(Remote) R. Takahashi, Nakagaki, Ikeda

This work is related to K-Log #37013.
We replaced the PR3 satellite box.

[Details]
1. We measured the forward voltage at the end of the conversion cable connected to the feedthrough.
   There was no significant change in the values, and all measurements were normal.

IM H3

(Reference value)IM H1

2. Next, we replaced the satellite box.
We placed the new satellite box (S1604893) on top of the old satellite box (S1504430) and reconnected the DSUB cables from the old box to the new one.
H2 and H3 were connected to the replacement satellite box.Fig.1
 

Detail

The first trial

First, we started using 1f signals for locking the DRMI. The signals were as follows:

• PRCL: REFL 45I
• MICHL POP 17Q
• SRCL: POP 17I

The gain settings of them were decided as follows:

• PRCL: Measured the relative gain between REFL 45I and REFL 135I when PRMI was locked using 3f signals. Multiplied the factor by the gain for PRMI 3f locking.
• MICH and SRCL: Estimated the relative gain from SRMI to DRMI using Finesse simulation as shown in Fig. 1 (15 for MICH and -17 for SRCL). Multiplied the factors by the gains for SRMI 1f locking.

The initial guess of the appropriate gains and the actual gains after tuning are as follows:

We built these gains beforehand and set triggers to trigger the output when POP90 and AS34 buildups exceeded a certain level. With these configurations, we could hold the interferometer in a good position for ~1 sec (Fig. 1).

Second trial

Next, we changed the PRCL error signal from REFL45I to POP17I, because in RSE, REFL45I is used for CARM and POP17I for PRCL. We tried w/o and w/ triggers, but we could not lock the DRMI, even tweaking the gains. The situation got worse. 

We noticed that the PRCL feedback was incredibly small, more than 2 orders of magnitude smaller than the others, so we increased the gain of PRCL, but the stability didn't change significantly. We decided to use REFL135I, which is used for PRMI3f lock, to control PRCL instead of REFL45I. The PRCL gain for the new signal was chosen to maintain the overall gain from POP17I to REFL135I. Soon after changing the signal, we could lock the DRMI in 26 seconds (Fig. 2). Something was wrong with POP17I.

Tuning gains, changing signals

Then we started further tuning of the gains and selecting the signals. We tried replacing POP17Q (1f) with REFL51Q (3f) for MICH control, and it seemed that 3f was better than 1f.  We also tried replacing POP17I (1f) with REFL51I (3f) for SRCL control, but in this case, 1f seemed better than 3f.

The final configuration of the selected signal, the control gain, and the upper and lower thresholds for each DOF are shown below:

With these conditions, we could lock DRMI in> 5 minutes (Fig. 3). We could barely measure the open-loop transfer function of each DoF, as attached (Fig.4 for PRCL, Fig. 5 for MICH, Fig. 6 for SRCL). Probably because of control instability, the measurement data have lower coherence, so it is not clear exactly what they looked like.

Issues

As reported here, the buildup of AS34 would be more than 5, whereas it is currently 0.5-0.6. It could be that the current locking point is not what we want due to the incorrect signs of the gains and the offset in the error signals. However, the POP90 buildup looked good (it's almost 1, as expected), suggesting the SRCL might be controlled at a weird point. On the other hand, the buildup during DRMI swings reached ~2 and never reached 5. It could be due to poor alignment of SRMI, but I'm not sure.

Another issue is the coupling between the DoFs. The open-loop transfer functions looked very weird in shape, especially in the 10s of Hz, and we suspect the couplings are the cause. At least, we know the demodulation phase of REFL51I is far from optimal (~7 deg rotated from the optimal angle, according to Ushiba-san). We have to tune the phase in SRMI or DRMI, and also tune the control gain again.

Fig.1 is OSEM signal timeseries in last night. The interval between cursors shows the duration when PR3 was in PAY_FLOAT state. As you can see, There seems to be the noisy timing and the quiet timing. 

In the noisy timing, the glitch occurres per 1-2 mins (fig.2). If this becomes noisy at once, the situation continues several hours.  On the other hands, any glitch seems to not occured over several hours if it is in quiet timing (fig.3).

Quick checks on the QPD signals at Pcal-X Rx module.

I checked the signals from a QPD that had been temporarily placed in the Pcal-X Rx module yesterday.
At that time, the Pcal beam had not yet reached the Rx module, so the QPD had not been properly aligned.

