Since it was installed, HAM7 ISI has been prone to trips if people are working nearby. Often walking by the chamber is enough, but work in the squeeze racks, HAM6 ISC racks or unplugging electronics from nearby outlets has caused CPS saturations. After talking to Jeff, we decided to try disconnecting HAM7 from the corner timing system and try running the ISI using it's own local 10khz timing clock signal. I pulled the P3 jumper from the H1 CPS and disconnected the timing DB9 that connects HAM7 to the timing system. Position readout seems unchanged before vs. after the change. This shouldn't affect anything else.

As I entered the VEA at X-End I noticed an out place noise, I tracked the noise source to be a noisy power supply at the vacuum rack.  Made the report to Filiberto and Richard, noise points to a fan issue.

Filed a FRS ticket, number 35689.

Tue Oct 21 10:08:56 2025 INFO: Fill completed in 8min 53secs

After my change to the laser power guardian (87545 87581) caused a problem last night (87598) , I edited it so that the power scaling is set to a function.  If the IMC is locked, it will use IMC-IM4_TRANS_NSUM_OUTMON , if not it will use H1:IMC-PWR_IN_OUT16.  

I also made a plot to check if the IMC throughput is dropping as we power up, it is dropping but only by 0.4%.  

Per FAMIS, VEA AHU fans were lubricated and swapped. Where AHU 2 previously operated both fans, and AHU 1 operated just 1 this has been flipped. AHU 1 is presently running both fans while AHU 2 runs just one. Ahead of this swap, the non-running fan in AHU 1 was SF1 (nearest the exterior wall).

E. Otterman T. Guidry

TITLE: 10/21 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Microseism
OUTGOING OPERATOR: TJ
CURRENT ENVIRONMENT:
    SEI_ENV state: USEISM
    Wind: 8mph Gusts, 5mph 3min avg
    Primary useism: 0.08 μm/s
    Secondary useism: 0.71 μm/s 
QUICK SUMMARY:

Detector was put in DOWN earlier due to very high secondary ground motion. Maintenance day is today.

Workstations were updated and rebooted.  This was both an OS packages update and a conda packages update.

In the CDS conda environment, the python package dttxml was updated to fix a bug, details here:  https://git.ligo.org/cds/software/dttxml/-/issues/17

Ryan S, Keita K, TJ S

After the lock loss at 0427 UTC (2027PT), Ryan S started an initial alignment. Green went as usual, but he couldn't get Xarm IR to go. I'll put more details below, but the issue was that the change I made earlier today to change the LASER_PWR guardian to use the IM4 trans channel (alog87581) upon Sheila's findings that gain scaling based on IMC REQUEST caused gain reductions after the power outage (alog87545), caused the power scaling to go to -1.4 after the lock loss. We reverted this change and will have to make it a bit smarter to use this channel in the future.

More details:

Ryan and I initially thought it was an odd alignment and chased that with the usual checking of suspension alignments, checking temperatures, etc. We just couldn't get any flashes to show up on LSC-TR_X_NORM_INMON. On the front FOM, POP90 was visible and noisy, both things that shouldn't be at this point of initial alignment. We eventually thought that maybe it was a dark offset issue since the TR_X channel wasn't quiet at 0, so we ran the dark offset script. This didn't help but I've attached the SDF screenshots. While looking at this channel more, Ryan noticed that it would dip below 0 frequently, almost as if it was flashing but into the negative values. Sign flip somehow? We tried to understand this for awhile and eventually gave up to call the top person on the call list, Keita.

While we were describing the issue to Keita and starting to look into where a sign flip could have happened, Ryan noticed that the power scaling was at -1.4. Since this should just scale with the IFO input power, it shouldn't ever be negative. I plugged in 2.0 and, then everything looked normal again. Looking at the last attachment, IM4 trans goes negative when we lost lock and powered down. The power scaling adjusted as such, and stopped adjusting when the rotation stage stopped moving, as it was supposed to do. If we want to use this channel, we'll have to add some logic to avoid this.

