Home Open Account Help 393 users online

Model Railroading > Runaway Trains & CV29


Date: 03/06/25 00:01
Runaway Trains & CV29
Author: tmotor

Some folks prefer to run DC on their layouts.  In order for these folks to enjoy some of the benefits of DCC (such as lighting and sound), decoders are designed to allow (decoder-equipped) locomotives to operate on a DC layout.  In this DC-compatible mode, if no DCC signal is detected, the power at the rails is sent directly to the motor.
 
We Have Lift-Off
On a DCC layout, there is a (slim) possibility that the decoder will NOT sense the DCC signal sent by the command station.  (This could be due to a short, fault, dirty track, malfunction, etc.)  It would be nice if the decoder double-checked, or even triple-checked this condition to make sure, but it doesn’t.  This can happen when first setting the locomotive on the rails, or after recovery from a short.  When the decoder is powering-up, if it thinks it is on a DC layout, it sends the full power from the rails (nominally 12 Volts), to the motor.  The locomotive immediately goes to Run 8 and takes off like a rocket.  :-O
 
Make It Stop!
The first instinct is to press the Emergency Stop button, but since it is NOT in DCC mode (anymore), that command is ignored.  The locomotive must be physically restrained (grab it!), or the power to the track needs to be killed.  Putting something metal (a quarter, house keys, a metal ruler, etc) across the rails to purposely short the power district would also do the trick.  (Not exactly an elegant solution, but the alternative is worse.)  Best case scenario, this is a frightening experience.  Worst case scenario, the engine runs into another train, or jumps the track at a curve and does a nose-dive off of the layout.  NOT Pretty!!!  All of this can be prevented by taking the DC option completely off the table.  In order to remove the possibility of a Runaway Train, CV29 needs to be told to NOT send raw power from the rails to the motor, EVER.
 
The Beaverton Railroad Club website has a nice explanation:
https://www.beavertonrailroadclub.com/runaway-trains.html

El Presidente’
The factory settings are convenient for DC customers, so I get why they are programmed that way out of the box.  However, odds are that I will never need to run my equipment on a DC layout.  For me, the main issue is if I take my equipment to La Mesa, and it decides to become a rocket, it could damage another club member’s equipment.  With my luck, it would be a prized train owned by the Club President.  :-O 



Edited 2 time(s). Last edit at 03/06/25 11:17 by tmotor.




Date: 03/06/25 00:02
Re: Runaway Trains & CV29
Author: tmotor

I was expecting CV29 to be changed, but CV50 was being changed as well.  From what I can tell, CV50 is an index, but I’m unclear about what it does.  (Searching the LokSound documentation didn’t help clarify CV50 for me.)  The factory settings for CV50 is 2, so I assume this is a bit of housekeeping just to ensure DC mode is disabled.  The good news is I don’t have to know the purpose of CV50, since LokProgrammer takes care of all that for me.
 
I’m trying to change as few CVs as possible, but leaving CV29 at the factory setting just isn’t worth the risk.  So, when a new locomotive arrives, the first order of business is to clear the checkbox for Enable DC analog mode (which will change CV29, and CV50), then take a snapshot of the decoder settings.  This will provide a baseline for this decoder.   If I have to reload it (or load it onto another locomotive), I’m sure DC mode is disabled.  My convention is to have ALL saved decoder files with DC mode disabled.
 
Dave
 



Edited 2 time(s). Last edit at 03/06/25 11:17 by tmotor.








Date: 03/06/25 06:47
Re: Runaway Trains & CV29
Author: KA7008

In most decoders "analog" means DC analog (aka USA old school variable 12v power packs).
Europe has had AC analog in addition to DC analog in the past - think Marklin.
ESU (being European and attempting to be backwards compatible there) uses CV50 to define "which" analog is being switched in CV29.
0=both "off"
1=AC analog "on"

2=DC analog "on"
3=both "on"



Date: 03/06/25 07:38
Re: Runaway Trains & CV29
Author: tmotor

Greetings KA7008!

> In most decoders "analog" means DC analog (aka USA
> old school variable 12v power packs).
> Europe has had AC analog in addition to DC analog
> in the past - think Marklin.
> ESU (being European and attempting to be backwards
> compatible there) uses CV50 to define "which"
> analog is being switched in CV29.
> 0=both "off"
> 1=AC analog "on"
> 2=DC analog "on"
> 3=both "on"
Ahh, OK.  
Thank you so much for the explanation in plain language.  :-D
One more reason to let LokProgrammer set the CVs for me.  Based on how buried the documentation is on CV50, there's no way I would have known to clear CV50.
(I have attached the CV50 description.)

Thank you so much for sharing your expertise.
Take care and God bless!
Dave

 



Edited 1 time(s). Last edit at 03/06/25 07:41 by tmotor.




Date: 03/06/25 07:55
Re: Runaway Trains & CV29
Author: KA7008

Interesting they changed CV50=3 to be Quantum engineer, thanks for updating that.
Their older documentation didn't show that.  I don't use analog so I never noticed the change.



