(Uchiyama, Miyakawa)

Ushiba-kun reported that GRY seemed to have mode hops, which introduced some fluctuation in the transmitted power of GRY.

We entered the PSL room and changed the crystal temperatures for all IR, GRX, and GRY. Yesterday, we had to reduce the temperature of both green by about 1.5 degrees in order to increase the temperature of IR by about 0.5 degrees. Today, we increased the IR temperature by 1.0 degree, from 39.4 to 40.4 degrees, and the greens followed: from 28.8 to 30.0 for X, and from 25.9 to 27.0 for Y. These green temperatures are close to the known values at which GRY and GRX had no mode hops. 

During the afternoon check, Ushiba-kun reported that there seemed to be no mode hops in either of the green lights.

Abstract:

I tuned PRC ASC and it seems to work well now, at least short term.
To check the long-term stability, I will keep PRFPMI_RF_LOCKED state today.

Detail:

As reported in klog35355, I investigated ASC.

First, I checked the phasing of REFL RF45 signals by injecting 13.125Hz sine signals from PRCL error point.
To avoid the unwanted coupling from DARM and MICH, I added resonant gain at FM4 of DARM2 and MICH1, which was enabled during the measurement.
Phasing seems fine except for REFL QPD1 RF45 SEG1, so I only change REFL RF45 QPD1 SEG1 phase from 68.396 to 108.396 (40 deg rotation).
Figure 1 shows the spectra of each segment after phasing, and all I signals have larger signals than Q signals.

Then, I measured the spectrum of REFL 17/45 QPD1/2 PIT/YAW and POP 17 QPD1/2 PIT/YAW signals to check which signals can detect each suspension motion largely.
Following tables are the rough summary of the measured data (original data can be found at /users/Commissioning/data/ASC/2025/1024/ with the file name of WFS_*).

Above values are the TF measurement from oplev signals of each suspensions when exciting to each WFS signals.
For PIT, REFL_QPDA2_RF17_I is not decoupled well for IMMT2 and PR3 but I ignored it because PR3 ASC gain is much higher than that of IMMT2, which reduce the effect of coupling.
For YAW, REFL_QPDA1_RF45_I is not decoupled well for PR3 and BS, so I decoupled it by adding REFL_QPDA1_RF45_Q signals to I signals.

Based on these measurement, I set the matrix element for PRC1 as shown in fig1 (PIT) and 2 (YAW).
The absolute value of the matrix element was decided so that the WFS signals are roughly calibrated in the unit of urad of PR3 angular motioin.
I also set the output matrix element from ORC1 to PR3 as shown in fig3 and 4.

Then, I modified the filter bank of PRC1_{P,Y} to achieve the 0.6Hz UGF loops for PRC ASC as shown fig 5 (FM1,2,3 was added for adjusting gain and the others were not changed because the integrator was already implemented).

After the ASC tuning, ASC seems stable at least for a short term.
So, I will keep IFO in PRFPMI_RF_LOCKED state to check the stability of new ASC loop.

Since PRFPMI_RF_LOCKED seems stable, I tried to lock PRFPMI_RF with 15W.
However, lockloss happened many times during increasing laser power, which can be passed in the previous time, so I gave up to increase the laser power today.
One possibility is glitches when changing IMC CMS IN1 gain during increasing laser power but not so sure.
Further investigation is necessary.

Note:

Since we replaced CMS (klog35394), it might be due to the indevisual difference of CMS.
So, we need careful check of the difference of the signals from the previous trial.

After replacement of IMC CMS, we haven't tuned IMC CMS cmp gain and IN2 gain, I tuned them.

Since IMC CMS IN1 gain was changed from 16 to 4 due to the change of CMS, I reduced IN2 gain from 0 to -12 dB as shown in fig1.
In addition, to effectively activate ofs comp function, I set ofs comp gain of 3.358 at K1:IMC-SERVO_IN2GAIN_OFS_COMP_GAINNEG12_OFS.
Since the other gains are not tuned, I set a gain of zero for all the other gain setting (fig2).

After that, I measured OLTF of ALS_CARM (fig1).
UGF is around 2kHz, so it seems fine.
Then, I tuned the crossover frequency between MCL and laser frequency, which is now 100Hz (fig4).
For the adjustment, I added the gain of 1.5dB at FM1 of MCL filter bank as shown in fig5.

In addition, IMC CMS IN1 gain comp ofe was measured from +4dB to -20dB for the laser power up.
Same as IN2 comp offset, I set zero where I did not measure this time as shown in fig6.

I also adjusted the common offset (K1:IMC-SERVO_OFS_COM_CALI_OFFSET) so that IMC ASC feedback signals are not saturated.
The new value is 0.936.