The rest of initial alignment went ok, I had to manual over the PRC offload check because the high useism. DRMI and PRMI aren't locking now, it looks like due to ground motion, but I'll have it try another initial alignment again and report back.

TITLE: 10/21 Eve Shift: 2330-0500 UTC (1630-2200 PST), all times posted in UTC
STATE of H1: Lock Acquisition
INCOMING OPERATOR: TJ
SHIFT SUMMARY: H1 was locked for most of my shift until the microseism eventually got too high. After that, I started troubleshooting issues with initial alignment that were eventually tracked down to be related to a commissioning change earlier in the day; TJ will have an alog with more on this. H1 is currently relocking up to DRMI.
LOG:

• 00:00 - Dropped observing from SQZ PMC unlocking
  • Guardians brought everything back automatically
  • Reapplied nominal damping gain for ETMX mode 1
• 00:05 - Observing
• 01:29 - Turned off ETMX mode 1 damping since monitor filter was ringing up again (see alog87585 for when this happened earlier today)
• 02:29 - Dropped observing to troubleshoot SQZ - alog87594
• 02:43 - Repplied nominal damping gain for ETMX mode 6 (it did not ring up again)
• 03:09 - Observing
• 03:22 - S251021u
• 04:27 - Lockloss - alog87595
  • Troubleshooting begins (see TJ's alog)

Follow up on the comissioning of the Length/Pitch OSEM estimator for PR3. 

I tuned up the pitch OSEM / Model filters and created some equivalent model filters for length. The functions that make the model and osem filters live in:
       blend_PR3_lengthv2_LP_est.m
       blend_PR3_pitchv2_LP_est.m

The files were committed to the SVN at (svnRoot)/sus/trunk/HLTS/Common/FilterDesign/Estimator/ under revision 12735. 

I saw that there is a new make_EST_blends function that will work nicely with these functions, so I will recommend using that one for getting these onto the real-time model.

Lockloss @ 04:27 UTC after 10+ hours locked - link to lockloss tool

Appears to be microseism-related. It's been coming up quickly over the past few hours, and have had several ETMY saturations off and on for the past 20 minutes while everything generally looked shaky.

I dropped H1 out of observing at 02:29 UTC so Sheila and I could try and touch up squeezing to hopefully improve the inspiral range.

We started by scanning the OPO temperature, but to try and stop the filter cavity from unlocking as it had been earlier today, lowered the scan amplitude on line 857 of the SQZ_OPO_LR Guardian from 0.015 to 0.007 and loaded the Guardian. Still, with just a couple seconds left in the scan, the filter cavity unlocked. The temperature didn't look like it needed changing though from the scan we did get, so we set it back to its previous setpoint. Then, we ran a SQZ angle scan with SQZ_MANAGER (after taking SQZ_ANG_ADJUST to 'IDLE') and decided an angle of about 165deg looked good, so we updated the angle servo setpoint on line 53 of sqzparams.py from 0 to -1 and loaded the SQZ_ANG_ADJUST Guardian. We set SQZ_MANAGER to do an alignment scan next, but forgot that the 'SCAN_ALIGNMENT_FDS' state does another angle scan to start. Since that scan ended with the angle in a quite different place that looked worse to us, we just put the angle back to 165 before the alignment scan started. After that, DARM high frequency BLRMs and inspiral range looked slightly better, so H1 resumed observing at 03:09 UTC.

Follow up to Oli's measurements on [LHO: 87362].

We fit the eight (yes, eight!) transfer functions needed for a full L-P M1 OSEM estimator for PR3. The measurements for PR3 were remarkably clean, so it was pretty close to just automated. I spent a lot of time cleaning up Ivey's code so it hopefully can be scaled up to other suspensions more easily.

The relevant fits are shown in the attached .pdf They are suspoint {L,P} to M1 {L,P} (4 fits), and M1 drive {L,P} to M1 {L,P} (4 fits). The estimator's control stability only depends on the M1 drive models, which are fit surprisingly well.

