Testing for bad decoder..

[lappdawg27]

Quick DCC question. I just sold my layout and my locomotived to somebody and wouldn't ya know, the first time he ran the train something failed. I went over to his house to check things out. The loco he was using was an Atlas Master B40-8 (w/ DCC built in). As soon as you place the loco on the rails, the DCC system shuts off, which normally means a short. You couldn't do anything to program it either, it would just show blank values then give you an error. Well I checked for obvious things on the engine itself, and didn't really see any problems. Oh yeah, I did run another engine around the layout and it worked fine on DCC.
Anyway, I brought the locomotive home and took the shell off. I got an electrical meter out and checked for shorts between the wheel sets and the pick-ups on the DCC board...nothing. The only place that showed continuity was between the motor leads (is that a prob?).
The next thing I did was take it to the local hobby shop. Upon placing it on the test track (while under DCC control) there was a short. I then changed the jumper on the DCC board to analog. The guy switched his test track to analog and it ran for about 10 seconds, then died. Turns out it blew the fuses to his controllers!!
Anybody have any idea what this could be? This loco was not used very much at all (maybe an hour at the most) and is about 5 years old?

 

[CanuckMark]

You'd get "continuity" of sorts at the motor since it is an active impedance and at rest is fairly low passive impedance, so that sounds normal. If there is no short at the wheels then it must be the decoder as you've figured. Is it a metal shell? Maybe metal truck sideframes touching something?

Another test you can do is bypass the decoder by connecting the motor to the wheels (with the decoder disconnected), test under DC and see what happens. Also check (with everything disconnected) that there is NO continuity between either motor terminal and either side of the wheels - that'll smoke a decoder quicker than something something (no creativity tonight ).

If it is the decoder then you're kinda beat unless you want to actually repair the decoder. The usual failure point I have run in to is one or both of the MOSFET chips (8 pin SOICs).

Mark

 

[UP2CSX]

Mark, you really know this electricity stuff. Sounds very much like a board level short since it won't run in DC analog mode either. The easiest way to prove this would be to swap out another known good decoder with the presumably faulty one. If the engine runs well with the swapped decoder, you know it's a bad decoder for sure. Unless you have Mark's electrical skills, it's a trip to the trash can for the bad decoder, especially since it's so old.