[Tanaka, Miyoki, Takase]

Summary

We recovered the BD chiller water circulation. So the FB output was set at the default current of 27.8A(~18W).

What we did

Firstly, we tried the water flow check of the Adjuster BD solely. We confirmed that the water flow is enough.

Secondly, we attached the the yellow tube to the chiller return and the transparent tubes from the PMC BD on the Adjuster BD in the space above the edge of the optical table after turn off the FB amp part (So the IR power was tiny round PMC). Then we put the Adjuster BD in the water basket and checked the water flow by operating the chiller. We confirmed the water flow in the all water paths and BDs. Actually, we sensed the BD temperature was cool by fingers.

The chiller was stopped again.

Thirdly, we replaced the 30W detectable power meter head with the recovered Adjuster BD. In this process, we found that the water tubes interfered with the PEM mike that was suspended from the metal mesh. So we pulled up the mike about 30cm to avoid interference. After that we also replaced the 50W detectable power meter with the original PMC BD at the rough position. Then we turned on the FB and got ~ 1W laser power. We realocated the PMC BD to have the beam be at the center. After that, PMC was locked, and we djusted the Adjuster BD height to introduce it at the center of the Adjuster BD. We also confirmed that the temperature of the PMC/Adjuster BD around was 20C by using the non-contacting type thermometer.

We restarted the chiller.

Fourthly, we increased the current of the FB upto 27.8A (~ 18W). If there was no water circulation, the temperature of the Adjuster BD became ~40C degrees when the FB output was 10W with in 30 minutes or so. After the recovery the chiller path for BDs, the temps of the Adjuster/PMC BDs were 21C just after the FB laser operation. After ~40minutes later, we checked the temps again, and got ~22C. So we recognized that the water  chiller system was working well.

Please Kimura-san ask somebody to fix it.

During the EY photosensor work, we found there is leakage of the air for clean booth at the connection port.
Attachment is a movie around the leakage point.

Yesterday 9:45, Hayakawa-san stopped the Y2300 water pump.

[Tamaki, Ushiba]

We measured resistance, forward voltage, and inductance of the newly installed photosensor/actuator cables from the feedthrough flange at EYV.
All measured values are consistent with those during health check before O4a and seem fine.
Following table shows the summary of the measurement (with the values before O4a).

Also, we measured the resistance of all the other connections like 1-2, 1-3, and so on.
All values are OL, so we confirmed there is no unexpected short among the new cables.

Date: 2024/5/8

I checked Pcal-X beam positions on ETMX with the suspension-aligned state.

Positions comparing to the results 25th Apr when we had the last beam alignment work:

I believed that a change of this magnitude could be adjusted by Picos, but today I leave it as is and monitor it. 

Ushiba, Tanaka, Takano

Abstract

We started to investigate the source of the strange coupling of ALS signal and IR REFL. We checked the slow DAQ of GrPDH of each arm with the shutter closed. We found that GrPDHX signal is noisier than GrPDHY above ~20 Hz, with many peaks and white noise floor,

and nonstationary peaks above 7 kHz were seen for both PDH signals.

Detail

For years we have suffered from the mysterious coupling between ALS signal and IR signal (ref: 11411). It might be due to a GND loop around REFL/POP/POS area and could be coupled to DARM sensitivity. To solve this issue once for all we started to investigate the source of this coupling.

First we suspected that some electronics for GrPDH might oscillate at some high frequency. To check the behaviour in high frequencies, we monitored the spectra of GrPDH slow DAQ, not at 16 kHz but at 64 kHz. Figure1 shows the spectra of GrPDHX and GrPDHY in terms of the input equivalent noise of the common mode servo. The spectra labelled noX are the measurements taken in different time and REF are the spectra with the input of the servo open electrically. It is certain that above 20 Hz the ADC noise is below the signal from the RFPD.