I will note that the L-P plant for PR3 is remarkably reciprocal, and I actually expect to see good performance of the estimators when we try them.

The fits were committed to the sus svn together with a script to install them under revision 12734.

The files live in (svnRoot)/sus/trunk/HLTS/Common/FilterDesign/Estimator/ and are named:

make_PR3_estimator_LP.m
fits_H1PR3_LP-2025-10-07.mat

_______________

These are the .zpk for the fits:

SUSPOINT TO M1
 
Suspoint L to M1 L fit
zpk([0,0,-0.223-4.338i,-0.223+4.338i,-0.04-7.231i,-0.04+7.231i,-0.122-15.233i,-0.122+15.233i],[-0.07-4.099i,-0.07+4.099i,-0.111-4.714i,-0.111+4.714i,-0.11-9.997i,-0.11+9.997i,-0.234-18.094i,-0.234+18.094i],-0.001)
 
Suspoint L to M1 P fit
zpk([0,0,0.954-6.919i,0.954+6.919i,-0.925-7.247i,-0.925+7.247i,-0.03-17.396i,-0.03+17.396i,-0.026-20.357i,-0.026+20.357i],[-0.083-4.115i,-0.083+4.115i,-0.09-4.669i,-0.09+4.669i,-0.107-9.984i,-0.107+9.984i,-0.288-13.181i,-0.288+13.181i,-0.275-18.054i,-0.275+18.054i,-0.084-22.109i,-0.084+22.109i],0.373)
 
Suspoint P to M1 L fit
zpk([0,0,-0.204-4.277i,-0.204+4.277i,-0.081-9.662i,-0.081+9.662i,0.016-17.983i,0.016+17.983i],[-0.044-4.018i,-0.044+4.018i,-0.102-4.661i,-0.102+4.661i,-0.028-9.938i,-0.028+9.938i,-0.018-18.044i,-0.018+18.044i],0)
 
Suspoint P to M1 P fit
zpk([0,0,-0.02-4.046i,-0.02+4.046i,-0.005-4.779i,-0.005+4.779i,0.017-5.716i,0.017+5.716i,0.003-9.719i,0.003+9.719i,-0.175-9.79i,-0.175+9.79i,0.075-13.176i,0.075+13.176i,0.001-22.098i,0.001+22.098i],[-0.01-4.018i,-0.01+4.018i,-0.06-4.157i,-0.06+4.157i,-0.06-4.653i,-0.06+4.653i,-0.007-4.797i,-0.007+4.797i,0-9.765i,0+9.765i,-0.017-9.957i,-0.017+9.957i,-0.036-13.12i,-0.036+13.12i,-0.019-22.115i,-0.019+22.115i],-0.001)
 
 
M1 DRIVE TO M1
 
M1 drive L to M1 L fit
zpk([-0.137-4.329i,-0.137+4.329i,-0.017-5.169i,-0.017+5.169i,-0.037-12.394i,-0.037+12.394i],[-0.098-4.128i,-0.098+4.128i,-0.09-4.66i,-0.09+4.66i,-0.103-9.985i,-0.103+9.985i,-0.245-18.089i,-0.245+18.089i],0.142)
 
M1 drive L to M1 P fit
zpk([1.391-5.654i,1.391+5.654i,-1.723-6.019i,-1.723+6.019i,-9,12.605,0.233-16.247i,0.233+16.247i,-0.042-19.777i,-0.042+19.777i],[-0.093-4.125i,-0.093+4.125i,-0.088-4.646i,-0.088+4.646i,-0.102-9.984i,-0.102+9.984i,-0.292-13.17i,-0.292+13.17i,-0.264-18.072i,-0.264+18.072i,-0.085-22.117i,-0.085+22.117i],0.119)
 
