I think the problem has been solved. Let me re-trace the steps:
First, I reset the decoder to default on the programming track by setting CV 8 to 8, then reprogrammed it. The programming track is on the opposite side of the layout room from where the problem surfaced, so I tested the locomotive near the programming track. I intentionally created both shorts and opens, and the loco recovered each time and kept running. Back on the branch where the issue first surfaced - it did again. Following an intentionally caused short or open it would not work, but it did on any other part of the layout. Note also that no other locos I ran on the branch were affected.
The particular branch line where the decoder "timeout" was occurring is maybe 80 feet or more from the command station and - I twisted the bus wires but hadn't terminated the track bus. I added termination to the bus as in http://www.wiringfordcc.com/track_2.htm#problem_blown_solution and the decoder functioned normally.
Again, no other locos were affected, regardless of the type of decoder, so SOMETHING about this specific decoder apparently made it more sensitive to the lack of bus termination.
Rob Spangler
Thanks Jim. I was suspecting a "hardware" type problem, but maybe a reset will help. Don't know why I didn't try that already.
Rob,
It could now have an 'intermittent' problem in it. Have you tried doing a 'Decoder Reset' on it and then key in the address again? Digitrax decoder usually can be reset by changing CV8 to a value of 8. Once that is done, program the address back into the decoder(it will be reset to '3' by the decoder reset). If the problem shows up again, it may be time to invest in a new decoder...
Jim
Modeling BNSF and Milwaukee Road in SW Wisconsin
Bump - anybody heard of this? Anyone? Bueller...?
Last night's operating session saw some odd behavior from a Digitrax DH 165 AO decoder mounted in a Kato mechanism. While this locomotive has been running fine since the decoder was first installed, last night it stopped responding after an operator shorted by running though a gap in a live frog turnout, or encountered a dead spot. After the short or open cleared up, the decoder wouldn't respond to any throttle input unless it was removed from the track (or tipped so the wheels on one side stopped making contact). Once back on the rails it was OK until another instance of a short or open. This happened about 4 times.
Has anybody encounted this phenomenon? If so, can it be solved without replacing the decoder?
*EDIT* I think I figured this out. See post #5 in the thread below.