A bad 40 minutes with DCC

kingpeta Jan 13, 2017

  1. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    Ok guys,
    I am so frustrated right now. I just spent 40 minutes of trying to run trains and gave up. I switched to DCC about 2 months ago. Prior to switching my layout was DC and I had no problems running. Tonight I had so many problems with engines stalling every few inches that I had to stop and do something else.
    My layout is N scale Atlas code 55 throughout. I have been in the hobby for decades but had a long spell of raising kids & stuff. Most of my locomotives are Atlas current generation. My chosen track cleaning method is to wipe down with 91% alcohol, burnish the track with an eraser type block designed for cleaning electrical commutators (much less abrasive than a clean boy), wipe again with alcohol, then spread NoOxid and follow up with a clean gun cleaning patch a day or so later. On my previous layout that proved to be successful for running trains.
    The only engine I can say runs almost flawlessly is a Kato SD40-2. The rest are Atlas 4 axle locomotives, I have disassembled and tuned every one and only my 1 Atlas SW2 is as reliable as the Kato. I even have a brand new Atlas RS-1 that runs poorly.
    So what is the problem? I think I have paid attention to proper track work. I have cleaned the wheels and all electrical contacts on the locomotives.
    Do most of you start an operating session by wiping down the tracks? I seem to have more issues on turnouts, but there are straight sections that give me problems too.
    My layout is in a finished basement area. We have pets ( 2 cats & a dog who isn't allowed in the basement), and no smoking. I just can't figure out where the problems are originating and I'm an engineer!


    Sent from my iPad using Tapatalk
     
  2. Greg Elmassian

    Greg Elmassian TrainBoard Member

    325
    62
    17
    I cannot find fault with your track cleaning regime, sounds great.

    So, how about wiring a small test loop of track and see what happens? Something is indeed strange. I would look at your system connections after you run a test loop. Perhaps you have a wiring issue that did not show up on DC but shows on DCC...

    So just wire the DCC system to the test loop only and see how the trains run. You will quickly determine if it is the locos or the track.

    Greg
     
  3. gcav17

    gcav17 TrainBoard Member

    1,065
    581
    30
    What dcc system are you using? What type of decoders are installed in your locomotives? Are your feeder wires to the rails attached well? Track cleaning is important. But not needed every day.

    Sent from my SM-G930V using Tapatalk
     
  4. Rocket Jones

    Rocket Jones TrainBoard Member

    783
    601
    18
    Start with the obvious stuff, even if it sounds silly. Did you clean the wheels of your locomotives? And definitely set up a test track if at all possible, those feeders should be checked as well.
     
  5. gatrhumpy

    gatrhumpy TrainBoard Member

    735
    131
    20
    1. Take a Digital Multimeter (DMM) and perform a continuity check using the ohms side between each and every track section. Anything higher than 0.10 ohms, solder the track sections together. You should probably do that anyways.

    2. Take a pink rubber eraser and clean the locomotive wheels. Sometimes the blackening gunk they use does not rub off easily. The pink eraser will get it off.

    3. Use a DMM to measure AC current to each section of track. Puteach lead on each rail and make sure you get a reading near your DCC's track voltage.

    4. If you're still have problems after soldering the track, cleaning the loco wheels with a pink eraser, and measuring the AC voltage and continuity, something is going on with the locomotives. Make sure the bronze pickup strips and contacting the nubs on each truck as it turns and goes up grades.

    5. If that STILL does not help, burn the layout. It's hopeless.
     
    kingpeta likes this.
  6. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    I am using an NCE Powercab. I have a variety of decoders, Digitrax, NCE and TCS. The feeder wires I attached by soldering to the outside of the rails.
     
  7. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    Yes, I did clean the wheels. I used the same commutator cleaning pad and actually burnished them to remove the wheel blackening stuff. They're as clean & shiny as could be. The feeder wires are soldered to the rails and connected to the bus using 3M Tee taps. I'll try setting up a test loop and see what that does. However, these same locomotive seem to run fine on the 3 foot section of programming track I have (which is Kato Unitrack, no ballast).
     
  8. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    I will run through these steps and see if I can uncover anything. I just hope I don't get to step #5.
     
  9. Greg Elmassian

    Greg Elmassian TrainBoard Member

    325
    62
    17
    I'm pretty convinced you don't have dirty wheels or track.

    Will be interested to see how the test loop works, since you had locos stopping every few inches on the main line.

    I'm suspecting a partial short or wiring issue in the track.

    By the way, it would be a good idea to turn on the "current" display (shows the amps you are drawing) on the PowerCab, it's in the powercab menu.

    Greg
     
    17fixer likes this.
  10. Carl Sowell

    Carl Sowell TrainBoard Supporter

    3,032
    8,079
    82
    kingpeta,

    First off, welcome back to the N scale world. Now, i am going to say something that will probably rattle some folks, but rest assured I am not trying to start a flaming episode. You mentioned that the only loco that runs fairly well is a KATO, not surprising, and the other 4 are Atlas (current generation).

