Runaway locomotive

[nickelplate759]

Bart - I'm no wizard with DCC (That would be @peteski ), but a the creeping behavior during programming is normal and expected.

Regarding the multiple throttles - if you set another decoder to same address as this one, does it also run away?

Having power across the motor leads when the throttle is set to 0 is NOT normal.  Either there is something wrong with the decoder or there is a spurious control signal reaching it (e.g. from another throttle that you aren't aware is sending signals).  The fact that your decoder shows that behavior even when removed from the loco (essentially you built a little decoder tester with a lamp instead of a motor) suggests that your installation in the model was not the problem.  There could still be a short or open circuit somewhere, but if there is it's within the decoder itself.

Test the decoder on an isolated piece of track (or with jumpers) with only ONE throttle attached.  If the problem goes away then there's a signal on the layout that's making it act up.  You could also try setting a known good locomotive to the same address as the decoder that is misbehaving and see if it also acts up.

I've seen similar behavior with a TCS CN-GP decoder - in that case the behavior was caused by a poor connection between the auxiliary board and the engine frame (i.e. an open circuit).

Good luck!

[Steveruger45]

Have replied below as addendums to your post.   Hope it helps.

Yes, I have more than one throttle on my layout. Visually, all are throttled down.

As Jbub suggested, I did take the decoder out and put the locomotive on the track. It sat perfectly still.

Comment- Ok then loco is ok.


I don't know if this was a valid test (my electronics knowledge is limited!), but I took the decoder out, carefully attached wires between my track and the pads that would touch the loco frame. I then used my old Life-Like track tester (basically 2 wires, a plastic housing and a lamp) and touched the wires to the pads that would make contact with the motor leads. The lamp illuminated. I then did the same with another decoder I had recently retired. When I touched the tester wires to the motor pads, no light.

Comment - hmmm. That behavior doesn’t sound right so I think decoder is still sending msg to motor to run and maybe a short on the decoder components ( rectifier bridge) or it’s not fully reset from previous programming.  When you did a reset did you interrupt the power afterwards, (That is remove it from track supply) ?

Am I right that power should only flow through to the motor pads when a signal is sent to the decoder to allow a certain amount of power to be sent to the motor? If that's correct, doesn't this also suggest that the decoder I am having the issue with is letting full power flow through to the motor pads (which would cause my engine to run wild?) Would this also suggest that the decoder got fried somehow?

Comment- see earlier comment but it could be a fried decoder.

Just throwing this out there - After I had swapped decoders between my 2 locos, I put this one on my programming track to read back the loco address. It came up with the address of the loco I removed it from. So I changed the 4 digit address to the new loco it was in. I noticed when it was writing the new address that the motor started for a brief moment and moved the loco maybe 1/4" to 1/2" on the programming track. Not sure I've seen that before. Might that have had something to do with my predicament?

Comment- ah, seems as if old loco address was still in it then it didn’t fully reset.  Suggest try another reset and fully remove power to the decoder after for a couple of minutes and try again before giving up on the decoder.


Thanks,
Bart

[Bart1701]

Have replied below as addendums to your post.   Hope it helps.

The sequence that happened was:
Swapped decoders between 2 decoders
Read address of NCE decoder.
It showed the address of the loco it had been in.
Wrote the new address in.
Noticed that the loco moved a bit on programming track while writing the new address.
Confirmed the new address by reading it back.
Loco was placed on the main and power applied to the main.
Loco becomes a runaway.

After it was suggested to do a reset, I did so.
Loco address now showed as 03.
I cut all power to the layout (includes the power supply to the programming track).
Loco is still a runaway.

I'll try to do another reset as suggested to see if it helps.

Thanks,
Bart

[Steveruger45]

Bart, good luck but it appears from your descriptions of what was done and it’s manifested behavior that the decoder is toast.
 BTW and as I think has already been mentioned, Locos moving a little bit (jerking) is normal when it is being read and written to during programming.  That said, I haven’t seen one lurch a half inch when programming,  that to my mind is a problem.

