Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Why does new decoder cause loco to run wide open?

2625 views
11 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    May 2018
  • 68 posts
Why does new decoder cause loco to run wide open?
Posted by Paul Schulze on Sunday, June 17, 2018 8:16 PM

I just installed a TCS CN decoder in an Atlas GP35.  The test run resulted in the rear light comming on and the loco running full open forward with no control.  The loco does not care what number is assigned in my Zephyr - even 00.  It will not read CVs on the program track and if you try to write to it anyway, it inches forward everytime I hit the write CV button.  

I have check the wiring at least four times and have done continuity checks on all wires.  I have checked and found no shorts.  Both front and rear boards make good contact with the frame halves.  I have tried to factory reset it but that does not work anyway because the programming track will not read the decoder.

Any help will be appreciated as this is my first full decoder install.

  • Member since
    March 2018
  • 688 posts
Posted by BNSF UP and others modeler on Sunday, June 17, 2018 8:43 PM

This is quite interesting. I have the exact same problem with a tcs decoder given to me that I decided to install (it was used, from a club member) but I did not think the burn in the insulation was a problem. It did the exact same thing that you described. I emailed him and he told me that it was a soldering iron burn and he had given it to me along with some junk locomotive parts of mine. It was in no way his fault or being cheap. I should have known better than to install that decoder under such circumstances. Does yours have a soldering burn? Either way, great news! If it is under one year old, TCS will give you a replacement for FREE. And if that one breaks or doesn't work, you can get a free replacement for that one too, and so on... You go to here: https://tcsdcc.com/warranty, watch the video, follow the instructions, and I believe that is all. Good luck!

I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.

Moderator
  • Member since
    June 2003
  • From: Northeast OH
  • 17,249 posts
Posted by tstage on Sunday, June 17, 2018 9:01 PM

The runaway issue sometimes happens on certain decoders.  The remedy is to turn off dual-mode in CV29 and operate your decoder in DCC-mode only.  This can be achieved by reading the value for CV29 with your DCC system and subtracting 4 from it.

If you are having trouble reading your CVs, the default value for CV29 is 6.  Change it to 2 (6-4=2) and your problem should go away.  Increasing CV29 by one will also reverse the direction of your locomotive.  This is a convenient way to correct that if it initially runs backwards.

Tom

https://tstage9.wixsite.com/nyc-modeling

Time...It marches on...without ever turning around to see if anyone is even keeping in step.

  • Member since
    May 2002
  • From: Massachusetts
  • 2,899 posts
Posted by Paul3 on Monday, June 18, 2018 1:21 AM

Folks,
What tstage said is correct.  Runaways are possible with DCC decoders unders certain circumstances; usually a short or a rapido series of shorts and/or power interruptions.  The solution is to program every decoder to be DCC-only by turning off DC Analog mode.  This is controlled by CV29, the configuration variable that controls a group of things.  One of those options is DC Analog Mode.

For most applications: for a 2-digit loco (from 01 to 127), change CV29 to a value of 02.  For a 4-digit loco (from 128 to 9999), change CV29 to a value of 34.

The reason why the decoders run away under analog mode is that DCC is a square wave digital signal.  Think of a sine wave in AC power, changing from + to - (at 60 Hz in the US), but instead of a constantly changing voltage (from -X to +X and back again), it does it abruptly.  The resulting wave form looks square as the voltage shifts in mere milliseconds from + to - and back.  The length of time the square wave lasts before switching back again indicates if it is a "0" or a "1" bit.  This is how DCC talks to the decoders.  With DCC, the power is the signal, which is one of the reasons why it is superior to analog systems.

Well, when you have a short circuit, voltage drops to zero.  Remember that voltage is the difference in energy between two conductors, and when both conductors are touching (shorting), the difference is zero because they are now both the same.

When the short is removed, the DCC square wave immediately starts up again, but the first initial burst of that very square wave before it can start switching back and forth is, technically, DC.  Some decoders are so sensitive that they can pick up that initial burst of the square wave and think they are seeing DC analog voltage. 

