These plots show that it is possible to completely damp out the fundamental L and P modes of the quad, that are causing the cavity pain. The damping method here is called 'full state feedback'. It is like modal damping in that it uses an estimator to compensate for a shortage of sensors around the quad. However, the controller is a matrix rather than a set of filters. The advantage of the matrix feedback is that you can solve the matrix for a set of desired closed loop poles. So here I have specified that the first two modes will have a Q of 0.67 in the closed loop. The others I specified between 10 and 50.

The price you pay for this controller is absolutely terrible noise performance. Also, it is not AC coupled. But, if the RMS test mass motion is a real problem for the cavity, this can be used to suppress it.