I increased the chilled water flow in the PSL crystal circuit in a repeat of a noise test here: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=30290 .  The results were similar, a noticable increase in DARM between about 300 and 1000 Hz. We might be able to turn down the chiller flow one or two lpms and still be in spec, my understanding is that the holdup is getting a flow meter onto the crystal branch of the crystal chiller water circuit.