Yesterday we performed the test as we made in November 2018 to check for any static charge on the Input Test Masses (NI and WI)
Due to the sistematical unlocks we experienced in the past while working and walking around the input towers on the platform, we decided to use a different approach respect to the one used for the End Test Masses: open only one arm of the Coils (using a dedicated relay box command via Giulio's GUI) and send a 15Hz line using the Virgo/Pisa DSP DAC. That way no current flows into the coil, and only electrostatic forces are exerted. The level of the signal written on the DAC was 1Vpk, then this signal goes through a "matrix" whose values are:
NI UL 0.979
NI DL -1
NI UR -0.9218
NI DR 0.9429
WI UL -0.9208
WI DL -0.9634
WI UR 0.9574
WI DR -1
Here the summary of the test (readings on LSC_DARM using DataDisplay with T=10S):
| COIL | real Vrms amplitude | 15Hz line | 30Hz line |
| NI UL | 0.6922 | 8e-6 | 1e-8 |
| NI DL | 0.7071 | 1.9e-5 | 1.4e-8 |
| NI UR | 0.6518 | 4.6e-6 | 1e-8 |
| NI DR | 0.6667 | 8e-6 | 1e-8 |
| WI UL | 0.6510 | 8e-6 | 1e-8 |
| WI DL | 0.6812 | 3.3e-6 | 1.2e-8 |
| WI UR | 0.6769 | 3.6e-6 | 1.3e-8 |
| WI DR | 0.7071 | 1.2e-5 | 1.2e-8 |
As a raw comparison with the old END mirrors measurements, we copy here the old values measured in LSC_DARM (they should be with the same T=10S)
| COIL | real Vrms amplitude | 15Hz line | 30Hz line |
| NE DR | 0.884 | 1e-5 | 1e-7 |
| NE UR | 0.871 | 7e-5 | 1e-7 |
| NE DL | 0.871 | 7e-7 | 1e-7 |
| NE UL | 0.871 | 4e-6 | 1e-7 |
| NE DL | 0.871 | 8e-7 | 1e-7 |
| WE DR | 0.877 | 3e-6 | 7e-8 |
| WE DL | 0.877 | 5.8e-5 | 7e-8 |
| WE UR | 0.877 | 3e-5 | 7e-8 |
| WE UL | 0.877 | 2.4e-5 | 7e-8 |
The "2f" quadratic signal (30Hz) should be relative to the actuation when the mirror is not charged, while the "f" linear signal (15Hz) should be relative to the charged mirror: the old test at END mirrors had a 2f of about 1e-7 while in this test we read values around 1e-8. We need a cross-check that sends back this (new) type of signal to the END mirrors again, in order to validate and calibrate both measurements (the old one and the new one). Unfortunately this has not been possible, because the relais dedicated to the arm opening at END test masses are "hard wired" using a jumper. Removing this jumper makes the relay free for this use, but first of all we have to change some lines of code in metatron or ITF it will never lock again (relais always open).
We also have sent a signal 0.5Vpk and 2Vpk of amplitude (WI DR only), results are:
| real Vrms amplitude | 15Hz | 30Hz |
| 0.35355 | 5.6e-6 | 3.2e-9 |
| 0.7071 | 1.2e-5 | 1.2e-8 |
| 1.4142 | 2.3e-5 | 4.9e-8 |
Some relevant notes:
- the 50Hz feedforward has been left OFF during these tests, and consequently the 50Hz line grow also because having one arm of the coil floating seems to pollute more than usual.
- during our test we tried to open one arm of a coil, then to repeat the test with the other arm opened and the first one closed (e.g.: open DL+ and inject in DL-, and after that we reversed the configuration): the signal is sent to each arm through the simmetrical DAC output (one arm is driven POSITIVE, the other arm is driven NEGATIVE), and we noticed that the "sign" of the global 50Hz in DARM flips. This maybe means that the origin of this 50Hz signal is upstream?
- we also noticed that during measurements some non linearities arose in DARM: 35Hz (50Hz - 15Hz) and 65Hz (50Hz + 15Hz)
Conclusions:
It seems that also the INPUT Mirrors are charged. Difficult to say now if they are at the same level of the END, because the "2f" line that should not depend from the mirror's charge has not the same amplitude. We need further tests.
attached the LogFile 20190826 Input mirrors charge.txt