I requested ETMX to go to the LOCK_ACQUISITION state and switched Pcal-X to high-power mode.
A clear signal was observed in the QPD sum channel (Fig. 1).

I also moved the setpoints of ETMX TM pitch and yaw.
Although small responses were seen in the QPD X and Y signals, no clear position response was observed (Fig. 2).

After the above checks, the ETMX TM pitch and yaw setpoints were returned to their original values.

[YamaT, Oshino, Nakagaki, Hayakawa, Sawada, M.Takahashi, Omae, Yamaguchi, Ikeda]

K1IOO1 was restored at approximately 16:00 on June 2, 2026.

Details
A new 120-meter HIB cable and a 150-meter MTP cable were installed between the computer room and the K1IOO1 rack.
The HIB cable was replaced, and the restoration of K1IOO1 was successfully completed.
The MTP cable will be retained as a spare. If the HIB cable becomes unusable in the future, it can be used by switching to a V2 IO chassis.

[Takano, Tanaka, Fujimoto, Saito]

The beam profiles of the main laser beam and the sub-laser beam incident on the RFPD were measured and fitted. The resulting mode-matching ratio was approximately 5%.

• The RFPD and the 50 mm focal-length lens placed immediately before it were removed. The beam profiles of both the main laser beam and the sub-laser beam were then measured downstream of the beam sampler where the two beams are combined. The fitting results are shown in Figure 1. The green and orange curves represent the fitting results for the main laser beam, while the red and blue curves represent those for the sub-laser beam. The origin of the coordinate system is defined at the beam sampler where the two beams are combined.

  The waist positions and waist radii obtained from the fitting are summarized below.

  Main Laser

  x-direction: Waist position = -611.3 ± 79.7 mm, Waist radius = 0.1246 ± 0.0120 mm
  y-direction: Waist position = -504.6 ± 98.5 mm, Waist radius = 0.1450 ± 0.0199 mm
  → Average: Waist position = -558.0 mm, Waist radius = 0.1348 mm

  Sub-Laser

  x-direction: Waist position = -194.3 ± 14.7 mm, Waist radius = 0.0944 ± 0.0031 mm
  y-direction: Waist position = -217.6 ± 19.2 mm, Waist radius = 0.1101 ± 0.0045 mm
  → Average: Waist position = -206.0 mm, Waist radius = 0.1023 mm

  The mode-matching ratio between the two beams is given by 4×R_1×R_2/((Z_1-Z_2)^2+(R_1+R_2)^2), where Z_x and R_x (x=1,2) are the waist position and Rayleigh range of each beam, respectively. Using the average waist positions and waist radii obtained from the fitting, the mode-matching ratio was calculated to be approximately 5%.

[Hirose, Tanaka, Takano Ushiba, Fujimoto]

Abstract

We performed the phasing of AS34-I during PRMI 3f lock.
As a result, AS34-I improved from ~0.14 to ~0.18. The value measured with SRMI (f1 resonant) was also ~0.18.
The agreement between these two values is consistent with theoretical expectation.
From these results, the AS34-I value expected for DRMI is estimated to be ~5.8.

Details

While PRMI was locked with the 3f signals, we adjusted the demodulation phase so as to minimize AS34-Q:

     K1:LSC-AS-PDA2-RF34-PHASE-R: -90 -> -126

With this adjustment, AS34-I in PRMI with 1.1 W input power improved from ~0.14 to ~0.18.

In addition, AS34-I was also measured to be ~0.18 when SRMI was locked with f1 resonant.
For PRMI and SRMI, the theoretical transmissivity of the f1 sidebands to AS34 are 3.0% and 3.1%, respectively, assuming PRM reflectivity of 90%, MI reflectivity of 14% for f1, SRM reflectivity of 85%, and ideal Schnupp asymmetry, PRCL, and SRCL.
Since the two transmittances are almost the same, the agreement of the two measurement results is consistent with the theoretical expectation.

The expected transmissivity of f1 to AS when DRMI is locked is 97%.
Therefore, AS34-I in DRMI is expected to be amplified by a factor of about 32 compared to SRMI, resulting in an expected AS34-I value of ~5.8 for DRMI.

Ushiba, Takano, Fujimoto, Hirose, Tanaka

## Abstract

We found that PR3 IM OSEM H3 has large gilitch frequently. This glitch shakes PR3 largely, more than 10 urads and disturbs PRMI/SRMI lock.

## Details

As reported in klog36978, PR3 moved largely again during today's commissioning.

