Laying Peco 3rd rail

Practical experience from Colin Duff

This is my experience of laying 24 real feet of third rail on my 00 layout using Peco IL-120 conductor rail chairs, IL-1X code 60 rail section, and joiners SL-210X.

The IL-120 pack contains sprues providing approx. 100 conductor rail chairs, plus “spacers”. As the chairs are designed to give the correct conductor rail height when using code 75 running rails the spacers are provided to get the height correct for code 100 running rails. In use the spacers look like a packing shim, or a very overscale extra insulator.

I am using code 100 track, basically because I bought it in bulk before code 75 was widely available (purchase to laying elapsed time being about 5 years!) but also because I have some Hornby Dublo and Triang stock of sentimental value that I want to run every now and then. It was my original intention to fit the chairs onto code 100 track without the spacers to give extra clearance. At the last moment I changed my mind and fitted them – I am glad I did as it would never have looked right.


First design your 3rd rail layout according to prototype practice. There is a dearth of literature available on this, but after a lifetime of observing the SR 3rd rail, plus looking at pictures and taking notes whilst travelling (knowing I was going to model the 3rd rail eventually), I planned mine on general experience. Knowing that the 3rd rail is generally, but not always, laid on the outside except in stations, and that alternate sections are laid on alternate track sides, plus the length of a gap without conductor rail is kept as short as possible – basically I put myself in the position of a pick up shoe and then designed the layout so that my chances of being in contact with the third rail without being knocked off sideways were minimised. The back of the Peco pack also gives advice with which I do not entirely accord because it shows the conductor rail running straight alongside the straight side of a point. My prototype track notes suggest that this is only possible if the conductor rail is split length-wise and overlapped (like an expansion joint in CWR) with the inside approach being ramped up. My modelling skills are not up to reproducing this so wherever possible I changed track sides through the point. There are a few places on my layout where I cheated and in reality a pick up shoe would be sideswiped off. At baseboard joints I was originally going to solder the third rail to a brass nail hammered into the baseboard end board then gap the rail, but I decided instead to represent a section gap at each baseboard end even if the section lengths are then unrealistically short. It does make the third rail layout look realistically “busy”!

Measure the length of a conductor rail run, cut the code 60 rail to length, lightly filing the rail-foot (I think that is what it is called) at each end to a slight point. It is easier to thread the chairs onto the rail if the end of the rail is slightly chamfered – I only realised this after a day and a half. Turn down the rail ends at about a sleepers gap length from each end. I did not find the need to pre-curve any rail except where I had no choice but to join two lengths of rail with a joiner on a curve. At one place (but only one place) my track radius goes down to 18 inches, but even then I did not have any difficulty fixing the conductor rail without pre-bending. Drill 0.8mm holes (I used a #66 drill) in the end of approximately every 4th sleeper. I was going to make a drilling jig but in the end I did every hole by mk1 eyeball with no real problems. It is realistic to vary the chair spacing every now and then – towards the end of a rail section I would suggest inserting an extra chair, and on straight long runs I (purely accidentally) sometimes only had a chair every 5th sleeper (it does get very tedious laying this stuff!). I found that a slightly larger hole in the sleeper end was better than a slightly too small hole because the spigot attached to the chair is quite flexible and will bend rather than go into a slightly too small hole.

Cut as many chairs off the sprue as you are going to need for the length being laid, plus a few extra. Ditto spacers if you are using them. Thread the chairs onto the rail, placing them in approximate position. Thread on a few spares – it is easier to remove excess chairs at the end rather than thread on extras once laying has commenced. These small plastic parts do sometimes adopt projectile tendencies. At the end of two days I had a very sore blister on the end of the finger I used to press the chairs onto the rail.

A close up view of the third rail on a previous layout whilst still under construction. 
Photograph by Colin Duff

Insert the spigot of the first chair into the first hole. I used a jeweller’s screwdriver or the end of a knife to manoeuvre the chairs into position. Then pressing down lightly on the rail at that position to keep it in place (I could not find any plastic solvent to bond this stuff) manoeuvre the next chair into final position and place it in the next hole. Go along the length of the rail progressively. On short lengths, particularly on curves, there is a tendency for the whole thing to come unfixed before you get to the end of the run unless you can find some way of holding the already fixed part of the conductor rail in place. Parts can fly everywhere – start again! If using the spacers, I suggest positioning the spacers between the sleepers at the sides of the running rails in advance of the holes. At this point it all begins to look like real-life track laying is in progress. First manoeuvre the spigot over the spacer, lifting the spacer onto the spigot with the blade. Then holding the spacer on the spigot with the blade manoeuvre the assembly over the hole and pop it in quickly.

To finish off I soldered thick layout wire to conductor rail sections, threading it under the track where appropriate, etc, to reproduce the power feeders which will eventually disappear into the trunking I have yet to model.

I then ran my Hornby Networker. I do not know how accurate the positioning and dimensions of the shoe beam on this model are, but the shoe ran a microscopic distance above the conductor rail head and tracked the position of the conductor rail exactly. Did I feel good about it – you bet! Subsequently I have found that the pick up shoes on my Ian Kirk 2Bil plus Hornby Class 92 and Eurostar models all track the third rail laid this way well. The Lima Class 73 needs its shoes filing down as they are set too low and lift the model off the track slightly. Unfortunately the bogies on Golden Arrow Productions’ Class 71 model are made to exact 4mm scale so do not track third rail laid on 16.5mm gauge track – in fact the shoe beams foul it so badly that the model will not run.

In all it took me about 16 hours painstaking but tedious work to lay 24 feet of conductor rail. The good thing is that you get quicker as you go along and perfect the technique.


So how has the conductor rail stood the test of time?  Be warned – not pre-bending the conductor rail on curves has proved to be a mistake as the rail ends have a tendency to want to straighten so putting the chairs under strain and eventually working loose.  All rail ends also have a tendency to get damaged with the same effect.  Thus periodically I have had to replace chairs, particularly on the end of curves, and refit the end of the rail – not an easy operation with the rail already laid.  Pre-bending of rail for curves I now think is a  necessity and soldering conductor rail ends to a track nail or brass screw to secure them would be useful albeit at the cost of then making the subsequent replacement of chairs of required very difficult.  I have used subsequently Peco conductor rail components for other projects – for instance the diorama used for photographs of models in this section.   However, for my next layout I am considering soldering the conductor rail to track nails or brass screws placed at the end of sleepers and and only using chairs – either these Peco components or alternative metal castings – cosmetically.