Slate wagon kit roundup

A quick look at the slate wagons kits available in 16mm/ft at the moment (Autumn 2021)…

‘uss_Raven74656’ on ebay – £8 – 3D printed model of the steel slate wagon

The Lineside hut – £9 – Laser cut plywood 3-plank
(A resin slate load is also avaialble for £5)

PDF models £9 – Plywood 2-plank
(bulk discounts available)

Binnie Eng (Approx £13, but not currently available) Colin and Peter’s original 3-plank wagon.
Peter’s FAQ says that he’s in the process of making new moulds –
How Colin Binnie made the originals

Resurgam £18-22, MDF or Plywood – FR 2-ton steel waggon

IP Engineering – £20 – Laser cut plywood 3-plank.
(Note: Search engines may give you this page, listing the Binnie wagon at £13. This page is quite out of date and these wagons are not available)

Slaters – £20 – £26 – – Dinorwic or Festiniog wagons, may contain brass etches.

Bole Laser – £30/£35 – 3D printed 2-bar as on Tallyllyn

Coast Line Models – from £17 – various kits for steel and timber slate wagons (requires Slaters’ wheels to complete)


This article explains how to make your own:

Posted in 16mm, Garden Railways, Trains | Comments Off on Slate wagon kit roundup

A list of preserved railways in Britain

Whilst on holiday we visited the North York Moors railway near Whitby. I wondered what the longest length of preserved railway line is in Britain, and does someone have a handy list. The answer seems to be no.

So here is my list, which is neither complete, nor comprehensive. For a complete list, see Wikipedia.

Standard Gauge lines – Name, where it runsMileskm
West Somerset Railway, Minehead to Bishops Lydeard (near Taunton)22.7536.6
Wensleydale Railway, Northallerton to Redmire2235
North York Moors Railway, Pickering to Grosmont (see below)18 29
Severn Valley Railway, Kidderminster-Bewdley-Bridgnorth1626
Mid Hants Railway “Watercress Line”, Alresford to Alton1016
Bluebell Railway, East Grinsted to Sheffield Park (near Haywards Heath)1118
Epping and Ongar Railway610
Narrow Gauge Lines – Name, where it runs
Leighton Buzzard Light Railway (2′ gauge narrow gauge)35
Welsh Highland Railway (Caernarfon-Porthmadoc) (2′ gauge narrow gauge)2540
Ffestiniog Railway (Porthmadoc-Ffestiniog) (2′ gauge narrow gauge)13.522

Some trains on the NYMR run through Grosmont into Whitby, adding another 7 miles to the route. This would make the NYMR the longest preserved line in Britain, as well as the busiest in the world.

Based on my very limited research, the longest length of preserved railway in Britain isn’t standard gauge, but is the combined WHR and Ffestiniog Railway. Thus the oldest surviving railway company in the world is also Britain’s longest heritage railway.

Posted in Uncategorized | Tagged | Comments Off on A list of preserved railways in Britain

Small Tool Caddy

How often have I been working on a project somewhere around the house or garden, and needed a box of screws, a few bolts and washers and some small tools.

During the project I invariably need a few smaller tools, but I only have two hands. So I make several journeys to the workshop or the tool cupboard to get the bits I need.

When it comes to tidying up, I don’t want to cart my entire toolbox to where I was working, but I end up making five journeys back and forth to tidy up.

What I need is a small tool-caddy. Like this:

  • Cost: zero. It’s made of 15mm ply because I have some. Plus a wooden batten for the handle, some wood glue, and two screws.
  • Tools used to build: hand saws, a cordless drill, and a band clamp.
  • Time taken: an hour here, an hour there, over a few evenings.

The lessons I learned here are about squareness. I didn’t really make much of an effort to cut the ends of the side pieces square … and the caddy is slightly wonky. Next time I will take more care to mark and cut square.

The base was cut to be oversize – about 3mm each size, then it was glued on and left to set overnight. I trimmed the base using a Japanese Pull-Saw and cleaned up the edges and corners with sandpaper.

Is it worth oiling it or painting it? Probably not. I’ve already found that it isn’t long enough for my hammer.

Posted in Tools, Wood | Tagged , , , | Comments Off on Small Tool Caddy

Lego Brackets

I got a new LEGO set recently.

I love discovering new pieces. I go all “Ooh, I’ve never seen one of those before”. Brackets are particularly interesting because they give you options for building things “sideways”

In this set, there are some lovely new right-angle pieces – a 2×2 L bracket, a 2×1 L bracket and a 2×2 T bracket. The 2×2 brick in the photo below I’ve seen before in 1×1 and 4×1 form.

I’m particularly interested in the 2×2 L bracket as the short side is the other way round from the ones I’ve seen before.

The ones I’ve seen before have the studs on the 2×1 side facing outwards –

The new ones have the studs on the 2×1 side facing inwards –

I’m having a go at building the marble run components invented by JK-Brickworks (see and Below, you can see two bridge pieces on the left and tower piece on the right. The bridge pieces use 2×2 L brackets with studs out on the 2×1 side

If I use the brackets I found in my new set, I can get the bridge pieces a lot closer to the tower. This might help the marbles flow more easily

Posted in Lego, Marble Run | Comments Off on Lego Brackets

Wooden loco

This is the result of my attempt to “3D print in plywood” – converting the 3D printer file to 2D, printing it onto paper and gluing it onto plywood. It’s not meant to be glamorous – it’s a first attempt. The chassis underneath is a combination of various kits.

