As was shown in measurements reported in entry 10262 it's possible to have an idea of the ground vibrations injected into the IM via the oplev by measuring the coherence between an accelerometer mounted on the oplev platforms and the IM degrees of freedom.
This report shows the effect of including notches in the DC feedback filters controlling the IM-P with the oplev signal.
- I added two filters to the filter bank: notch_oplev and notch oplev2.
- The channel names of the accelerometers and their physical locations are given in entry 10405.
The figure shows the coherence between IM-P and the accelerometers. The blue trace was measured using the new notches and the black one is without the notches. Many peaks are lower with the notches. It's worth pointing out the following:
- The large coherence peaks at 16-18 Hz reported in entry 10262 don't appear in the new measurement. I wonder what might have changed.
- The height of some coherence peaks seem to change from one measurement to another.
- I began measuring with 30 mHz resolution and when I changed to 10 mHz some peaks appeared or were higher. Therefore it's better to work with higher resolution.
In general, removing individual peaks with notches seems to work but likley we should do it on an as-needed basis rather than for sport. Another strategy would be to roll-off the control filter but that requires more care.