Need help with troublesome engine

[DWS3]

Hey everyone!

I have an older Atlas n-scale engine that is giving me fits and I'm hoping someone has some ideas. Here is the setup:

Atlas GP40-2, purchased in 2000. It is DCC ready (split frame) and had a Digitrax non-sound drop-in decoder installed when I bought it. It says made in China on the bottom and has Atlas 1997 molded into the side of the metal frame. It ran fine for over 20 years. After sitting in storage the last couple of years, it refused to run more than a couple ties and then would quit. I thoroughly cleaned and lightly lubricated everything to include the trucks but no luck. After a few more attempts to run the loco, it completely quit working and my NCE programming track wouldn't even recognize there is a decoder.

Just to make sure the engine was still good, I put the original light board back in and ran it on a test loop of DC track for a couple hours and it ran great, if a little loud. It was never a particularly quiet motor. Since it works well on DC, even at low settings, I bought an ESU Lokpilot drop-in decoder (59751). It fit but was fairly loose and again, the DCC system wouldn't even recognize a decoder. I tried it again on DC and it still ran great. Since the fit was loose, I soldered wires to the brass pickup strip for input and to the motor tabs and then soldered the respective ends to the Decoder pads. The DCC system still doesn't recognize there is a decoder. Just to check the DCC system, I put several other engines on the programming track and they all came up just fine in the cab and DecoderPro. I even tried removing the motor, brass strips and decoder from the engine frame and attached the brass strips directly to the individual programming track rails, just in case it was a engine frame issue. Still nothing.

At this point, I could try buying another new decoder but hate to spend the money if it is something else, especially on an engine that I like OK but was never ran as well as my Katos or even my Bachmann Steam engines. I have a matching engine that I bought at the same time and has a hardwired Lenz decoder. It still runs just fine. Ideally I would like to consist them. At this point, I'm not sure if I just got bad ESU decoder as well.

Any thoughts? Thanks!

D.W.

[tehachapifan]

Sounds like it may have a short somewhere. A motor-to-frame short is fatal to a decoder but may not affect DC operation with a stock analog light board. I would check to make sure the motor leads are not making contact with the frame.

[sizemore]

More than likely not a decoder issue. I'd place the issue with the motor, more likely brushes as they age they get brittle and the motor gets "tired". DCC programing often depends on motor load, and if the brushes are going, that would lead to failures on the programing track. Combine that with the spotty operation on the main, sounds like motor/brushes.

In past couple of years I've been pulling my fleet out of mothballs (storage) and there have been a few times where the performance mimics the behaviors youre describing (on Programming and Main). Replacing the motor has been the ticket for few problematic engines I've put back into service. I went searching for Atlas replacement brushes, but the closest thing is Kato brushes which I havent tried. You can also roll your own brushes from graphite/carbon rod, but its hard to cut and naturally brittle.

HTH,
The S.

[peteski]

I also suspect that the problem is with contacts either between the decoder and chassis, or even more likely between the decoder and the motor.  As the S mentioned, without good motor connection your DCC system will not be able to get feedback from the decoder.  These models are infamous for poor contact between the metal strips coming up from the motor and the decoder's motor pads.  You could even try to press the decoder down in the area the motor contacts are, then try to run or read the decoder. I remember friend of mine added a block of styrene over that area so when the shell is down, it presses the decoder down for better contact with the metal strips coming from the motor.

If you think you remember the address the decoder was configured for, and you are sure decoder makes good contact with the chassis (you did solder wires directly), then address the decoder on your throttle and turn the headlight on. Look at both front and rear headlights in case loco is running in reverse. If the headlight turns on, the motor is not making a connection with the decoder.

[jdcolombo]

So let me make sure I understand.

1.  The engine runs well on DC, without a decoder?
2.  You have soldered wires from the phosphor bronze pickup strips that contact the trucks to the power input pads on the decoder?
3.  You have soldered wires from the motor output pads on the decoder to the phosphor bronze motor pickups?

If all the above are true, then do the following.

1. Turn the layout power off.
2. Check to make sure the wires you have soldered from the truck pickups to the decoder input pads have electrical continuity - use a multimeter's continuity function, put the engine on the track, and then put one of the multimeter's pickups on the rail and the other on the end of the wire where it is soldered to the decoder pickup pad.  Check both wires for proper continuity.  This will also tell you if the wheels are properly picking up power from the track and sending it to the wire.  Got continuity?  Then go to step 3.  No continuity?  Fix the pickup problem and verify continuity.
3.  If you have proper continuity from step 2, then make sure that the phosphor bronze motor pickups and the wires you soldered to them are not hitting the frame or some un-insulated component on the decoder.  This is most likely where the pickup strip for the motor gets closest to the frame, and particularly likely with respect to the bottom motor brush.  It is easy to slightly deform these pickup strips when you solder a wire to them, and then one or both could be touching the frame.  This wouldn't affect DC operation, but WOULD create a short and cause the decoder not to respond.  You can verify that the motor pickup is NOT touching the frame by using the continuity function on a multimeter, with one lead touching the frame, and the other multimeter lead touching the motor pad that you've soldered a wire to on the decoder.  You should NOT have continuity.  If you do, then you've got a short.  Fix it.
4.  After verifying that you do not have a short in step 3, verify that your soldered wires have continuity to the motor pickups.  Again, use a multimeter, touching one lead to the pad where you've soldered the wire and the other lead to the phosphor bronze motor pickup.  If continuity is fine, go to step 5.  If not, fix it.
5.  If all is well from steps 1-4, take the decoder out of the engine.  Verify the decoder is working by using wires from your decoder's inputs to the track outputs of your command station (e.g., you're going to hard wire the decoder to your command station).  You shouldn't need a motor load for the command station to recognize the decoder, but just to make sure, I'd also run wires from the decoder's motor outputs to the motor tabs on your engine.  Yes, what you're going to do is solder say a foot or 2 feet of wire from the deocder power inputs and then connect those to the command station track outputs, and then use a couple more feet of wire from the decoder's motor outputs to the motor tabs on the engine.  If the decoder now operates, then something is messed up with steps 1-4 above.  Re-install the decoder and try again.  If the decoder is not recognized and/or will not operate the motor, you have a defective decoder.  Send it back to ESU for a replacement (free).

