The death and resurrection and death and resurrection of a DT SDN144K0a

[OldEastRR]

OK, so I was running the KATO E8 with this decoder, shell off. It stalled, I touched the loco on the board with my finger to give it a nudge -- poof of white smoke. OK, I fried it somehow (I thought). I put it on the dead line and told my friend who has lots of DT decoder experience what happened. He said he wanted to look at it, took the decoder out, found one of the motor contacts had broken off (KATO part). He replaced that, carefully put Kapton anywhere anything could short, re-assembled the loco -- and it ran! Sound and all functions!
At least ran for a little while. Then smell of burning electronics and another stop. Removed from tracks, let set, put back on tracks ...works! Except now it will either stop every fraction of an inch and restart the sound sequence, OR, run smoothly along without a hitch. (might be just dirty track, tho)
WTF is going on?

[cgw]

Simply put...Its toast!    First sign of smoke means a short on one of the components that heated up enough to char and vaporize the encapsulant on a circuit component .  Eventually the decoder will completely fail!  The spastic performance  is from various circuit components heating up and failing.     I would pack it up and send it off to tech support at Digitrax   It may be time to upgrade ther decoder to one of the newer ones on the market.  Loksound 4 will blow the sdn out of the water in sound quality and speed performance.         

[daniel_leavitt2000]

Blew one of the capacitors right? That would "sort of" still be functional but with no power reserve for even the smallest intermittent drop outs.

[OldEastRR]

Hmmm.... could be. This decoder has a huge-a$$ "can" capacitor which doesn't seem to be much of a keep-alive. Looks too big to blow up but .... it would explain the intermittent start-stops. Are there capacitors on the board itself? Where did the smoke and smell come from?

[cgw]

Most likely it was one of the transistors that blew.  look for some white/grey smoky appearance on the surface of the black chips.  Usually when the ic fails it leaves a very tiny pin hole that can been seen under small magnification.  when caps blow they tend to do some physical damage to the surrounding area and are real easy to spot.  The big round electrolytic cap is for the sound circuits and not the drive.  If you look at the top (opposite end where the wires attach) and it is bulged or ruptured the it has failed.    You would have heard a phaaaat or pop when it failed.   Can you post a close up picture of the failed board?

[OldEastRR]

Can't see anything wrong with the board
[/img]

[peteski]

Unfortunately that photo is neither sharp enough or close enough to the components to be able to tell you if there are any damaged components there.  Sure, some components fail spectacularly with very obvious damage, but in most cases there is only a faint fracture line on the black resin case (the encapsulating resin) and sometimes a dark color liquid might be visible where the leads come out of the case.

Often it takes a very bright lighting and magnification (and examining the parts from multiple angles) to spot a failed (fried) component.

Pushing down on this type of decoder might cause a short between the metal chassis and one of the copper traces on the bottom of the decoder board.  Yes, the decoder is coated with a green-colored solder mask which is an insulating coating, but if there is a sharp metal edge on the chassis under the decoder and your press the decoder board down hard in that ares, that protective coating can be penetrated.  I like to cover the bottom of the decoders in this type of install wit Kapton tape for extra protection.

[djconway]

Get your nose in there and sniff,  I've found quit a few toasted electrical by the smell of the magic smoke.

[peteski]

Get your nose in there and sniff,  I've found quit a few toasted electrical by the smell of the magic smoke.

I have used that method back in the '80's when I was an electronic tech repairing computer boards.  But back then the through-board-mounted  DIP chips were much larger than today's tiny and densely packed SMD components. I doubt I would be able to pinpoint one of those by puttign my nose to the board. Unless maybe I could somehow sniff them using one of those skinny coffee stirrer straws.