Virgo Logbook

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

mpinto, mwas, sposito - 23:12 Wednesday 26 July 2023 (60998)

ISC shift: recovery of the ITF + test with RAMS servo ON and OFF

This afternoon we continued the recovery of the interferometer.
In the initial part of the shift we didn't manage, again, to go past the CARM_MC step as we were unlocking sistematically whenever we were approaching 40 mW of arms power after the CARM handoff. Later on, towards the end of the shift, we recoverd higher steps of the acquisition (up to CARM null 3f for few seconds).
-------

Michal noticed an excess of noise present at the DRMI 3f step (bump-like shape from 10 to 100 Hz), wrt to the working conditions pre-RAMS before its installation.
----------------------------------------
01.18 UTC jul 25: GPS during the night in which no bump was there.
----------------------------------------
We decided to open the loop, and see if there were visible differences.

15.30 UTC Gianmatteo went to the Lab to open the loop. At 15.32.29 UTC ALL has been switched OFF.

15.42 UTC we went back to DRMI 3f.

By opening the loop, we didn't notice any difference, as the bump at DRMI_3f stage was still present. So we continued the lock acquisition. In the attached picture, comparison of the Error signals and Correction signals spectra of the main DRMI and LSC signals is reported (purple: before RAMS; blue: After RAMS).

On the way to 40 mW of power we noticed that some parameter was a bit off-tuned, so in order to smoothly get past the ramp, we decided to go to 40 mW by hand, and try to adjust the loop gains step by step.
We put 10 mW of power for the ramp after CARM MC handoff (4.47 in MC_SET units).
We unlocked a couple of times due to 8 Hz oscillation (looks MICH gain too low).
So, we increased MICH gain up to an UGF of 20 Hz (up of a 30% = from 1800 to 2300 in the 1f/3f ratio).

While I was increasing by hand the CARM_MC_SET in order to reach 40 mW I injected also noise in order to measure the CARM_MC open loop. From the measurement, nothing unusual on the shape appeared. (MC_RFC Noise from 16.36 to 16.39 UTC)

MC SET increased with the following steps: 4.47 (10mW), 5.5 (15mW), 6.5 (20mW) and 7.5 (around 30 mW). ITF unlocked few seconds after reaching 30 mW at 16.39.54 UTC.

17.22 UTC We put the ITF in Single Bounc NI, in order to try to lock the OMC on the 56MHz and then on the 6 MHz to investigate if the RAMS servo is working as it's supposed to.
Once locked on the 56 MHz, we took a reference of the SDB2_B1_DC spectra, and then we switched on again the RAMS servo.

18.04 UTC RAMS servo back ON.

By looking at SDB2_B1_DC FFT, the effect of the RAMS (in terms of attenuation) is clearly visible.

Then, we locked the OMC on the 6MHz. In such condition, the effect of the RAMS while locked on the 6 MHz was not totally evident; at 18.30.25 UTC we increased the 8MHz mod depth from 0 to 6 dBm but no clear difference on the SDB2_B1 spectra appeared.

More details about the performed tests will follow.

.----
After dinner we kept going with the 'recovery'. We were able to go past CARM to MC step, but we unlocked at different steps of the acquisitions, e.g. 100 and 200 mW.

19.56.35 UTC unlocked at 200 mW (15-16 Hz oscillation)

a step 2 reduced PR TY TX gains and 169 MHz phases (of 0.3 rad) and reduced SRCL gain from 0.2 to 0.18

20.14.36 UTC unlocked at 100 mW (same 15-16 Hz oscillation)

at step 1 decreased SRCL gain from 0.25 to 0.22 and reduced of 0.3 rad the 169MHz phase

At the next attempt (20.28.41 UTC) we unlock on the way to CARM STEP 3 of 3 (400 mW) due to the same 15-16 oscillation. We tuned accordingly the parameters as for the steps before. (SRCL gain and 169 MHz phase)

----

20.46.14 UTC we unlocked on the way to CARM null 3f!

at that step we tuned the CARM B4 6MHz phase (reduced it of 60 deg accordingly to the delta applied yesterday on the other 6 MHz based signals).

at 21.04.50 UTC we recovered CARM null 3f. After 45 Seconds at 21.05.35 UTC we unlocked.

