Yesterday morning some modifications of ITF control and working point have been performed:

- WI mirror have been shifted by 860 um horizontally (fig 1) and 480 um vertically (fig 2), in order to zero the corresponding dither error signal. The other six dither signals are in loop (NE dither --> NI alignment, WE dither --> WI alignment, NI dither --> COMMp reference), so now the optical axes of the long cavities are centered on the mirrors. In principle the WI dither could be kept at zero closing a slow loop with WI displacement. If needed, it will be done.

- The amplitude of dither lines has been reduced by a factor of 10, going back to the standard value used in low noise 3. The up-conversion visible in DARM at 10-20Hz has been removed.

- A full bandwidth angular control of BS through B5_QD has been implemented. In this way we get rid of the BS local control glitches.

- A new balancing of BS marionette longitudinal actuation has been impemented, obtaining a further improvement of BS angular stability.

In the following plots the big improvement in many relevant signals is shown.