Jennie W, Masayuki N

Daniel set up some model changes last week in h1lsc to allow us to use the JAC-TRANS_A DCPD to lock the cavity. This is a workaround of what is intended to be the normal operation where the locking is done with the JAC-REFL_A PD on IOT1. Rolling this table up to the chamber would make it hard to install the optics on that side of the table so we need this workaround to see the beam transmitted through the cavity that we can use to align the beam from JAC into the IMC.

This PD is just a DC PD so we cannot use PDH to lock as we would for the REFL PD. Yesterday and today Masayuki and I worked on an offset lock, where we use the transmitted power offset to the side of the fringe after normalising the error signal by the maximum transmitted power on resonance.

Use JAC Overview->DITHER. The demodulation is set up so the SIN gain is one and connects to I filter bank which also has a gain of 1. Q filter bank is terminated at the output.

First step is to use the ramp on JAC Overview -> JAC-PZT_Driver by pressing "ENABLE" in right-hand corner, the slider on the left side needs to be near -10V for this to work. Measure the max peak height of the TM00 mode - transmitted power off resonance.

The normalisation is done using the high pass filter bank block at the left of the screen, this opens up H1JAC-DITHER_PD_IN. The gain should be set to the normalisation constant, 1/(transmitted power on resonance - transmitted power off resonance).

The actual servo shaping is done using H1JAC-DITHER_SERVO, which is linked on the right of the 'DITHER' screen. Set the offset to -0.6 to shift the normalised error signal so the zero crossing is just below halfway up the fringe. Check the input and output of the filter bank after the servo are turned on, this is linked on the right of the 'DITHER' screen and is called 'PZT'.

Make sure the PZT ramping is turned off by pressin 'DISABLE on the  JAC-PZT_Driver screen.

Turn on the UGF100 and LP10 filters in the 'H1JAC-DITHER_SERVO' filter bank. These have a gain of 3000 and a low pass filter which rolls off at 10Hz respectively. This latter one is to invert the PZT response.

The gain in this filter bank should be on, use the slider on the 'DITHER' screen to tune to a TM00 peak and then turn on the servo output. Once locked you can turn on the Int1Hz filter to add an integrator below 1Hz and the boost to give an extra 5x gain.

For a better lock we also implemented a proper DITHER SERVO.

The dither frequency is set to 2.5KHz and the demod phase is set at 0 degrees. The input signal no longer had normalisation in H1JAC-DITHER_PD_IN as the error signal will now be the beat between the PZT dither and the cavity transmitted power and so has a zero crossing at the centre of resonance.

There is a high pass filter in  H1JAC-DITHER_PD_IN to avoid up-conversion of noise below 1 kHz. After the demodulation both 'I' and "Q' filter banks have a low pass filter at 1kHz to prevent 2.5kHz and harmonics from getting magnified by the loop.

Both I and Q phase now have a low pass filter at 1KHz to prevent harmonics of the dither frequency appearing in the error signal. The locking technique is the same after this point as the filters needed in the 'SERVO' block do not change as the actuator response has not changed.

Masayuki has set up both these locking processes in the JAC guardian but this is still WIP.

RyanS told me yesterday that he could re-isolate the ETMX ISI. I finally got to looking at it today and something was off with the CPS. Using the weekly CPS spectra script I got ASDs for the CPS from last night and a bunch of the the CPS spectra were way off, first image.

I went to EX and looked at racks, SEI chassis all looked fine. I unplugged a couple of CPS probes at the chamber and did not get the expected behavior. One CPS went to the near limit rail (-32k) the others hardly changed at all. This made me suspect timing, so I traced out the timing sync wires, found the 71Mhz CPS timing fanout and the lights were all off. Looked at the power strip and only the -18v led was on. Turned off the fanout chassis, took some pics and found Marc and Fil.

Marc came to the end and we looked at power supplies in the front of the building and the +18 for the rack was dead. We swapped both power supplies for spares, and CPS timing seems to have come back, I was able to isolate the ISI. Marc said the fans seemed a bit stiff and smelled weird, so maybe another fan failure.

Per WP12970

ISC-R1 +18V supply failed this afternoon.  This set was slated for replacement, so we replaced both legs.

+18V S1300287 was replaced with S1202002 with new ball bearing fan installed.
-18V S1300275 was replaced with S1201995 with new ball bearing fan installed.