[Bart1701]

Additional follow-up...

I changed the loco address to 004 and checked it. (Did this since the address was already 03 and wanted a way to confirm the reset that was going to be done worked)
I did another decoder reset.
Removed from track for a few minutes.
Checked address and it was now 03 (shows me the reset was successful).
Put on the main and applied power.
I still have a runaway.

It is starting to sound like the decoder got fried somehow when I moved it from one loco that it was working in and moved it into the other locomotive. It wouldn't be the first decoder that I have fried (done this once or twice before out of 50+ installs    ). Does this sound plausible?

Thanks,
Bart

[Maletrain]

With only this decoder (not others) making a loco run away at both an old 4 -digit address and the reset short address of "3", it seems pretty clear that the problem is not another throttle on the layout that is sending a speed command.  With the decoder out of the loco and still showing power to the motor (actually, the substituted light bulb), it seems pretty clear that the decoder is somehow sending high power to the motor.  Since the decoder has already been reset, it seems unlikely that it is going to be fixed, unless you can see some sort of physical problem on the board, itself.  Did you do any soldering on the board that might have accidentally joined two circuits by accident?  Are you sure the outputs were properly connected - have you tried turning off the headlight with a DCC command to see if the light bulb substituted for the motor goes off or remains lit?

[Steveruger45]

Additional follow-up...

I changed the loco address to 004 and checked it. (Did this since the address was already 03 and wanted a way to confirm the reset that was going to be done worked)
I did another decoder reset.
Removed from track for a few minutes.
Checked address and it was now 03 (shows me the reset was successful).
Put on the main and applied power.
I still have a runaway.

It is starting to sound like the decoder got fried somehow when I moved it from one loco that it was working in and moved it into the other locomotive. It wouldn't be the first decoder that I have fried (done this once or twice before out of 50+ installs    ). Does this sound plausible?

Thanks,
Bart

Sorry to say but yes.

[davefoxx]

Just throwing this out there - After I had swapped decoders between my 2 locos, I put this one on my programming track to read back the loco address. It came up with the address of the loco I removed it from. So I changed the 4 digit address to the new loco it was in. I noticed when it was writing the new address that the motor started for a brief moment and moved the loco maybe 1/4" to 1/2" on the programming track. Not sure I've seen that before. Might that have had something to do with my predicament?

Thanks,
Bart

If the locomotive moved a fraction of an inch, that's okay.  It means the decoder is being reprogrammed.  That's normal.

Did you try putting the decoder back into the original locomotive to see what happens?

DFF

[davefoxx]

Sorry to say but yes.

If so, destruction by static charge?

DFF

[MK]

Sounds like a bad decoder to me.

This happened to me a while ago with a brand new out of the bag Digitrax  decoder.  Exact symptom as yours, loco runs away as soon as it touched the tracks.

I purchased two identical decoders so I tried the other one and everything was fine.  So I started to examine the components on the defective one.  Turns out there was a fine solder bridge between adjacent components.  It was either the bridge or the amplifier component.

I took a very fine soldering iron and broke the solder bridge and everything was back to normal.  No permanent damage.

As a last resort check the decoder for any burn marks from fried components.  I doubt a solder bridge was on yours since you said it was previously working on another loco.

[lajmdlr]

MK is right except NCE decoders come with yellow tape.
See photo.
You want only the Frame tabs to touch the frame. You can apply the tape to the underside of the decoder but you would have to make a cut out so the motor tabs can contact the motor terminals.  I usually stick the tape to the frame and trim once stuck in place.  (Attachment Link)
Is there a chance of the motor terminals being pressed sideways and touching the frame.  If yes, then wrap some caption tape around the frame in those places too.

By "caption tape" do you mean "Kapton tape"?

[Steveruger45]

Ha ha it’s auto correct doing it’s  very annoying correction of my always perfect spellinh 😂