Yesterday profitting of TCS CO2 lasers being shut down during 44125, we took some data with both NI and WI chillers powered on IPS line. We used two different IPS plugs located in NI gallery and WI gallery (kindly provided by Massimo D'Andrea).
19 December, the interferometer was NOT locked:
- 7:41 UTC (aprox) both lasers OFF
- 7:53 UTC both chillers OFF
- 8:09 UTC WI chiller ON on IPS
- 8:13 UTC NI chiller ON on IPS
- 8:28 UTC both off
- 9:52 UTC both on (back on UPS)
An effect is visible in the mains, magnetic and also electric sensor (ITF was not locked):
- Sidebands of mains harmonics at 100Hz and 200Hz disappear from CEB UPS, CEB magnetometers and the Electric probe (measuring voltage noise btwe the WI DSP rack ground and the Virgo ground): Figure 1 (100Hz sidebands) and Figure 2 (200Hz sidebands)
- In particular we are talking about the external sidebands (peaking at 98.6Hz and 101.4Hz) .... while inner sidebands (peaking at 98.9 and 101.1Hz) remain (these are instead associated to devices in EER, see 43858).
- When chillers are powered on IPS a corresponding noise appears in CEB IPS sensors as well in MAGNETOMETERS and also the ELECTRIC sensor (Figure 3); nothing instead is visible in the UPS sensors (Figure 4)
Preliminary conclusion:
- powering the chillers via IPS lines does not change the noise coupling to magnetic sensors. Instead it cleans the noise that was present in UPS sensors: this noise is transferred to IPS sensors.
