The first plot shows the current contribution to Hrec of the four crossbar modes at 48 Hz. It is obtained rescaling four spectra of Hrec measured during the noise injections performed in October (some peaks are slightly shifted). The applied scaling factors are about 100 times larger than the ones needed in order to project the supposed current flowing in the actuators in clean conditions, therefore the level of the peaks is still not explained. Moreover, the largest peak at 48.5 Hz does not match perfectly the projections as the other peaks, so we should not exclude a different source for that.

The second plot is the same with a different zoom. Trusting in it, one could assert that the contribution of the modes to the residual excess noise in that region is not so relevant, and a significant reduction of the mode quality factors, by the application of mechanical dampers, would not improve so much the situation. But again the mechanism which produces the main noise could be associated to the one which excites the modes, as it is for the one which has been removed today, so the investigation has to continue in the same direction.

"It is worth noting that the motors drivers crate is OFF since ever (it is powered via a "smart plug" which receive a remote command), but this type of smart plug has an internal relay that interrupt only one wire of the two (phase and neutral), and which of the two is open is a coin launch (the plug can be inserted on both ways). This could explain the differencies we found between WEB and NEB"

Today we did a test on nine (out of ten used) smart-plugs that are used in Virgo (one for each tower) to remotely turn on and off the motors crates (in addition to the suspension crates).

Of these nine, eight are connected in order to interrupt the phase of the mains (we do not see the phase when the outlet is switched off). Only one (the BS) seems inverted (we see the phase on the output), but the result in this case is not very clear (we also detect about 70 Vrms on the remaining and supposedly interrupted neutral wire *). All measurements are performed with a TrueRMS multimeter connected to the ground.

The important thing is that there are no differences between WEB and NEB from this point of view: by chance these smart-plugs both act by interrupting the "hot" wire.

-------

* In the BS tower we also did a test on a mains line that does not pass through the smart-plug, and also in this case we read 230Vac on the hot wire (the phase) and about 70Vac on the cold wire (the neutral) with respect to the safety earth wire (yellow / green) and also with respect to the body of the tower. We suspect some problems in the wiring, we will check as soon as possible with the help of the electrical manager.

BS electrical system verified today with Massimo D'andrea, nothing relevant was found (and also we were unable to replicate the behavior ...)
 

today we checked the effect of replugging each single cable of the LAST STAGE MOTOR CRATE of WE that were left unplugged since the activity of Friday.

(details of the times are in the attached log files: the first file refers of the replug tests done on MONDAY 20 Jan, the second one is of today)

NOTES:

• (1) the WE LAST STAGE MOTOR CRATE is  smart-plug' switch is OFF (while the switch in the crate fron panel in ON) (same for all similar MOTOR CRATES, since before O3)
• (2) all 48Hz dumpers were OFF (Paolo)
• (3) just before the start of this test all motors cables of NE crate were reconnected (Federico)

reconnection tests:

• only MH-RH (test done on Monday 20Jan) no effect
• only MV-RH (test done on Monday 20Jan) no effect
• only MH-F7-TX (test done on Monday 20Jan) no effect
• only MH-F7-TZ (test done on Monday 20Jan)  it seems to have an effect on both 48Hz and 47.8Hz
• only MA.F7.U - no effect
• only MA.F7.D - no effect
• only MV.F4 - no effect
• only MV.F7 - no effect
• only MH.MA.TX  - large effect on 48Hz and not only (read below)
• only MH.MA.TZ - large effect on 48Hz and not only (read below)
• both MH.MA.TX and MH.MA.TZ  (read below)
• keeping MH.MA.TX and MH.MA.TZ we UNPLUGGED the crate from the power cord (read below)
• we left things in this condition

Figures 1, 2 and 3 show the effects in hrec observed TODAY.

We noticed a large effect when we connected the two cables of "marionetta balancing motors" (MH.MA.TX and MH.MA.TZ). The effect loooks similar for both: the 48Hz peak increases by almost a factor 6 (precisely from 1.5E-9 to 9.E-9). There is also some more broadband noise increase, and a couple of small peaks appear around 41, 42 and around 53 and 54 which resambles some sidebands around 48 (to be checked).

When BOTH the two cables are connected the noise effect in hrec at 48Hz and around it seem to add linearly (so total increase of 48Hz ia about afactor 10) ... but this is to be better checked.

When we UNPLUG the power cord of the CRATE the noise reduces at the same level of when the cables are unplugged (!)

When UNPLUGGING the ITF UNLOCKED (... or better: the two events were quite simoultaneous).