EY followed EX's lead and also came back to life this morning at 10:34 causing our second VACSTAT alarm for today. EY had been flatlined since yesterday 06:23, close to EX's time.

I restarted VACSTAT at 10:45 to clear the alarm.

10-32x0.375" SHCS that was blocking the access to one 1/4-20 screw was replaced with a low profile 10-32 SHCS.

"Issue 2" in alog 88862 was solved.

See picture, Mitch found a 10-32x0.5" SHCS with a low profile head. 0.5" seemed to be OK in that it's not too long, but we used two washers to make sure that the scrwe doesn't bottom out.

At 10:11 the EX BT-Ionpump gauge came back to life, causing a VACSTAT ALARM. This gauge has been flat-lined since yesterday, 07:09 Fri23Jan2026.

Could be because today is a sunny day after a long period of cloudy weather (the EX_BT system is solar powered)?

At 10:18 I restarted VACSTAT to clear the alarm.

Sat Jan 24 10:10:15 2026 INFO: Fill completed in 10min 11secs

Jennie W, Sophie M, Sheila D,

First we checked that the MC mirrors top mass OSEMS matched their position at 00:00 UTC on December 19th. This was the last time when the IMC was locked in air after HEPI was locked on HAM2 before JAC install.

MC1 -678 mrad P, -1365 mrad Y.

Sheila and Sophie went to align in chamber using JAC_M3  (mirror right before JAC HAM1 output periscope) and JM2 (mirror right after JAC).

I looked with the beam card between the upper and lower periscope mirrors at the beam spot.

Sheila and Sophie walked the beam in pitch and the beam started to clip badly. They undid this and Sheila came to check the beam on the table and realised that it looked much dimmer than expected compared to the beam leaving the chamber.

She suggested it might be a ghost beam coming onto the table.

After lunch we checked this with the power metre and we had 40mW into HAM2 but 0.46 mW onto the bottom periscope mirror in IOT2L.

Sophie managed to unclip this mainly with pitch (in the JM2 basis) beam walking, with very slow changes on both mirrors in turn and checking the REFL output and the power onto the table we eventually got ~ 20mW, measured after the bottom periscope mirror.

We eventually made enough change that the beam was getting to both MC REFL WFS but not hitting the diodes.

Sophie did more beam walking in yaw on the two mirrors while looking at MC REFL PD and maximising the power on this. The power metre at the end of the day showed 23 mW just after the lower IOT2L periscope mirror but we were still not hitting the WFS PDs. The beam is low and right on MC REFL WFS B.

The workstation CDS conda environments will be updated by Saturday Jan 24.  This is a major update that affects many conda packages.  Many packages have been updated and some have been dropped.

Python is updated from python 3.10 to python 3.11.  Yes, 3.11 is old already, but the CDS environment tracks the IGWN environment, and important packages have not been push beyond python 3.11.

A complete list of changes in the new environment can be found here:

https://git.ligo.org/cds/packaging/cds-conda-distribution/-/wikis/Environments#version-2026-01-06-01

TITLE: 01/24 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY: Another productive day to wrap up the week.

• HAM7 alignment is essentially finished and irises are being removed from the table this afternoon
• The viewport assembly was removed from HAM5 in preparation for the relay tube to HAM7 to be reattached, which will happen next week in laser safe conditions (see alog88876)
• Alignment in HAM1 continued for the JAC, including downstream pointing into HAM2.
• Work progressed on the to-be-installed JAC EOM in the optics lab
• Jeff took some transfer functions on the OPO suspension and claims they're "good enough to say the suspension looks healthy" after finding issues with open-light values (see alog88872)

LOG:

In prep for HAM7 closeout, the temporary viewport assembly was removed from the HAM5 relay tube port. RV-1 was closed, and the volume vented with viewport covers still installed, with dry N2. I monitored the corner pressure (PT-120B) during venting, to make sure there were no leaks through the gate valve. No change in pressure, so I continued with viewport removal.

Viewport was removed first, then the pump-out tee. The mitered spool piece and bellows assembly was installed on HAM5 side. Lines were marked across the relay tube assembly prior to removal, so the install orientation remained the same. 

