OMCS was touching.
Rahul made OMCS transfer functions and found that something was touching. He, Koji and Betsy went to the floor to center the OSEMs. T1 was touching, which was fixed. Depth of OSEMs were adjusted, too.
Something is funny about OM3 damping.
Rahul also made TFs for OMs and thought that they were fine. However, OM3 PIT damping was oscillating at 1.75Hz. I flipped the sign of H1:SUS-OM3_M1_DAMP_P_GAIN (nominally -1, now +1) and it damped fine. I hate this.
1st attachment red shows the OLTF of OM3 P damping loop at around the offending resonance. This was of course measured with the flipped damping gain, and the OLTF looks absolutely strange. The 1st UGF is ~1.73Hz with -42deg phase, and the 2nd UGF is ~1.77Hz with -180deg phase.
If you draw a Nyquist diagram, you will find that this is a stable TF. OTOH, with the "correct" damping gain, this will become unstable. So we know why this works and why the correct sign doesn't, but we don't know what changed and why.
Alignment converges.
With the OMC suspended without any biases, we adjusted OM1/2/3 biases to align the beam into the OMC (i.e. centering both of the OMC QPDs).
At some point we found that the ASC-AS_C was totally off-centered, used SRM to recenter, and continued the alignment.
2nd attachment shows the status as of now. Both of the OMC QPDs are reasonably centered, ASC-AS_C too. BOSEM DAC output for OM1 is ~11k max. OM2 and OM3 are at ~7k max and ~9k max, respectively. These are OK though OM1 is not super.
Current status of the LVEA etc.
It's laser hazard. SQZ manager was set to DOWN. Koji closed small HAM5 GV and HAM7 -X curtain.
Remaining work (Monday).
- Fix the balance mass for OMCS.
- Check the beam positions on OM1, OM2 and OM3 (again).
- Adjust the WFS / OMCR paths.
- OMC trans video path?
- Restore the shroud panels.
- Restore the OMC trans beamdump on the -Y side.
- Check the centering on the irises on HAM6.
OM3 damping mystery solved, everything back to normal.
The sign of the OM3 P damping gain got back to the nominal -1 (negative 1) after yesterday's boot fest (alog 75627) and OM3 was happily damping everything.
Turns out that the problem was a rogue filter in all of OM3 M1 COILOUTF channels, stealing about 7dB and ~61deg at 1Hz, 12dB and ~73deg at 2Hz, which was fixed after booting things.
In the attached ndscope screen, at 2024/Jan/26 17:25:11 UTC, OM3 M1 BIO status request was changed from 2 (nominal, hardware LPF ON) to 1 (cannot be used for OMs as hardware LPF seems to be always ON). I don't know how this happened.
FM1 in OM3 M1 COUIOUTF was turned on, that's a digital LPF used when hardware LPF is off, but the hardware filter is always on, causing mismatch. The same filter was turned on for OM2, but it was set back to normal after ~30 minutes. After rebooting things yesterday, state request for OM3 coils got back to normal.
Today I made OM3 P damping TF and it made sense (attached dtt, red is now, blue is when COILOUTF was in a bad state). Taking into account the sign flip between the red and blue, clearly the blue is delayed about 60deg at 1Hz and maybe 80deg-ish at 2Hz.
The blue showed a huge L coupling, which went away in today's data, I don't know why. (Green trace is the L2L OLTF just to show you where the L peak is.) I haven't bothered to change the state request back to 1, flip the damping gain and repeat the measurement just to see if the L coupling comes back.
I don't know how this happened.