M1 drive P to M1 L fit
zpk([1.425-5.579i,1.425+5.579i,-1.619-6.117i,-1.619+6.117i,-9.724,13.511,0.26-16.388i,0.26+16.388i,-0.122-19.664i,-0.122+19.664i],[-0.092-4.129i,-0.092+4.129i,-0.097-4.664i,-0.097+4.664i,-0.081-9.967i,-0.081+9.967i,-0.298-13.165i,-0.298+13.165i,-0.27-18.058i,-0.27+18.058i,-0.08-22.14i,-0.08+22.14i],0.115)
 
M1 drive P to M1 P fit
zpk([-0.068-4.408i,-0.068+4.408i,-0.012-5.489i,-0.012+5.489i,-0.108-10.011i,-0.108+10.011i,-0.023-21.169i,-0.023+21.169i],[-0.089-4.125i,-0.089+4.125i,-0.079-4.666i,-0.079+4.666i,-0.104-9.96i,-0.104+9.96i,-0.29-13.181i,-0.29+13.181i,-0.084-22.141i,-0.084+22.141i],82.922)
 
 

This is basically a continuation of 86624. Sheila gave Jeff a few more channels to look at regarding the SQZ ASC comparison before and after the satamp swap. The channels I looked at were H1:ASC-AS_{A,B}_RF42_{PIT,YAW}_OUT_DQ.

I looked around and found the best times for before and after the swap, but it is possible there are better times, and the secondary microseism was changing a lot over that period of time, so the lower frequencies are all over the place.

The results I got honestly seem a bit too good of an improvement (comparing them to the amount of improvement we saw in 86624).

Comparison Times
Before swap: 
- 2025-08-02 01:44 UTC; 0.01 BW, 46 averages (PINK) OR
- 2025-08-02 09:23 UTC; 0.01 BW, 46 averages (RED)
After swap: 
- 2025-08-07 07:21 UTC; 0.01 BW, 46 averages (GREEN)
- 2025-08-09 12:52 UTC; 0.01 BW, 46 averages (BLUE)

Comparison Results
Pitch
ASC-AS_A_RF42_PIT_OUT_DQ
ASC-AS_B_RF42_PIT_OUT_DQ
Yaw
ASC-AS_A_RF42_YAW_OUT_DQ
ASC-AS_B_RF42_YAW_OUT_DQ

There does seem to be a decent decrease in noise between 0.8 and 4 Hz for all four ASC dofs, but again, it could just be that the before times aren't as good as they could be. The increase in noise below 0.5 Hz (especially in the green traces) is due to the increase in secondary microseism that I mentioned  in 86207.

TITLE: 10/20 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Observing at 151Mpc
INCOMING OPERATOR: Ryan S
SHIFT SUMMARY:

IFO is in NLN and OBSERVING

The day started and we were locked until planned commissioning from 8:30AM PT to 11:30 AM PT. Commissioning was somewhat light.

15:34 UTC - No SQZ Time Start, Start of COMMISSIONING

16: 04 UTC - No SQZ Time End (30 mins)

16:09 UTC - Matt ran ISS Injections (unfinished)

16:35 UTC - Lockloss (alog 87578)

16:51 UTC - Initial Alignment after bad flashes

During Reacquisition: TJ updated INP1_P and CSOFT_ P Gains from -1 to -2 and 25 to 30 respectively, which should avoid ringup locklosses. (alog 87582)

During Reacquisition: TJ switched the LASER_PWR node input to use the IM4 TRANS signal (post PSL). (alog 87581) - this will result in kappa C being off by 2-3% until we update gains, which is planned for tomorrow.

18:09 UTC - NLN Reacquired
18:10 UTC - Oli Estimator commissioning

18:31 UTC - Back to OBSERVING, End of COMMISSIONING

Start of SQZ Issues (total time out of OBSERVING 1hr 15 mins)

18:59 UTC - Dropped out of OBSERVING to optimize SQZ (range was 145 and not usual 151). This caused a lot of issues due to FC unlocking due to opo temp guardian seemingly causing FC to drop out, requiring me to trend and reset FC1, FC2 and ZM3 P and Y. (alog 87587)

19:39 UTC - Back to OBSERVING to see if changes were fixed - they weren’t - range was now ~143. 