The ported spool piece will be installed next week during Laser Safe conditions. Bellows flanges were covered with foil for protection.

J. Kissel, R. Short

Trying to chase down the issues Ryan Crouch was seeing in his attempt at health check TF measurements of the OPOS (88851), I found that there were no OSEMINF OFFSETs and GAINs to compensate for the OSEM sensors' "open light" current (OLC). 
Trending back, this suspension has hasn't had OLC compensating OFFSETs or GAINs since we migrated it from HAM6 to HAM7 circa 2021.

However, there are TFs in the measurement library (list ${SusSVN}/sus/trunk/OPOS/Common/MatlabTools/plotallopos_tfs.m) in 2022, documented and compared to matlab model in LHO:64275, and they look great. This sadly makes sense, as the former open light current compensating gains were close to 1.0, and the "centering" offset to make the "flag is centered within the range of the OSEM sensor" equal to 0 doesn't really matter for a driven transfer function or ASD, since these both don't care what the DC value of the sensor is, as long that sensor doesn't saturate during the measurement (i.e. the raw ADC counts are somehwere around +15000 [ADC ct]).

There's several reports -- over the years as HAM7 has been occasionally open -- how incredibly inaccessible the AOSEMs are within the suspension, and thus every attempt to gather new open light current values are thwarted by "we don't have time for that" or "I tried and it'll require interfering with way too much other stuff." Examples of this are  the 2021 attempt by Keita / Camilla (LHO:59939) and last week's undocumented attempt by Rahul (while he was in HAM7 to relieve the yaw offset on ZM4; LHO:88788). 

Even worse, prior to the move to HAM7, during the installation in HAM6 circa Feb 2018 (LHO:40727) the new OSEMs AOSEMs saturated the ADC at olc = +32768 [ADC ct], so we installed relatively meaningless compensating OFFSETs and GAINs of -olc/2 = -16384 [ct] and 30000/olc = 0.916 [ct/ct], for all but one -- the V3 sensor which was set to -15481 [ct] and 0.973 [ct/ct]. In Apr 2018, we found that the V2 AOSEM's flexi-circuit V2 OFFSET and GAIN was changed to -14553 [ct] 1.031 [ct/ct] in Apr 2018 (see LHO:41468, LHO:41470 and LHO:41503.

So this has been left as wrong for 5 years. And really, prior to that, inaccurate.

Citing me from 2021 -- the OPOS is using the same AOSEMs and sat-amps as it did while it lived in HAM6, the only difference in the open light current will be from the cable length difference between HAM6 to SUS-R4 and HAM7 to SUS-R4, which I suspect is negligible.
So let's at least restore the "inaccurate."

But in short, assuming no LED light lumin decay, I think the "correct" values -- and the values that I've now re-installed -- are
   OSEM      OFFSET      GAIN
    H1       -16384      0.916
    H2       -16384      0.916
    H3       -16384      0.916
    V1       -16384      0.916
    V2       -14553      1.0307
    V3       -15481      0.973
These have been accepted in the SAFE and OBSERVE.snaps 

So, the question that now remains whether the AOSEM LEDs have decayed significantly, but that -- as we every suspension we ever try to assess -- is horribly confused by alignment shifts (intended or not), making it essentially impossible to assess. The OPOS is no different.
The flags physical relationship to the OSEM sensor / LED is wherever they are, with the chamber in-air at the moment. Comparing the last relatively pre-chaos period to now, 
        2025-12-04    NOW
        Under UHV     In-air
H1      7904          8488          
H2      8831          7431
H3      8482          3106
V1      14934         14088
V2      13534         11864
V3      12618         9824

In between the "reference" times above, there have been a series of major alignment shifts of the suspension (see last attached), as reported by the OSEM ADCs, for the following reasons:
   2025-12-04 18:03 UTC H567 Mega-clean room turned on, causes all the OSEMs to drift over ~7 days.
   2025-12-04 20:23 UTC Site Power Outage, brief outage, but 
   2025-12-10 00:46 UTC HAM7 Door pulled off
   2025-12-10 18:50 UTC OPOS locked down in prep for removal of OPO cavity for crystal swap
   2025-12-12 23:41 UTC Removal of OPO cavity for crystal swap

