Short Circuit = melted ties

[Scarpia]

I had the oddest thing (well, I'm new to this, so it may not be that new) but I had a 2-8-0 jump a switch, and in the time it took for me to walk around the layout to get it, it shorted the track enough to not only produce flame, but melt the ties of the turnout. You can see it here just in front of the frog

 

 What's even more odd is that I've had many, many successful hours running over this turnout with no problems, I'm not even sure why it jumped.  

I'll have to figure that part out on my own, but is this kind of shorting normal? I've had things derail in the past and short, but not to the point that I could smell burning plastic. Is there something I'm missing in my wiring?

I run CVP's Easy DCC, and have a new dual zone booster unit. The turnout is a peco insulfrog, (all of them are), and the loco is a Bachmann 2-8-0 spectrum (that's run well for weeks). I have a 16 gauge bus, and 18 gauge feeder wires, the layout is gapped as well.

Thanks for any input.

[MisterBeasley]

Does the booster have a circuit breaker on it?

You definitely should have tripped a breaker on this.  It might be a good idea to try the "quarter test."  Put a quarter down on the track.  If the breaker doesn't trip, you've got a problem.  Since you said the booster is new, it's quite possible that the breaker is defective.  You might also have a wiring error.

Was this turnout right at a zone boundary, by any chance?  I suppose you could get some kind of weird effect if the engine derailed while crossing between zones.

[cacole]

Something is not right with your power booster -- its internal circuit breaker should have turned the power off as soon as the locomotive derailed and caused a short.

You need to contact CVP and send this unit back for analysis and repair/replacement before it causes a more serious problem such as a house fire.

[retsignalmtr]

i've had the same thing happen with a peco insulfrog switch. in my case i had a loco derail there and didn't know it until i smelt burnng plastic. apparently the derailed loco was bridging the insulation on the frog and was causing a short but it did not draw enough current to trip the circuit breaker. my empire builder is a five amp unit. the ties melted all the way across to the stock rail. this happened not long after i installed the dcc system but fortunatly hasn't happened again. no damage to the loco but the switch was toasted. you don't need a large current draw to cause enough heat to melt the plastic tie. i'll give you a prototypical example of this. i worked as a signal maintainer for the NYCTA for 31 years. one day a piece of the third rail broke off and came into contact with the stockrail on a switch. this stock rail was also a signal rail which was insulated from ground. the current being drawn by this occurance melted the switch operating rods, and part of the switch machine, jumped several insulated joints and then melted the steel rails in them and also melted the rails on the frog up to the point. all this while never tripping the circuit breaker which was set at  around 8,000 amps. our layouts are just the prototype in miniature.

[Scarpia]

@Mr. Beasley,

I'll run the quarter test this weekend, thanks. It isn't at a boundry, actually, so I don't think that's the problem.

When you suggest a wireing error, could you be more specific? If you look at the picture above, you'll see the feeder wires soldered to the rail, and red is to red, etc. I'd appreciate it if you had an idea, as this is my first attempt at wiring a layout, so i could have munged it up someplace.

@cacole,

I hope you're not right and its me, but I guess I'll find out with the quarter test.

@retsignalmtr,

Do you think I need to replace the switch, just by default?

[jeffrey-wimberly]

I'm glad my EZ-Command system has a good breaker and the booster has one as well. I also use the taillight trick that I learned from Joe Fugate.

[rolleiman]

I would definately look into the circuit breaker thing.. That shouldn't have happened. I don't know what the current capacity of the CVP is but if I were to guess at 3 amps (for example) and 16 volts, a dead short is going to have to disappate 48 watts. That's a lot of heat running over those plastic ties (more than the soldering irons some of us use). If you have to add a circuit breaker, make sure to size it properly or it won't work either. If 3 amp is the output capacity of the CVP, then go for something like a 2.5A circuit breaker to help avoid really overloading the unit. You'd have to be running several sound locos to exceed that.

[MisterBeasley]

Boosters are generally in the 5-8 amp range.  So, that's even more wattage going out.  I would assume that a "dual booster," without knowing the make and model, would have 2 channels.  Each of these should be protected with its own breaker.

As for possible wiring errors - Each zone should be completely isolated from the other, on both rails.  The insulated rail joiners (or gaps, if that's what you're using) should be directly across from each other, to minimize the period when a locomotive is bridging the gap.  The sort of thing I was thinking about would be feeding one rail from one side of the booster, and the other rail from the other side, or missing a gap so that the two zones are permanently joined together on one rail, but isolated on the other.