19:42 UTC - Back to COMMISSIONING. Assuming that the FC locking issues were fixed, now I could optimize angle and opo (in that order). 

19:47 UTC - Back to OBSERVING. I had run the SQZ_ANG_ADJUST and it worked but didn’t improve range. If anything, the range dipped to 139 now… Maybe this was a thermalizing thing and I was getting ahead of myself. So I waited the remaining 30 mins until we thermalized fully.

20:17 UTC - Back to COMMISSIONING. Then, I ran the SCAN_OPO_TEMP guardian but once again, it unlocked the FC and the values got reset and had to be adjusted by hand. This is where I realized that the OPO_TEMP unlocking the FC might be unrelated and that I may have to adjust the OPO_TEMP manually. (went back to observing accidentally for a minute or two)

Back to OBSERVING for rest of Shift

21:08 UTC - Back to OBSERVING for final time. After relocking SQZ by trending back values for FC2, FC1 and ZM3, was able to bring them back to normal. Range is now around 144, which means technically nothing worked. However, I think any further attempts to play with the opo temp may risk FC lockloss and more time out of OBSERVING. As such, we should just optimize what we can tomorrow/pick an opportunistic time for leaving OBSERVING.

LOG:

FAMIS 31108

PMC REFL has been on the rise and seems correlated with a drop in ISS diffracted power. No large change in PMC TRANS, though. Other than that, no major events in the past week.

The satellite amplifiers for FC1 and FC2 were swapped out on August 5, 2025 (86207). I checked the filter cavity LSC and ASC signals to see if we can see improvement in the noise before vs after the swap.

For the LSC signals, I looked at H1:SQZ-FC_LSC_DOF2_OUT_DQ, since that is the LSC channel that is on when we are fully locked. 
For the ASC signals, I looked at H1:SQZ-FC_ASC_CAV_{POS,ANG}_{P,Y}_OUT_DQ, since those are the ASC signals inside the filter cavity.

Even just looking at the ndscope of the average noise levels of these channels before and after the satamp swap (swap at t=0), all the ASC channels seem to drop a bit in noise after the swap. It's hard to tell anything from the LSC DOF2 channel.

For comparing spectra, I looked in the range of a few days before and a few days after the swap and found before and after periods of 50 minutes each where the ASC noise looked to be at its lowest. These times and their measurement settings were:
Before: 2025-08-05 03:00 UTC; 0.01 BW, 46 averages (BLUE)
    - Bonus Before for SQZ-FC_LSC_DOF2_OUT_DQ: 2025-08-02 09:23 UTC; 0.01 BW, 46 averages (GREEN)
After: 2025-08-09 12:52 UTC; 0.01 BW, 46 averages (RED)
I also tried out a couple other before times, trying to get the absolute lowest noise before the satellite amplifier swaps, but this before time had the lowest noise that I could find. The bonus Before time for LSC DOF2 is because I was able to find a time where the noise below 0.6 Hz was decently lower than the noise from the default Before time.

Comparison results
ASC
Pitch
SQZ-ASC_CAV_POS_P_OUT_DQ
SQZ-ASC_CAV_ANG_P_OUT_DQ
Yaw
SQZ-ASC_CAV_POS_P_OUT_DQ
SQZ-ASC_CAV_ANG_Y_OUT_DQ
Most of the improvement is seen around 1 Hz, which looks really good!

LSC
SQZ-FC_LSC_DOF2_OUT_DQ
When only comparing to the default Aug 05 03:00 UTC (BLUE) before time, we see broadband improvement almost everywhere below 3.5 Hz. When comparing to the bonus before time (Aug 2 09:23 UTC), however, it looks like the noise between 0.3-0.55 Hz is much better before. I believe this is due to lower ground motion at the time of the green before time (ndscope).