Virgo Logbook

AdV-ISC (Commissioning up to first full interferometer lock)

cifaldi, gherardini, mpinto, palma, spinicelli, tringali - 1:30 Wednesday 01 November 2023 (62340)

ISC shift: SR scan in LN2

During the first part of the afternoon, the lock acquisition was not working correctly due to the bad weather.

However, starting from around 17.30UTC we could start to reach carm null easily. We had few unlock due to the saturation of the PMC_PZT (fig. 1) and DARM that went away (fig. 2).

After some time, we could also reach LN2 but unlocking after few minutes of lock (see fig. 3 and 4), with the BS correction to increase. We then relocked in LN2 with both BS_TY_SET servo and MICH_SET servo open. Closing them manually one per time, we noticed that BS_TY_SET went positive together with an instability of the lock. We re-open the loop, and manually moved it toward negative values. As an effect, Hrec improved as well as the DARM_SSFS_LF. We left it at -1e-3. To be noted: we leave the BS_TY_SET servo commented out in Metatron, since it has to be checked.

On the other side, the MICH_SET loop appears to work correctly, we thus leave it automatically closed in the automation.

Profiting of the stability of the lock. we then started to test the misalignment of the SR_TX (see fig. 5). When we reached +1um of misalignment, the ITF became unstable, we came back to the original value but eventually we unlocked.

We leave the ITF in safe autorelock.

Images attached to this report

Comments to this report:

spinicelli - 4:15 Wednesday 01 November 2023 (62341)

In the following lock in LN2 with the BS_TY_SET servo loop open, the sensitivity was very bad.

Since in the previous lock, the behavior of the servo seeemed to follow the opposite direction, I tried to change the sign of the servo gain and close it.

For the time beeing, the loop is working properly, zeroing the LF CMRF, increasing the DCP and globally improving the sensitivity. It is maybe worth to notice that, in the absence of BS_TY_SET, the MICH_SET went down to -13, while, once the BS_TY servo ON and stabilized, the servo brought MICH_SET fast toward positive values (see Fig. 1).

As the servo seems to work, I re-enable it in the automation for the next lock. In case of problems, just comment out line 5638 (" SC.BS_TY_SSFS_LF_Q_ENBL = 1") in ITF_LOCK.py and then reload ITF_LOCK node. I also saved into the .INI file the gain with the corrected sign.

Images attached to this comment

mwas - 9:28 Wednesday 01 November 2023 (62344)

This data with very bad CMRF gives a nice opportunity to check if the current frequency noise projection method works.

Figure 1 shows the noise budget (with intentionally underestimated frequency noise) where the detection losses were increased to 50% and the SR round trip losses increased to 7%, in order to match the measured optical response of DARM.

Figure 2 shows that noise budget with in addition the mystery noise and the SSFS noise projected by rescaling the model frequency noise coupling by a transfer function with a 1/f and flat components (two free parameters), adjusted to match the coupling at 227Hz and 1111Hz lines. The projection works relatively well, which means that SSFS unnorm Q is providing a good measurement of the SSFS sensing noise.

Images attached to this comment