[Ushiba, Smith, Hirose, Saito]

The optical layout was slightly modified to make the alignment procedure easier. The beam profile of the sub-laser was measured, and the beam waist position and waist radius were determined. Comparing these values with those of the main laser measured in klog:36730, the mode-matching ratio was estimated to be approximately 89%. In addition, the alignment of the sub-laser beam incident on the SRC was performed using two irises. When the sub-laser PZT was driven, flashes were observed with the OMC REFL PD (K1:OMC-REFL_DC_OUT_DQ). By adjusting one mirror to maximize the signal while the PD gain was set to 40 dB, the maximum signal reached approximately 478 counts. Further improvements in alignment and mode matching are expected to increase the signal to approximately 1700 counts.

• First, the optical layout was modified slightly to facilitate alignment (Figure 1). Using the sub-laser beam, the alignment was adjusted to avoid beam clipping. Next, the beam profile of the sub-laser was measured at locations far from the beam waist using a beam profiler, and a fit was performed (Figure 2). The resulting waist positions and waist radii were as follows:

  Sub-laser

  x-direction: Waist position = 290.1 ± 9.8 mm, Waist radius = 0.0728 ± 0.0025 mm
  y-direction: Waist position = 237.5 ± 8.9 mm, Waist radius = 0.0686 ± 0.0022 mm
  →Average: Waist position = 264 mm, Waist radius = 0.071 mm

  For comparison, the main-laser beam profile measured in klog:36730 was used. The results were:

  Main laser

  x-direction: Waist position = 264.3 ± 5.1 mm, Waist radius = 0.0526 ± 0.0025 mm
  y-direction: Waist position = 253.6 ± 2.3 mm, Waist radius = 0.0555 ± 0.0013 mm
  →Average: Waist position = 259 mm, Waist radius = 0.054 mm

  Using these results, the mode-matching ratio was calculated to be approximately 89%.
   

• Next, two irises were used to ensure that the optical paths of the main laser and the sub-laser coincided. The alignment was adjusted using the BS and a nearby mirror. The main laser was then blocked so that only the sub-laser beam was present in the SRC. The sub-laser PZT was driven with a sinusoidal signal generated by the Moku:Lab function generator, and the OMC REFL PD (K1:OMC-REFL_DC_OUT_DQ), which had been installed by Tanaka-san(klog:37057), was monitored for flashes. Initially, no flashes were observed. However, after adjusting the mirror associated with the BS, flashes became visible (Figure 3). By adjusting one mirror to maximize the signal, the signal amplitude oscillated between approximately 305 counts and 478 counts.
   

• Assuming that the optical power of the sub-laser incident on the SRC is 500 mW, the optical power reaching the OMC REFL PD at resonance is expected to be approximately 16 μW. The responsivity of the OMC REFL PD (Thorlabs PDA100A2) is approximately 680 V/W, and the PD gain was set to 40 dB. Therefore, the output voltage at resonance is expected to be approximately 1.1 V. Using the calibration factor of 610 μV/count, this corresponds to approximately 1700 counts at resonance. Therefore, by using the second mirror to further improve the alignment, the signal should be increased further.

KAGRA Pcal-Y updates (2026/06/11)

Workers: Dan Chen, Misato Onishi, Seiya Matsuo

We performed monthly Pcal-Y calibration on 2026/06/11.

After the calibration, we updated EPICS parameters related to the Pcal-Y system. No issues were found.

Detailed report.

Before starting the alignment work, we confirmed that all Type-A suspensions went to LOCK_ACQUISITION.

Xarm alignment

We followed the MEDM procedure for the Xarm alignment. First, we requested IRX_LOCKED, and then requested GRX_LOCKED_WITH_IRX. At this step, it took some time to lock the arm. Therefore, PR3 was slightly moved to improve the locking condition.

After that, we requested ALIGNING_XARM. Since the GRX transmitted power was not good enough, we manually adjusted ETMX and ITMX. As a result, the transmitted powers became approximately GRX ~ 0.96 and IRX ~ 0.95.

Then, we requested RECORD_GOOD_VALUES_XARM and performed OFFLOAD.

Yarm alignment

We also followed the MEDM procedure for the Yarm alignment. We requested IRY_LOCKED, GRY_LOCKED_WITH_IRY, and then ALIGNING_YARM. The alignment condition was already not bad, so we did not perform any manual adjustment.

Then, we requested RECORD_GOOD_VALUES_YARM and performed OFFLOAD.

OMC alignment

For the OMC alignment, we followed the MEDM procedure as well. We requested IRY_LOCKED_FOR_OMC and then INCREASE_LAS_POW_FOR_OMC. We waited until the Guardian comment showed READY.

