Robert, Anamaria

As previously reported, there are some variable peaks in DARM around 550, 820, 1100 and sometimes as low as 270 Hz (see especially comment 4566 and its plots, eg this). We now think the PZT is also an unwilling witness and not to be blamed. 

Some days are worse than others for this noise in DARM, March 14 being quite bad for the coupling, see summary page plot attached.

The magnetometer in the ISC rack (called LSCRACK) was already seeing this noise very well, so we took a temporary magnetometer, scaled it to match the LSCRACK one and moved it around to find the largest location.

By putting paper on the IO chassis fan intakes of LSC and ASC, we were able to move their frequency, so we confirmed that that's where they're coming from, see plot attached. Their power is already separated on its own power strip in this rack, though it may connect to others father back towards the DC supply racks.

Robert has worked a bunch on IO chassis stuff, and it seems his work at the end stations holds for the corner as well. The IO chassis fans put noise on the grounding so much (in the sense of what a magnetometer sees) that I spent half an hour chasing ground wires around, thinking they would lead me to a different source. 

We have tentative approval from Keita to work with Richard next week and power the LSC and ASC IO chassises (or just their fans) from a supply plugged into the wall such that there is a good few 10s of ohms between its ground and the other sensitive electronics.