Operator's name: Sawada, Miyakawa

Shift time: 9-17 JST

Check Items:
VAC: No issues were found.
CRY cooler: No issues were found.
Compressor: No issues were found.
Temp check: There is a temperature change at 4PM in FIELD_IYA (-0.40 deg)
VIS GAS filter output check: No issues were found.

IFO state (JST):
9AM Start of shift, STANDBY remains
5PM End of shift, LOCKING

There were 4 lock loss of IMC after working yesterday in 20 hours. 3 of 4 were kicked by seismic motion.

It seems to be not bad continuously.

EX duct shield cryo-cooler BRT side:
 The cold head of the EX BRT-side duct shield cryocooler is showing a gradual temperature rise trend.
Consequently, the temperature of the BRT-side duct shield is also showing a gradual rise trend.
It is necessary to monitor whether the temperature of the BRT-side duct shield reaches 150 K. 
 

Date: 2025/10/23

Member: Misato Onishi, Keisuke Sakanoue, Dan Chen

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.

(Uchiyama, Miyakawa)

We entered the PSL room and adjusted the laser crystal temperature to lock PLLs and to avoid mode-hop.

First, we changed the IR temperature from around 25 degrees to 38.7 degrees. This temperature was adjusted to the current green temperatures to lock the PLL, but the PMC transmitted power was fluctuating, and it indicated mode-hop of IR. We again changed the IR temperature to 39.3 degrees, and both green temperatures from 30.2 to 28.7 degrees for X and 27.2 to 25.8 degrees. IR temperature changed to higher, but it hopped to another mode, so the actual frequency got lower, and green temperatures were matched to the lower frequency.

Then we increased the IR current from 6.4A to 7.7A, which produces ~30W output power. We aligned the PMC again, and the transmissivity reached ~88.7%.

It seemed to be very stable for 16 hours last night!

[Kimura and Yasui]
 We have examined the temperature changes over the past few weeks for each cryostat's duct shield cryo-cooler.
Please see attached graphs.
From the graphs:
IX duct shield cryo-cooler BS side:
Temperature increase trend continued but shifted to a decrease about 4 days ago.
 

EX duct shield cryo-cooler BRT side:
Shows a slow temperature increase trend.
The temperature of the EX duct shield cryo-cooler BRT side has reached a state requiring  
 

IY duct shield cryo-cooler arm side:
Temperature is stable except for a peak about 10 days ago. It appears unrelated to the compressor oil temperature rise.
 

EY duct shield cryo-cooler arm side:
No particular changes. It also appears unaffected by the cooling water trouble on October 15th.

 

Ushiba, Tanaka

Abstract

We repleced IMC CMS to gain modified one and also replaced EOM amlifier to x20 gain one in order to enlarge the actuator range. For now, IMC seems to keep the lock for 2 hours. We left IMC lock to check the stability. 

What we did

Replacement IMC CMS to the modified one

We brought a modified IMC CMS by Moriwaki-san to a mine. Moriwaki-san modified a gain of a slow gain filter in a SLOW PATH of IMC CMS from -0.1 to 1 and also modified a gain in FAST PATH of IMC CMS from 1 to 10 (The detail modification is reported in klog35396).

Before the replacement, we confirmed some TFs of the original IMC CMS. Fig. 1 and Fig. 2 show OLTF of the IMC loop and TF between EOM and PZT loops, respectively before the replacement, Current UGF and cross-over frequency are 123.7 kHz and 14.3 kHz, respectively.

Then, we replaced the CMS from the original one to the modifed one. After that, we modifed the CMS setting and the related script in IMC guardian as follows.

• K1:IMC-SERVO_SLOWFILTER: ON -> OFF 
• K1:IMC-SERVO_SLOWPOL: - -> + (to compensate the sigh flip due to the turn off the slow filter)
• K1:IMC-SERVO_IN1GAIN: 24 dB -> 4 dB (to compensate the gain increase of both slow and fast path by x10)

We confirmed IMC could be locked with this setting. However, since the offset was changed due to the CMS replacement, the normalized trans. power did not reached to ~1. Ushiba-san adjusted K1:IMC-SERVO_IN1GAIN_OFS_COMP_GAIN4_OFS not to change the value of K1:IMC-SERVO_MIXER_DAQ_OUT when IMC CMS IN1 opens and closes when IN1GAIN was 4 dB, and then tweaked K1:IMC-SERVO_OFS_COM_CALI_OFFSET to maximize IMC trans. power. (So you should perform the same adjustment if you want to change IN1GAIN from 4dB.) At any rate, IMC trans. power was maximized.

