We performed an analysis on the mis-centering of cavity beam on PR before and after the interventions that modified the Y centering of the PR mirror between the 26/04/2022 and the 29/04/2022 (see entry #55627).
As usal, as monitoring signal we used LSC_PRCL_CORR_50Hz demodulated at the PR angular dithering lines . We performed the demodulation offline with Matlab by multiply the CORR singnal for the properly fitered marionette signals (Sc_PR_MAR_TX_CORR_50H, Sc_PR_MAR_TY_CORR_50Hz). The optimal demodualtion phase was ~ 0 as expected.
The analysis has been done for all the locks in CARM NULL 1f longer than 1600s (minimum duration to get a certain steady state on the demodulated signals) during the period before (from 21/03 to 25/03 include) and after (from 30/04 to 11/04) the change. Some of the locks have been discarded due to the presence of large jumps and will be further investigated.
Fig 1. shows the centering signal evolution corresponding to each selected lock before (top) and after (bottom) the change respectively for Y (left) and X (right).
By considering for each lock the averaged value between 1300 s and 1600 s from the lock start. We found:
Before shift | After shift | |
Y | 4326 ± 412 | 1029 ± 321 |
X | -4890 ± 416 | -6835 ± 636 |
Units are arbitrary. By comparing the average before and after the intervention, we can estimate the relative change caused by the intervention.
In conclusion, the changes appear to have improved the Y-centering at CARM NULL in absolute value by a factor of 4, but have slightly deteriorated the X-centering by a factor ~1.4.
Next steps:
- Improve statistics for locks after the change (not a lot of stable locks after the 29/04)
- Calibrate the signals
- Check discrepancy with respect to CITF/CARM NULL
- Investigate origin of jumps/changes in the centering signals