Robert, Nutsinee

Yesterday we had a nearly 30-min lock at 50W that broke around 20:00 UTC when a fire drill started (43 seconds after the alarm sounded to be exact). Right before the lockloss I observed a noise at high frequency (+1000Hz) creeped up on DARM spectrum before it lost lock. I was curious if the fire alarm were responsible so I investigated a little. The coupling seems to be the highest around 2000-5000 Hz region by about 3 orders of magnitudes. The DARM RMS up to 1000 Hz doesn't seem much different so I'm afraid we cannot blame the fire alarm for causing the lockloss. The timeseries shows OMC DC PD was already becoming unstable over time. The thermal drift is known to be a problem. So this lockloss *could* just be a coincidence with the fire drill.

The coupling information itself worth an alog post though. According to Robert we have never done any acoustic coupling injection this loud (4 order of magnitude higher than the noise floor). I wouldn't want to be in the PSL when the fire alarm goes off.

Below are spectrum of DARM and a PSL periscope accelerometer, before and during the fire drill. Uncalibrated.

Zoom in from 0-1000 Hz