   2026-01-20 21:34 UTC OPO is reinstalled, OPOS is unlocked
   2026-01-21 19:02 UTC On-platform adjustment of optics, reasonable to expect alignment shifts
   2026-01-22 22:15 UTC More on-platform optic adjustments, reasonable to expect alignment shifts
All this makes it tough to say what the "right" position of the OSEMs should be, and whether e.g. having a raw OSEM count of 3000 [ADC ct] on the H3 OSEM is "acceptable."

Acknowledging that there's very little range on the H3 OSEM right now, I'll try reducing the excitation amplitude.

However, while we have the chamber open, I recommend re-centering the AOSEMs.

Other "interesting" aLOG references.
2022-08-02 LHO:64275 Post HAM7 install health checks in Feb, May, and Aug 2022 show a reasonable, and coherent set of TFs that match the model. Even without OFFSETs or GAINs.
2021-10-27 LHO:60422 Bug identified in OPOS OSEM2EUL/EUL2OSEM basis change matrices. Identified, but never implemented.
2021-09-17 LHO:59939 Attempt to measure open light currents after installation into HAM7, but inaccessibility of H3/V3 OSEMs thwarted effort. 
2021-08-26 LHO:59741 OPOS Lands in HAM7
2021-08-17 LHO:59752 OPOS Front-end Infrastructure Migrated from HAM6 (h1susopo) to HAM7 (h1sussqzin)
2021-08-17 LHO:59652 h1susopo model is brought down for above migration to new model -- HERE'S WHERE OPOS OLC OFFSETs and GAINs were lost.
2018-02-26 LHO:40727 Generic 16384 ct OFFSETs and 0.916 GAINs installed because new AOSEMs saturated the ADC during open light.
2018-04-17 LHO:41470 V2 OSEM replaced, new OSEM has non-saturating OLC values

FAMIS 39748

Laser Status:
    NPRO output power is 1.835W
    AMP1 output power is 70.39W
    AMP2 output power is 139.0W
    NPRO watchdog is GREEN
    AMP1 watchdog is GREEN
    AMP2 watchdog is GREEN
    PDWD watchdog is GREEN

PMC:
    It has been locked 14 days, 20 hr 48 minutes
    Reflected power = 26.68W
    Transmitted power = 104.1W
    PowerSum = 130.8W

FSS:
    It has been locked for 8 days 2 hr and 16 min
    TPD[V] = 0.3187V

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
    FSS TPD is low

Both RLF_QPDs A and B are centered on both PIT and YAW, at low seed launch power (2.3)

The CLF transmission PD on SQZT7 was centered.

Fri Jan 23 10:09:29 2026 INFO: Fill completed in 9min 25secs

Note I am using a new format for the trend plot which reduces the number of log divisions on the top plot's y-axis.

The problem: during the winter months the TCs start with a positive temp, dropping to around -70C when LN2 is flowing. To show the TCs along with the discharge line pressure and LLCV on the same plot the TC signals are inverted and then all are ploted logarithmically. This means the baseline (positive)temps are now negative which is not shown on a log plot, and when crossing zero they force many divisions in the autoscaled log axis.

The solution: I modified cp1 overfill IOC to serve modified TC[A,B] channels with max temp limits. Max temp is currently -12.3C, settable in the configuration file. Any temp greater that -12.3 is kept at this value for the channels H1:VAC-CP1_OVERFILL_TC_[A,B]_TEMP_PLOT_DEGC

We have identified two mechanical issues with the JAC EOM and its mount.

Issue 1. Crystal isn't captured in the EOM assembly when we install it in a certain way, but installing it in another way to capture it might (or might not) put overly strong force on the crystal.

No I haven't dropped the real crystal, I did drop alumina piece with the same outer dimensions as the RTP crystal while testing the installation procedure.

It's probably hard for you to understand this issue if you're not familiar with the EOM assembly (D2500130) and the assembly/test procedure (google doc), look at my super-simplified cartoon (ideal.png).