John C.

[DWS3]

Thanks everyone for all the inputs!

John C, yes you have it correct. I have already tried most but not all steps you suggested. Unfortunately, I'm on a work trip for the next few days but will try the rest when I get back. Will ESU swap a bad decoder for free even if there is solder on the pads? I am reasonably skilled at soldering and pretty sure I only got solder on the correct pads. in any case, I'll try the last couple suggestions and see if they make any difference.

D.W.

[jdcolombo]

Never had a problem with ESU replacing a defective decoder, even if I had soldered wires to pads. 

John C.

[peteski]

John, when the command station in service mode identifies a decoder it does  that by reading certain CVs.  Without the motor as a load for the decoder to respond with acknowledgment current pulses, the command station will fail to identify the decoder.  That is why you often see the model moving slightly when the decoder is being read or written to.

[jdcolombo]

Hi Pete.

Ah, yes, the OP did say he put other engines on a programming track.  I didn't pick up that he was using programming service mode to ID the decoder.  I thought (incorrectly, it seems) that he was simply trying to address the decoder in normal operations mode.

So I guess I need to amend my troubleshooting instructions.  D.W., if you get to step 5, and you don't know the address of the ESU decoder, then instead of connecting the decoder to the command station's track outputs, you should connect it to the programming track outputs and see if it is recognized in programming service mode.  And Peteski is correct: you WILL need to add a motor load to do this.  And you can use any 12v DC motor for this - if you happen to have a spare motor sitting in a parts bin, solder wires to the brush terminals (carefully - you don't want to melt the plastic where the brush holders are - most motors have some kind of "cap" on the brushes that you can remove to do the soldering) and then connect them to the motor terminals of the decoder.

John C.

[peteski]

Since this was a working loco before it was stored for some time, the quick and dirty test would be to simply put a finger on the decoder board over where the motor contacts are and apply some pressure.  There is a good chance that the model will come alive. If not, then the more advanced troubleshooting would commence.

[jdcolombo]

Hi Peteski.  Normally, I'd agree with the "try pressure on the decoder" approach, but according to D.W.'s original post and his response to my long troubleshooting steps post, he's already hardwired the motor contacts and the decoder power input pads.  So I don't think putting pressure on the decoder iis going to help. 

John C.

[peteski]

Hi Peteski.  Normally, I'd agree with the "try pressure on the decoder" approach, but according to D.W.'s original post and his response to my long troubleshooting steps post, he's already hardwired the motor contacts and the decoder power input pads.  So I don't think putting pressure on the decoder iis going to help. 

John C.

Yes, I missed that - I though he only wired the track pickups.

[DWS3]

Hey everyone! Just wanted to provide a quick follow-up after having some free time this weekend to work on the engine again. After going through all the troubleshooting steps offered, everything seems to be working fine in terms of continuity. The DCC system just won't recognize the decoder, either through Decoder Pro or the NCE cab.

A couple quick follow-up questions:

1) I put the continuity test of the multimeter on both tabs for the engine and it showed continuity. I'm 99.9% sure that's correct but wanted to make sure I'm not missing something.

2) I never soldered on the front and back LEDs. I can't imagine why this would matter, but again, trying to make sure I haven't missed something simple.

3) If I send the ESU decoder back for repair, do you think they would substitute the hardwire version instead? They are about the same price and since I know the drop in version still doesn't fit great I'd rather just hard wire it. I am also planning to add one of the TCS micro keep alives with a pin so it can be disconnected for programming.

Otherwise, since I already have the TCS keep alive, I might just try a TCS decoder since I really only need motor control and forward/backup lights.

Thanks again for all the help!

D.W.

[peteski]

Thanks for the follow-up.  Yes, decoders (of any brand) can experience sudden death for many possible reasons. I have witnessed it myself few times.  The headlights no being attached are irrelevant.

Not sure if ESU will repair (exchange) if for another type of decoder. I assume you asked.  But for a wired-in decoder you might need to make some frame modifications (as the nanos or micros with wire adapter are rather thick). ESU no longer makes those with wires soldered directly to the decoder board.

[Sumner]

1) I put the continuity test of the multimeter on both tabs for the engine and it showed continuity. I'm 99.9% sure that's correct but wanted to make sure I'm not missing something.

Not sure I followed you there but...

If you have the meter on the engine tabs (a lead to each side) you need the meter on the lowest reading.  Then most of the ones I've tested show 25-40 ohms of resistance through the motor if I remember right.

If you go from a motor tab to the frame and have zero or close to it then you have a short which could cook the decoder.

Sumner