At last, we measured TFs after the replacement. Fig. 3 and Fig. 4 show the OLTF and the TF between EOM and PZT loops, respectively, after the replacement. There seems to be no significantly strange point in OLTF. On the other hand, there seems to be a small dip around the cross over frequency in the TF between EOM and PZT loops compared with before. However, the phase seems to be large enough around the cross over frequency. So they may be fine.

Replacement of the EOM amplifier from the x10 one to the x20 one

To enlarge the actuator range, we replaced an amplifier for a wideband EOM from x10 one (F10A, FLC electronics) to x20 one (made by Miyoki-san, klog35274). Miyoki-san told us the installation procedure of this amplifier. I followed the procedure.

1. confirm that a dial on a regulator (DP1001), which is used as a 200V power supply, is 0. Then, connect a plug of the regulator to a power tap. 
   1. This time, I connected the plug to UPS in PSL
2. set a current limit of the regulator to 60 mA.
3. connect a plug for a fan in the amplifer to a power tap.
   1. This time, I connected the plug to the other power tap, not UPS.
4. Turn on the regulator. And confirm 15 V are shown in a indicator on the regulator. Then, confirm that +/- 15 V is applied as an output voltage with a multimeter or something.
5. At once, turn off the regulator before cabling
6. Connect the regulator output port to a DC input port behind of the x20 amplifer box
7. Connect the F30PV output port to the signal input port on the amplifer box
8. Connect the signal output port on the box to EOM
9. Turn on the regulator
10. increase the applied voltage up to 190 V by rotating the dial on the regulator. (If the lamp on the upper right in the front panel of the regulator illuminated, reduce the voltage to 0 immediately. Then, contact with Miyoki-san)

Fig. 5 shows the amplifer box and Fig. 6 shows the regulator after the installation. Following this, we modified the CMS setting and the related script in IMC guardian. Because Miyoki-san's amplifer is an inversed amplifier, we flipped the sign of K1IMC-SERVO_FASTPOL from + to -. And we reduced K1:IMC-SERVO_FASTGAIN by 6 dB to compensate the gain increasing by x2 (=20/10) due to the replacement. After that, we confirmed IMC could be locked. Moreover, the red light did not illuminate during the lock acquisition and the lock.

Then, we measured the OLTF after the EOM amp. replacement. Fig. 7 shows the results. The gain seems to be not changed. On the other hand, the phase was rotated slightly by Miyoki-san's amplifier. However, since the phase margin is large enough at UGF, the loop is fine. (This time, UGF was not changed after the EOM amp. replacement so we did not measure the TF between EOM loop and PZT loops.)

---

IMC lock is keeping for 2 hours for now. We left IMC lock this night to check the stability.

(Uchiyama, Miyakawa)

We swapped the master laser inside the box of neoLASE.

Preparation was done yesterday, so we started shutting down the power amplifier and confirmed the beam position on the camera. We shut down the master then, and loosen the screws to remove the aluminum board with the laser head. After removing the board, we kept the head position relatively using bars and clamps (Fig.1). After swapping the master head to another one, which was used in another neoLASE at Toyama Univ., we put the aluminum board with the new head into the box. However, we noticed that a big misalignment (~2 mm off at the EOM hole) had happened. It seems that some washers were used like a shim to adjust the angle of position of laser head, but it was too difficult to recover the same setting. So we decided to align using mirrors inside the box. This trial might destroy the alignment to the amplifier, but we had a recorded position on the camera, and we maximized the EOM transmissivity, then the beam reached the amplifier, and then we performed the fine alignment to the amplifier. Output power and PMC transmitted power became the same as before, with the same 6.4A. We measured the frequency noise of the neoLASE with the new master laser (Fig.2). It is a bit better than the orignal master at 1kHz but worse above 2kHz, and anyway worse than the FB laser.

We will see the long-term stability tonight. We keep PMC and IMC locked over the night.

Operator's name: Sawada, Miyakawa 

Shift time: 9-17 JST

Check Items:
VAC: No issues were found.
CRY cooler: No issues were found.
Compressor: No issues were found.
Temp check: There is a temperature change at 10:00 in FIELD_PSL (-0.80 deg), FIELD_PSL2 (-1.50 deg), TABLEPSL_PT05 (0.38 deg), and at 16:00 in FIELD_PSL (0.50 deg).
VIS GAS filter output check: No issues were found.

IFO state (JST):
09:00 Start of shift, STANDBY remains
17:00 End of shift, STANDBY remains

We accepted the following SDF changes which were made by Pcal-Y calibration reported klog35391.

