Virgo Logbook

Environmental Monitoring (Environmental Monitoring)

Paoletti, Fiori, Tringali, Bucci, Ruggi, Mours, Seglar-Arroyo, Nanadoumgar-Lacroze - 23:05 Monday 12 February 2024 (63246)

Another hunt of the 25m glitches source

Today we had an ENV shift dedicated to finding the possible source of the 25m glitches affecting Hrec since on year.

We started right after the Dailly Meeting by openeing all the relays of the NI coils, leaving all four NI coils completely floating. The 25m glitch occurred. The relays of the coils were set back to the standard state.

After this test we turned off Ncal (@ NEB) and Pcal (@NEB & @WEB), waiting for the glitch, which came anyway. We can therefore rule out these devices, which will be switched ON tomorrow morning during the maintenance.

Also: the RGA (Residual Gas Analyzer) between NI and BS was switched OFF at 16:25 UTC. Its cable is still connected anyhow, and will be unplugged on the first occasion during one unlock.

We then turned off IMMS box #7 (ENV_IMMS7 located in the CEB DAQ room) that manages the temperature sensors in the "mirror" area of WI and NI (see attached file for exact name and details). The glitch recurs even when the box is turned off, but we noticed a curious behavior: a glitch-like signal in Hrec at the exact moment we turned the IMMS box back on. To be sure of this, we repeated the test, and again we noticed the same "little glitch".

After that, we moved to WEB. Here we performed a sequential action: ENV_IMMS12 was turned off; then CH13 (WE_MAR_COIL_FL_TE), CH14 (WE_MIR_COIL_UL_TE) and CH15 (TCS_WE_RH_TE) were disconnected from this IMMS box; after a few minutes, they were reconnected; finally, the IMMS box was also turned on again. For each of these operations we noticed a small glitch in Hrec like that seen acting in CEB, or (in the case of disconnecting and reconnecting the cables) a 50Hz line disturbance in Hrec. Again we saw that the main 25m glitch was present during the IMMS shutdown, so this system is not the source we are looking for.

Finally we moved to NEB by performing the same actions on ENV_IMMS41, but adding in this case also the NN array shutdown. Here again we noticed the small glitches inHrec as seen in WEB, but the "main" 25m glitch was always recurring.

All actions are described in detail in the attached TXT file.

Conclusions: we think that the actions performed today (turning off and on and disconnecting/reconnecting the cables of the devices that are in front of, or near the mirrors), recreate the old "electrostatic actuator" situation: the mirrors are charged, and varying the potential of the metal objects in front of them creates an electrostatic force that moves the mirrors.

That said, we still need to find the source of the glitches and the path by which they enter the ITF.

Images attached to this report

Non-image files attached to this report

envnoise_shift_12_02_2024.txt

Comments to this report:

verkindt - 14:55 Wednesday 14 February 2024 (63276)

The attached plot compares the sensitivity while NCal and PCal were switched off (purple) to sensitivity the day after, when NCal and PCal were back on (blue).
The difference in BNS range is small, as expected, within the usual BNS range fluctuations of about 1 Mpc.

Images attached to this comment

direnzo - 18:56 Wednesday 14 February 2024 (63278)

The artificially generated glitch at 18:56:01.574 UTC looks similarly "blippy" (Gravity Spy jargon) in a time-frequency map (Figure 1) but its time series is quite different than that of the 25-minute glitches. The latter are all very similar: they look like an "elbow" in the row data (Ref. to #63006) and an inverted sinc in the whiten or high-passed time series (Ref. to #61944). The whiten time series of this glitch is shown in Figure 2.

To stress the difference between the two glitches, I've also run the omicron scans for this artificial glitch and one of the 25-minute family: artificial glitch results, 25-minute glitch results. In the former, the simultaneous presence of a glitch in V1:ENV_WEB_MAG_*, V1:ENV_WEB_MIC_* and V1:ENV_WE_ACC_* is clearly visible. For the latter, there are no equally good witness channels: all the glitches simultaneously present are in unsafe channels (LSC_*, Sc_*, SDB2_B1_PD*) or just coincidences (PSL_ML_DC, BsX_*, ENV_DT_CT_FINGER_ACC_*, etc.).

I' let the experts to comment if they notice some interesting correlation in the (safe) channels for the 25-minute glitch.

Images attached to this comment

Paoletti, Cavalieri - 11:30 Thursday 15 February 2024 (63284)

Taking advantage of the interferometer locked in LN3, we went into the CEB DAQroom to turn off all the IMMS boxes that read suspension chain temperatures (AD590), and also the one that reads environmental signals from the clean rooms.

The sequence of operations is detailed in the attached file.txt

Basically, this operation did not eliminate the glitches that regularly happened even with the boxes turned off.

We later turned them back on, and everything is now back to standard condition.

p.s.

we may have made a mistake in analyzing the data, but comparing Hrec at 20000 and Hrec at 16384, we noticed a different behavior with respect to glitches. We attach the two spectrograms.

Images attached to this comment

Non-image files attached to this comment

file.txt