M. Pirello, J. Warner

Tue Jan 13 10:09:38 2026 INFO: Fill completed in 9min 34secs

TITLE: 01/14 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY: Another busy day of upgrades; JAC installation and HAM7 beam alignment continued while the move/upgrade of BS DACs began today and should wrap up tomorrow morning. Jim also discovered a failed power supply at EX which looks to be the reason the chamber can't be recovered, so he and Marc are starting to take action there.
LOG:                                                                                                                                           

Sheila, Kar Meng, Elenna, Jim

Yesterday Jim and I went to the chamber in the morning to work on cables. We got a solution for the cables after some work, but one peek zip tie luanched itself across the chamber.  We spent a few hours searching for this; I climbed into the chamber on the -X side and could them see it on the bottom of the +Y door where Jim was able to reach through the port and retreive it.  

In the afternoon Kar Meng and I went back to the chamber and found that the beam was blocked on the cable clamp that Jim and I had added earlier.  The previous cable routing is shown here, the peek cable clamp is on the bottom of the pillar so that it doesn't block the beam.  We removed the new alumium clamp and I tried to move it lower, the second from the top bolt hole is not the right size so we need to place this in the 3rd from the top hole.  I found that difficult to reach and eventually decided to leave it for another day or ask for help from someone with longer reach.  

After this we saw that our beam did arrive on AS_C in HAM6 but wasn't aligned onto AS_A or AS_B.  

This morning Elenna, Kar Meng and I went to the chamber to see if we could improve the alignment.  We walked B:M3 to align onto the second iris on SQZT7, and B:M4 to align onto the iris in front of ZM4, which seems degenerate with the iris at the bottom of the SQZT7 periscope.  This worked well and we could see light on AS_B and AS_C although nothing on AS_A.  We then went to the control room and moved ZM6 to center on AS_C better, then we could engage the AS centering loops.  

This afternoon Kar Meng and I had another look at alignment from the control room.  We have set the OMs back to their alignments during the run, and we still have the beams on AS_C, AS_A and AS_B although we are using a lot of the range on ZM6.  

Still to do before finishing in HAM7:

• adjust alignment to relieve range on ZM6
• add cable clamp on bottom of OPOsus pillar near B:M4, without blocking beam
• Photos of cables as we are leaving them
• unlock VIP
• photos of iris locations and beams on them
• power budget measurements:  with power meter at various points on VIP, seed reflected power and seed power to homodyne path
• realign beam onto FC QPDs 
• check alignment onto other photodiodes in chamber
• remove irises
• unlock ISI, SUS and ISI checks

Masayuki, Jennie, Rahul

Given below are the list of lenses and mirrors I have placed on the ISI table and roughly positioned (without the beam for now) them as per following drawings  - D2500344 and D0901821.

Lenses - L1, L2 and L3: Masayuki confirmed from the vendor that the arrow mark on the lens point towards the convex side. For lens L1, arrow is pointing against the PEEK ring (in the lens holder). For Lens L2 and L3 the arrow is pointing towards the PEEK ring. The PEEK ring for all three lens are facing towards the HAM2 chamber. In other words, curved surface of L1 is facing towards the PSL and for L2-3 towards HAM2.

I re-cleaned L1 lens after finding some dust particles on the curved surface, post cleaning the lens looked better.

Reflecting mirrors - JAC-RM1, JAC-RM2, JM2 and M3.

Masayuki and I also added a temporary Siskyou mount acting as JM3 (for Tip Tilt, just like JM1). We will replace both of them with the Tip Tilt next week.

Next, we proceed with aligning/pointing the above optics with the beam. L1 and JM2 are already decently pointing (although without a locked JAC) towards JM3.

J. Kissel
ECR E2500296, E2400409
WP 12962

D2300401 (for susb13) and D2300383 (for susb2h34)

We begin the major upgrade of the H1SUSB123 and H1SUSH34 SUS and SEI Systems converting them to SUSB13 and SUSH34 a la G2301306 today. We're focusing on upgrading the DACs in the IO chassis () and all the downstream surrounding impact of that  analog electronics This will take down the following computers, and they will be resurrected with new names as follows:
    FORMER NAME           FORMER SUS              NEW NAME           NEW SUS
    h1susb123             ITMX, ITMY, BS          h1susb13           ITMX, ITMY
    h1sush34              MC2, PR2, SR2           h1susb2h34         BS, MC2, PR2, SR2, LO1, LO2
    h1susauxb123          ITMX, ITMY, BS          h1susauxb13        ITMX, ITMY
    h1susauxh34           MC2, PR2, SR2           h1susauxb2h34      BS, MC2, PR2, SR2, LO1, LO2

As such, I've 
    - brought the ITMX, ITMY, BS, HAM3, and HAM4 SEI systems to ISI_DAMPED_HEPI_OFFLINE (so we don't risk any "hard" trips of HEPI during all this).
    - brought all impacted SUS gaurdians to AUTO mode, then to the SAFE state
    - Increased the bypass time on all impacted software watchdogs to bypass time to a large number (90000000 secs), and hit BYPASS

TITLE: 01/13 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: 5mph Gusts, 3mph 3min avg
    Primary useism: 0.02 μm/s
    Secondary useism: 0.49 μm/s 
QUICK SUMMARY: CDS upgrades this morning; moving SUS in BSC2 over into the sush34 model and upgrading to LIGO DACs. JAC installation and HAM7 beam alignment will continue as well today.