Then, we requested ALIGNING_TO_OMC. We monitored K1:OMC-TRANS_DC_SUM_OUT16. The value was around 30, which was acceptable. We note that a value larger than about 60 may be dangerous for the OMC DC PD.

Finally, we requested RECORD_GOOD_VALUES_OMC. After recording the good values, the OMC alignment state automatically went down.

PRMI alignment

For the PRMI alignment, we requested ALIGNING_PRMI following the MEDM procedure. At this step, ADS was engaged for IMMT2, PRM, and BS.

After the alignment became stable, we requested RECORD_GOOD_VALUES_PRMI and then performed OFFLOAD.

With Yokozawa-san

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

[Smith, Hirose, Saito]

The optic that was identified as a BS in the previous experiment (klog:37052) was replaced with a mirror. In addition, while attempting to correct beam clipping on the edge of a mirror, it was found that the beam was not passing through the center of the 150 mm focal-length lens. Since the beam was laterally displaced, an attempt was made to move the lens sideways. However, due to the mounting configuration, the lens could not be translated laterally. Modifying the mount to allow lateral adjustment would likely cause interference with a nearby mirror and shift the lens from its original position. Therefore, it will be necessary to move the mirror instead. Furthermore, because two lenses were removed during the alignment procedure, mode matching must be redone. In addition, the short distance between the 150 mm focal-length lens and the nearby mirror, as well as the difficulty of adjusting the mirror near the lens, have led us to consider redesigning the optical layout.
 

• First, the optic at Spot 1 in the photograph, which had been identified as a BS in the previous experiment (klog:37052), was replaced with a mirror. (Please note that although a mount labeled "NON POL 50:50" was used, the optic itself is a mirror (HYD-10).) The alignment of the main laser was then readjusted, and the alignment of the sub-laser was performed using two irises.
   
• However, when the sub-laser power was measured at Spot 2 in the photograph, it was found to be approximately 830 mW. Therefore, Spot 3 was inspected, and it was discovered that the beam was clipping on the edge of a mirror. The alignment was then adjusted so that the main laser beam was centered on the mirror. However, under these conditions, when the beam was adjusted to pass through the center of the 200 mm focal-length lens, it clipped on the side edge of the mirror at Spot 4. To correct this, the alignment was adjusted using mirrors located upstream of the two lenses so that the beam would pass through the centers of both the lenses and the mirrors. However, this was not successful. It was suspected that the presence of the lenses was making the alignment process difficult, so the two lenses were temporarily removed. With the lenses removed, the beam was aligned to pass through the centers of the mirrors, and the 150 mm focal-length lens was then reinstalled. However, the beam became laterally displaced. An attempt was made to adjust the lateral position of the lens, but this was not possible because of the way the lens was mounted. Furthermore, modifying the mount to allow lateral adjustment appeared impractical because the lens was located very close to a mirror, making it difficult to reinstall the lens in its original position.
   
• Therefore, it will be necessary to move the mirror located before the lens. In addition, because the two lenses were removed, mode matching must be performed again. Considering this, together with the short distance between the 150 mm focal-length lens and the nearby mirror and the difficulty of adjusting the mirror in that region, we are considering redesigning the optical layout.

To align the Aux. beam axis with SRC axis more precisely, we would like to use the flash transmitted from SRC. We estimated the flash power on AS PD when 1W Aux. input power was injected from the SR2 AR side. The power just after trasmitting SRM seems to be ~65 uW and this beam is transmitted to AS port via OSTM (T = 3%). Then, the estimated power on AS DC PD is ~2 uW. Therefore, it maybe too low to obtain the flash with AS PD.

So we put new DC PD (Thorlabs, PDA100A2) in OMC REFL path for monitoring the flash instead of AS PD. The channel for OMC REFL DC is already existed (maybe it was used previously but now it is not used.). So I connected the new DC PD to the channel (K1:OMC-REFL_DC). I confirmed the PD response with DGS.

However, the alignment work on POS table seems not to be completed (Maybe Saito-kun reported detail later). So I put the beam dumper in front of the PD and blocked the beam by the dumper for now. 

Some notes on the measurement setup:

1. For the single-bounce configuration, the following suspension states were requested:
PRM: MISALIGNED_BF
ITMX: MISALIGNED_BF
ETMY: MISALIGNED
SRM: MISALIGNED
All others: LOCK_ACQUISITION

2. In the case of locking with 7.8 mW and 16.2 mW on DC PD B, OMMT2 and OSTM were actively controlled to optimize their alignment using the in-vacuum OMC QPDs, whereas no alignment control was applied during locking with 1.6 mW.
This is because the alignment control using the in-vacuum OMC QPDs did not work well due to the very low power on the OMC QPDs; however, this is expected to have little impact on the results.

