We tried to add in extra boosts to the C/DHARD P/Y loops so that we could reduce the amount of bilinear noise couplings whose low freq coupling coefficients were modulated by alignment fluctuations. We added the boosts without increasing the loop UGFs, so that we would not increase the amount of ASC noise contaminating DARM.

In the first two plots we show the new boost filters we used. The first one is for C/D HARD PIT and the second one is for C/D HARD YAW. Mostly they are designed to invert the 10 W sus plant (but with lower Q for the resonances than those from the sus plants; hopefully this should make them more robust against power fluctuations which modifies the plant).

Those filters are named as "boost10W" and are engaged in the LOWNOISE_ASC state.

=============================================================================

We also tried to compare the test masses' angular motion between now and O2 using oplev's.

Please see the third plot for the comparison of the OPLEVs' outputs, and the forth plot for the ground motions. The reference traces (pink/cyan) are from O2 (8/8/2017 11:00:00 UTC), and the red/blue ones are from today.

Notice that the ground moves x10 times more below 0.2 Hz now than O2 due to the high microseismic motion. On the other hand, from the oplevs the test masses' pitch motion are comparable between now and O2. For XARM we actually improved the pitch motion. For ETMY it moves about factor 2 more than the O2 reference, yet given the fact that the input noise is 10 times higher, our current ETMY residual motion is still improved. For IY, the excess motion seemed to be consistent with the excess input motion. This is also the case for YAW, where the current excess motion is consistent with the excess ground motion input.

Overall, our current ASC + ISIFF performance should be satisfying.