EOM comprises the face plate, RTP crystal and everything else. Everything else is already assembled. The task is to sandwich the RTP between the electrode board (which is a part of "everything else") and the face plate. You'll put the face plate on a table, put RTP on top of the face plate, carefully lower everything else until the board touches the RTP along its entire length, and you bolt the face plate to the side panels. Simple.

It can be much different from that if the board is not orthogonal to the side panels, look at the second attachment. Here, the board is crooked, the distance to the face plate on the output side is smaller than on the input side (this cartoon).

Suppose that I choose to make the face plate contact with the output side plate, and bolt the face plate down. The face plate is squared up relative to the side plate, not the board, and the "everything else" part will rotate away from the face plate (or face plate rotates away from everything else) and RTP is free to move. I think this is what happened consistently (10 or so trials) in the lab today, no matter what we did, the alumina piece (i.e. fake RTP for excercize) slid out of the assembly but only after tightening the screws.

On the other hand, if I choose to make the face plate contact with the input side plate (notsoideal.png), the rotation will be in the opposite direction, the face plate is not reqlly squared up but will put a pressure on the RTP, securely captureing it. After switching the side where the face panel contacts in the lab, the alumina piece (fake RTP) never dropped (3 trials so not much statistics but 3/3 success is much better than 10/10 failure).

The problems are that we don't have any control over how much force is applied to the crystal. Moreover, in our "successful" trials, the board might not be touching the alumina piece along its entire length. Look at the picture, this is after the last (3rd) "successful" attempt with the alumina piece still in, the board and the face plate are not parallel with each other, we might be pinching the alumina thing at the corner closest to the output side. Not sure if it's always like this but it's likely.

Stephen and Michael suggest that instead of bolting the face plate firmly to the input side, there could be a gap on each side, the face plate is supported by the tension of the screws, making sure that the face is parallel to the board. I'm not necessarily a fan of the idea but we'll try.

Issue 2. One of the bolts for the EOM base is blocked by another screw head.

See the last picture. One 1/4-20 screw cannot be tightened because the 10-32 bolt head just above that blocks access. We might replace the offending bolt with 10-32x0.375" pan head screw AND use the ball head Allen key for 1/4-20, that might work.

Other issues.

We can move the mount in PIT/ROLL and YAW, but somehow it's very difficult to cause pure PIT motion, it always couples to ROLL.

When we use set screws to tilt the base and then back them off, the mount won't return to the original position, I have to push down the pivot plate toward the base plate firmly, then there's a metal clicking sound and the mount goes back. This might be related to the fact that, as of now, the dowel pin (part #14 of the assembly drawing) is a REALLY tight fit for the pivot plate as of now. 

Finally, the cable strain relief post could not be set at a desired angle using the supplied slotted washer (the only difference I was able to make is either bad angle or 180 degrees off of the bad angle). But using other washers on top of the slotted one(s) I was able to manage good-ish angle.

Jennie W, Jenne D, Masayuki N

Today Jenne and I went into HAM1 and onto IOT2L.

Our first priority was to check the input alignment to the JAC after alignment efforts yesterday to check we were not locking to a HOM mode. We locked the JAC with a power of 0.1 on the TRANS_A_LF_OUT channel.

I checked after lens JAC_L3 which is right before the steering mirror into the HAM1 output periscope. Here the beam looked like a 10 mode (two lobes in the horizontal axis). Further upstream (just after JAC cavity this was not obvious due to the beam size.

We used the steering mirror between the input periscope and JAC + the PSL periscope PZT mirror to improve the alignment. To do this we changed yaw in the closer mirror and pitch in the PSL PZT mirror as the input periscope switches the P and Y around.

We recovered a value of 0.22 on TRANS_A_LF_OUT which matches with the values we got on friday the last time we had a good TM00 lock on the JAC cavity.

After this Jenne went to the table to look at the beam we found yesterday which is at the edge of the MC REFL periscope mirrors in yaw and clips on the BS1 optic.

