Evan, Kiwamu,

We had a difficulty in locking the IMC in this afternoon after the FSS activity (see alog 17875). It turned out that the signal in the IMC trans PD was too low to trigger the IMC-MCL filter. Note that the upper and lower thresholds have been set to 100 and 90 cnts respectively for the nominal 3 W operation. We could have simply increased the output digital gain of this path, but at the same time we noticed that the analog signal was pretty low to begin with. It was about 20 cnts at the ADC when the IMC was locked on a 00-mode. We went to the IOT2L table and checked whether there is a clipping or some stupid things, but we did not find anything obvious. Since I did not like an analog signal to be as low as several counts at the ADC, I decided to crank up the analog gain of the PD (PDA100A) from 0 to 30 dB. Now it detects more than 200 cnts when the IMC is locked on 3 W. Evan then adjusted the digital gain such that it keeps the same conversion. We did not get a chance to subtract the dark offset yet, but the IMC is now locking fine.

Note that it seems that the trans PD has been in this low-ADC counts situation since approximately Feb-25 at which point the signal in IMC-TRANS decreased dramatically for some reason. See the attached trend. Could this be due to the LSC/OMC model split (alog 16893) ????