The attached plot compares the CHARD and PRCL signals at the beginning of the lock and at intervals during 3 hours:
Coherences of CHARD_Y in the noisy condition (3 hours in lock) vs less noisy (beginning of lock)
The increased noise in the 5-20 Hz region is due to... HAM1
Trained new FF filters from HAM1 to CHARD_Y, PRC2_Y, INP1_Y and DC1_Y.
Some improvement in all signals, except DC1, so I turned off the DC1_Y FF.
PR3 damping and IM1 damping dominates the low frequency motion in CHARD_Y