A scan of BS vertical position has been performed, while a line at 26 Hz was injected on BS TX, and it was visible in DARM. It has been found that BS_TX --> DARM coupling can be reduced if BS is rised a lot. The scan has been stopped after a shift of 5 mm, but the minimun coupling was not yet reached. The projection of the zero crossing says that 2.7 mm were still missing (fig 1).
We use to consider the angular to longitudinal coupling as a good measurement of the beam centering on the mirror, so it would be better to change BS vertical position in order to have the mirror correctly centered. Yesterday BS has been put back almost in the previous position, in order not to change the tuning of the vertical attenuation stages. The correct way to do the shift is by the top stage screw (chiocciolone). It is worth to remark that no relevant changes has been observed in the sensitivity during the scan: the only thing which clearly changes is the coupling of BS TX angular noise, but it is currently negligible.
The optimal BS vertical position is a bit too high, for the constrains determined by F7 sensing/actuation. But also NI and WI setting are a bit high with respect to F7 sensing, so there would be the possibility to align everything in a different position of the input beam (let's say: 2 or 3 mm lower). The impact on DET has to be checked.
Last remark: also PR is working in a very low position with respect to F7 sensors. For PR, a simple scan of the vertical position is impossible, because any small shift requires an adjustment of the input beam inclination. This means that currently we do not have a clear confirmation of the centering correctness of PR, and an activity devoted to verify/optimize that is a bit long and invasive (one entire shift, at least). But I think October stop is the good time to do it.