[Vail and Southwestern RR]

This may be a demonstation of why larger wire is good for busses.  If there is enough resistance through the whole path, bus wires, feeders, any connections, the less than perfect contacts made at the short, there may be enough voltage drop through the circuit to limit the current to less than the breaker rating.  In that case, you end up with almost full current running through the short.  The quarter test will help to determine if this is a problem, by finding out if the breaker will detect a short anywhere on the layout. 

[jfugate]

Ah, the Peco insulfrog short problem!

Here's how you fix Peco insulfrog turnouts so you never get that problem again.

The problem is you have a piece of rolling stock or a loco that bridges the two diverging frog rails. A short on the track with a DCC system where the short is not removed quickly will act like applying a soldering iron to the rails and just leaving it there -- and so the melted ties you experienced!

The other problem is your layout short management. On a DCC layout, you need something to manage shorts so they don't cause this kind of problem. Ordinarily, your booster should have tripped. But it's also good insurance to include an additional layer of short management on a DCC layout like I show here in this video clip.

[DanLW]

On the DCC side of the club I attend, we have small light bulbs in series with the main track power line.  I'm not sure what the specs for them are, but they're about the size of those big old Christmas light C9 bulbs.  They work great in that they serve a dual purpose.  First, they limit curent draw in case of a dead short.  Second, when they light up, you know there is a short somewhere.  Very helpful when multiple operators are running and you're wondering why your train just stoped for no apparent reason.

[Scarpia]

Ah, the Peco insulfrog short problem!

Here's how you fix Peco insulfrog turnouts so you never get that problem again.

The problem is you have a piece of rolling stock or a loco that bridges the two diverging frog rails. A short on the track with a DCC system where the short is not removed quickly will act like applying a soldering iron to the rails and just leaving it there -- and so the melted ties you experienced!

The other problem is your layout short management. On a DCC layout, you need something to manage shorts so they don't cause this kind of problem. Ordinarily, your booster should have tripped. But it's also good insurance to include an additional layer of short management on a DCC layout like I show here in this video clip.

 

Thanks for that Joe. I'll add a bulb this weekend, and start grinding down my insulfrogs. You know this is a bit irritating, as I chose the insulfrogs as I thought they'd be the least headache to wire up.

cheers, and have a great weekend. Thanks to everyone for their input.

 

[UpNorth]

Also put the quarter in the area encircled.  Because this could be the cause of the short not being detected and not tripping the breaker if they are wired to different power boosters.

[locoi1sa]

  The short radius right hand Peco turnout. Peco being a British prototype sort of, does not have the same flangeway spec as an NMRA turnout. On all of mine I shim the right hand (sometimes the left too) flangeway to guide the wheels away from the frog point. All it takes is a .010 to.015 depending on how wide the flangeway is to guide the wheels away. Styrene works great for this and lasts a long time.

  Did the decoder get fried? Your wireing is probably fine and the circuit breaker is probably good too. The turnout melted due to the small guage jumper wires instaled by Peco. Those small wires must have been glowing like a light bulb under that turnout. Joes 1156 bulbs work great 90% of the time but every now and then you will get the short at the right spot that does some damage.

    Pete

[betamax]

This may be a demonstation of why larger wire is good for busses.  If there is enough resistance through the whole path, bus wires, feeders, any connections, the less than perfect contacts made at the short, there may be enough voltage drop through the circuit to limit the current to less than the breaker rating.  In that case, you end up with almost full current running through the short.  The quarter test will help to determine if this is a problem, by finding out if the breaker will detect a short anywhere on the layout. 

Heavy wire is really important.

If the current suddenly shoots up, it is a short, and the circuit is opened protect everything. You could be drawing 4.5A, but if it suddenly spikes to 4.9, it's a short. It can even act on the change from 1 amp to 3, because it happened faster than it would under a normal load.

Therefore, excessive resistance is bad, because it will limit the current's rate of change, and prevent the protection circuit from detecting a short.

Old ideas that worked with low current DC don't belong with high current DCC systems. The quarter test establishes that the protection system is working, and that there are no wiring issues preventing that from happening.

Proper wire and protection devices meant for DCC applications are still cheaper than melting a locomotive or two. That can happen real fast, especially in N scale.

[UpNorth]

I was reading up on your web page. You are running the Zonemaster dual. It has 7A per channel and 30A of surge current. That would be enought to liquify the ties.  Zone A (or channel) has adjustable trip voltage from 1.5A to 7A. Zone B is NOT ajustable at all. 

You certainly need current limiting on this setup, via bulb or solid state.  A single bulb will not be sufficiant if you are running both power zones from the dual power booster, you will need 2 bulbs. Plus the layout could still drop dead on you with single bulb, defeating the use.

