Some of our members are using switchable bandpass filters in their stations. In particular the fellows who operate SO2R or multi-op. The objective is to be able to listen or transmit on two separate bands with reduced interference between the radios on the two or more bands.
I bought two Dunestar filters from one of the MWA members. At the time we made the deal I had been advised that one or two of the bands did not seem to operate correctly. I accepted that condition as a part of the purchase.
The curves for the filters show that 10m on one filter set and 15m on the other has been affected by previous operation. This could happen if the antenna input to the filter is significantly different than 50 ohms and the operator uses an internal tuner to adjust the radio output so that the SWR seen by the transmitter is very close to 1:1. Such an adjustment only causes more power to be sent to the filter and the result can be destroying the fixed capacitors in the bandpass filter because they are operating a excessive voltage or current.
In many radios it is possible to turn on the internal 'tuner' and it will sense SWR and automatically correct it. That is just fine for the radio's output, but might be devastating for the bandpass filter. If you are operating SO2R or multi-op it probably should be one of the things on your operating checklist to turn off the internal tuner on all bands before beginning the contest.
The SWR curves of my Dunestar filters were made using a precision 50 ohm, non-inductive resistor as the 'antenna'. The curves show that that even if the antenna load is 50 ohms on 160 and 80 meters, the SWR rises at the high frequency end of the band. If the actual antenna also has SWR increases at the high end of the band the power supplied by the transmitter may be sufficient to cause the high SWR filter to be damaged. In the cases where the filters go to an amplifier input generally the input will be 50 ohms and one might feel there is no risk in operating at the high end of 80 or 160, but there is because the filter even with a 50 ohm load has high SWR at the top end of those bands. The risk with ICE filters is less because they are broader. Because they are broader they are not as good as the Dunestar filters in suppressing out of band signals.