J. Oberling, R. Short
Our PMC Refl signal has been slowly increasing over the last several months; by extension this means that PMC Trans and the amount of power available to the IFO has been decreasing. Ryan and I went into the PSL enclosure to take a look at things.
We began by using an IR viewer to look at the PSL optics. All looked as expected except for M11, which was a bit brighter than we both recall; this is the 1st of the two turning mirrors that steer the beam into the PMC. We noted to take a close look at this optic once we had things turned off. We next took several power measurements around the PMC, all done with the ISS OFF; with the ISS OFF the power moves around a little bit more, so I've included the min and the max readings at each location:
- PMC In: 126.5W - 127.7W
- PMC Refl:
- Locked: 20.3W - 20.8W
- Unlocked: 125.9W - 126.2W
- PMC Trans: 106.1W - 106.6W
- Amp2 Out: 136.8W - 138.3W
- ISS Diffracted Beam: 3.8W (@ ISS offset of 4.3)
Things here match our PD calibrations OK. One thing we noticed, the amount of ISS diffracted power was a little low for the amount of variation we were seeing in the output of Amp2; we therefore adjusted the ISS offset a little to bring that bank back up a bit, to ensure we have enough power available to counter the swings in Amp2 output. We finished here with an offset of 4.7, and 4.2W in the ISS diffracted beam (again, this was with the ISS OFF). The ISS now defaults to ~3.0% diffracted power when it is OFF (this change was accepted in both SAFE and Observe SDF).
This done we moved on to inspecting optics with our green flashlight. With the PMC unlocked and the PSL external shutter closed, we looked at all of the optics between the external shutter and the PMC. Sure enough, all optics looked good except for M11 (see first picture). 8 in situ, very careful drag wipes later the optic looked much better, and better yet no signs of damage (see 2nd picture). Still some small dots on it that I couldn't remove, my guess is these are small coating defects that the green flashlight is really highlighting. We then opened the shutter and viewed the mirror through the IR viewer, and while it was still a little bright it did look much better. The PMC relocked without issue, but PMC Refl was higher than when we started (likely due to the PSL environmental controls being on and slight alignment shift during drag wiping M11).
Last, we checked centering on the PMC locking PD and the PD that reads PMC Refl. These PDs were well centered, with no improvement made when we tried to adjust the alignment onto the PD. This tells me that we have not seen an alignment shift between Amp2 and the PMC, which tracks with the fact we have acheived almost zero improvement to PMC Trans/Refl by tweaking beam alignment. Since we already had a mutlimeter set up, we took a visibility measurement for the PMC:
- Unlocked: -0.463 V
- Locked: -0.038 V
- Visibility: 91.8% (!!!)
The visibility measurement looks good, really good, and does not match what we see with PMC throughput. Using our power measurements from earlier, we're only transmitting ~83.7% of the incident light through the PMC, nowhere close to what our visibility measurement suggests. This reminds me of the old loss issues that plagued the original glued aLIGO PMCs at LLO, and spurred the development of the all-bolted PMCs we're using now at both sites. Could our issue be loss build-up in the PMC? This suggests that is a likely possibility.
At this point we left the enclosure to consider next steps. We do have a spare all-bolted PMC that we could swap in and see how it performs. This is, theoretically, a quick process, as the all-bolted PMCs have magnetic ball mounts that help preserve PMC alignment upon removal (we tested this a little when we installed this PMC in April 2018, and it seemed to work well). For now, the PMC Refl signal is still higher than when we went into the enclosure, by almost 3W. This isn't entirely surprising, as I did clean mirror M11 and the PSL environmental controls were on for a couple hours. We're watching things for time being, hoping that PMC Refl comes back to "normal" once the enclosure thermalizes; if it doesn't then either Ryan S. or myself will do a quick alignment tweak at an available opportunity (or next Tuesday if no opportunistic time pops up between now and then). Probably more to come here.
This closes LHO WP 11881.