Workstations were updated and rebooted.  OS packages were updated.  Conda packages were not updated.

Daniel, Ryan S, Erik, Jonathan, Dave

In preparation to tomorrow's work on h1susauxb123 we decided to move the EDC from here to h1susauxh56. We did this at the same time the DAQ was being restarted for:

1 Daniel's new h1lsc model

2 Adding Ryan's JAC_LOCK guardian node to EDC

The EDC move had a few configuration issues which were quickly resolved. There were no issues with the h1lsc restart.

TITLE: 01/13 Day Shift: 1530-0030 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Planned Engineering
INCOMING OPERATOR: None
SHIFT SUMMARY: Laser safe in the LVEA this morning for some craning and cabling, both in-chamber and out, but now back to laser HAZARD for more JAC installation and SQZ beam alignment this afternoon.
LOG:

(Travis, Dave, Gerardo)

Dave noted and reported a rise on vacuum pressure at X-End.  Since, Travis and I were in the control room, we quickly determined that the controller for X2-8 ion pump was not working.  I drove down and found the high voltage disabled/off at the controller, and on on the screen the following note:"An excessive arcing condition has been detected (Error10)".  No guidance found on manual, snapped a photo to submit later to vendor.  I restarted the system and the HV started, controller hummed and crackled and was only able to get up to 600 Volts, all while making noise, and one point some funny smell came out of the inside of the controller, HV was disabled, time to replace the unit.

[Jason, Jennie, Betsy, Keita, Sophie, Masayuki]

Initial alignment (1/8)

• For the initial JAC alignment, we need to scan with laser frequency since we didn't have JAC PZT actuation. The laser frequency was swept by changing the laser crystal temperature from −0.13 K to −0.05 K with a period of 30 seconds.

• The alignment was adjusted to suppress higher-order modes resonance. The actuators used were the PSL PZT and the fixed Siskiyou-mounted mirror that replaced the JM1 tip-tilt suspension. I worked inside the chamber while Jason controlled the PSL PZT via MEDM. Beam walking was performed by moving the PSL PZT first and optimizing alignment with the other mirror in the chamber. Note that due to the HAM1 periscope rotating the beam axis by 90 degrees, the pitch and yaw of the PSL PZT are effectively swapped in the JAC coordinate basis.

• As a result, moving the PSL PZT by approximately 1000 counts in pitch (corresponding to yaw in the JAC basis) achieved an alignment good enough to observe the TEM_{00} mode. Further alignment will be performed using the transmission PD signal.

JAC wiring (1/8)

• After discussion, we decided to delay installation of the reflection path. Since PDH locking using the reflection RF PD was therefore unavailable, we proceeded to align the TRANS DC PD and aim for shoulder locking using this signal.

• In the nominal design, the TRANS PD receives light from the JAC leak port (~0.3% of the input power), picked off by a laser window with 0.7% reflectivity. However, with the current PSL power of ~100 mW, the available light was insufficient. Therefore, the TRANS PD was configured to directly receive light from the leak port without a laser window.

• Due to a mistake, a post that was 0.5″ too short was prepared for the DC PD mount, and no suitable replacement was readily available. As a temporary configuration, the PD was mounted in a base–dog clamp–post–dog clamp–PD stack, using a 0.25″-thick dog clamp as a spacer.

• In-vacuum cables were also wired. From the feedthrough (D4F10), a DB25 cable (D2500336) splits into three branches: TRANS PD, PZT & thermistor, and heater & thermistor. Since the ALS beam runs adjacent to the JAC, cable routing clamps were used to avoid interference with the ALS beam (see attached photos).

Wiring confirmation (1/9)

• To enable locking using the TRANS PD, we confirmed that the corresponding signal path exists in the CDS infrastructure added by Daniel. In the H1LSC model, ADC2 channel 10 is assigned to the DC PD and routed through the dither locking module to the JAC PZT output.

• In the analog path, the DCPD output from the JAC interface chassis is connected to the auxiliary PD concentrator, and its monitor port is routed to the DAC via the D-sub patch panel. Both the fast and slow channels were confirmed to be correctly connected using the actual PD signal.

