There's some evidence that something strange is happening in the ETMY drive electronics and it's associated with the range drops. The ETMY L2 coil monitors (the noisemons) get quieter by about a factor of two around the same time as the range drops, but there's also excess noise that's not accounted for by the drive signal.

The correlation of strange behavior of the noisemon is not perfectly correlated with the drops in range, and the excess noise in the noisemon does not have the same shape as the excess in DARM and is not coherent with it. However, this does point to further problems with drive problems at EY. So would it be possible to switch control over to EX as Sheila suggested before? The range loss due to this problem is dropping the event rate by at least a factor of 2 (20% range loss cubed), so we should do anything with a chance of solving it. Power-cycling the EY electronics might also help, or checking any fast readbacks of the driver (the version in frames go only up to 1 kHz so if there's high-frequency junk we can't see it.

Attached are three plots. The first is a comparison of the noisemon during the reference (Jan 13 1 UTC) and the bad time (Jan 13 14:45 UTC). The second is a comparison of the noisemons with the drive signals (L2 MASTER UL) subtracted. The transfer function is measured from the data, since it is clearly changing. The transfer functions have the same shape and phase, but are different by a scale factor, about 0.42. There's excess noise during the bad time that is clearly above the noise floor of the coil monitor. The final plot is the RMS of the monitor channel during Jan 13 (ending at the last lock), showing that it starts to drop around the same time as the range. On Jan 8, the RMS is steady for the whole day, while on Jan 12 it also shows a drop during a time when the DARM range drops.