From these measurements we found these characteristics;

• For GrPDHX the floor level is almost flat above 20 Hz, 6e-7 V/√Hz, whereas the floor level of GrPDHY has some weird frequency dependence. At 100 Hz GrPDHX is about two times larger than GrPDHY.
• For GrPDHX many peaks are seen, at 60, 120, 180, 240, 300, 480 (power line), 68, 136, 234, 260 (mysterious) Hz.
• For both spectra, many peaks are found above 7 kHz. The center frequency of these peaks also changes in time (measurement no2 was taken 5 min after no1 measurement).

Anyway, it seems that the condition of GrPDHX is worse than GrPDHY.

Next

We will check the spectra of the related electronics one by one in high frequencies. Especially, we are interested in

• The original RF signal from RFPDs
• The demodulated signal above 32 kHz to check whether RFPDs are oscilating or not

Tanaka, Miyoki, Uchiyama, Yuzu, Takase,

[Summary]

The stuck existed in the Adjuster BD.

[What we did]

After the STM2 cable cut and the STM1 cable fixing, we restarted the search for the stuck position in the water path in the BDs chiller path in the PSL room.

We replaced the Adjuster BD with 30W detectable power detector. Because the water tubes that are attached to the Adjuster BD were fixed on the ceiling metal mesh above around the center of the optical table, we scarcely could move the Adjuster BD just above the edge of the optical table.

At this position, we removed the yellow tube that directly went to the chiller return side above a water basket. We found there was almost no water in the yellow tube. When we injected air by using air duster can into this yellow tube, the return side had also slight water and air. In this sense, this yellow tube had no stuck.

After that, we checked the Adjuste BD itself. We injected air by using the air duster can to the water input where the transparent tube was connected (this tube came from the PMC BD), and checked whether the air could come out from the other side water input where the yellow tube was connected or not. Consequently, there was no airflow could path through the Adjuster BD even though I tried 3 times. After that, I retried the same air injection from the other side. The first trial failed to the air pass through. However, the second trial succeeded in letting the air pass through the BD with some dirty pieces as shown in the photo. After this first air conduction, all air injection trials have succeeded in passing through the BD. So we concluded that, at least this Adjuster BD had stuck.

After that, we injected air to the transparent tube that was connected to the PMC BD to check the airflow to the output side of the chiller through the shutter BD and several tubes. Then we obtained the expected airflow! So we concluded that there was no stuck in PMC BD and Shutter BD and their tubes.

After this klog I found to need one more input port, the all required number of ports for WFSf3 is 10 ports. 
I checked the IOO rack availability today and there were 10 ports available in the IOO0 rack and 1 port available in IOO1.

The following is a summary of the different information when compared to ADC/DAC channel assignment.

• IOO0, ADC0, ch8-ch12: Cables unplugged. There is still a filter for Refcav monitor on the model. (FIG1, FIG2)
• IOO0, ADC1, ch20-ch23: "MCL DMD I&Q" cable is plugged in. No corresponding signal was found in k1IOO0.
• IOO1, ADC1, ch20-ch23: "MZM DC SERBO ADC" cable is plugged in. No corresponding signal was found in k1IOO0.

To reduce electrical noise at the ADC, I need to avoid cabling across the rack. Now I have the following as a plan for the ADC inputs for WFSf3.

1. Plug the ADC into the 10 available input ports in the IOO0 rack.
   → In that case, we need to consider whether to clean up or leave the Refcav monitor filters that remain in the model. I cannot decide now if Refcav will be reinstalled in KAGRA in the future.
2. Add a new ADC to the IOO1 rack to supplement the 8inputs. Also plug it into one empty input on IOO1. Also unplug one unused port on IOO1 and plug it in.
   →However, when cleaning up unused filters and unused signals in the ADC of the IOO1 rack, we need to confirm that we are not reading those signals in the Guardian. 
   The MZM, ISS, and ALS signals are coming to the ADC in the IOO1 rack. At the moment, there is a proposal to use the input port of the ADC used for MZM since MZM is unlikely to be used in the immediate future, and also a proposal to borrow the QPDB of ISS since it is not in use.

