Not sure what is going on, but ETMX Stage 2 performance is pretty bad, over a broad frequency range, first attached image. I don't thing its something like a clean room is pushing on HEPI, because the HEPI L4Cs at both ends seem similar motion, second image. I took l2l tfs and it might be rubbing or some interaction with EFM. The third attached image are the stage 2 l2l transfer functions, and the vertical stage 2 resonances (~1.5hz) in particular look pretty ratty. I didn't see this at close out, because it's hard to get good resolution of these resonances in air, especially with the purge going. I also checked for rubbing right before close out of the chamber, fixed the issues that I could find, but it's hard to do much from the under side of the ISI. The response of one of the GS13s, V3, is a bit low at low frequency, not quite factor of 2. Last plot are the St1 tfs, and other than showing the St2 vertical resonances still look bad, I don't see much to be concerned about here.
I can confirm that this problem is still there, and provide a nice calibrated template for anyone to check on this.
/ligo/home/jeffrey.kissel/Templates/DTT/
H1ISIETMX_ST2_RXRY_Performance_ASDs.xml
Can @DetChar tell us when the problem started?
Sadly, the summary pages are pretty limited as far as showing non "interferometrically important" degrees of freedom. And also strangely, this doesn't appear to be significantly contributing to the longitudinal motion.
Hi Jeff,
It looks to me like the noise in these channels has been high since isolation was restored to the ETMX ISI on 6/13/18. Here's a very quick look at some supporting data and I can look more tomorrow.
According to the SEI:ETMX summary pages, ETMX ISI isolation was mostly on until 2/14/2018 and then there was no isolation all the way until late on 6/13/18 (then there were some days with no isolation on 6/27-6/29). I didn't do deep alog crosschecking but this appears to match with EX vent work. For the plots below I plotted times when the isolation was on and the other EX SEI channels looked ok.
The first plot below (link) includes these times:
- O2 - 2017-07-20 03:11:30 (1184555508) len: 100s.
- post vent - 2018-06-16 12:05:30 (1213185948) len: 100s.
- post vent - 2018-06-30 14:11:30 (1214403108) len: 100s.
- pre vent - 2018-02-12 12:11:30 (1202472708) len: 100s.
- pre vent - 2018-02-03 12:11:30 (1201695108) len: 100s.
Sorry that the plot legend is in GPS not UTC. The post-vent times (June 16 and June 30) are both bad, while O2 and both pre-vent times are good.
So now let's focus on times all post vent (and one pre-vent reference) (link):
- pre vent: 2018-02-03 12:11:30 (1201695108) len: 100s.
- post 2018-06-14 11:05:30 (1213009548) len: 100s.
- post 2018-06-24 17:11:30 (1213895508) len: 100s.
- post 2018-06-30 15:11:30 (1214406708) len: 100s.
- post 2018-07-03 15:11:30 (1214665908) len: 100s.
All four random-ish times post vent have elevated noise.
You too can follow along here: FRS Ticket 11021.
I should have looked at this before, but I think that rubbing is looking more likely on ETMX. For ETMX, with the platform in Isolated-Damped (i.e. St1 fully isolated, St2 damped only, HEPI position loops on (first attached image)) Stage 1 (green) is moving less than HEPI (red), but Stage 2 (pink) is moving as much as HEPI up to a couple hz. For ETMY in the same configuration (second image, same colors), Stage 2 moves a bit more than Stage1 around the stage resonance, but moves less than HEPI, above .25hz. The coherence between between HEPI and the stages is also a bit different at each of the endstations (when both are isolated-damped) with ETMX Stage2 a little more coherent (third image) with HEPI than ETMY (fourth image). The detchar pages show that ETMX is moving a lot more that the other chambers especially above 1hz in RY, compare:
https://ldas-jobs.ligo-wa.caltech.edu/~detchar/summary/day/20180706/plots/H1-ISOLATED_EE66AC_SPECTRUM-1214870418-86400.png
https://ldas-jobs.ligo-wa.caltech.edu/~detchar/summary/day/20180706/plots/H1-ISOLATED_CE5C7B_SPECTRUM-1214870418-86400.png
