-get the times of 2 hours of glitches using the command
omicron-print channel=V1:Hrec_hoft_16384Hz gps-start=1266249618 gps-end=1266256818 freq-min=25 freq-max=35 snr-min=15
Taking the difference between showed alternating intervals of roughly 83 and 269 seconds. Since two intervals were too long, 2 fake glitches were inserted at the approximate position
-generate a square-wave signal that flips sign at the moment of the glitches
-do a brute-force correlation between this signal and all channels in trend.ffl. The winning channel is LSC_NI_HB_moni_mean, which shows perfect correlation. This channel is likely is some monitor of the current sent to the heating belt that is used to actuate on the NI etalon. Fig 2 shows that this is indeed correlated to glitches in the Hrec (band-passed from 25 to 35 Hz). Fig 3 shows that the signal is indeed going crazy since a few days. To be understood by the experts if this is some numerical issue with the digital control loop, or with the actuator itself.
