SD9's back in service

After inquiring with Digitrax on the stall issue (see last post), they got back to my email in less than an hour!  What was confusing me was the fact that one loco would stay running while the other stalled, even though they were MUed together...  So they recommended resetting Ops Sw 36 and 38 on the DCS100.  I thought that would do nothing, but after resetting they appeared to be back to normal?  Thinking that it was just coincidence, I kept running the locos MUed together for an extended period.  Never stalled when hooked together again.  So then moved to running them independently, but then each was stalling again.  I tore them both down again and re-cleaned the wheels and needle bearings.  After inspecting closer with a 10x magnifier, it appeared there was maybe oxidation or residue that the alcohol missed last time around.  This time I used brake cleaner and burnished the contacts followed by alcohol.  There is a heavier current draw for the sound decoders, and it looks like they are way more prone to drop-out as compared to conventional decoders.  For now, they are running much better, but still a bit temperamental (more testing needed).  I suspect for these two units that meticulous cleaning on regular intervals is going to be necessary, but have not found the need to do so on any of my other sound units...  I have the nod of approval to return these decoders for exchange if needed, as they mentioned it is possible for a weak spot on the decoder.
Final Analysis at this time:  inconclusive