• The polarity of the PZT wiring was verified. For this test, the DB9 connector of the PZT & thermistor cable (D2500336) on the JAC side was disconnected, and gold pins were inserted into pins 5 and 9 of the feedthrough-side female connector (the PZT inputs). A 100-count offset was applied to the JAC_SERVO filter bank output driving the PZT, and the voltage between pins 5 and 9 was measured with a multimeter. To avoid high voltage during this test, the HV amplifier was powered by an 18 V supply instead of the nominal HV source.

• Toggling the 100-count offset resulted in −10 V (off) and +9 V (on). The same behavior was observed when measuring the PZT driver output directly, confirming correct cable connections and no risk of reverse loading the PZT.

• These voltage levels were not considered reasonable for normal operation, so Daniel was consulted. The likely explanation was operation with the 18 V supply instead of the HV source. After reconnecting the HV supply, injecting the laser into the JAC, and applying a ramp signal to the PZT, motion of the PZT was clearly observed in the transmission signal.

• The TRANS PD was re-aligned, and the signal observed on MEDM was maximized.

• The thermistor connection was also verified, yielding a reasonable reading of approximately 22 °C. Heater testing has not yet been performed, but functionality of the PZT, thermistor, and PD has now been confirmed.

Optics preparation (1/9)

• In parallel with wiring checks, preparation of the optics on the JAC output side was carried out.

• An issue was found with a newly fabricated ISC post (8-32 variant) intended for the lens mounts; it was confirmed that the standard D1000968 post works without issue. Keita and Sheila located suitable posts in the staging building, and three lens mounts were assembled using these posts.

• Attempts to peel the First Contact from the lenses were unsuccessful, as the FC layer was extremely difficult to remove, and we gave up. Applying FC to spare optics is considered the fastest path forward, and for the stuck FC, reapplying FC to soften it may be necessary.

• HR mirror preparation was also performed.

• Further details of the optics preparation will be posted shortly by Keita.

[Jason, Rahul, Jennie, Betsy, Masayuki]

Summary

JM1 was swapped from a tip-tilt suspension to a fixed Siskiyou mount to improve beam stability for alignment, profiling, and JAC control. Mode matching to the JAC was measured using a beam profiler, yielding <1.5% mismatch in both axes, well within the current requirement. The JAC pedestal and body were installed in HAM1, and initial beam injection showed clear HoM resonances at the transmission port. One discrepancy in body mode damper mounting holes was identified and will be tracked via FRS.

Details

• After discussion, we decided to replace the JM1 tip-tilt suspension with a fixed Siskiyou mirror mount. This provides easier alignment, beam profiling, and more stable JAC control compared to using a suspended mirror. As reported in the previous alog, JM1 had already been aligned to the target irises with a good angle, so the position of the JM1 suspension was marked with dog clamps to allow easy recovery of the alignment after installation work. The Siskiyou mount was aligned using the same target irises.

• After switching the mirror mount, a beam profiler was set up in the reflected beam path from the fixed JM1 to measure mode matching to the JAC, as shown in the attached picture. The z-origin was defined at the position where the beam profiler cart reading was set at 16 cm along the rail. Due to cable length limitations, beam size measurements near the waist were not possible, but this was not considered a significant issue.

• The measurement results are shown in the attached plot. The target and measured beam parameters are:

  • w0_target = 548.00 µm, z0_target = −4.318e−01 m

  • w0_x = 572.01 µm, z0_x = −5.277649e−01 m

  • w0_y = 603.50 µm, z0_y = −3.736913e−01 m

• The resulting mode mismatch was 1.363% (x) and 1.455% (y). Our current target for mode mismatch is 10%, which is sufficient for achieving a reasonable JAC lock. Therefore, this result is excellent and no further adjustment is required at this stage. Fine tuning may be performed later after JAC lock, possibly after HAM1 pump-down.

• The JAC pedestal was installed and secured to the table, and the shim was placed on the pedestal.

• The JAC body was then brought into the chamber. The procedure was recorded on video. A handle (not Class-B cleaned) was temporarily attached to the JAC; it was wiped down, and aluminum foil was inserted between the handle and the JAC body. During installation, Jason and Rahul supported the body from the sides while I handled the lift into HAM1. The end caps protecting the mirrors remained installed throughout this process.

• During installation of the body mode dampers, we found that one of the screw holes specified in the drawings was missing. The location of the missing hole is shown in the attached picture. As a result, one body mode damper was shifted by one column of screw holes (a 2″ offset). We discussed this with Stephen and believe it should not cause an issue, although it will be monitored. An FRS ticket will be submitted to track this discrepancy.

• After removing the end caps, the beam was injected into the JAC for a quick alignment check. A higher-order mode (HoM) resonance was immediately observed at the transmission port, confirming successful initial alignment. The next steps are fine alignment of the input and reflected beams, followed by in-vacuum wiring.