    You did not mention if the decoders were Atlas installed or installed by yourself, nor did you mention if they are drop in or hard wired decoders. The reason I am asking is because Atlas has quite a reputation for very poor contacts between the decoder and frame. Seems that many decoder boards are too thin and have problems plus, if I remember correctly, the Atlas motor contacts have to make contact with the decoders underside. I have read many threads on TB and The Rail Wire regarding this issue and most all the time the solution is to solder small decoder wire to the motor tabs and then to the decoder.

    Not flaming Atlas, but just stating the facts. BTW, I agree that your track / wheel cleaning regimen are sufficient.

    Regards,

    Carl
     
  11. Atani

    Atani TrainBoard Member

    1,460
    1,698
    36
    I had the same problem on one of my Atlas GP9s with a NCE drop in decoder. The board was a bit thinner, not a whole lot but enough to not reliably get contact with the frame. I solved it by dropping a thin bubble of solder on the four contact points with the frame and carefully reassembling so the frame cut into the solder "pads". I have seen others unfold the motor pads but that seems a bit extreme to me (though it likely does solve a different problem that I am not facing).
     
  12. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    That could very well be the problem. All my decoders I installed myself. All but 1 are drop in type boards. The 1 I had to solder in seems to run pretty good. I thought I paid attention to where the frame grips the boards, but it could be that's where the problems lie.


    Sent from my iPad using Tapatalk
     
  13. Another ATSF Admirer

    Another ATSF Admirer TrainBoard Member

    849
    56
    21
    It would also be worth testing just one loco at a time. DCC locos parked on the track not moving do draw a little bit of power. Not much, but when fault finding, eliminate all the variables! :)

    I have used little metal shims between decoders and the frame on several of my locos, because the "drop in" PCB is so thin it doesn't make good contact.
    My worst install, the body shell pinched the frame somehow. It ran fine without the shell when testing, but never with the shell on. Had to pull that one all the way apart and start again.
     
  14. emaley

    emaley TrainBoard Supporter

    327
    44
    9
    I had a similar intermittent issue with a Kato F7 which turned out to be the motor contacts. Once I soldered them , problem solved. Sounds like you are getting closer. Good luck.

    Trey
     
  15. mu26aeh

    mu26aeh TrainBoard Member

    270
    170
    18
    Along with where the frame makes contact with the board, also the motor contacts that touch the underside of decoder. These need to be sprung/bent properly to make sufficient contact.
     
  16. kingpeta

    kingpeta TrainBoard Member

    69
    14
    13
    Well I disassembled the engine tonight. I soldered a little pad on the decoder boards to make sure they had good contact with the split frame halves. Put it back together and had a short. Disassembled again, no short but still had intermittent contact issues. I tried it on the programming track and it ran crappy. I tried to program it and got 2 different manufacturer numbers, one of which isn't even registered with the NMRA, #232. The second time I tried to program it it gave me #153 which is TCS. Did I fry the decoder? I now think this is either a locomotive or decoder issue since it runs poorly on even a 3 foot programming track. The decoder was swapped from a previous locomotive to this one.
    I do thank everyone for their input.


    Sent from my iPad using Tapatalk
     
  17. Atani

    Atani TrainBoard Member

    1,460
    1,698
    36
    When you assembled the frame halves, did you put the tape on the halves to electrically isolate them from each other around the motor area?

    I am not sure you did fry it, if you did it wouldn't likely respond at all. More likely it is still getting intermittent connection to the track. The way the base station reads the values from the decoder is via motor pulses, similar to morse code in a sense. If one bit of the value gets mangled or lost it will throw off the value that the base station reads.

    Have you tried it without the shell on the loco? I had one loco that had poor contact that worked great with the shell off but not with it on. After taking it apart again, adding small solder blob on the pads and reassembly I had better luck. I also found that the wipers for the frame sides were not maintaining proper contact with the nub on the trucks. After adjusting the wipers a bit the issues more or less went away, even with the shell put back on carefully (watch the wipers as they can easily move out of alignment).
     
  18. woodone

    woodone TrainBoard Member

    959
    51
    24
    DCC is very touchy to dirt, bad electrical connections ETC.
    One problem with decoders into Atlas loco's is the motor brush to decoder contact place- Some decoders only use a mechanical connection from the motor to the decoder- This can be poor in most installs- TCS uses a small slot in which the motor tab is fit into. VERY hard to get the tab into the slot- I can do it but I still solder that connection. With other decoders the motor tabs just push up against the decoder an a contact pad- I use a very small wire and make a jumper and solder the motor tab via the jumper wire to the board. As others have stated the contact from the frames to the decoder are sometimes to blame- All four tabs. Build up the tabs on the decoder with some solder and slide into the frame- you will have to push hard to make sure that the decoder is making contact with the frame. If you have too much solder-file down the tab so it fits tight. Make sure you use Kapton tape between the frame and the motor tabs, so they don't make contact with the frame.
    One other thing- make sure that the truck tabs are making contact with the two strips on the frames- I have see time where the truck tab gets trapped behind the strip and does not make contact.
    Do one at a time and you will WIN! Can be frustrating for sure, but you will get the hang of it.
     

Share This Page