First, GAS LVDT signals seem to become glitchy at this moment. We turned off only GAS controls and monitored the local sensor signals. Although GAS LVDT signal seems to be not changed, PR3 moved largely (fig.1). So GASs are not gulity.

Second, We turned off TM OLDAMP controls after restoring GAS controls. Similarly, PR3 moved largely (fig.2). So, TM oplev seems to be innocent.

Third, we turned off IM DAMP. Actually, we requested PR3 guardian to be the PAY_FLAOT state. Then, we found that IM OSEM H3 has large glitch, ~10 cnts? in the view from K1:VIS-PR3_IM_OSEMINF_H3_OUT (fig.3). This time, TM oplev seems to be stable(fig.4). IM OSEM H3 seems to be used for IM DAMP T and Y. We turned on the PAYLOAD controls (actually we requested the guardian to be LOCK_ACQUISITION) and then we turned off the IM DAMP T and Y. In this state, TM OLDAMP started oscillate 1.15 Hz. Except for this 1.15 Hz oscillation, PR3 seems not to move largely (fig.5). 

Finally, we closed the laser shutter in order to confime whether this glitches are caused by the scattered light from IFO. The glitch is still in the OSEM signals even though the laser shutter was closed (K1:PSL-BEAM_SHUTTER = 0) (fig.6). This indicates this glitch comes from OSEM itself.

According to OSEM2EUL matrix, the coefficient from DGS cnts to urad? in Y direction is 5.8. So This 10 cnts glitch is relavant with ~58 urad.  

OSEM giltch seems to appear in only one direction, that is in negative direction in K1:VIS-PR3_IM_OSEMINF_H3_OUT. we suspect that somethin ex. OSEM LED intensity? or LED current? gets weak.

We left PR3 in the PAY_FLOAT state overnight to check how frequent this glitch occurrs.

> The sequence stopped at INCREASE_LCS_POW_FOR_OMC. At that time, the IMC TRANS output stayed around 8.1 W.

I fixed a bug in the INITIAL_ALIGNMENT guardian, and also changed the way to increase the input power. Now the guardian monitors the input power and increases it until it exceeds 8.5 W.

With Jiahui Xiong

Today, we checked the circuit and cabling for the QPD readout at Pcal-X.

First, we checked the circuit diagram: DocDB 9608 . From the diagram, this circuit seems to provide the 15 V power supply to the PDQ80A and to convert the QPD signals, X, Y, and Sum, from single-ended signals to differential signals.

Then we checked the setup at the Pcal-X area, and found the cable connecting the PDQ80A and the D-sub 9-pin connector.
We also tried to check the actual circuit, but it was difficult to remove the cables, so we could not directly inspect the inside of the circuit box.

However, after connecting the PDQ80A, we could confirm that the QPD signals were visible in the Pcal real-time model. Therefore, the basic signal readout seems to be working.

Although the alignment was not available today, we fixed one PDQ80A behind the mirror inside the Rx module as a temporary setup.

If possible, we would like to align the ETMX suspension tomorrow morning and adjust the position of QPDs.

I'm afraid that currently SMI ADS doesn't work welo, so DO NOT USE SRMI for aligning SRM until we confirm it works well. Instead, the conventional SRY ADS is fine for SRM alignment.

I offloaded the F3 GAS with the FR using the standalone stepper driver.

We compared two GIF animations (titled BEAM SPOT and MIRROR): one generated from two images acquired on 2026/04/22 (see 36793), in which the beam spot position was shifted, and another generated from two images acquired on 2026/06/01 (see 36984), in which the mirror position was changed. For the characteristic speckle-like structures indicated by the red circles, no apparent motion is observed in the BEAM SPOT, whereas a clear vertical displacement is observed in the MIRROR animation. This indicates that these structures are associated with the mirror.

With Shingo Hido

Summary

We performed the initial alignment for Xarm, Yarm, and OMC.

There were two points to be noted during the work:

• First, the IMC lost lock a few times. Most of the lock losses were recovered automatically, but once the IMC did not recover by itself. In that case, we manually changed the PZT offsets to recover the IMC lock, and then gradually returned all PZT offsets to 75.
• Second, the initial alignment sequence once stopped at INCREASE_LCS_POW_FOR_OMC. At that time, the IMC TRANS output stayed around 8.1 W. The PSL HWP had reached 169, while the REFL HWP was 152.266, slightly different from the target value. It is possible that the final fine adjustment of the HWP did not complete properly in that trial. After retrying the sequence, it passed this state and the alignment was completed.

