Reports 1-1 of 1 Clear search Modify search
AdV-DET (Commissioning)
bonnand, eisenmann, gouaty, mwas, tacca - 0:21 Thursday 22 November 2018 (43687) Print this report
OMC mode matching tuning in LN3 (first trial)

Today at 18:00 utc we started the activity foreseen for the evening shift, which was dedicated to the tuning of the OMC mode matching in low noise 3. During this activity the interferometer unlocked twice. It was very difficult to relock the first time, and impossible the second time. Neithertheless, we could observe that the optical gain of the north cavity could be slightly improved by moving SR along the longitudinal axis.

After the Low noise 3 was recovered, we started by increasing the height of the permanent calibration line at 62.5 Hz in order to reduce the fluctuations of the north arm optical gain down to the 0.1% level. The line amplitude was increased in several steps between 18:02 and 18:09 utc, from its initial value equal to 1.3e-7 V to its final value equal to 2e-6V.

Then we started scanning the longitudinal position of SR:

  • at ~18:21 utc we moved SR by +1000 um, with a too short ramp time (only 5s). It seemed that the optical gain increased during this step.
  • at 18:26:48 UTC we started to move SR by another +1000 um, with a ramp time set to 30s. Unfortunately the ITF unlocked during this ramp.

During the relocking attempts we noticed that the B2 beam was no longer centered on the B2 QD1 and QD2 quadrants (the galvo loop were not functioning any more). Thus we recentered the beam on both quadrants by acting on the picomotors B2_M7_H: -100 steps, B2_M7_V: -500 steps, B2_M8_V: -700 stps.

We also undid the last motion of SR with -1000 um (ramp 300s) at 19:57:33 UTC.

Several relocking attempts failed, and we noticed that at least some of the failures were due to an oscillation in the MICH loop (see for exemple unlock at 19:36:30). Thus we tried to readjust the gain of the MICH loop: we reduced it from 350 to 150, and this allowed to relock the ITF.

We also took the occasion of one of the relock to readjust the SPRB bench TX setpoint from 80 to 76 urad, in order to better center the beam on the B4 photodiodes, as we had noticed on the camera that the beam was too high (and the beam was also clipped on B4_PD1).

Finally the ITF was locked again at 20:31 utc and we performed the following measurements:

  • 20:42:46 UTC: SR moved by +1000 um (ramp 300s)
  • 20:49:41 UTC SR moved by -2000 (ramp 600s): this scan is shown in the attached Figure. One can clearly see that at the beginning of this scan (near position +1000 for Sa_SR_F0_X) the optical gain was smaller. Then it improves during the scan, reaches a kind of plateau, and finally it seems to decrease again when Sa_SR_F0_X is near to -1000 um (but this should be confirmed with a longer scan). We can also notice that the beam size on the B1p camera changes during this scan.
  • At 21:03:00 UTC we tried to move the SDB1 bench along Z, but unfortunately this induced some alignment problem and the ITF unlocked.

It was then impossible to relock the ITF. At some point we could start to lock the OMC1, but we observed large fluctuations in the dark fringe power which made the OMC1 lock fail.

We decided to leave the ITF in recombined and autorelock mode for the night.

The SR position has been put at +800 um with respect to its initial position (that is to say Sa_SR_F0_X is now at -200), which seems to maximize the optical gain sa obsered in the attached Figure.

Note that we left the MICH gain at 150 (it has been written in its configuration file).

We also left the calibration line at its high level (to be lowered tomorrow morning).

Images attached to this report
Comments to this report:
mwas - 8:01 Thursday 22 November 2018 (43689) Print this report
Put back the calibration line to the nominal amplitude of 1.3e-7 at 7:00 UTC this morning.
mwas - 9:31 Thursday 22 November 2018 (43692) Print this report
Looking at violin modes during these times is not useful for assessing mode matching.
Alignment fluctuations where causing B1s1 to saturate from time to time, but more importantly the violin modes are much smaller than two weeks ago.

The initial idea of using violin modes was done using data from Nov 7, 1 day after steps in the suspension were applied to excite the violin modes.
Looking at the violin mode time evolution Nov 7 evening the violin modes were already damped by 2 orders of magnitude compared to the excitation.
However now the violin modes are another 2 orders of magnitude down.

Figure 1, shows the situation on Nov 7 with B1 and B1s1 on the same scale (B1s1_PD1 that sees 8% of the powermultiplied by 6 to compare with B1s1 that sees 50% of the power), where the violin modes are less than a factor 10 smaller in B1s1 than B1.
Figure 2 shows the situation on Nov 12, the violin modes a 100 times smaller, and only a factor 5 above noise (even with 300s long FFTs)
Figure 3, shows the situation last night, the violin modes are 50% higher than on Nov 12, but still too small to be useful

For the next attempt of OMC mode matching in LN3 we need to excite the violin modes, maybe just adding noise in the DARM loop at the violin mode frequency would be enough to excite them despite the notches in the control filter.
Otherwise just temporarily removing the notches might be needed (but no need to do steps in the marionetta).
Images attached to this comment
Search Help
×

Warning

Error

The present report has been modified outside this window. Please check for its integrity in the main page.

Refreshing this page will move this report into drafts.

×