> Please note that we confirmed that the current drawn from the independent power supply approximately doubled (~1.4 A for +18 V and ~0.8 A for −18 V) after connecting the two boards.
> Correspondingly, the current meter in the computer room showed a reduction of more than 1 A.

Photo. 1 and 2 show the current values on +18 V power supply before and after changing the power supply for the whitening filters. You can see the current reduction before and after the power supply change. 

This was a question about the PSL table that was suggested I post here.

Robert Schofield saw the presentation on beam jitter at LVK and wanted to ask if the legs of the PSL table had been grouted. He said he had discussed this with some people from KAGRA a few years ago and wanted to ask if there had been any further discussion about this.

LIGO grouted its PSL during the initial LIGO phase and it reduced PSL table coupling to ground motion. It involves glueing the table legs to the floor with a cement called hydrostone. It can apparently be done without moving the table as you build a dam for the cement around the leg and just pour it in situ. 

[Takahashi.R, Washimi, Hirata]

We confirmed that IRM damper works well. After that, we locked suspention again.

1. Release suspention
BF, IRM, IM, RM and TM are released. IP and Topfilter were locked.

2. Check IRM damper position
Fortunately, IRM damper position looks almost same as before releasing suspention.

3. Check motion of IRM damper
First, we tried to check the damper motion, but it seemed that the damper was not working.
Takahashi-san inspected the area outside the SRM rack and found that the cross‑cable used between the feedthrough and the D‑sub cable was not appropriate.
After we replaced the cross‑cable, Takahashi-san confirmed that the IRM damper was working properly using the OSEM signal.

4. Transfer function
Takahashi-san will report later.

5. Lock the suspention
BF, IRM, IM, RM, TM were locked again.

6. Fasten feedthrough flange
We tighten feedthrough flange bolts(K-4-2). 10Nm→15Nm, 2times.

Today's photo is here: https://www.dropbox.com/scl/fo/ljozph0rohm3n46sp2ewj/ACaU31NZxhAETiqBB0v-Ras?rlkey=akuabdr9t2kroq02a5ae85t08&st=wy6asdyg&dl=0

*Sorry, I don't have any information about the relation of the pins and the coil wire-winding directions now. I will ask Takahashi-san.

I have performed an overall test the these check scripts: link

I did not find any issue on the scripts, but there are many issues on the test data.
We need the reconstructed data made by updated offline pipeline.

[Tanaka, Ushiba, Komori]

Abstract:

We successfully mitigated the coupling issue by using the independent power supply to drive two whitening boards for the PDA1, PDA2, and PDA3.

Detail:

As a continuation of yesterday’s work (klog:36538), we compared the ALS DARM spectrum during IFO flashing under the following configurations.
The results are shown in the attached figure.

• Turning on PDA1 (brown):
  One of the PDA1 RF outputs was whitened using a whitening board that was already driven by the independent power supply.
  The spectrum became slightly noisy above 100 Hz, but it did not become noisy in the 10–100 Hz band, as expected.

• Using the conventional cable for PDA3 (magenta):
  The spectrum became noisy above 100 Hz, similar to the case with PDA1 turned on, but it did not become noisy in the 10–100 Hz band.

Finally, we measured the spectrum with PDA1, PDA2, and PDA3 all turned on, with the two whitening boards for these PDs powered by the independent power supply (red).
The noise level is close to the noisy reference (black) above 100 Hz, but it remains clearly lower than the reference in the 10–100 Hz band.
In other words, the total RMS is still smaller.
In fact, the DARM, MICH, and PRCL signals appeared to become quieter during arm flashing with the PRMI locked.

Please note that we confirmed that the current drawn from the independent power supply approximately doubled (~1.4 A for +18 V and ~0.8 A for −18 V) after connecting the two boards.
Correspondingly, the current meter in the computer room showed a reduction of more than 1 A.

In addition, after these modifications, we found that the dark offsets for PRCL and MICH had changed, likely due to the change in the power supply.
Therefore, we performed a dark offset compensation.

Tomorrow, we will check the stability after opening the shutter in front of the REFL QPDs.
If the IFO is not stable, we will power one of the QPD whitening boards with the independent power supply and check the stability again.
Moreover, we found that the REFL HWP before PDA1 and PDA2 was not functioning, so we will restore it.