Detials

Preparation

Before starting the initial alignment, we did the following checks and preparations.

• SRM and PRM were set to MISALIGNED.
• We confirmed that all Type-A suspensions could reach the LOCK_ACQUISITION state.
• The GRX and GRY shutters were open.
  • TR-GRX power was about 1.
  • TR-GRY power was about 1.
• PLLX and PLLY were confirmed to be in PLL_LOCKED.
• FIBX and FIBY were in SAFE due to the known DGS issue, but we proceeded.

Xarm alignment

For Xarm, the IX TCam image looked strange, possibly because of a different zoom or field of view. Therefore, we decided not to rely on the IX TCam image.

We requested the following states one by one:

• IRX_LOCKED
• GRX_LOCKED_WITH_IRX
• ALIGNING_XARM

During this process, the IMC lost lock about two times, but it seemed to recover automatically.

At ALIGNING_XARM, the GRX power was about 0.92, which was slightly low. We manually adjusted ITMX and ETMX as follows:

(IX-P, IX-Y, EX-P, EX-Y) = (9.4, -10.9, -12.0, -11.3) -> (8.4, -12.1, -12.8, -11.3)

After this adjustment, the GRX power increased to about 0.98. The IRX power stayed around 1.0.

Then we requested:

• RECORD_GOOD_VALUES_XARM
• OFFLOAD

Yarm alignment

For Yarm, we requested the following states one by one:

• IRY_LOCKED
• GRY_LOCKED_WITH_IRY
• ALIGNING_YARM

The alignment looked good, so we proceeded without manually adjusting the BS.

Then we requested:

• RECORD_GOOD_VALUES_YARM
• OFFLOAD

OMC alignment

For the OMC alignment, we first requested:

• IRY_LOCKED_FOR_OMC
• ENGAGE_OMMT2_TRANS_CENTERING
• ALIGNING_TO_OMC

The lock was lost during this process. We tried again by requesting:

• IRY_LOCKED_FOR_OMC
• ENGAGE_OMMT2_TRANS_CENTERING

but the lock was lost again, and this time the IMC did not recover automatically.

To recover the IMC, we manually changed the PZT offsets based on the DOF outputs.

• For DOF4-P:
  • The output had been around 23, but it was around -7.
  • We increased PZT1-P from 75 to 104.
• For DOF5-P:
  • The output had been around -46, but it was around -71.
  • We increased PZT2-P from 75 to 100.
• For DOF4-Y:
  • The output had been around -30, but it was around 0.
  • We started changing PZT1-Y from 75 toward 45.
  • During this adjustment, the IMC locked.
• DOF5-Y was not adjusted.

After that, IO reached PROVIDING_STABLE_LIGHT. We waited for a while and then gradually returned all PZT offsets to 75.

After the recovery, we requested IRY_LOCKED_FOR_OMC again. The IMC immediately lost lock once, but recovered automatically. The system soon reached IRY_LOCKED_FOR_OMC.

Then we requested:

• ENGAGE_OMMT2_TRANS_CENTERING

The sequence stopped at INCREASE_LCS_POW_FOR_OMC. At that time, the IMC TRANS output stayed around 8.1 W. We checked the HWP values and found:

PSL HWP  = 169REFL HWP = 152.266

The REFL HWP was still 0.266 away from the expected value, so the final fine adjustment may not have completed properly.

We retried the sequence again. Then, the sequence passed this state. In the successful trial, the HWP values were:

PSL HWP  = 169REFL HWP = 152

Then we requested:

• ALIGNING_TO_OMC

The OMC transmitted power, OMC-TRANS_DC_SUM_OUT16, reached about 30.5. The ground motion became somewhat noisy during this process, but it calmed down after a while, and the guardian state was successfully reached.

Finally, we requested:

• RECORD_GOOD_VALUES_OMC

The initial alignment up to OMC was completed.

[Fujimoto, Tanaka, Takano]

We could lock DRMI with a mixture of 3f and 1f signals. Currently, the lock is unstable, and it keeps locking up for several minutes. Further investigation is necessary.

• Fig. 1: When DRMI was locked
• Fig. 2: DRMI got unstable due to a glitch (probably the control instability)
• Fig. 3: measured OLTF for PRCL
• Fig. 4: measured OLTF for PRCL
• Fig. 5: measured OLTF for PRCL
• Fig. 6: Filter settings
• Fig. 7: Trigger settings 

The details will be reported later.

