Looking at the subtraction for today, I noticed that the PEM mainsmon (PEM-EY_MAINSMON_EBAY_1_DQ) was coherent enough to subtract noise at frequencies other than the power lines. This is pretty unusual, so we're trying to think up things to check, like whether the ETM is much more charged than it used to be. TVo and Sheila are looking at BruCo to see if anything else looks suspicious.
I'll take a quick look to see if other mainsmon channels are also highly coherent, but this may be an interesting avenue to look at for why we have this extra noise now.
Of note is that I didn't see this mainsmon subtraction for data from July 15th (alog 37590).
The mainsmon channels also seem more noisy than in the past. Is this something that we've seen before?
The attached spectra show that the mainsmons are noisier now than they were a few days ago. Spot checking, the noise was there on July18th, but not July17th. It's not clear from the time series trends that there is anything different going on.
EDIT: Spot checking more spectra, it looks like the noise may have started on 17July2017, between 20:00 and 21:00 UTC.
Josh, TJ
Can you explain how you're finding the broadband coherence between the MAINSMON and h(t)?
The bruco run from today doesn't show anything other than 60 Hz + harmonics:
We did some spot checking and don't see significant broadband coherence throughout the day, it looks consistent with other days from the last few weeks:
https://ldvw.ligo.caltech.edu/ldvw/view?act=getImg&imgId=172856
Blue: 2017-07-19 03:30:00
Yellow: 2017-07-19 01:30:00
Green: 2017-06-30 14:30:00
Red: 2017-07-04 14:30:00
For the record, the analog CDS team reverted the ESD power supplies on July 11th -- see LHO aLOG 37455.
Hmmm. Looking at coherence on DTT, I'm also not seeing much. I was inferring that there would be coherence based on the subtractability of the noise. As Kiwamu pointed out, perhaps it's a series of glitches or something, where the coupling is constant but the noise isn't, so when you look at coherence averaged over long times, it doesn't show up?
EDIT: It looks like Kiwamu was right, that there was a glitch, probably in the power lines. I re-did the subtraction in sections of 256 seconds rather than a full 1024, and the first sets were fine and normal (no broad subtraction with mainsmon), and the last set is pretty significant. So, maybe this is just a regular thing that happens, and I just caught it by accident. The attached is a spectrum during the time of the glitch. I assume that the glitch must be on the power lines, since I get such good subtraction using them as my witness.
Sheila and I ran BruCo:
https://ldas-jobs.ligo-wa.caltech.edu/~thomas.vo/bruco_July19/
During this GPS time (1184486418) around 25-35 Hz range, there is a lot of coherence between H1:ASC-OMC_A_PIT(YAW) to DARM. But spot checking the few hours after with DTT, this seems to go away so maybe there's some transient stuff going on during this time.
Looking at a spectrogram of the MAINSMON channel, there are two broadband glitches near the end of the 1024 second stretch from your original plot:
https://ldvw.ligo.caltech.edu/ldvw/view?act=getImg&imgId=172923