Deeper than just saying clear out the grease. 2 in 2 days, first input selector of an amp , not explored in depth, second vol control of a music centre.
First had radial contacts that were slightly staggered and the fixed contacts were in line , second was exact radial contacts and staggered fixed contacts. On the second I took some R measurements before and after. There are 3 contacts, one common and the other 2 staggered to pick up CW or CCW rotation. Just measuring between the 2 sense contacts as at some point they are cross connected, measured 2.5R each of the posistions but very tiny position to get this reading. Ladled out and then dissolved the grease, reassembled and took readings again . Now 1.5R each bridging posistion and a lot easier to find that posistion, ie more latitude. I imagine from the consistency of the R readings that something to do with deforming or block under one of the contacts reducing the contact to a very small area and then only marginal connection posistion bridging the 2 contacts. I assumed the electronic sensing was one line before the other but perhaps duration of both contacts on simulataneously, or not, is also something to do with normal sensing