History and Design
The purpose of this blog is to chronicle my third attempt at creating a branch line terminus “shelf layout” in N gauge.
My first attempt was closely based on the real branch line terminus at Hemyock in Devon. All track was Peco, and all six turnouts were DCC-controlled with SMAIL slow-motion switch machines. I had to mount the switch machines sideways due to the low clearance under the benchwork, and this led to connectivity issues on the turnouts. To keep it all working for a couple of years, I spent most of my modeling time fiddling with the piano wires that connected the SMAIL machines to the turnouts. I relied on the SMAIL machines to power the “frogs” on the turnouts, and when they began to lose their ability to switch the polarity on the frogs I knew it was finally time to start over.
My second attempt was still based on Hemyock, but I made a few minor changes and then added a Peco turntable in front of the engine shed. All track (except for the turntable) was Kato Unitrack, and all six turnouts were DCC-controlled with Switch-Kat decoders. The turntable was driven by a stepper motor and programmed for DCC using an Arduino and DCC shield. Unfortunately, I couldn’t figure out a way to properly mount the motor directly to the underside of the turntable’s well, and this caused frequent misalignment between the bridge and approach tracks. I tried mounting it to an adjustable plate and then a 3-D printed bracket, to no avail. This past winter, I became frustrated enough by the turntable that I decided to start over again without it.
After taking a hiatus for a few months, I started planning. I used the free version of AnyRail software to design the layout. My current plan still resembles Hemyock, at a glance, although I’ve made substantial changes that have altered the way it will operate. I suppose this means that Little Snoring has come into its own.
All track is Kato Unitrack, and all curves (including the divergent tracks on turnouts) have a 15° radius over a length of at least 718mm. In my opinion, this gives the layout a flowing, cohesive appearance. Of course, using long curved sections takes up a lot of space on a small layout, but I’m not interested in cramming. My mantra while designing this version was “less is more.” To that end, I’m including five turnouts rather than six. On both previous versions, there was a turnout for an added spur between the runaround track and the goods shed. With the runaround so close, and with the secondary track below the station to store rolling stock, the spur was unnecessary.
When it comes to Kato Unitrack, “simplicity” also means “economical.” As most straight and curved sections come in packs of four, I tried to minimize the number of packages I’d have to buy, and spent less than $150 USD for all track (after applying a 10% discount code to my order from Model Train Stuff.
Previous versions of Little Snoring required benchwork that varied in length and width. The first version was 15″x 72″, and the second was 16″x 68″. The benchwork for both was dimensional softwood with a 3/4″ plywood surface. The plywood sheets were expensive and difficult to cut precisely with handheld power tools. To minimize benchwork costs this time around, and to reduce the amount of cutting I’d need to do, I decided to make the layout fit on a 12 x 72 stain grade panel, which cost less than $20 USD. Including the dimensional softwood, the entire benchwork cost less than $30 USD ($35 USD if the adjustable feet are taken into account).
DCC control and track power are provided by the excellent NCE PowerCab throttle that I’ve used on both previous layouts. I am reusing my Switch-Kat decoders to provide DCC control to turnouts, and of course I’m also using my Dapol GWR 2-8-0 engine for services to and from the station. I recently retired my Graham Farish GWR 0-6-0 after dropping it on a concrete floor, but I hope to replace it with the Dapol GWR 0-6-0, which was released a couple of years ago. The 0-6-0 will handle station switching and preparing outbound consists for the 2-8-0. Should be fun!
I should note here that I don’t have a proper workshop. My family and I live in a rather small farmhouse, and our shed doubles as a one-car garage in winter. I have a few handheld power tools, but I don’t particularly enjoy using them and it probably shows. So I use manual tools whenever possible, and I plan my woodworking projects accordingly.
This time, as noted above, I started with a 12 x 72 precut panel and three 72″ 1 x 3 planks. Simple and elegant… sort of. Like all dimensional lumber, the panel and planks had been milled to size before being kilned. After that, they were cut roughly to length. The kilning process shrunk the wood, though, so they are thinner than advertised: the panel is only 11.25″ wide and the planks are only 2.5″ wide. I had to account for this in my calculations.
With a miter saw, I cut down the ends of two planks because they were slightly longer than 72″. I cut two 9.75″ lengths from the third plank, to be used on the each end.
I printed a template of the faceplate for NCE’s PowerCab, which I found here (I lost the template that came with my PowerCab years ago). Using the template, I measured a slot in one of the 72” planks to make room for the printed circuit board and connectors, which the faceplate secures and conceals. I drilled a couple of small holes at the bottom corners then cut down into the plank with my miter saw, ending at the holes. I used a chisel (actually an old screwdriver that I sharpened) across the grain between the holes and removed the material.
