Monday (Dec. 19) evening we replaced the L2 coil driver for ETMY in order to check if it was the source of glitches, as suggested by DetChar. By Tuesday morning there had been a run of a couple of hours with the new driver, and I found at least 3 glitches with similar spectrogram/omegagram signatures during the run with the substitute chassis. Figure 1 shows one of these glitches during the period when the substitute chassis was used. Since typical glitches were still common, we put the original chassis back in during maintenance on Tuesday to avoid re-calibration. When the interferometer came back, the glitch rate was lower. It is thus unlikely that the original chassis was the source of the glitches. One possible reason for the increase/decrease in glitch rate would be a decrease/increase in humidity as the outside temperature rose, similar to what was found by Paul S. and others for O1 (https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=28534). It would be interesting to check how well glitches of different SNRs correlate with humidity, and if the correlation is better for any of the "RELHUM" channels located at multiple sites in the 3 buildings.