Recent Laser Setpoint Values Different From Lock To Lock: Keep ACCEPTING or NOT MONITOR?
Over the last few days it has been noted that there have been TCS Diffs from lock to lock & they have simply been ACCEPTED in SDF. These Differences have been tiny changes (e^-15) for the (H1:TCS-ITMY_CO2_LASERPOWER_SETPOINT & ...CO2_LSRPWR_SET_POINT_OFFSET).
TCSx Servo Changes Knocking Us Out of OBSERVING Tonight: Known Issue
Tonight, we have been bumped out of OBSERVING twice to some automatated servo changes being made to the TCSx. Drops from Observing were:
- 0:17:05 - 0:19:30
- 0:23:44 - 0:25:26
- Channels in question: H1:TCS-ITMX_CO2_CHILLER_SERVO_GAIN, H1:TCS-ITMX_CO2_PZT_SERVO_GAIN , H1:TCS-ITMX_CO2_PZT_SET_POINT_OFFSET
Basically it looks like the Input, a couple of filters, & output for two TCS ITMX CO2 filter banks are toggled, and then the offset for the TCSx PZT setpoint is ramped.
(Currently messaging Nutsinee & she mentions this is a known issue & has been brought to Aidan/Alastair's attention.)
Yes this was seen before. Changes to values in these channels should not be a reason for the IFO to be taken out of observation.
When the TCS lasers are locked the power output is monitored and fed back to the PZT and chiller temperature. Because we want to lock the lasers to the nearest stable point at their current temperature, the Guardian script scans the PZT and maps the power output, then it calculates the gradient and uses this to set the gain for the feedback to the PZT. There are a couple of entries in matrices that are altered as well, though we could in theory have this all in one value. The gist of this is that there are values that are meant to change every time the laser is locked.
I think we should remove the monitoring of changes to these channels as a reason for dropping out of observation mode.