F. Matichard, K. Venkateswara

This is related to Sheila's alog (15607), where they had trouble locking MICH_DARK. This was being caused by excessive RZ motion as described in the comment to that post.

1. Fabrice and I then tried to understand what was producing this motion, because the ground is very quiet in RZ under normal circumstances. To do so , all the isolation loops were turned OFF (red) and then ON (blue) at ITMX. The result is shown in the first pdf. This led us to conclude that all the RZ/Yaw being produced is self-inflicted. RZ is increased by a factor of ~50 from normal.
2. To see where this comes from, we looked at the coherence between RZ and X, Y  and Z channels. As seen in the second plot, it looks like Z dominates (through the known Z-RZ coupling) near 0.1 Hz, and Y and X cause the rest at lower frequencies.
3. This excess RZ contributes to the problems locking MICH as seen in the third attachment. Note the high coherence between BS_CPS_RZ and the MICH_OUT at very low frequencies.
4.  One possible solution is to close the RZ loop, as seen in the fourth pdf. First page shows the CPS_RZ, second shows the T240_RZ and the third shows the OpLev_Yaw. Green curve is with no sensor correction, blue is with Z sensor correction to HEPI and the red is with Z sensor correction to HEPI and RZ loop closed with a high blend (T750)  (along with Z to Rz subtraction at Stage 1).

Ideally, we should not even need Z to RZ subtraction and a simple position sensor loop ought to be sufficient to hold RZ constant. We will test this configuration soon and look at other chambers as well.

Currently, the RZ loop is engaged on the Beamsplitter.