With everything cut to size, I drilled pilot holes 3/4” from the edge of the panel, spaced roughly 6” apart. Then I simply started screwing the panel down onto the planks with 1.5” wood screws. With 3” wood screws, I screwed the ends of the long planks into the sides of the short planks for added stability.
Finally, I drilled a 1/2” hole in the center of the short plank on the right end. This would allow me to run power to additional modules in the future.
The result was a shallow box requiring very little measuring, even less cutting, and no sanding.
I test-fit the PowerCab panel, making sure the circuit board had plenty of clearance.
I added four adjustable feet. In retrospect, it would have been much easier to add them before assembling everything!
It was time to start setting things up to lay track.
Kato Unitrack sits on a plastic roadbed of “ballast” that is unsuitable for a railway station because it is so tall. The ties should be near ground level. I couldn’t remove the roadbed, of course, so I had to bring the level of the “ground” up closer to the height of the ties. To do this, my first instinct was to use foam project panels, which are almost 0.375” thick.
My thought was that foam would be much easier to cut than plywood or hardboard, and foam would react well to a static grass applicator, helping the grass to stand on end.
So after laying out the track, I cut three foam panels to fit the top of the module.
Then I measured the thickness of the module’s profile with the foam included. The thickness was 3.375”.
Because the track would need to extend to the edge of the fascia on the right-hand side of the module, I had to cut out a recess into the fascia before I could continue laying track.
I cut several strips of tempered hardboard to a thickness of 3.375” (roughly). I cut one strip to the width of the module, then adhered it to the right-hand side of the module with Liquid Nails construction adhesive, keeping it proud of the module surface by at least 0.375”. I clamped it down for a couple of hours, then cut a notch for the roadbed of the track to fit through.
After tracing the shape of the track on the module’s surface, I marked the locations where the feeder and turnout wires would be. Then, I drilled a hole at each location and countersunk each one to ensure that the wires wouldn’t cause any heaves above ground.
I laid the foam panels down on the module then laid out the track again, making sure the end of the tack on the rightmost side was flush with the face of the hardboard panel. I traced the track with a pencil onto the foam, then cut along these lines with a hobby knife. The shapes that remained were placed around the track to make sure everything fit.
As I worked with the foam, I began to realize how fragile it was. If I glued it down, there was no going back, and I thought about how difficult it would be to work on the surface of the layout without damaging it.
I picked up some “Sandeply” that was nearly the same thickness as the foam. I used the foam cuts as stencils and then I went to town with my jigsaw. And I—who always disliked power tools—actually started to enjoy it. The Sandeply didn’t splinter the way most plywood does, and it cut evenly and easily, like balsa wood.
Once cut, I test fit everything, made some adjustments (shaving a bit here, adding a shim there) then glued it all down with construction adhesive. I filled some of the gaps with Drydex spackle and sanded the surface flush. You’ll see that I also added the hardboard fascia across the front of the module, which enabled me to mount the PowerCab panel.
I turned my attention to the station platform, which would start as a simple raised section beginning at the tracks and filling the upper-left corner.
First, I had to figure out how tall the station platform should be in relation to the tracks. That would give me the elevation for the entire section. A quick search on RMweb brought me to understand that N gauge platforms should be about 6mm taller than the top of the rail.
With a few different thicknesses of balsa wood, I worked out which combination was closest. It was almost 1mm too tall, but from what I had been reading on RMweb, adding a bit of height can appear more realistic when viewed from above.
I cut the “platform” to fit the space, then glued it down. I forgot how much balsa warps when wet, so I put some heavy weights on it—even so, some sections were pulled away after it had dried. I filled those gaps with spackle and sanded the top of the platform flat.
It was time to start wiring. I dropped Unijoiner feeders (Kato-manufactured electrical feeders fitted to the section connectors) to ensure that each section of track was directly connected to at least one feeder. I didn’t want to rely on the Unijoiners to provide power from one section to another. That was overkill, of course.
At the same time, I dropped the turnout control wires. Due to some feeders and turnout wires not having much room to bend to reach the holes I had drilled under the track, I ended up having to countersink some larger openings at the top of each hole.
In this way, the track was able to lay perfectly flat above the feeders.
Underneath the layout, I ran bus wires from terminals at one end to terminals at the other, and connected them to the PowerCab circuit board. Then I started dropping Unijoiner wires as I connected and glued down each section of track.