After the calibration work, we moved the RxPD back to its nominal position and found that its output had changed from the pre-work value. We investigated the cause and found that the alignment between the integrating sphere and the plate underneath it had slightly shifted. We realigned the integrating sphere with the plate, and the RxPD output returned almost to its original value.

[Tanaka, Hirose, Saito]

A 10 dB attenuator was added, reducing the beat signal level from 23 dBm to 10 dBm. The open-loop transfer function of the PLL was then measured. The results suggest that the sub-laser PZT responds with a frequency change to the frequency-difference signal from the PFD, rather than responding with a frequency change to a phase-difference signal. In addition, the alignment of the sub-laser beam incident on SRY was performed using two irises. The sub-laser PZT was then driven, but no flashes were observed at the AS port. This is likely because one of the optics expected to be a mirror was actually a BS, the beam was clipping on the edge of a mirror, and the observation was being made at the AS port. In the next experiment, the BS will be replaced with a mirror, the alignment will be redone, and a PD will be installed at the OMC reflection port to observe the flashes.
 

• First, a 10 dB attenuator was added to reduce the input power to the PFD. As a result, the beat signal level decreased from 23 dBm to 10 dBm. Next, the 1 MHz low-pass filter in the SR560 used for feedback to the sub-laser PZT was replaced with either a 100 kHz or a 10 kHz low-pass filter. Under these conditions, the approximately 260 kHz oscillation previously observed in the error signal (klog:37046) disappeared. The open-loop transfer function was then measured, revealing a flat region. This was unexpected. We had assumed that the feedback loop operated by changing the sub-laser frequency via the PZT to eliminate the phase difference, which should behave as an integrator. However, the observation is consistent with the PFD producing a frequency-difference signal and the sub-laser PZT responding with a frequency change. Therefore, a 10 kHz low-pass filter was applied in the SR560 feedback path to the sub-laser PZT, and the gain and integrator cutoff frequency in Moku:Lab were varied. The measured open-loop transfer function then closely matched the designed filter response (Photo 1).
   

• Next, the mirrors were adjusted so that the main laser beam followed the optical path intended for the sub-laser beam entering the interferometer. Two irises were then installed along this path and aligned to the main laser beam. The sub-laser alignment was subsequently adjusted so that it also passed through the two irises. The main laser was then blocked, leaving only the sub-laser beam incident on SRY. The sub-laser PZT was driven with a triangular waveform generated by the Moku:Lab function generator, and the AS port was monitored for flashes. However, no flashes were observed. This is likely because the optical power incident on SRY was too low. The laser powers measured at Spots 1~5 in Photo 2 were as follows:

  Spot 1: approximately 130 mW
  Spot 2: approximately 390 mW
  Spot 3: approximately 780 mW
  Spot 4: approximately 1.0 W
  Spot 5: approximately 1.1 W

  From these measurements, the laser power incident on SRY is estimated to be approximately 130 mW, which corresponds to only about 0.25 μW at the AS port. Such a small signal is likely buried in noise. The reduction in laser power between Spot 2 and Spot 1 occurred because the optic between them was a BS rather than a mirror. In addition, the reduction in power between Spot 4 and Spot 3 may be due to the beam clipping on the edge of the mirror.
   

• Therefore, in the next experiment, the BS between Spot 2 and Spot 1 will be replaced with a mirror, and the alignment will be adjusted so that the beam does not clip on the edge of the mirror between Spot 4 and Spot 3. However, even after these modifications, the power at the AS port is expected to increase only to approximately 0.97 μW. Therefore, a PD will be installed at the OMC reflection port, and flashes will be monitored there instead.

KAGRA Pcal-X updates (2026/06/09)

Workers: Dan Chen, Jiahui Xiong, Misato Onishi

We performed monthly Pcal-X calibration on 2026/06/09.

After the calibration, we updated EPICS parameters related to the Pcal-X system. No issues were found.

A CAL Tcam session was performed to obtain beam position information necessary for Pcal. The parameters have already been updated, and SDF has been accepted.

Operator: Dan Chen

Update Time: 2026/06/09 05:54:18

Update Time: 2026/06/09 05:54:58

[Tanaka, Hirose, Saito]

Using the 20 dB RF amplifier and the 40 dB RF amplifier, the beat signal amplitude was increased to 23 dBm. After replacing the phase detector with a Phase Frequency Discriminator (PFD), it was confirmed that the beat signal frequency followed changes in the LO frequency. However, the error signal appeared to oscillate at approximately 260 kHz. This oscillation may be caused by a resonance of the sub-laser PZT.
 