CALEY

K1:CAL-PCAL_EY_{1,2}_PD_BG_TX_V_SET
K1:CAL-PCAL_EY_1_PD_BG_RX_V_SET
K1:CAL-PCAL_EY_{1,2}_OE_{T,R}_SET
K1:CAL-PCAL_EY_{1,2}_RX_V_R_SET
K1:CAL-PCAL_EY_WSK_PER_{TX1,TX2,RX}_SET
K1:CAL-PCAL_EY_2_INJ_V_GAIN
K1:CAL-PCAL_EY_CURRENT_TIME

KAGRA Pcal-Y updates (2025/10/22)

Workers: Takahiro Sawada, Dan Chen

We performed monthly Pcal-Y calibration on 2025/10/22.

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

I forgot to upload a picture of moving intensity noise.

If the laser diode module can be detached, it might be a good idea to first check whether the proper DC supply is reaching the laser diode section with the current power source before trying 220 VAC or similar inputs.

How about using 220V AC line to wake up them?

At present, the EOM path in the IMC loop (common mode servo) uses two HV amps, F30PV and F10A. F30PV has a 10 times gain and 35V max out, while F10A 10V max in, 10 times gain and 100V mazxout. And the signal path is 1) common mode servo out -> F30PV input -> F30PV output -> F10A input -> F10A output -> EOM.

If the output of the common mode servo for the EOM path is over 1V during IMC control, the output of F30PV will be over 10V, which is over the max input for F30PV. 

So, It is better to check the output of the common mode servo or F30PV by using T, whether it is over 1 or 10V in the case of neoLase, because neoLase has larger frequency noise around the frequencies that EOM should manage.

It is better to check whether the F10A is healthy or not. (If F10A just cuts the over 10V by using a diode or something, F10A can be healthy, I guess.)

On the other hand, PA-85 can accept 35V input. (Of course, the output will be cut at ~ 200V which was limited by the operation voltage for PA-85)

(Uchiyama, Miyakawa, remote: Tanaka)

First, we swapped the fiber laser setting to the neoLASE setting this morning. A mirror on the X-axis stage to merge the light was shifted to select the neoLASE path. We aligned to PMC and locked PMC and IMC to the neoLASE laser with 20W input. Tests I show below will be the final test of the current master laser of neoLASE.

1. First, we measured the frequency noise of neoLASE. What we measured on 10/8 seemed to be with ~9kHz PMC UGF. Fig. 1 shows the comparison between today and 10/8. It looks that today's data shows lower frequency noise above 1kHz. It was measured with 4kHz PMC UGF (Fig.2). 10dB gain between IN1 and IN2 in the  PMC circuit exists, so the gain at -10dB shows the correct UGF. I guess we forgot the 10dB gain on the 10/8's measurement. When I added another 10dB gain on the servo, which made 9kHz UGF correspond to the 10/8's noise. Anyway, they correspond below 1kHz and are still 3 times noisier than the FB laser.
2. We looked at the effect of the scattering light by shaking a sensor card around the amplifier, especially around the bright location looked at by the sensor card. However, we did not see any frequency noise enhancement on both PMC and IMC error signals.
3. When we looked at the spectrum of intensity noise of neoLASE, we saw moving peaks in the intensity noise as we had seen at Toyama Univ (Fig. 2). However, it was difficult to see the coherence between the position of peaks and the frequency noise.

We left the trials above because they can be seen after swapping the master laser. And, we started preparing for swapping the master laser of neoLASE. I put two bars to keep the position of the board below the laser head (Fig. 3), and set a camera to keep the beam position for swapping (Fig. 4).

We are ready to swap the master laser, but we did not have enough time to finish it today. So we leave ~20W neoLASE on and keep PMC and IMC locked over the night to see the final stability of the current master laser.

[Uchiyama, Ushiba]

Abstract:

We replced pump LDs on the 2nd fiber amplifier to the first one.
Unfortunately, the 1st FIB amplifier doesn't emit the light even if we enabled the FIB amplifier from the GUI.

Detail:

We replaced pump LDs from the snd fiber amplifier to the 1st one to recover the 1st FIB amplifier.
What we did is as follows:
1. Disconnect screws for power supply for LDs (red circle in fig1).
2. Disconnect screws to fix LDs to the cooling plate (cyab circle in fig1).
3. Disconnect fibers between LDs and amplifier (yellow circle in fig1).
4. Remove LD units from the 1st FIB amp. and install LD units from the 2nd FIB amp.

After replacement of te pump LDs, we tried to operate the 1st FIB amp. but it didn't work.
Since the situation doesn't change from that before the replacement (the FIB amp outut doesn't increase to 1W even if the FIB amp. is enabled), the power supply unit for LDs might have a problem for the 1st FIB amp.