J. Oberling, R. Short, R. Savage
Today we went in to finish the power measurements and PD calibrations we were unable to finish last week. We began by measuring the IOO EOM output, with the ISS ON and OFF and PMC Refl with the ISS OFF:
- EOM Out
- ISS ON: 109.3 W
- ISS OFF: 108.5 W
- PMC Refl
This all matched up with the MEDM readings (at the time), so we next went to measure PMC Trans (to check that we did/didn't mess something up with that calibration). While we were taking the measurement, the PMC became really noisy and transmitted power dropped to ~22W. We tried to re-lock it a few times and it behaved the same each time. Turns out that when coordinating our work with CER power supply swaps on Monday, we missed that one of the power supplies being swapped was powering the RF sources for ISC, which of course took down the 35.5MHz RF for the PMC and the 21.5MHz RF for the FSS RefCav; with no RF we can't lock the PMC. Oops. At this point I decided to check if there was any contamination on the PMC mirrors, and found the main PMC_Out window to be fairly dirty (no laser damage spots, thankfully). We drag wiped until clean; it took roughly 8 attempts, dragging lightly to (hopefully) not disrupt PMC alignment. RF was still down, so while that work was finishing up we moved on to measuring more power in leakage beams between Amp2 and the PMC:
- HPA Rejected beam #2: 524 mW
- M10 Trans: 0.96 mW
- M11 Trans: 214 mW
- M12 Trans: 2.51 mW
Combining the above with last week's measurements of Amp2 Out (137.8 W), HPA rejected beam #1 (0.6 W), ISS diffracted beam (4.4 W) and PMC IN (127.9 W), we are still missing ~4.1 W of power somewhere (we should have ~132W at PMC In, instead of 127.9 W). Again, we found no leakage/ghost/etc. beams or otherwise with 4W in it, everything looks as it should and all are dumped accordingly. Total mystery at this point.
By this time the RF was back, so we went back to the PMC. Something really odd happened that we cannot explain. Upon first relock, the PMC was only outputting ~92W and the output beam shape was not Gaussian; it almost looked like a 00 with a 20 superimposed on top of it. The PMC reflected power was also really high at ~30W, but the change in output beam shape I have never seen before and cannot explain. We called Rick to bounce some ideas around, and while we were talking the PMC transmitted power slowly improved and the output beam shape began looking like a 00 again. It took almost 1 hour before the beam looked like it did before we lost RF, with ~107.8W transmitted at the end. During this time we touched almost nothing; we did not tweak beam alignment, or touch mode matching lenses. Thinking that maybe the drag wiping moved the PMC, we did push it around in its mount a little (it wouldn't move, as designed) and saw no change. None of us had any explanation for this behavior. Once the beam shape and output power had stabilized, we checked the PMC In mirror and decided to clean it as well; it had a decent bit of scatter and a bright spot on it. Cleaning it reduced the brightness of the spot a little and was scattering less, but otherwise changed very little. We also checked beam alignment through the EOM (reduced power using IO_MB_HWP1) and found it unchanged (so the PMC did not move). We then took one last measurement of EOM Out at 107.2 W (we recalibrated that PD upon leaving the enclosure, as the MEDM was reading ~3W higher than our power meter measurement).
To finish up, we measured the output voltage of the PMC locking PD (beam alignment into the PD was checked and found good), with the PMC locked and unlocked:
- Unlocked: 0.463 V
- Locked: 0.020 V
- Visibility: 95.6%
Highest visibility we've measured yet, but when we compare PMC Trans and Refl with PMC In we're only transmitting 84% of the power and reflecting 13%, implying ~3% loss in the PMC. Unclear why the reflected power is higher than the visibility would indicate. We also adjusted the ISS PD DC voltages, as they were reading higher than we like (10.3V on PDB, when we want 10V or slightly less); after adjustment PDA is reading 9.08V and PDB is reading 9.89V. We then re-centered the Bullseye PD, and left the enclosure.
Due to the PMC oddness, we looked at the TF, and found the UGF at ~2kHz with ~60 degrees of phase margin. This is a little higher than we like (the new all-bolted PMC is supposed to have a UGF closer to 1kHz). We lowered the PMC gain slider from 3dB to 0dB, which gave us a UGF of ~1.6kHz with ~59 degrees of phase margin (see attached picture). To get closer to a 1kHz UGF we have to lower the light level on the PMC locking PD. As the maintenance window was over and it seemed to be happy where it was we did not do this; if things become unhappy at this operating point we can go back in and lower the locking PD light level. Back at the control room we recovered the FSS and recalibrated the EOM Out PD. Ryan tweaked the beam alignment into the RefCav which increased the TPD from 0.76V to 0.87V; we'll have to go back in and tweak the FSS beam path alignment to get the TPD higher. The ISS was re-engaged and its RefSignal adjusted until we had ~2.5% diffracted power (giving a RefSignal of -1.99V). All SDF diffs were accepted; Ryan has screenshots that he will post as a comment to this alog. We will monitor the PMC over the coming days to make sure all is well; it seemed pretty happy when we finished up. This completes WP 11164.
