Yesterday there were quite a lot of unlocks, both during the first part of the lock acquisition (due to the glitchiness of the green) but also quite a few in the last stages of the lock acquisition. I have tried to take a look to those (see the times in the table below), and it seems that all of these are triggered by one of the NE glitch. I attach the plots of the unlocks.
| 02:46:22 - 27 sept | 08:09:32 - 27 sept | 08:46:31 - 27 sept | 15:09:03 - 27 sept | 18:57:06 - 27 sept |
| LN2 | LN2 | LN2 | LN2 | AC. LN3 |
These are not NE glitches. These are times when B1s exceeds the 45mW threshold that triggers the detection system safety and closes the fast shutter.
Figure 1, 2 and 3 show a few examples. Figure 3 is the most clear with an oscillation at ~400mHz visible in B1s that eventually makes the power go above the safety threshold. Manuel has seen that this 400mHz oscillation is most visible in BS TX, and starts when BS TX is enabled in full bandwidth.https://logbook.virgo-gw.eu/virgo/?r=65219
Note that BS TX has been in drift control in LN3 since Friday Sep 27, which might be explaining why the data was more glitchy above 100Hz on Sep 28 (figure 4).
It is crucial to understand the reason why BS TX cannot be engaged in full bandwidth anymore, and find how to fix it.
