Is The Resampler Really Capable Of Switching Reference Channel?

[James_McN]

We have had an issue where an application we have written based on the power quality examples was switching reference channel unexpected.

The switching was caused by a bug in the example channel switcher which I have reported to NI directly.

However it highlight an interesting issue - when the reference channels switchs there are a lot of spurious events generated - I believe due to filters in the resampling algorithm taking time to resettle (we see these for about a second).

The question is, is there a way to use this and switch reference channels without causing filter effects? What other affects might this cause? We see ripples on dips which I believe to also be caused by the filters - could this be true and if so again, what could be done to remove these?

I attach a screenshot of one such dip generated in the lab.