I while back a freind of mine who worked for the CSRIO (massive research company in Australia) borrowed a giga hz storage oscilloscope worth 10's of K$ and we started to do some experiments.
On all the best pots mentioned in my first post, what we found was that during fast pulsed millisecond transisents at cd levels, audio band squarewaves, then stored and reviewed frozen on the scope, the leading edge was getting corrupted sometimes ringing, this leading edge has a very fast rise time > 100's volts per microsecond, and what we found was that the wiper on all of the pots were getting bounced, if you like off their tracks during these fast transients, (it was as though the contact area of the pot had become a diode and it was trying to rectify the ac signal) if we pushed on the wiper to give it harder contact to the resistive track the corruption stopped. Naturaly this would wear out the track in no time with this pressure. This also happened to switched resistor pots and also input switches, though to a lesser degree.
