optimized using MC2, but not perfectly:
  • Some TEM10/01 content was intentionally kept to aid robust peak identification.
• Data acquisition:
  • Channels recorded:
    • JAC-PZT_DRV_OUT_DQ (frequency proxy)
    • IMC-MC2_TRANS_NSUM_OUT_DQ (IMC transmission)
  • Each scan was recorded for 200 s.
• Pre-EOM reference scans were taken with three alignment conditions:
  • Nominal alignment (reference) GPS: 1453668961
  • MC2 pitch returned to the initial (bad-alignment) value: GPS: 1453672393
  • Pitch restored, MC2 yaw returned to the initial value: GPS: 1453677200

After the EOM installation, a representative post-EOM scan was taken with the same scan settings. For this measurement, the IMC transmission whitening gain was set to 30 dB. Unfortunately, the GPS start time for the post-EOM dataset was not logged at the time of measurement.

Notes

• The IMC is free-swinging during the scan; the JAC PZT signal is used as a proxy for the instantaneous laser frequency.
• Intentional yaw/pitch misalignment in the pre-EOM scans enhances TEM10/TEM01 as expected, confirming reliable mode identification and peak labeling.
• The IMC scan after the EOM installation exhibits noticeably stronger higher-order mode resonances than the pre-EOM reference scans.
• Calculation details can be found in the LIGO Git repo.

Result

• In the pre-EOM scans, the p-polarization resonance (π-shifted from s-pol) is visible at the \sim 0.1\% level, indicating a small but measurable mis-polarization component as shown in the second attached plot. The wider peak at the center is the p-pol resonance. The MC mirrors has higher transmissivity for p-pol the finesse for p-pol is lower than s-pol peaks.
• After the EOM installation, this p-polarization signature is no longer apparent, indicating that the mis-polarization is strongly reduced. It might be due to the filtering effect of the EOM. The mispolarized component caused by the input periscope is filtered by the EOM birefringence.
• In contrast, the post-EOM scan shows a substantially larger higher-order mode content (notably a large 2nd-order component ~10%), indicating significant mode mismatch in the current post-installation layout.
• A dedicated mode-matching calculation based on the as-built post-EOM layout will be reported separately in the near future.

Sun Feb 01 10:08:16 2026 INFO: Fill completed in 8min 12secs

Sat Jan 31 10:06:56 2026 INFO: Fill completed in 6min 52secs

Jennie W, Jason O, Masayuki N, Keita K, Jim W,

Summary: To check the function of the new EOM in chamber we made a measurement of the modulation indexes by locking the IMC and aligning the beam to AS_C with SR2. We couldn't get a good measurement of the sideband heights but this is probably due to the RF power being down by a factor of 100 from nominal. Will check with EE/Daniel on Monday.

First order of business was checking for stray beams at 100mW input power. Jason moved the BD we already placed for JM2 as we had moced this mirror position yesterday. We also put a new beam dump right after unused JAC port (output side there is tranismission through the curced mirror).

After this we turned the power up to 1W.

Lastly we found a stray beam exiting the table in the -Y direction, this was traced to the JAC REFL path. The REFL beam was hitting the side of a beam dump (near the -X side/PSL of the table) which is meant to cath a beam reflecting off the SEPTUM plate. This beam reflected off the beam dump causing a stray beam. We re-aligned the REFL path so the beam does not do this and instead bounces off the three REFL path steering mirrors and heads into a previously placed beam dump for this purpose on the -Y side of the table. This path will have to be re-aligned in order for the beam to get onto the IOT1 table.

No further stray beams were found so Jason de-energised the waveplate.

Photos to come.

After realising that the beam did not reach the AS_C,A and B QPDs yet we came to the control room to re-align to the output port. This is with the ITMY, PRM and SRM mirrors mis-aligned to allow us to mode scan the OMC in the 'single bounce' IFO configuration.

After Jim re-isolated HAM4,5 and 6 and BSC2 for us we were able to use SR2 to bring back the alignment to AS_C and then turn on the DC centering loops for AS_A and AS_B.

The OMC ASC did not work at first as the suspensions were railed. I cleared the ASC history, this did not help. We cleared the locking filter banks for OM1-3 and this unrailed the outputs and allowed us to turn on the OMC ASC.

We took an OMC scan at 1W input power, shown here. Roughly calibrated into MHz with the known FSR.

We cannot identify the 45MHz or 9MHz peaks, but after checking we realied that these RF driver powers were lowered 15 days ago. See image.

We will come back to this Monday.

Jason put back the rotation stage, locked it out and closed the light pipe.

In the control room, the alarm handler was complaining about opticsd lab dust monitor.

It seems that it surged up and stayed there for about half an hour. I acknowledged the alarm (the dust conts are already back to normal, it seems).

Not sure what happened but this seems to happen kind of regularly.

J. Wright, J. Oberling

As part of WP 13006, we added another beam dump and moved one to ensure all stray beams that we could find were properly dumped.

We added the required beam dump on the unused output port of the JAC, to dump the unused beam here.  We also moved the beam dump that catches the leakage beam through JM2 so that it continues to dump this leakage beam in JM2's new location (needed to move JM2 as a result of EOM installation).  See the 2 attached pictures.

While scanning for other stray beams, I noticed a beam on the outside of the large beam dump that was temporarily relocated to the +X side of the new JAC input periscope.  This was traced to the JAC REFL path.  Turns out that when the REFL path was installed, the beam was erroneously reflected off of this beam dump on its way from JACR-M1 to JACR-M2.  To fix this we had to loosen the dog clamps for JACR-M1 and physically rotate it until the beam was centered on JACR-M2 (there wasn't enough range in the yaw actuator).  We then had to loosen and physically rotate JACR-M2 so that the REFL beam was once again captured in its temporary beam dump (dump located on the -Y side of the ISI).  Both mirrors were re-secured to the table.  Unfortunately, this means that we will have to revisit REFL path alignment to the new IOT1 table.

(Randy Thompson, Corey Gray)

In preparation for the upcoming Bigger BeamSplitter Suspension (BBSS) Installation at BSC2 in a few weeks (~midMarch), a new type of cleanroom infrastructure was needed too allow for more vertical space for the cartridge to be craned in/out of BSC2.  Instead of using our standard large mobile cleanroom (which are used for HAM & BSC chamber work), a smaller clean space will be attached to the elevated work platform for BSC2.  There will be a "cleanroom tent" on this platform (here's the platform [white] with 2-sets of 4 "tent poles") and a new array of (qty-4) HEPA Filters Fans (same type as the ones on top of the large mobile cleanrooms which blow down) will be on the side of this cleanroom tent to push air flow horizontally for this elevated workspace for the BBSS install.

Randy designed a unistrut framework for this 4-HEPA fan Array (weighs 377lbs).  This morning this assembly was carefully craned into place on top of 3 unistrut brackets attached to the BSC work platform (in between BSC1 & BSC2).  For added support, this afternoon, Randy added a "leg support" on both ends of this HEPA Filter Fan Array (here is a photo of one of these legs that is in the Beer Garden).