• First, the alignment was adjusted to maximize the beat signal. The signal level under various RF amplifier configurations was as follows:

  No RF amplifier: -51 dBm
  40 dB RF amplifier only: -5 dBm
  20 dB RF amplifier only: -20 dBm
  40 dB RF amplifier followed by the 20 dB RF amplifier: 8 dBm
  20 dB RF amplifier followed by the 40 dB RF amplifier: 23 dBm
   

• Therefore, lock acquisition was attempted using the phase detector with the 20 dB RF amplifier followed by the 40 dB RF amplifier. A 1 MHz low-pass filter was applied in the SR560 used for feedback to the sub-laser PZT, while a flat filter was implemented in Moku:Lab. Various gains and integrator settings were tested. However, no behavior indicating that the beat frequency was being pulled toward the LO frequency was observed.
   

• Next, the phase detector was replaced with a PFD. With a filter consisting of an overall gain of -6 dB and an integrator providing 6 dB of gain at 1 Hz (Photo 1), it was observed that the beat signal frequency followed changes in the LO frequency. In the lower plot of Photo 1, the red trace represents the error signal and the blue trace represents the feedback signal. Since the feedback signal changed when the LO frequency was varied, the LO frequency and beat signal frequency were measured for feedback signal levels of approximately 0 V and ±1 V. The beat signal spectrum exhibited a shape with a dip at the center and peaks on both sides (Photo 2). Therefore, the frequencies of the two peaks were measured, and their average was taken as the beat signal frequency. In Photo 2, the light red trace corresponds to a feedback signal of approximately +1 V, while the red trace corresponds to a feedback signal of approximately -1 V.

  The results are summarized below:

  Feedback Signal           LO Frequency    Beat Signal Frequency
  Approximately -1 V      85.066 MHz         85.16 MHz
  Approximately  0 V      88.066 MHz         87.99 MHz
  Approximately +1 V    89.666 MHz         89.74 MHz
   

• In addition, when using a flat filter and gradually increasing the gain, the error signal began oscillating at approximately 260 kHz, and the width of the beat signal increased. This oscillation may have been caused by a resonance of the sub-laser PZT. The corresponding error signals and beat signals are shown below:

  Flat filter with a gain of -6 dB (Photo 3), corresponding beat signal (Photo 4)
  Flat filter with a gain of 0 dB (Photo 5), corresponding beat signal (Photo 6)
  Flat filter with a gain of +6 dB (Photo 7), corresponding beat signal (Photo 8)

Date: 2026/06/08

Member: Dan Chen, Misato Onishi, Seiya Matsuo

We performed our usual WSK calibration at UToyama.

The results look no problem.

Results

Comparison with Previous Results

Comparing with previous results, no significant issues were found.
Attached graph is the result summary including the latest measured data.

I offloaded the BF GAS with the FR.

[Tanaka, Fujimoto, Saito]

The alignment was adjusted to maximize the beat signal, resulting in a beat signal level of approximately -55 dBm, which was the same as in the previous measurement (klog:37031). It was also confirmed that the sub-laser PZT was operating properly. The mixer currently in use (ZX05-1-S+) was then replaced with a phase detector (ZRPD-1+) purchased by Dan-san, and lock acquisition was attempted. However, no lock was achieved. Furthermore, when an RF amplifier was inserted after the high-pass filter while using the phase detector, the signal was amplified without waveform distortion, unlike the behavior observed in the previous experiment (klog:37031).
 

• First, the alignment was adjusted to maximize the beat signal. The resulting beat signal level was approximately -55 dBm. Next, a signal from the Moku:Lab function generator was applied directly to the sub-laser PZT, and the corresponding movement of the beat signal was confirmed. Therefore, the PZT was verified to be functioning properly.
   

• Next, the 1.9 MHz low-pass filter located between the mixer and the SR560 was moved from immediately before the SR560 to immediately after the mixer. PLL lock acquisition was then attempted using Moku:Lab with a 10 kHz low-pass filter while varying the gain and adding an integrator. However, no behavior indicating that the beat frequency was being pulled toward the LO frequency was observed. Lock acquisition was also attempted without using Moku:Lab by varying the gain of the SR560 low-pass filter, but no lock was achieved. In addition, a 10 kHz low-pass filter was applied to the SR560 used in the feedback path to the sub-laser PZT, while a flat filter was implemented in Moku:Lab. However, lock acquisition was still unsuccessful. The mixer currently in use (ZX05-1-S+) was then replaced with the phase detector (ZRPD-1+) purchased by Dan-san. Various Moku:Lab filter configurations were tested in the same manner as before, but lock acquisition was again unsuccessful.

• Next, while continuing to use the phase detector, an RF amplifier was inserted after the high-pass filter. Unlike the result obtained in the previous experiment (klog:37031), the signal amplitude increased while maintaining a clean waveform (Photo 1). The error signal also increased in amplitude and became saturated. Therefore, the gain of the SR560 was reduced from 200 to 5. Under these conditions, the error signal amplitude was approximately 384 mVpp (Photo 2).