If DC analog mode is turned on, the decoders are programmed to take whatever voltage is in the rail and put it straight through to the motor...which in DCC (HO) is like 14 volts.  Off your locos will go at full throttle until they reset or crash.

For a good CV29 calculator, try this site:
http://www.2mm.org.uk/articles/cv29%20calculator.htm

  • Member since
    May 2018
  • 68 posts
Posted by Paul Schulze on Monday, June 18, 2018 1:22 AM

No burn, but thanks for the replacement info.

  • Member since
    May 2018
  • 68 posts
Posted by Paul Schulze on Monday, June 18, 2018 1:30 AM

Paul3, I cannot access any CV.  I tried 29, 8. 1, all of them.  How does the loco see anything when I can enter any loco number and it does the same thing?  Is it see ing sqaure wave rise for any DCC loco?  I think the decoder has a problem I will not be able to solve.

 

Paul S

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, June 18, 2018 8:48 AM

 That leads me to believe either the decoder is fried or the motor tabs are touching the frame (or both, if the motor tabs are contacting the motor terminals on the decoder AND the frame, it's putting track voltage back into the output of the decoder which pretty much fries any of them.

                                     --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, June 18, 2018 10:15 AM

Paul3
When the short is removed, the DCC square wave immediately starts up again, but the first initial burst of that very square wave before it can start switching back and forth is, technically, DC.  Some decoders are so sensitive that they can pick up that initial burst of the square wave and think they are seeing DC analog voltage. 

If DC analog mode is turned on, the decoders are programmed to take whatever voltage is in the rail and put it straight through to the motor...which in DCC (HO) is like 14 volts.  Off your locos will go at full throttle until they reset or crash.

why should the loco take off at full speed in DC mode if DCC is alternating and has an average DC voltage close to zero?

is it possible that one motor terminal is connected to a rail (or frame as Randy siggested) and the other to decoder ground; wouldn't the motor see a DC voltage ~half the time? 

could the decoder be fried such that the two h-bridge paths providing rectified DC are on?

greg - Philadelphia & Reading / Reading

  • Member since
    May 2018
  • 68 posts
Posted by Paul Schulze on Monday, June 18, 2018 10:50 AM

I have checked the continuity between the and gray and orange motor wires at the decoder board and the two motor halves and they are all open with no shorts.

  • Member since
    February 2002
  • From: Mpls/St.Paul
  • 13,892 posts
Posted by wjstix on Monday, June 18, 2018 11:11 AM

Best bet would be to try another decoder. If it works, then it was the decoder that was bad; if not, it's something with the engine. (I'm assuming the engine ran OK on DC power before the decoder installation??)

Stix
  • Member since
    November 2005
  • From: Phoenix, AZ
  • 693 posts
Posted by woodone on Monday, June 18, 2018 1:45 PM
If you check between the orange and gray wire you should see resistance. Some thing amiss here!
  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Monday, June 18, 2018 2:41 PM

gregc

 

 
Paul3
When the short is removed, the DCC square wave immediately starts up again, but the first initial burst of that very square wave before it can start switching back and forth is, technically, DC.  Some decoders are so sensitive that they can pick up that initial burst of the square wave and think they are seeing DC analog voltage. 

If DC analog mode is turned on, the decoders are programmed to take whatever voltage is in the rail and put it straight through to the motor...which in DCC (HO) is like 14 volts.  Off your locos will go at full throttle until they reset or crash.

 

why should the loco take off at full speed in DC mode if DCC is alternating and has an average DC voltage close to zero?

is it possible that one motor terminal is connected to a rail (or frame as Randy siggested) and the other to decoder ground; wouldn't the motor see a DC voltage ~half the time? 

could the decoder be fried such that the two h-bridge paths providing rectified DC are on?

 

 Most definitely. It seems most of the time, the drive transistors fail shorted, same with the function drivers, which is why a blown function tends to be on all the time and you can't shut it off. If two of the 4 H bridge transistos fail closed, then the motor is going to see DC in one direction for half the DCC wave and 0 for the other, net somewhere around 50% of the track voltage depending on the exact packets being sent, so it will take off - in one direction only. If all 4 drivers are failed closed, the decoder should be a dead short.

                                             --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!