Yesterday after the maintenance some new structures appeared in the sensitivity (84 Hz and a few peaks above 150 Hz), very well visible also in PRCL (fig 1). The band rms of PRCL at 84 Hz has the shape of the microseism and has a not negligible coherence with that along north direction, while is not coherent with the west direction (fig 2, fig 3). It could be an indication that the peak is generated by a mechanical resonance, which is periodically excited following the main ground motion.

In order to understand the effect of the alignment on this noise, some offset has been applied to the error signals produced with the dither lines. Those actions drive slow adjustment of angular dof which in principle should have small effect on the sensitivity, but as soon as the adjustment started the peaks increased a lot, in particular when vertical offsets were applied. In general, when a new stationary working point was found, the peaks were back to the normal amplitude.

Finally a good marker of the misalignment actually correlated to the peak amplitude has been found: it is ASC_B5_QD2_H_56MHz_Q, the error signal used for BS_TX loop. In fig 4 the range has been predicted using the formula:

    Hrec_Range_BNS = 19 - 50000*(ASC_B5_QD2_H_56MHz_Q-0.005)^2