New DCC Layout Short Circuit

[AndrueC]

Oh that is good news! How did you diagnose it in the end? The steps you took could help other lost souls in the future.

[Dungrange]

2 hours ago, Quikbrew said:

Well, that was a LOT quicker to fix than I expected!!! Faulty set of points! (pictured) Thanks for the early Christmas present guys!!!

 

Have got it powered-up, just DC, to test and my little Hornby 0-4-0 tank engine is whizzing around like a good 'un!

 

So, do I still need the isolated rail joiners after each set of points then?

 

Great to hear that you've found the problem and now have a working layout - especially if you were hoping to play trains over Christmas.  Do you still need the isolated rail joiners after each set of points then?  Not if it works.  However, it depends on what you plan to do with the faulty point, which I assume is the one highlighted with the arrow.  If you plan on replacing it, then you won't need insulated joiners.  However, if you can't afford to replace it, then you could reuse it provided you fit insulated rail joiners on both rails from the V in the same way as you would if it was an Electrofrog point.  The short will be caused because the two crossover wires within the frog, which should be electrically isolated from one another obviously aren't - ie a manufacturing defect that won't be obvious with a visual inspection.  That means that both switchblades, closure rails and exit rails from the common crossing will be the same polarity and it therefore causes a short with the droppers in your sidings.  In one position, the short will be in one siding, and when you throw the point it will be the other siding that is causing the short.  However, with insulated joiners, that shouldn't be a problem.

[Michael Hodgson]

Personally with that layout I would wire the inner circle and outer as two separate circuits with double rail insulation in mid-crossover.  I would feed each circle through its own double pole on-off switch, so that I could isolate either one to detemine which one had the fault. 

 

Once you've got it all wired and working, you know it's not a wiring problem unless you've accidentally done some damage, but the problem can still raise its ugly head.  Operationally it is a real pain when the whole lot goes dead because something trivial like a screw has fallen onto the track somewhere inconvenient.  Or a can of coke!  Sod's law applies.  So if you know which half of the layout to check, you'll find it quicker. 

 

PS - glad to see you've found the original fault

Edited December 23, 2022 by Michael Hodgson

[Pete the Elaner]

It sounds like you've made some good progress, both with your layout & knowledge.

I've always felt that rail joiners are a weak point for conductivity. This goes back to an old loft layout of mine when I was a teenager:

Trains got to the back of the layout then slowed quite noticeably. Surely rail-friction join-fishplate-friction join-rail cannot be as good as wire-solder-rail? ... It isn't.

Since you have fed every rail, insulating after every point will make the section fed much smaller: no chance of the section being fed from an unsuspected location. The smaller the sections the easier it is to troubleshoot & the less chance there is of something going wrong.

[Quikbrew]

On 23/12/2022 at 19:15, AndrueC said:

Oh that is good news! How did you diagnose it in the end? The steps you took could help other lost souls in the future.

AndrueC

 

I'd already desoldered all the points in a previous act of desperation(!), so it was just a matter of pulling out and checking each set of points with a mulitmeter each set of points until I found the offending one. It was the second set I tried.

 

I'm about to wire them all back in, but will make sure that I check the circuit after I add each one.

[Quikbrew]

Thanks for all the tips guys. You learn a lot with this hobby!

[Pete the Elaner]

2 minutes ago, Quikbrew said:

 

I'd already desoldered all the points in a previous act of desperation(!), so it was just a matter of pulling out and checking each set of points with a mulitmeter each set of points until I found the offending one. It was the second set I tried.

 

I'm about to wire them all back in, but will make sure that I check the circuit after I add each one.

 

 

On 23/12/2022 at 23:53, Pete the Elaner said:

It sounds like you've made some good progress, both with your layout & knowledge.

I've always felt that rail joiners are a weak point for conductivity. This goes back to an old loft layout of mine when I was a teenager:

Trains got to the back of the layout then slowed quite noticeably. Surely rail-friction join-fishplate-friction join-rail cannot be as good as wire-solder-rail? ... It isn't.

Since you have fed every rail, insulating after every point will make the section fed much smaller: no chance of the section being fed from an unsuspected location. The smaller the sections the easier it is to troubleshoot & the less chance there is of something going wrong.

 

Make many smaller sections rather than fewer larger ones then troubleshooting will be much easier. By all means test each as you lay it, but isolate them from each other then feed them from below (because connections from below can easily be snipped & re-made).

Don't join the rails together. Rail joiners are a weak point & nickel silver rail is a poor conductor compared to copper wire.

[Michael Hodgson]

There is a case to be made for using screw connectors to connect the droppers to the DCC bus, rather than soldering at that end of the droppers.  When you've got it all correctly wired and tested, in principle it should never need to be tested again - you can assume any short is probably caused by a derailed wheelset, a faulty piece of rolling stock or some foreign body like a pair of pliers left somewhere silly.  However layout wiring faults can appear over time, for example rails expanding in hot weather and gaps that should be insulated closing up.  Such intermittent faults are a b*gger to find.  

 

 

[AndrueC]

11 hours ago, Michael Hodgson said:

There is a case to be made for using screw connectors to connect the droppers to the DCC bus, rather than soldering at that end of the droppers.  When you've got it all correctly wired and tested, in principle it should never need to be tested again - you can assume any short is probably caused by a derailed wheelset, a faulty piece of rolling stock or some foreign body like a pair of pliers left somewhere silly.  However layout wiring faults can appear over time, for example rails expanding in hot weather and gaps that should be insulated closing up.  Such intermittent faults are a b*gger to find.  

When I was building my current layout I toyed with the idea of using isolating joiners everywhere instead of metal fishplates. If I'd done that I would've used some form of screw connector to the bus rather than suitcase connectors. In the end I decided not to bother.