Sheila, Elli

This morning I noticed the IR buildup in the arms as meaured by LSC-TR_X/Y_QPD_B is read out as 10% lower in the x arm than in the y arm.  This is unexpected because we think the loss in the x-arm is lower than in the y-arm.  Because of this we wondered if the normalisation of these QPDs is a little off.  To check this, I adjusted the gains and offsets of these QPDs to get the single-arm buildup of these equal to one.

Firstly the dark offset on the LSC-TR_X/Y_QPD_B QPDs was not correct (LSC-TR_Y offsets were of by 15% of the single-arm buildup power).  It looks like there is a slow drift in these offsets, and it looks like they have been corrected multiple times in multiple places.  To remove the dark offset, we unlocked theIMC so there was no light on the arms and adjusted the LSC-TR_X/Y_QPD_SUM_SEG?_OFFSET for segment so that it read zero.    The changes to the X_TR_A_SEG?_OFFSETs are:

Once the QPD segment offsets were set to zero, we were able to remove a few rouge offsets from other places.  We turned off offsets on H1:ASC-X_TR_A_SUM_OFFSET (was 8), H1:ASC-Y_TR_A_SUM_OFFSET (was 1), H1:ASC-X_TR_A_SUM_OFFSET (was 4), H1:LSC-TR_X_QPD_B_SUM_OFFSET (was 0.5), H1:LSC-TR_Y_QPD_B_SUM_OFFSET (was -0.2).

We changed LSC-Y_TR_A_LF gain from 0.017 to 0.0188 and LSC-TR_Y_QPD gain from 0.219 to 0.231  so that the y-arm single arm buildup is equal to one.  The x-arm buildup was close to 1 so we didn't change it.