I started changing the tilt of the beam through HAM2 by changing the pitch of the mirror right before the HAM1 output periscope. This did not really change the beam position so we moved to a further away mirror (JM2) to translate the beam.

During this process we also became unable to lock the JAC using the guardian. More details at the end.

After some iteration we were able to see a beam on the MC REFL PD but have not been able to walk the beam to see any signals on the MC REFL WFS orflashing from the IMC on the MC TRANS PD.

After a break we went to the control room and did some checking of the MC sus alignments. We found problems with the alignment of MC3 - Dave and Ollie debugged this.

Jenne then did some walking of the MC mirrors and by moving MC1 she could get a larger signal on MC REFL. We might be able to use this tomorrow by moving MC1 to this 'wrong' ( ie. not consistent with the nominal IMC alignment before JAC was installed) place and walking MC1 back while changing the in-vac fixed mirrors in HAM1 to reover the beam on MC REFL PD.

During the day an electronics chassis was swapped that does the whitening for the TRANS PD A. The signal stopped getting to the diode so Daniel gave us the go-ahead to bypass this - this might need to be fixed properly at some point, it is level 10 on the rack closest to HAM1, next to the PSL enclosure. I took put IN3 and OUT 3 cables and connected them with a TNC connector.

This afternoon and evening we tried to trouble-shoot the JAC locking.

There was also an incorrect gain that got reset by the model restart for h1lsc earlier today but fixing this (JAC_DITHER_PD_IN gain was set to 200 and should be 4) did not allow us to lock.

After changing the servo gain in the dither lock Jenne noticed that this had no effect on the lock level/noise on the TRANS PD.

After checking with Masayuki the problem seems to be that the fast channel to drive the PZT is not connected somehow at the racks but the Beckhoff controller can be used to drive the PZT - will consult with Daniel/Marc tomorrow.

[Sheila, Karmeng]

We offloaded the saturation on ZM4 and ZM6 while maintaining the alignment to HAM6 QPDs.

Power budget: we measured 0.75mW at the OPO output, after SFI1, and after thin film polarizer. The power dropped to 0.74mW after BM3. And further to 0.71mW at the viewport/SQZT7 periscope and HD.

Scanned the IR and green transmission of the OPO, the green transmision (orange trace) will need further adjustmnet.

We have intalled the POP periscope stiffener.

Some dog clamps in the REFL path as well as the cable bracket for PM1 were relocated to accomodate. 1st and 2nd picture are "before" photo. 3rd one is after PM1 cable bracket relocation but before installing the stiffener.  There are also three "after" photos showing how things look on the table.

POP beam clearance:

I took the picture of the bottom periscope mirror through dichroic (HR for IR, transmission for green) to see the beam clearance. In the first such photo (POP_beam_clearance.jpg), the camera is close to the center of the dichroic and the short stiffener beam is close to the edge of the optic but not occulting the mirror, so we're OK. Just to make sure that we're absolutely safe, I moved the camera closer to the -Y edge (right on the picture) of the dichroic (POP_beam_clearance_extreme.jpg) and it still looks OK.

If it's hard to understand what was done, look at the annotated photo (the last attachment After_stiffener_installed_annotated.jpg), the cellphone camera was inserted to "Camera" position.

REFL beam path:

I confirmed that the long stiffener beam doesn't interfere with the motion of the REFL beam diverter. Also, when the REFL beam diverter is open, I looked into the last steering mirror for the REFL air path from the viewport position to make sure that the short stiffener beam won't occult the REFL path. 

Some hiccups:

We used D2500433 -11 variant S/N 4 and -1 variant S/N1 even though page 11 of T2500339 suggests it should be -10 and -2 variant, respectively.　We didn't have -10 variant, and -2 variant was absolutely too short.

B&K

We performed B&K hammer measurements before/after the stiffener installation for POP periscope. Before, there was a 70Hz-ish peak. After, it was pushed higher up in frequency. The transducer was attached to the ISI table and Jim hammered the top of the periscope.

Likewise we did B&K test for the input periscope of the JAC even though it was not absolutely necessary.

We haven't done B&K for the JAC output periscope because it's not even fully clamped down (we will move it).

Jim will post the data.