This is a followup on Ed's alog22918 where he reported seeing MICH glitch around the same time DHARD magnitude became large. Miquel found that the MICH "breathing" was actually a result of bad FFT windowing but it could have hidden the 8.6 Hz line that became apparent in DARM the morning of October 28th (and disappeared on the next day). Plot 1,2 and 4,5 attached show spectrum of DHARD Pitch, MICH, and ground motion near HAM2 during the time DHARD oscillation time reported in alog 22918 and alog 22875 compare to their nominal. Plot 3 and 6 are the timeseries of DHARD ptich and yaw during those time. DHARD oscillation frequency is ~21mHz on 10/28 plot and ~39 mHz on 10/27 plot. To show that this oscillation has nothing to do with MICH 8.6 Hz glitch I've attached a set of spectrum and  timeseries during the time 8.6Hz began to make its appearence in DARM (7, 8, and 9). The low frequency oscillation isn't there and 8Hz peak shows up clearly in every channel.

I will continue to look for any environment factor that might have caused the DHARD instability. Although Jenne did mentioned that a wrong gain could have caused the loop to become unstable. I was told the that gain hadn't changed for months until somebody lowered it recently to help with the oscillation (Kiwamu?).