[Satoshi, Jenne]

This morning we balanced the test mass actuators to each other (ETMX to ETMY and ITMX to ITMY) and then balanced the actuation of the ITMs relative to the ETMs for soft / hard decoupling.  Unfortunately this didn't help the mysterious phase situation in DHARD_P at high power.

To balance the ETMs to eachother, we used the ASC oscillator that goes through the output matrix, and pushed on the ETMs in pitch with coefficients +1*ETMX and +1*ETMY to actuate in common pitch. The line was at 15.9Hz with an amplitude of 30k counts.  We then adjusted the gains in the L2 LOCK filter banks (H1:SUS-ETMX_L2_LOCK_P_GAIN) to minimize this line in AS_45_Q_Pit.  This oscillation was stopped, and then the output matrix was changed to push on the ITMs in pitch, again with coeffs +1 and +1.  For the yaw balancing, we repleated the same test, but pushed +1*ETMX and -1*ETMY to be actuating in common yaw.  Similar +1,-1 coeffs for the ITMs in yaw. 

The L2 coefficients for the test masses are different from 1 by about 5%, so not very much, and also not very different from what they used to be.

Once that was done, we then actuated in DSOFT (again using the ASC oscillator, at 15.9Hz), and adjusted the elements of DSOFT to minimize the line in AS_45_Q.  For pitch, the elements changed from 0.87 to 0.74.  For yaw, the elements changed from 0.87 to 0.72.  

The DHARD OLG at 2W didn't change when we changed the output matrix elements.  Also, when we powered up to 20W, we still see the not so great mysterious phase situation, so this balancing did not magically fix that stuff.  We are able to power up to 30W and do most of lownoise_ASC, but as has been true in the past few days, the DHARD_P lines of code are commented out.  Trying to implement them by hand (lowering the gain, and engaging a cutoff) made the 1.1Hz motion start to ring up, so we reverted them back to the high bandwidth situation.  Frustrating.