Reports 1-1 of 1 Clear search Modify search
AdV-COM (AdV commissioning (1st part) )
ruggi, majorana - 0:54 Saturday 16 September 2017 (39493) Print this report
Many changes in suspension and alignment control strategy

- Marionette reallocation for NE and WE starting from the beginning of lock acquisition

- GIPC

- Mixing of alignment DIFF and COMM

- PR_TX and PR_TY controlled full bandwith with QD

- alignment control filters improved

 

Plots will follow.

Comments to this report:
ruggi - 14:28 Monday 18 September 2017 (39497) Print this report

Fig. 1 --- Error and correction signals of PR alignment control.

In the old configuration (blue) a blending of quadrant and optical lever was used; now (red) only the quadrant signals are used. The gain and the shape of the control filters are different. Concerning TY, the accuracy is much better, but the level of correction above the UGF (~2 Hz) is much higher. A noise budget of this d.o.f. should be done in order to understand if an improvement of the roll-off is needed. Concerning TX, there is a significant increase of gain at ~0.8 Hz, but the main improvement is the reduction of noise at the resonance (2.4 Hz) and at the UGF (~3 Hz). This is due both to a reduction of sensing noise and the removal of a big gain peaking.

Fig 2. --- Improvement of longitudinal error signals.

A large disturbance was visible in MICH and PRCL, due to PR_TX bad behavior at 2-3 Hz. Now those signals are much better, and a good effect is visible also in DARM.

Images attached to this comment
ruggi - 17:13 Monday 18 September 2017 (39498) Print this report

In figure, the error signals of arm alignment. Blue is Saturday Aug 19 at 23:20 UTC; red is Saturday Sep 16 at 23:20 UTC. The microseism at the two gps was about the same amplitude. The improvement is the overall effect of many changes at different levels of the control loops.

DIFFp_TY is working quite well. DIFFp_TX is not working well at low frequency, for unknown reasons.

Images attached to this comment
ruggi - 16:21 Wednesday 20 September 2017 (39524) Print this report

An undesired side effect of GIPC has been the disabling of fModErr loop (the tidal control of common arm length which keeps MC length tuned on the modulation frequency). This means that since a few days we are working with fModErr slightly mistuned. It cannot be excluded some impact of this defect on the sensitivity. The bug will be fixed as soon as possible.

ruggi - 17:55 Friday 22 September 2017 (39551) Print this report

The problem of fModErr loop has been fixed yesterday evening. Now GIPC can be used without affecting fModErr.

Search Help
×

Warning

×