[Ushiba, Haoyu, Carl]
As we observed a lot of frequency noise coherent with the MCO angular drive in the beam jitter coupling experiments klog32591 we decied to decouple IMC optics angular drive from length.  To do this the interferometer was locked in the PRX configuration, this gives a better signal to noise ratio in mode cleaner length error signals.  The mode cleaner length error signal K1:LSC-MCL_IN1_DQ was used as a witness of mode cleaner length noise when driving mode cleaner optics in angle. 

We drove MCO,MCI pitch with 50 counts, MCO and MCI yaw with 20 counts and MCE pitch and yaw with 1 count at 34Hz in the test filter bank.  The the actuator alignment decoupling matrix elements (like K1:VIS-MCE_TM_ACTALIGN) were then tuned to minimise the lines seen in the mode cleaner length witness channel.  MCO and MCI matrix elements could be stepped in 0.005 incriments without casuing glitches, MCE had to be stepped in incriments of 0.001 and still caused glitches.

We achieved a factor of 1000 reduction in coupling at 34Hz, we then checked (for MCO only) the coupling at 7Hz.  The matrix element had to be increased by 10% to minimise the line.  However at 7Hz the coupling was still a factor of 100 reduction when compared to the original coupling at 34Hz.

The matricies oringially had 0 in the off diagnal, the new matricies are attached, the largest element is about 0.25.