Last night a lot of glitches happened in the magnetometers which are also visible in HRec around 50 Hz.
The magnetic glitches seem to be stronger inside than outside.
Further investigation required.
The glitches in the magnetometers are a few seconds after the glitch in Hrec.
Seems to be the consequence of a compensation mechanism.
these glitches look like excitation of structures around 48Hz.... a few are very loud and excite several structures
Indeed. I did some further investigation about the time delay and it seems to be an issue of the spectrogram using Hrec sampled at 20.000 Hz.
Attached a spectogram of a time where there were two glitches just after each other and there is clearly no time delay when using Hrec sampled at 16384 Hz.
Also clearly visible are the 'arches' around 50 Hz
An other view of this double glitch, with the noisy 50Hz between them.
And the plot shows also that there is no delay between the spectrograms made with the dataDisplay, whatever hoft_xxxxxHz is used.
I write down some comments and attach some plots; no clear conclusions so far ...
These glitches seem to appear mostly during the nights, and we have noticed them clearly in our sensors since March the 4th
Our best witnesses are the external magnetometers located outside the northeast corner of the CEB, where nearby and underground are the mains Medium Voltage (15kV) coming from the local supplier, the mains Medium Voltage (15kV) sent to the NEB and the mains Low Voltage (400V) sent to the first 600 meters ot North arm, if I'm not mistaken
During one of these glitches we can see that "S" mains line increase, "R" line remain more or less the same, while "T" line decrease. This happens in ALL Virgo buildings. These voltage variations are a couple of Volt and, related to the MV mains, could correspond to a variation of less than 100V over 15kV (if really coming from the local energy supplier ENEL)
These glitches last about 2 seconds, and they are mostly "double"
What puzzle me is the fact that, even if most of the sensors detect these glitches, there are some noticeable exceptions:
- the ENV_MCB_MAG_N magnetometer see quite nothing
- the ENV_EER_MAG probe that is over the Mode Cleaner pipe (and should mostly sense the current flowing into it) see they very well, maybe better (or larger in amplitude) than the external magnetometers
Conclusion: we still are unable to say that these glitches comes from the external mains company (ENEL) or instead they are locally generated due to some problems in Virgo (maybe in our MV to LV systems). Detecting them in the magnetometer over the MC pipe is very strange, and I have no clear explanation for that.
I forgot to say that, as pointed out by Irene and reminded by Maria, the glitches in the mains create a drop of the horizon in Hrec because they mainly excite the family of peaks of the crossbars (e.g. 48Hz @ WEB)