During the migration work of k1script, I found a plot range of finesse plot on the web was broken as shown in Fig.1.
It came from a change in the format of a history of finesse values.

I'm not sure when this change was made, anyway, an unknown header line had been added to the beginning of the file.
For now, I've modified the plot code to properly handle header rows.
Note that the character used to indicate comment lines varies by language and plotting tool.
So this is just an ad-hoc fix.

I checked the actuation of the IRM damper using the IM OSEMs. The DC actuation was confirmed with the offset (10000 count) to each COILOUTF (H1, H2, H3). Transfer functions from each actuator to the IM OSEMs were measured.

I tuned the FLDACCs by adding counterweights towards resonant frequencies around 0.2Hz.

Ushiba, Komori, Tanaka with Takahashi-san's and Hirata-san's help

This afternon, we found that ALS DARM often lost the lock after PRM aligned. We investigated the cause of the lock loss and noticed only GRY tran. power dropped largely by oscillating SR3 pitch after PRM aligned (fig.1). Also, we found that SR3 IM local damping control with OSEM, especially L, got noisy after PRM aligned. Therefore, the scattered light from somewhere (maybe SRM-GV) injected into OSEM sensors and disturbed the SR3 control.

We tried to change the situation by moving SR3 with the setpoint of GAS and IP, or by misaligning SR2 and PRM. However, the situation was not changed.

Ushiba-san pointed out that IM actuator was saturated at high frequencies. So we disengaged the whitening filters in IM coil drivers. SR3 oscillation got smaller and ALS DARM could keep the lock. 

[Kimura and Yasui] 

 On the afternoon of March 11, the vacuum evacuation pumps for MCE, #7 (between IXA and IXC), and #8 (between IYA and IYC) were switched to Ion pumps, and TMP and Dry were stopped.
This may cause a change in temperature within the central mirror room.

Tomorrow, we plan to switch the vacuum exhaust pumps on the 1F of the X-end and Y-end from TMP to Ion pumps.

By the way, can you make sure to the relation of the pins and the coil wire-winding directions?

This is additional information: SRM IRM damper pin assignment is as follows.

H1(-Y side, Coil#1) : 1-6
H2(-X side, Coil#2) : 2-7
H3(+X side, Coil#3) : 3-8

I temporarily modified the watch dog threshold for geophone signals from 10,000 to 40,000 at READY state in Type-B params to avoid suspension TRIP during the hardware work.
This should be reverted after finishing the hardware work.

Though the threshold parameters are common for all Type-B suspensions, the new threshold is only applied to SRM because I only loaded SRM guardian.

Several additional information during the work:

1. LSC_LOCK guardian was modified so that ISC WD threshold is set at 80,000 (150,000 before modification) to avoid large kick of the suspension when the lockloss happens.

2. After the blackout, nominal HWP position seems to be changed. Since the power on PDA1 at IFO REFL  is 1 mW when HWP angle is 155 degrees (nominal value before the blackout) while it should be 9 mW for PRMI_3F lock, it is necessary to rotate HWP to 145 degrees to lock PRMI_3F now.

[Ushiba, Komori]

We observed frequent excitation of the IX TM oplev pitch signal at 1 Hz.
This appears to be caused by glitches injected into the GAS filters, which excite the 1-Hz length mode of the cryo-payload, as shown in the figure.
According to Ushiba-san, the glitches in the GAS filters occur when the GAS is at a specific operating point.

[Tanaka, Ushiba, Komori]

Abstract:

We continued the investigation of the GR–IR coupling issue.
We successfully identified a configuration without the coupling, in which the DC power for the whitening board is supplied by an independent power supply.
A possible cause of the issue might be instantaneous saturation of the power supply during IFO flashing.

Detail

This work is a continuation of yesterday’s investigation (klog:36533).

First, we inserted an RF transformer (T1CA) between PDA3 and the demodulator after measuring the transfer function of the transformer (gain ≈ −1 dB, phase delay ≈ 1 deg).
However, the noisy behavior did not change after inserting the transformer.
The related measurement results are shown in Fig.1.

Next, we checked whether the noisy spectrum depends on which board the cable is connected to.
The spectrum did not become noisy when the cable was connected to the demodulator, but it became noisy when the cable was connected to the whitening board.
Therefore, we tried using an independent whitening board outside the rack with an independent power supply.
With this configuration, we confirmed that the spectrum remained quiet.
In contrast, the spectrum became noisy when we used the power supply inside the rack to drive even the independent whitening board (Fig.2).