Date: 03/06/25 08:26
Re: Runaway Trains & CV29
Author: funnelfan

Programming CV's directly seems rather archaic. I use JMRI to do most programming, even ESU chips. I only use the Lokprogrammer for more advanced stuff with ESU chips. When I first program a locomotive, I turn off anything that is not DCC protocol. ESU chips have a number of things like RailCom and Brake on DC turn on in factory settings, and I turn those off so that it only runs on DCC commands. NCE systems also make it super easy to program the loco to be DCC only as they have a prompt for it under decoder configuration.
When I get a new loco, I first test run it at factory settings to make sure there is no inherent issues, then read all the settings into JMRI and save it to the laptop. I then go to town changing a whole bunch of the settings to my preferred settings, remapping functions, sound levels, ect. When that is done, I will speed match it with the other locomotives I intend to run it with.

Ted Curphey
Ontario, OR



Edited 1 time(s). Last edit at 03/07/25 20:28 by funnelfan.



Date: 03/06/25 10:58
Re: Runaway Trains & CV29
Author: tmotor

Greetings KA7008!

> Interesting they changed CV50=3 to be Quantum
> engineer, thanks for updating that.
> Their older documentation didn't show that.  I
> don't use analog so I never noticed the change.
Thank you for the clarification.
I'm just glad LokProgrammer lets me click a checkbox, and it figures-out the CVs.

Take care and God bless!
Dave



Date: 03/06/25 11:16
Re: Runaway Trains & CV29
Author: tmotor

Greetings Ted!

> Programming CV's directly seems rather archaic.
Agreed.
Way too many changes for me to make rookie errors.


> I
> use JMRI to do most programming, even ESU chips. 
Many folks prefer JMRI

> I
> only use the Lokprogrammer for more advanced stuff
> with ESU chips.
For some stuff, LokProgrammer is the only thing that can program a LokSound decoder.

> When I first program a locomotive,
> I turn off anything that is not DCC protocol. ESU
> chips have a number of things like RailCom and
> Brake on DC turn on in factory settings, and I
> turn those off so that it only runs on DCC
> commands.
Thanks for the heads-up on those.

> NCE systems also make it super easy to
> program the loco to be DCC only as they have a
> prompt for it under decoder configuration.
I will have to give that a look.  I'm all for anything that is more push-button.

> When I get a new loco, I first test run it at
> factory settings to male sure there is no inherent
> issues, then read all the settings into JMRI and
> save it to the laptop. I then go to town changing
> a whole bunch of the settings to my preferred
> settings, remapping functions, sound levels, ect.
> When that is done, I will speed match it with the
> other locomotive I intend to run it with.
I'm beginning to understand the flexibility DCC has to customize the Functions and features.  The stock settings are a best guess by ScaleTrains, but I'm no longer afraid to change them.  As long as I have a snapshot of the decoder's settings, I can always reload them.
Eventually I will get into the sound options.  I like that they can be slightly modified so the tone is not exactly the same for each locomotive, even though they are all using the same sound file.

Thanks for the Pro Tips.  :-D
Take care and God bless!
Dave




Edited 2 time(s). Last edit at 03/06/25 16:07 by tmotor.



Date: 03/06/25 19:33
Re: Runaway Trains & CV29
Author: wabash2800

That would be worse than the old, "Who's got my train! with DC and manual blocks.

Victor Baird



Date: 03/06/25 21:06
Re: Runaway Trains & CV29
Author: railstiesballast

When I first started with DCC I had not changed CV 29 out of enabling DC.
I had a DC layout and Intergrated Signal Systems automatic block signals "Twin T" detectors to control them.
They used the "twin T" diode principal that sensed the small forward voltage drop in the track power.
Then I experienced runaway trains on occasion when using DCC.
What I think happened is that some of the detectors would have a diode fail and the single remaining diode gave full voltage half wave DC to that signal block and the decoder responded as described above.
The cure has been to use the "tombstone" inductive coupling detectors where one line of the track power in each block passes through the hole of the transformer, which senses the presence (or absence) of the DCC power.
 



Date: 03/07/25 22:35
Re: Runaway Trains & CV29
Author: tmotor

Greetings Victor!

> That would be worse than the old, "Who's got my
> train! with DC and manual blocks.
Indeed!
The Ghost In The Machine :-O

Take care and God bless!
Dave



Date: 03/07/25 22:40
Re: Runaway Trains & CV29
Author: tmotor

Greetings railstiesballast!

> When I first started with DCC I had not changed CV
> 29 out of enabling DC.
> I had a DC layout and Intergrated Signal Systems
> automatic block signals "Twin T" detectors to
> control them.
> They used the "twin T" diode principal that sensed
> the small forward voltage drop in the track
> power.
> Then I experienced runaway trains on occasion when
> using DCC.
Yikes!

> What I think happened is that some of the
> detectors would have a diode fail and the single
> remaining diode gave full voltage half wave DC to
> that signal block and the decoder responded as
> described above.
> The cure has been to use the "tombstone" inductive
> coupling detectors where one line of the track
> power in each block passes through the hole of the
> transformer, which senses the presence (or
> absence) of the DCC power.
Glad you were able to resolve it.
A lot of DCC is fly-by-wire, where a computer is accepting sensory data, and our inputs, and then decides what to do. 
If some sensors are receiving false data, then things can get sideways pretty quick.  :-O

Take care and God bless!
Dave



[ Share Thread on Facebook ] [ Search ] [ Start a New Thread ] [ Back to Thread List ] [ <Newer ] [ Older> ] 
Page created in 0.1486 seconds