Virgo Logbook

Environmental Monitoring (Environmental Monitoring)

fiori, paoletti, genin - 9:59 Monday 03 April 2017 (37121)

EM noise from the EER conditioner

Eric noticed that a noise line around 80Hz (from time to time switching to 85Hz) is affecting the BPC signals: See Figure 1. This noise has a 24h cycle. The same noise is present in magnetic probes and in the ENEL voltage monitors (all three phases): see Figure 2. This made us suspect of some sizeble 3-phase engine powered on ENEL, .... we thought of something related to air conditioning

More precisely, the fundamental frequency of the noise is at 15-20Hz. (Figure 3) and it produces sidebands of 100Hz and harmonics up to 400Hz or so (Figure 4).

The noise correlates well with the EER temperature (Figure 5). We suspect the dual split air conditioner installed in EER in January, maybe it is its compressor unit.

It might be useful to investigate why the similar device installed in IER appears not to produce such evident EM noise...

Images attached to this report

37121_20170403092055_noise80hzspectrobsxx50110604800.jpg

37121_20170403094319_noise80hzspectroenvcebipsvolts060604800.jpg

Comments to this report:

fiori - 17:58 Thursday 06 April 2017 (37181)

In these days the "80-85Hz noise" in BPC signals is around 85-90Hz: see the weekly spectrogram in Figure 1.

The EEroom aircon splitter has been switched off today at 16:20 LT and switched back on at 16:58 LT.

It is evident that the 90Hz noise in BsX and associated 10Hz noise in IPS_VOLT disappears at the time of the switch off: Figure 2. While its reappearing when the splitter was switched back on is not so evident.

Images attached to this comment

37181_20170406175319_noise80hzswitchofftestairconeer.png

Fiori, D'Andrea, Paoletti, Soldani - 11:03 Wednesday 19 April 2017 (37295)

Tuesday 18/4 (maintenance) we unplugged the IPS mains of the EE_room Airmec inverter & compressor unit from its original supply, and we made a temporary connection to a line that run outside the CEB.

The suspect was that the current flowing into the IPS mains cables create a magnetic field that couple someway with the BPC signals, due to the fact that the Airmec and the whole INJ UPS system are plugged into the same supply cabinet and use the same cable tray.

Unfortunately the noise in the BPC signals seems remained at the same level (see the 86Hz line in the attached spectrogram)

Today 19/4 around 08:00 UTC we also changed the splitter's fan (EE-room) from "auto" to "mid", just to see if the noise we have is somehow related to the fan's speed variations.

Images attached to this comment