Finally, we measured the spectrum using the original whitening board inside the rack but powered by the independent power supply.
The spectrum remained quiet even without the RF transformer (Fig.3).
This result indicates that the coupling issue can likely be resolved by using an independent power supply for the whitening board.

We suspect that a possible cause of this issue is instantaneous saturation of the power supply during IFO flashing.
We found that the whitening board draws a relatively large current, approximately 0.7 A at +18 V and 0.4 A at −18 V.
In addition, we checked the rack power supply located in the computer room, which currently outputs approximately 17 A, while the protection threshold is set to 21 A.
Therefore, such saturation could plausibly occur.

Tomorrow, we will measure the spectrum with the PDA1 and PDA2 cables connected to their corresponding whitening boards, which will be powered by the independent power supply.

[Takahashi.R, Washimi, Hirata]

We have completed the installation of the IRM damper’s mechanical parts and in-vac cabling. Suspention is still locked, so we have not operated IRM damper yet.

1. Exchange dummy plates to magnet plates
IRM has 3 dummy plates. We removed them(pic1). After that, we assembled the plates with the IRM, which already had the magnets glued on by Takahashi-san. (klog 36517)(pic2)

2. Install coil base
Only coil base part are installed first.(pic3)

3. Install coils
3 coils are connected to 1 cable. Pins and sockets are covered by PEEK shrink(pic4). 3 coils are installed at the same time. (pic5)
The cables got tangled, and it took  a while to untangle them. The relay connector was mounted on a nearby pole, and cable from the coil was connected to it.(pic6)
We checked resistance at the relay connector.

・Coil #1: 6.3 ohm
・Coil #2: 6.5 ohm
・Coil #3: 6.3 ohm

4. Coil alignment
Coil base can move 3 degrees of freedom, and coil 's angle can be adjusted using 6 screws. 3 Coils are aligned with magnet using those mechanics.

5. Exchange Feed through panel
We replaced the feedthrough panel, which previously had only thermometer connected, with 4 connector feedthrough. (pic7)

6. Continuity check
After connecting all in-vac cables, we did continuity check at the outside of feedthrough.(pic8)

・Coil#1: 7.6 ohm
・Coil#2: 7.9 ohm
・Coil#3: 7.7 ohm

We checked thermometer cables too.
・Thermometer1-6: 111.1 ohm
・Thermometer2-7: 111.5 ohm

Today's photo is here: https://www.dropbox.com/scl/fo/0lwieoxtf72h3d59xim6g/AMqNUPbwqT_vkmvjlaAGhMk?rlkey=scefj78rth7kpcsnq352ibedp&st=6qtyal01&dl=0

[Alex, Yokozawa]

We entered the PSL room to conduct a tapping test on several mirrors to help identify peaks due to beam jitter in DARM. We confirmed that the peak around 215 Hz comes from M17 and that the group of peaks around 450 Hz comes from the periscope. We also identified the peaks around 350 Hz as coming from M18 and the peak around 150 Hz as coming from M16. We had issues connecting to the ipad which made peak identification more difficult so we plan to try again tomorrow.

We then did a tapping test on the IMC REFL Table and found that the 150 Hz comes from the beamsplitter that splits the light off for the WFSs from the main path.

M17 - 215 Hz

Periscope - 450 Hz (2 peaks) or four peaks (one either side of the main two) when tapped at bottom. Also saw some of the 350 Hz peaks when large excitation

M16 - 150 Hz , 250 Hz

M18 - 350 Hz two peaks

REFL Table

150 Hz peak is from beamsplitter that splits off the QPD path from the PD path

[Kimura, Yasui, M. Takahashi, H. Sawada, R. Takahashi, Hirata and Washimi]

 On March 9, from 8:30 to 9:00, the pressure inside the SRM was increased from 87 kPa (k-log 36520)  to atmospheric pressure (~98 kPa).
After pressurization, the top flange and two side flanges of the SRM were opened.

I offloaded the following GAS filters with the FRs.

• ITMX: F0 GAS
• ITMY: F0 and BF GAS
• ETMX: F0 and BF GAS
• ETMY: F0, F1, F3, and BF GAS