Summary

The modulation phase of POP17 was optimised using SRMI.

Detail

For a long time, we haven't tuned the demodulation phase of POP17. We optimised it by checking the POP17I and POP17Q signals when SRMI was locked using them (POP17I for SRCL and POP17Q for MICH).

We injected the excitation signal at 50 Hz, which would be outside of SRCL control band, at K1:LSC-SRCL1_EXC, and measured the ASD of SRCL_IN2 and MICH_IN2. We should not see any signal in MICH, i.e. POP17Q, so we changed the demodulation phase of POP17 and minimised the excess noise in MICH_IN2. Fig. 1 shows the results of the optimisation; the injected noise in MICH_IN2 was minimised when the modulation phase was set to -141 deg (by default it was -137 deg).

[Ushiba, Tanaka, Fujimoto, Saito]

Following klog:36974, we attempted to achieve lock by adding an integrator to the low-pass filter, but locking was not successful. We also tried changing the feedback sign, changing the frequency relationship between the main laser and the sub-laser, and using the low-pass filter of SR560 instead of Moku:Lab. However, none of these approaches resulted in lock acquisition. In addition, the measured beat signal amplitude was approximately 0.2 mVpp, whereas the value predicted from the laser powers was approximately 0.1575 V, representing a large discrepancy. Therefore, in the next experiment, we plan to investigate the alignment, mode matching, and polarization conditions to determine whether the beat signal amplitude can be increased.
 

• First, a 10 kHz low-pass filter was applied, and the gain was gradually increased. Oscillation occurred once the gain exceeded a certain level. We then inserted an integrator with a corner frequency of 1 Hz while keeping the gain just below the oscillation threshold and tested whether lock could be achieved. However, no behavior indicating that the beat frequency was being pulled toward the LO frequency was observed. In addition, when the integrator was enabled, the beat frequency was observed to move further away from the LO frequency. We also changed the feedback sign and tested both cases in which the sub-laser frequency was higher than and lower than the main laser frequency, but lock could not be achieved in any configuration. Furthermore, we attempted to use the low-pass filter of an SR560 instead of Moku:Lab and varied the gain, but lock acquisition was still unsuccessful.
   

• In addition, when the gain was increased with the 10 kHz low-pass filter in place, the frequency of the error signal at the onset of oscillation was approximately 900 kHz. When the beat signal was observed on the oscilloscope, its amplitude exhibited temporal fluctuations, and the measured frequency was also approximately 900 kHz. This amplitude-modulated signal disappeared when the sub-laser beam was blocked, but remained unchanged when the main laser beam was blocked. The same 900 kHz signal was also observed on the spectrum analyzer. Its level was approximately -43 dB with the noise eater turned off and decreased to approximately -80 dB when the noise eater was turned on. However, changing the state of the noise eater did not enable lock acquisition.

• The beat signal amplitude, converted from the spectrum analyzer reading in dBm, was approximately 0.2 mVpp. At present, the sub-laser power incident on the RFPD is approximately 1.78 mW, while the main laser power is estimated to be approximately 17 μW. Since the responsivity of the RFPD (1611FS-AC) is approximately 525 V/W, the expected beat signal amplitude is (525V/W)×2√(1.78×10^−3×17×10^−6) which is approximately 0.1575 V. This predicted value is about 788 times larger than the measured amplitude. Therefore, in the next experiment, we plan to investigate the alignment, mode matching, and polarization conditions to determine whether the beat signal amplitude can be increased.

[Kimura and Yasui]
 On the afternoon of June 1, we filled two helium compressors (EYC P-53 and EYC P-55) with G-1 grade helium gas to a pressure of 15 bar.
As a result, maintenance work on the EYC cryocooler for the radiation shield has been completed.

[Kimura, Yasui, Tanaka (Edwards) and Okada (Edwards)]
 On the morning of June 1, two Edwards technicians conducted an inspection of the Gamma ion pump power supplies used on the X and Y arms.
This inspection involved connecting a high-resistance dummy pump to the power supply’s high-voltage terminals and verifying that the voltage (~6 kV) and current (~60 μA) were being output stably. (Photos 1~3)
As a result, it was confirmed that the five Gamma ion pump power supplies used on the X and Y arms were operating without issues.
 Subsequently, the Edwards technicians inspected the Gamma ion pump power supply at the X-end that had generated an error. (k-log 36915)
(Photos 4~5)
No loose connections were found in the wiring forming the interlock.
Therefore, it was decided to monitor the situation for the time being.