We left the ITF locked on the IR.

Images attached to this report

Comments to this report:

mwas - 8:22 Thursday 27 July 2023 (60999)

A few figure to illustrate some of the observations of yesterday.

Figure 1 shows the intereferometer with arms locked and Michelson fringes swinging now, and figure 2 before the RAMS servo isntallation. The top of the fringes seen on B4 is a factor 1.53 higher for the 12MHz and 1.67 higher for the 112MHz. A similar increase is visible with the CITF locked. This contradicts measurment of the modulation depth with the OMC scan, where the 56MHz depth increased from 0.17 to 0.19. So one would expect an increase of only a factor 1.25 for the 112MHz.

Figure 3, 4 and 5 have the same reference time (purple). Jul 25 at 1:18, before the RAMS servo installation. They all show DRMIT 3F lock data. After the RAMS installation both with (figure 3) and without (figure 4) enabled there is a broad bump between 100Hz and 500Hz on ITF audio channels (in particular B4 and B2). This was not present the last time the RAMS servo was put in place (figure 5).

Do we have an air flow set to high left over somewhere? Lights left on?

The made some single bounce checks to verify that the RAMS servo and new EOM work as expected.

Figure 6 shows in purple OMC locked on the 56MHz USB, with the RAMS loop open, and in red with the RAMS loop closed. The total power on the PD is ~0.8mW in that state. And with the RAMS loop open the fluctuations are at the expected level of a few 1e-7, and with the loop closed the signal is limited by the PD electronic noise. In green is shown with the OMC locked on the 6MHz USB, half of the power on the PD is the tail of the carrier TEM00. The RAMS servo is enabled. In blue is with the OMC stabilized in temperature at the 6MHz frequency and the 6MHz modulation depth reduced by 7dB. Note that moving away the OMC lock from the 6MHz makes this noise that has a pole at ~500Hz shape disappear. So this noise is some interference between the 6MHz and tail of the carrier light. Changing the 8MHz modulation depth by a factor 2 had no impact on it. In any case, it is not much higher than the RAM noise when the servo loop is open, so it shouldn't be a problem for locking.

17:56 (5min) OMC locked on 56MHz USB and B1 PD1/PD2 open. RAMS loop open
18:05 (4min) OMC locked on 56MHz USB and B1 PD1/PD2 open. RAMS loop closed

In between there are seval minutes of data with B1 PD1/PD2 shutter
closed but Vbias enabled.

18:18 (10min) OMC locked on 6MHz USB and B1 PD1/PD2 open. RAMS loop closed

8MHz mod depth increased from 0dBm to 6dBm

18:30:35 (2min) OMC locked on 6MHz USB and B1 PD1/PD2 open. RAMS loop closed. 8\
MHz at 6dBm

Figure 7. Note that the change in the spectrum of the power at the output of the PMC have a corresponding vhange in the power reflected by the IMC. And this has been changing by an order of magnitude since at least a week, so is unrelated to the locking troubles.

Images attached to this comment

60999_1690433597_Screenshot from 2023-07-26 17-42-15.png

60999_1690433601_Screenshot from 2023-07-26 17-41-45.png

60999_1690434556_Screenshot from 2023-07-27 07-05-03.png

60999_1690434560_Screenshot from 2023-07-27 07-06-08.png

60999_1690434564_Screenshot from 2023-07-27 07-08-53.png

60999_1690434812_Screenshot from 2023-07-27 06-59-53.png

60999_1690438977_Screenshot from 2023-07-27 08-22-44.png

mwas - 12:17 Friday 28 July 2023 (61021)

Figure 1. After bypassing the RAMS for the 6MHz and 56MHz, the bump between 100Hz and 500Hz, most visible on B2 and B1p, during DRMI 3F seems gone (or at least reduced by a factor few).

Images attached to this comment