Kiwamu, Matt, Stefan

We were still suspicious of the 2-omega RF buildup signals, so we decided to put a line on the PRM length at 303Hz, and demodulate the line. To avoid loop feed-back we turned on the existing "notches" FM8 filter in PRCL1. POP9_I was demodulated at 303Hz in H1:LSC-LOCKIN_1_DEMOD_3_I_OUTPUT.

Assuming good overlap, we would expect the gain to scale as sqrt(carrier recycling gain * 9MHz recycling gain).

Interestingly, the 9MHz sideband drops by about the same amount as the carrier, but with a timne constant of about 10min (vs about 1 min for the carrier).

The attached plot shows the power increase to 40W, without the soft offset adjustment that usually happens right away. All scales (except purple) start at zero at the bottom of the plot.

Conclusions:

- POP18 is not reliable - it suggest a larger gain drop than the optical gain.

- The optical gain (RF9 & carrier) does drop by the same 25% as the carrier, but on a much longer time constant.