I discussed these issues with Yamamoto-san and Ikeda-san.
We have no idea if we will need MZM and Refcav in the future, so we think Plan 2 is appropriate.

From around 14:00, Ushiba-kun tried to LSC lock of IMC and it suceeded. However, ASC did not work even though it seemed to work in the morning time. After several checks of IMMT1 trans and so on, he suspected the laser beam was hidden by something at somewhere. Finally, we found the aluminum foil between IFI and IMMT was shading the beam! The aluminum foils inside seemed to roll inside because of no fixing and its rolling seemed to shade the beam.

Ushiba-kun, Hitrata-san and I fixed this nasty aluminum foil coveres. After that the IMC lock with LSC and ASC was successfully done.

> At the same time, the cleaning in the water vessel of the original chiller was done.
I attached the photos for the record. [before the cleaning] [after the cleaning]

Attached picture shows the current situation inside the IFI chamber.
STM2 pico cables are cut just below the heat-shurink tubes and tied at the STM2 pole with peek tie (red circle in the picture: sorry but cables themselves cannot be seen because they are on the behind of STM2 pole.).
STM1 pico cables are fixed on the optical table with aluminum clamp by M6 screw with vent hole (yellow circle in the picture).

[YokozaWashimi]

We took 4π photos inside the IFI, IMM, and PRM chambers.

https://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=15766

The tube connection information is https://klog.icrr.u-tokyo.ac.jp/osl/?r=22249.

Precisely speaking, Chiller output -> transparent tube -> Shutter BD -> yellow tube -> PMC REFL BD -> transparent tube -> Adjuster BD -> yellow tube -> Chiller return.

So, 

• remove the transparent tube at adjuster BD, then check the water flow.
  • If OK, the adjuster BD and its downstream yellow tube have stuck.
  • If no flow, the stuck point is at upstream.
    • remove the yellow tube at PMC REFL BD
      • If OK, the PMC REFL BD and its downstream transparent tube have stuck.
      • If no, the stuck point is at upstream: Shutter BD and its upstream yellow/downstream transparent tube.

By the way, do you know why the water-cooling type beam dumper was selected? The air cooling type beam dumper for 50W and 150W is enough for our purpose.

[miyoki, tanaka, yuzu, takase]

We replaced the PMC REFL BD with the 50W detectable power meter. the PMC REFL BD was just reset at other place near the original position on the same optical table in the PSL room.

Adter that, Tanaka-kun set the FB laser power at 10W to investigated the lock of PMC and IMC. PMC and IMC was successfully locked, and ASC seemed to work because the IMC output increased after angaging.

At the same time, the cleaning in the water vessel of the original chiller was done.

[Ushiba, Hirata, Tanaka, Miyoki]

After the recovery of the IMC LSC ASC control, ushiba-kun checked the infkuence of the cable touch for the STM2 on the IMMT1/2 trans and pop forward etc.

According to Ushiba-kun opinion, the alignment of the STM2 seemed to be sensitive to the weight load of the cables, not the trembling of the cables themselves. When he lifted up the cables, the STM2 alignment moved temporally, but came bacck to the original position even after stopping to lift up. This implies that we can keep these cables if we can properly fix them on the optical table of the pole. However, it seemed to be quite difficult to perform such thing without mialignment of the STM2 mirror. So we decided that we cut these cables below the pin connections.

While, not so remarkable influence of the cable touch on the STM1 was observed. This is maybe because the a part of cables were winding on the optical table and their weight seemed not to pull the STM1, then they fell to the bottom of the vacuum chamber. So we decided to fix the cables by aluminum clamps on the optical table with a kapton sheet for protection.

I estimated the transfer function from the base plate vibration to the optical table vibration, by
TF(impact@base -> ACC@table) / TF(impact@base -> ACC@base)

I also did the same calculation for the Ground (z) ->Table case, but the results became over 1 for many frequencies. I need to investigate whether my analysis is correct or not...