Reports 1-1 of 1 Clear search Modify search
AdV-COM (AdV commissioning (1st part) )
bersanetti, chiummo, swinkels, gouaty - 4:10 Friday 24 March 2017 (37018) Print this report
Locking shift

This is the summary of today's locking shift:

  • the goal was to try and lock the OMC, if the lock at dark fringe was stable and long enough; unfortunately, this was not the case, as initially the dark fringe could not be reached in a stable and reproducible way;
  • after some failed attempts, we found out that removing the PR_F0 X and Y offsets during the MICH offset reduction helped reducing the excess noise, and this made us reach the dark fringe; this however highlighted a bigger problem, that I found out last thing tonight: due to the way things are coded now, an unlock before dark fringe would not bring such setpoints to their original values, causing drifts of the beam position on the PR mirror. I disabled these offsets, but the beam position on the PR mirror should be checked more extensively; as a consequence, what follows should be taken into account with a grain of salt;
  • at this point, when we reached the dark fringe we had no alignment loops closed, as the signals that were used were already been found earlier to be not good anymore;
  • so we tried investigating the error signals for the various loops in order to see if we could close some AA loop:
    • the first thing we noticed is that the gain for the DIFFp_TY loop had changed sign, as this caused a rapid degradation of the DC powers and a misaligned position of the NE mirror after an unlock;
    • we tried to find a good set of setpoints and gains for the DIFFp and the PR loops:
      • we managed to close the PR loops from MICH_SET =0.7 onwards, and the DIFFp loops once in dark fringe; the PR loops helped to keep the mirror aligned a bit during the transition, as we noticed that PR_MAR_TY was moving quite more than the other angular DOFs;
      • we could find a decent configuration for the DIFFp loops, but all the DC powers kept fluctuating quite a lot: the best lock was ~18 minutes long (see Figure);

 

The new settings are temporarily hardcoded in ITF_LOCK, in the states LOCKED_PRITF for the PR loops, and in the state LOCKED_PRITF_DF for the DIFFp loops. I leave the ITF in the LOCKED_SSFS state.

Images attached to this report
Comments to this report:
ruggi - 18:35 Friday 24 March 2017 (37027) Print this report

The actual working point of PR_X and PR_Y is not determined only by the value written on the DSP set point. In recombined mode, the position is adjusted by a control loop adding the output of an integrator to the local error signal. In this way, the correct working point is always recovered, even if the DSP set point is changing.

Search Help
×

Warning

×