Yesterday, at the beginning of the afternoon activity, there was some difficulty to acquire the lock because B4 beam was not centered anymore on the photodiode. The centering was recovered tilting in TY the POP (PR_F7_TY) and putting B4 power in the middle of the plateau. Then the lock was acquired, but the sensitivity was bad. We found that the position of B4 on the camera was not yet the standard one; doing the final adjustment of the POP the sensitivity improved a bit (Fig. 1).
Later in the afternoon we explored some different working point for recycling cavity alignment. We found soon that a small adjustment of BS set point had a visible effect on the noise, namely the structures associated to SDB1. We found also that the idea of driving PR cavity towards a higher sideband power cannot be used to optimise BS alignment, because both for TX and TY the noise increases soon moving in that direction. Finally we found a new set point for TY which seemed to improve a bit the range (Fig. 1), but this has to be confirmed, and a good procedure to define the optimal BS aligment has still to be found. The last point to be underlined is that after a change of BS alignment the DIFFp set poit has to be adjusted: probably we are using a signal not really diagonal with respect to diff arm alignment.
We checked also PR set point. For that, it seems that the good alignment can be defined as usual, i.e. the one which miximise the sideband power and minimise its sensitivity to PR angular fluctuations.
The exploration confirms that in general we cannot relay on the zero of the AC signals obtained from the quadrant, in order to define a good working point for the most critical angular d.o.f.. The way to found the correct offsets to add is not yet well defined, even if there is a kind of euristic procedure which can be applied from time to time.
