Sheila, Jenne, Rahul
I went through all the monitor filters for the violin modes to confirm the following,
The band pass filter has a gain of 120db and the bandwidth is 0.02Hz. The bandwidth at 80db is 0.04Hz.
With this boundary condition, I found 4 modes (ETMY 1,6 and ETMY 18,20) which are very close to each other (or are pairs), i.e. within 0.04Hz. The bode plots for all 4 modes are attached below.
In fact mode 1 and mode 6 and only 0.011 Hz apart and they are sitting on the flat top of the band pass filter (which is not good, since they are creating beating and hence Guardian cannot effectively damp them). Sheila suggested that we can use notch filters to reject signals in this case.
...adding a few more lines to bring clarity to the above alog,
The aim was to hunt for pair modes in the monitor filter which are very close in frequencies and hence could be crossing over into the narrow band pass filter of each other. Since the gain for the narrow band pass filter was defined as 120db, we looked for modes which are crossing over at 80db and above. The bandwidth has been user defined and is consistent for all the violin mode monitor filters. Below 80db, the beating of the two modes won't be that significant and hence can be ignored.