On your 4 X 8 layout, the wire size you have (16G and 18G  feeders)  is quite adequate.

[Scarpia]

Thanks, everyone for your input.

The quarter test worked all over; lay it down and the breaker will snap. I had forgotten that the Zonemaster dual has a variable output, as I'm just using zone A at the moment, and not running more than 4 locomotives, I'll cut it down.

The decoder on the loco is fine, at least it runs without problem. Per the wire gauge, I actually used what CVP specified.

To correct this in the future, I'm going to be doing both the modifcations to the turnouts, and add a bulb.

Cheers!

[DanLW]

Therefore, excessive resistance is bad, because it will limit the current's rate of change, and prevent the protection circuit from detecting a short.

Resistance only resists current flow, but it does not resist the rate of change of current flow.  In a purely resistive circuit, current change is instantaneous throughout.  The thing that actually resists a change in the flow of current is an inductive circuit.  In an inductive circuit, the resistance in change of current flow is caused by the magnetic field generated around the wires in the coil.  As current decreases the field collapses, but the collapsing field induces a current flow within the coil thereby slowing, or resisting, that change in current flow.

And in the interest of completeness, a capacitive circuit resists a change in voltage.

The simple way to remember, as taught by my electronics teacher is ELI the ICE man.  E (EMF, or Voltage) comes before I (current) in a L (inductive) circuit.  I comes before E in a C (capacitive) circuit.

[betamax]

Remember, this is AC, not DC, so things are not always as they appear.

I've seen an incorrect connection throw a very accurate system out of balance, just because of the additional resistance of a length of 10 ga. wire. Once the load's connection was changed, everything worked just fine. And the current flow wasn't any more than your average DCC system would have. But the resistance was enough.

[UpNorth]

The quarter test worked all over; lay it down and the breaker will snap.

Did you try the it in the zone I circled in your turnout picture ? Bridging those two points can happen if the engine jumps the frog.

[Scarpia]

Yes, sorry, It shuts down the breaker there as well. (it was actually the first place i tried).

 

I'm thinking the wiring is ok, I need to look at fixing those points, installing a bulb, just in case, turn down the booster, and like everything on this layout, learn from the experience.

 

Thanks again. 

[larak]

Resistance only resists current flow, but it does not resist the rate of change of current flow.  I

Dan is absolutely correct. AC, DC doesn't matter, the differential equations are the same. Parts just fall out for DC.  Component characteristics don't change. Inductance limits rate of change in current, resistance limits maximum available current. So larger wire will more reliably allow overcurrent protection to work.

Karl 

[UpNorth]

Yes, sorry, It shuts down the breaker there as well. (it was actually the first place i tried).

I'm thinking the wiring is ok, I need to look at fixing those points, installing a bulb, just in case, turn down the booster, and like everything on this layout, learn from the experience.

Thanks again. 

Dick Bronson (RR CIRKITS) has a variation on the bulb fix that drops the bulb output during a short to a less bright (read burning) bulb.

http://www.rr-cirkits.com/Notebook/short.html

 

[DanLW]

So larger wire will more reliably allow overcurrent protection to work.

Just don't let Monster Cable catch wind of that.  Otherwise they'll start trying to sell model railroaders 10AWG oxygen free charged dielectric "railroad cable" at $5/foot!

[larak]

So larger wire will more reliably allow overcurrent protection to work.

Just don't let Monster Cable catch wind of that.  Otherwise they'll start trying to sell model railroaders 10AWG oxygen free charged dielectric "railroad cable" at $5/foot!

Great one! 

[UpNorth]

So larger wire will more reliably allow overcurrent protection to work.

Just don't let Monster Cable catch wind of that.  Otherwise they'll start trying to sell model railroaders 10AWG oxygen free charged dielectric "railroad cable" at $5/foot!

Would that correct my sound decoder hickup !... 

 

[60YOKID]

I am an electrical consultant and inspector.  It seems to me there may be two serious design problems described here.

1. The system has a surge rating of 30A. This is not at all desirable and in my opinion is a serious fire hazzard just waiting to happen.  For our model railroads, there is little or no advantage to a high surge rating.  This can be overcome by adding the tail light lamp, or another add-on like a faster acting circuit breaker.

 2. The Peco turnouts have a built-in jumper with a much higher resistance than the rails. This presents a possible hot spot if higher than expected current happens to flow.

Most accidents happen when two or more unexpected things occur at the same time.  You should write to both manufacturers and describe what happened.  These things need to be looked at and possibly changed by the manufacturer.