During this intervention (not really coordinated with locking group) it has been forgot that there was an Alp loop trying to stabilize TCS power. Therefore while channels were disconnected, the servo changed the set point of TCS rotator, shining a lot of CO2 power on the input mirror.
This had two side effects:
- as visible in the attached plots the input mirror temperature changed a lot and this triggered a very large etalon effect. The sensitivity is much worse than before, as known in this cases. We expect to have to wait at least 1.5 days before stabilizing.
- theoretically, if TCS powers are the same as before, we should stabilize back a equilibrium to a good etalon point. However, changing ADC already shown in the past to introduce offsets. It seems indeed, looking at B5_2f_ACq which did not come back to original value, that we lost again the reference of TCS power and we are no more using the same power as before.
Again, I think it is important to stress that such actions, which might seem 'transparent', should be better discussed and coordinated with the commissioning crew, to avoid side-effects caused by a partial knowledge of the control systems. For example, the problems of these days could have easily avoided simply informing somebody of the locking group who would have switched off AlpTCS loops for the duration of the action.
To solve the situation now, we should wait some more time to see where we're going to stabilize, and then try to search for a good TCS working point again. To be optimistic, things might go back to normality for the end of the week.
