Adrian Helmling-Cornell, Robert Schofield
We wanted to check whether blip glitches were produced by data corruption or other problems with the data acquisition system so we used Omicron to look for glitches in an four day period (22:00:00 UTC Aug 13, 2020 to 12:00:00 UTC Aug 17, 2020, 56502) when there was no light on the DC photodiodes.
We wondered whether H1:OMC-DCPD_A/B_OUT_DQ would have blips during this period, as those channels consistently experienced glitches similar in morphology to blips in DARM during operation (see figs. 1-3 for an example). Blips in these channels when the light was off could be indicative of blips being produced somewhere in the DAQ or CDS systems. From the set of Omicron triggers in the OMC channels during the time the light was off, I only considered those which were similar in duration and ending frequency to the population of O3 blips. I also only considered glitches which had an SNR greater than or equal to 6. A total of 18 distinct trigger times satified these conditions. Of these remaining triggers, none were similar in appearance to blip glitches (see highest_a/b for the Q transform of the highest SNR trigger or figs. 6 & 7 for more examples). I could not find evidence of blips occurring in the OMC photodiode channels when the IMC was shut down.
You probably should add a DC offset to properly mimic the locked state, like it was done at LLO.
@Anamaria I think this is a different test than the LLO test. We looked at multiple days with Omicron for blip glitchs from the DAQ system, not giant glitches from the OMC electronics, as was done at LLO. We dont claim that this clears the DAQ system, just eliminates a lot of the possible mechanisms, but does not eliminate DAQ system mechanisms that depend on the RMS or spectral shape of DARM.