Dew point measurement taken this morning read at -43.8 ^oC.  Measurement taken before any activity inside chambers.  Soft covers were on HAM5 and HAM6, HAM7 is isolated and a bit pressurized.

Tue Jul 30 08:09:44 2024 INFO: Fill completed in 9min 40secs

FAMIS 21010

PMC REFL hasn't been rising as much over the past week, but it certainly still is. Everything else looks fairly nominal.

Workstations were updated and rebooted.  This was an OS packages update.  Conda packages were not updated.

TITLE: 07/30 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Corrective Maintenance
CURRENT ENVIRONMENT:
    SEI_ENV state: MAINTENANCE
    Wind: 3mph Gusts, 2mph 5min avg
    Primary useism: 0.01 μm/s
    Secondary useism: 0.05 μm/s
QUICK SUMMARY: Continuing work today on the FS assembly swap in HAM5.

Betsy, Camilla, Sheila, Jennie W, Keita, Robert

21:30 UTC Camilla and Betsy figuring out new way to lock diwn the OFI.

22:05 Placing peak sheet to shield ofi while they unscrew bars from FS assembly.

22:40 Took mount with bar and beam dumps still attached out of HAM5. One of the bolts for the bar was stuck in the mount so they unscrewed it chamber side.

22:42 Sheila dis-assembled FS wedge mount

Sheila left at 23:00 UTC

Camilla and Betsy taking photos and videos of damage and wiping beam dumps and mounts.

We found indium foil in the top screw hole of the mount which would sit on top of the wedge.

23:00 UTC Robert came to take photos between HAM5 and 4 from outside chamber.

23:14 Keita joined.

Lots of discussion about what to do about mounts as there was lots of the black coating in the screw holes and so Betsy and I washed the holes on the FS mount with isopropanol and she tapped the holes again.

00:10 Camilla and I put beam dumps back onto OFI.

Mount and old FS wedge covered up chamberside and cover back on.

Everyone out by 00:15 UTC.

Camilla to add pics.

TITLE: 07/29 Day Shift: 1430-2330 UTC (0730-1630 PST), all times posted in UTC
STATE of H1: Corrective Maintenance
INCOMING OPERATOR: Tony
SHIFT SUMMARY:

The wind started to pick up around 1700UTC, which brought increasing dust counts in the PSL anteroom and LVEA. Smoke could be smelled inside the OSB as well from outside.

The KTP was replaced in chamber during the afternoon and the chamber work continues (FS assembly).

LOG:

Looking at the OFI TEC servo we noticed that the response was much slower than what one could expect from the given ugf frequency. The gain was increased fomr -1 to -15, whereas the ugf frequency was decreased from 50mHz to 10mHz. This resulted in a response time around 10s, which is about 3x faster than when we started. Increasing the ugf beyond this can make the servo unstable.

As a side note: The applied voltage at the driver was about 3x higher than the voltage readback from the TEC, indicating a rather large voltage drop in the cable.

Noticed that GV5 had hard closed on Saturday at 19:59 utc time, no human intervention.  It appears as if the instrument air pressure left for soft close GV5 was enough to hard close it.  According to PEM instruments near GV5 made lots of noise, see attached plot, accelerometer and microphone data.

I checked the dial on the gauge to measure the instrument air at GV5, and indeed there was only about 10 psi, this value is within the "usual" pressure used for soft clossing pneumatic gate valves and per procedure.

No interaction is required at this moment from the vacuum team.

Tagging ISC, since sensitive equipment is hanging near this gate valve, like cameras and OpLevs.

After discussing with Jenne this morning, I have flipped the sign of the ETMX ESD lock bias voltage from -450V to +410V to reduce the charge build up in the optic. I will keep this settings for a week or two and measure the charge on the optic. If the charge gets reduced, then will continue with this until we are ready for locking the IFO - will revert the settings to nominal values at that point. Attached below is a trend of the lock bias offset and ESDAMON_DC_OUT, which shows the changes I have made today.  

Sheila, Keita, Betsy, Camilla

In the last power measurements in 79313 we were loosing 25% through OFI and another 15% from lost through incorrect polarization through the TFP.

Today's measurements:

• Before OFI after SRM, 1) mean 5.212mW, stdev 40uW over 10sec 2) lights out mean 5.134mW st 35uW 3) lights out mean 5.098mW stdev 39uW, mean
• Input to HAM6 1) mean 5.173mW stdev 32uW, 2) mean 5.199mW stdev 40uW 3) mean 5.066mW stdev 23.4uW
• SQZ beam power to ZM6: 1) mean 10.09uW stdev 78nW, Blocked/Background 1.175uW stdev 0.3nW, difference 8.915uW 2) mean 9.92uW stdev 64nW, Blocked 1.16uW stdev 0.5nW, difference  8.76uW  3) mean 9.944uW stdev 76nW, Blocked/Background 1.185uW stdev 16nW difference 8.759uW
• Max power out of laser collimator at iris with 220mAmp setting, 38mW+13mW dumped ~50mW

From the mean of these measurements Sheila and Louis calculated the rejected ratio (power sent to ZM6) is 0.17 ± 0.002%. The transmission through to HAM6 is 99.96 ± 1.45%. The sum of these is 100.13%, which is within the margin off error.

At 16:08 Thu 25 July 2024 the continuous wave psinject excitation on h1calinj stopped running.

After investigation, I found that this process could not restart because of an expired leapseconds file on h1hwinj1. From the end of June onwards it would not have been able to restart, but was eventually restarted on Thursday 25th July. We currently do not know why the process restarted on Thursday.

The fix was to manually edit the /usr/share/zoneinfo/leapseconds file and change the lines:

#Expires 2024   Dec     28      00:00:00
#expires 1735347600 (2024-12-28 00:00:00 UTC)
#       File expires on:  28 December 2024
 

I had disabled monit during this investigation. After fixing the leap seconds and undid all my temp changes and performed a monit controlled restart of psinject with no problems.

Dew point measurement taken this morning read at -43.7 ^oC.  Measurement taken before any activity inside chambers.  Soft covers were on HAM5 and HAM6, HAM7 remains isolated and a bit pressurized.

FAMIS 26468 -Capture Monthly--Trend HEPI Pump Pressures and Drives

There was a number of spikes over a few days, that I expect to be associated with the venting and jumping on the HEPI platforms.

Mon Jul 29 08:11:00 2024 INFO: Fill completed in 10min 56secs

Gerardo confirmed a good fill curbside.

Keita, Sheila

Summary:  We chipped the spare KTP crystal this morning, and this afternoon installed the back up spare that Rodica had shipped.  We've adjusted the roll, and the pitch angle seems better than the crystal that we removed.  So this should be ready to go into the chamber first thing Monday morning.  We have many photos from today, which will be attached and annotated to this alog later. 

Following on From Jason's alog 79319

With this placement of the damaged KTP in it's assembly, where the assembly is bolted perpendictular to the beam path, and a wave plate to mix the polarizatoins upstream of the KTP,  we have 13.5mW in the p polarized beam (see first attachment which is a diagram Paul sent us this morning), 16.3mW in the s polarized beam, and 2.3 mW in the AR surface.  This is not set up with the AOI that the AR coating is designed for, which is why the AR beam is large, we can take advantage of this to set the pitch by setting the AR reflection parallel to the table.. 

With the original KTP in place, the AR reflection isn't parallel to the table it is pitched up by 10/(530 +405 mm) about 11 mrad. The optic lab beam is hitting the KTP a little below the burned spot, which would have been on the +X side in chamber.  The burned spots are on the side closer to our alignment laser.

Before taking it out of the mount we took a variety of photos showing the damage spots position.  We did not find any indium foil in the mount, Paul had warned us that there might be indium between part 5 and part 9 of D2000038. 

As we were placing the new crystal in the mount, we left the set screws (11) that hold the peek loose, and held the piece upsidedown while we screwed in the bolts labeled #10 to hold on the aliuminum plate.  It chipped as we were tightening the bolts labeled 10 in D2000038. After some discussion with Rodica, Paul, Gabriele, we went back to the lab to take photos of the chip and try to understand how the chip happened.  We have photos showing that the crystal was proud of the top of part 5, the orientation of the black metal piece.  After we removed the black front plate, we could see that the KTP was firmly stuck in part #5, with the wedge stuck.  Keita had to push on the back of the crystal to get it loose. 

Keita made a series of measurements using calibers:

Chipped KTP: (agrees well with the drawing: E1900284)

• height on thicker side: 0.8850 inches
• heght on thinner side: 0.8845 inches
• width (center to center): 0.8870 inches
• width from one corner to the other (longest width): 0.9360 inches
• thickness: 0.5055 for thicker side

part 5: https://dcc.ligo.org/DocDB/0171/D2000567/001/D2000567_A%2B_OFI_KTP_Wedge_Holder.pdf

• side to side width of the opening for the crystal (measured on the side where the black part attaches) 0.9225 inches
• depth of the slot 0.515 inches on the deeper side.
• depth of the slot 0.458 on the thinner side
• height of the slot, including space for peek, 1.003 inches

peek:

• thickest part 0.1105 inches
• thinnest part: 0.1095 inches

total height (max): crystal + peek = 0.885+0.1105 = 0.9955 inches, smaller than 0.2mm than the height of the slot.

cratered KTP:

• height of crystal at thick side 0.8855 inches
• height of crystal at thinner side: 0.8835 inches
• width (center to center): 0.887 inches
• width diagonal: 0.9345 inches
• thickness of thicker side: 0.5095 inches

We then did a trial run (using the chipped KTP) of an approach that we think avoids the problem of letting the KTP get wedged in the holder.  We set the KTP on several wipes on the bench, and set the peek underneath it.  We then slid the holder (part #5) over this whole thing, which went smoothly, and tightened the set screws.  After flipping it over, we can see that the KTP is not proud of part #5, and it sits nearly flush to the front of the holder.  We could then easily attach the black front plate.  We decided to do this with the final spare, which had 2 small chips on the edge. 

With that spare installed in the assembly, we adjusted roll using the alignment laser that Camilla and Jason set up.  We wanted to adjust roll so that the transmitted beams were level to the table, since we know that the laser bea we started with were parallel to the table.  The roll adjustment was sticky, and tightneing the set screw tended to lower the transmitted beams.  The beams transmitted are off in yaw compared to the ones from the cratered KTP, by about 7mm/870mm + about 8 mrad.  We had difficulty seeing the AR reflected beam, but we can see a tiny beam with all the lights off (this suggest that the reason the beam was so bright before was the coating damage, not the AOI of the crystal in our set up.). At 820 mm from the KTP, the AR beam was about 1 mm too low, so it's pitched down by 1mrad. 

We think this is about as good as we can do by adjusting with this mount, so we are satisfied and will leave the KTP assembly under wipes and foil to be reinstalled on Monday.