Posted in 16mm, 3D printer, Garden Railways, Wood | Tagged , , | Comments Off on Wooden loco

3D printing on a 2D printer

One of the great things about having a 3D printer is the vast online library of things that one can print. Sites such as thingiverse contain everything you never knew you needed – from an airbrush stand to Lego bricks that don’t exist, via drill-bit sharpening jigs for a Dremel, and an articulated slug.

I found a 3D printed locomotive. I encountered two problems: some of the STL files contain multiple parts, and are too big for my printer; and I can’t find a way of splitting the parts so they can be printed separately (I tried Meshlab and OpenSCAD, and my brother tried Fusion360). Additionally, this model involves printing some parts that are just flat plates of plastic. There are more efficient ways of making these, such as using styrene sheet. However, it occurred to me that, if most of the parts are flat, maybe I could make them out of plywood.

So maybe the title of this article should be : 3D printing in Plywood

To do this, I need a 2D drawing of the parts (to glue onto a sheet of plywood so I can cut it out). But STL files are 3D models. However, it is possible to convert them. The process involves loading the 3D model in a CAD program and projecting the 3D model onto a 2D surface, then exporting as a drawing. I’m sure it can be done in one piece of software, but I used a 2-step process. I found the answers here (project and convert from STL to DXF) and here. (open DXF onto a drawing).

Step 1: Convert STL to DXF.

I like using OpenSCAD. It’s free, and it allows one to build up a 3D model using “Constructive Solid Geometry” – adding and subtracting cylinders, cubes and other shapes. It also supports fonts, and loads of other amazing features, but it’s not a pointy-clicky-draggy-droppy software. You have to type stuff.

  • Launch OpenSCAD and put this text in the editor on the left (using your own location to your STL file):
import("/Users/Path/files/cab_2.stl", convexity=3);
  • Render the model (F6, or on the menu choose Design >> Render)
  • Use Export >> Export as DXF
  • It looks like this in OpenSCAD.

Next is to open the DXF file in a CAD program and print it. There are loads of CAD software packages out there, ranging from free to expensive commercial. AutoCAD seems to be the most popular, and I’ve heard people have great success with TurboCAD and Dassault. I haven’t purchased a license for any of these commercial tools, and don’t need to. The free open source “FreeCAD” does the job. The documentation is pretty good. In essence I did the following:

  • Load the DXF file.
  • Switch to the TechDraw workbench (another toolbar appears)
  • Press the “New Pick” button (after moving the menus around to find the orange bit)
  • I chose the A4_Portrait_ISO7200TD template. (TD = Technical Drawing)
  • In the tree view, I selected all 608 shapes. Shift-click is your friend
  • I changed the properties as per the instructions, then adjusted the X and Y values to put the parts in the right place on the drawing.

the result looks like this in FreeCAD:

It remains to do File >> Print or File >> Export PDF which results in a lovely piece of A4 paper which I can glue on the plywood to cut it out with a fret-saw.

Update: the end result is visible here.

Posted in 16mm, 3D printer, Garden Railways, Trains, Wood | Comments Off on 3D printing on a 2D printer

How to fail at painting something to look like steel.

The short version: don’t use car touch-up paints for large flat areas.

I bought a wagon for my garden railway a few years ago. I thought it was an oil wagon but it turns out to be a model of a World War One water-tank wagon. The real thing has been restored and is on the Apedale Railway Staffordshire. The model is produced by Swift Sixteen.

I was originally going to paint it black, and experiment with weathering techniques to make it look messy and oily. But when I realised it wasn’t an oil tank, I thought I’d try painting it to look like the real thing. How does one paint things in steel? The model itself was already built when I obtained it, so I can’t take it apart. It’s made of resin, and already had some brown paint on it.

This is my first attempt at painting with metallics. It involved a trip to Halfords. They sell touch-up paints for cars. I thought “if a car is nice and shiny, then this should be a breeze”. I was wrong. This stuff is meant for touching up scratches. It doesn’t do large areas neatly. The brush inside is tiny. I tried a wider brush, with better results, but I found that it softens existing paint underneath it, and I couldn’t clean the brush – not even with white spirit. I’ve now thrown away three brushes.

And this is what the results look like on the wagon. I’m not satisfied with this result. Time to reconsider.

So where did I go wrong? How does one paint metallics? What should I use instead? Watch this space.

Posted in 16mm, Garden Railways, industrial archaeology, Trains | Tagged , , | Comments Off on How to fail at painting something to look like steel.

A 3D Printer diary – Slicer settings

I was getting poor quality prints on my 3D printer. There are a number of possible factors:

  • The surface texture of my prints is not as good as I would like. That’s the subject of this post.
  • Sometimes the first layer doesn’t stick very well to the base, and occasionally gets knocked by the print head. This is usually overcome by means of a “raft” or a “brim”.
  • Objects stick really solidly to the base when they are done, and I have to use a knife to get them off.
  • The filament is old, and has probably acquired moisture, so it is brittle. If I haven’t used the printer for a week, the filament between the spool and the print head snaps. I just feed the end of the filament into the top of the print head by hand, and the prints seem to come out OK.

There is plenty of useful advice for calibrating a 3D printer including this video, this blog at TeachingTech. and this page at All3DP. You can get calibration cubes, temperature towers, a test boat, and other prints from Thingiverse. These blogs look at dozens of issues that might be causing poor prints.

My printer is a Dremel 3D20, and I use Cura for my slicer. From Cura, I save the sliced gcode file to an SD card. I then plug the SD card into the printer and print from there. In reading the blogs mentioned above, I realised that I had been using the “Draft” settings in Cura rather than the “Fine” or “Normal” settings. Using the latter, and printing off some calibration cubes, I found that the results are actually quite good.

Lesson 1: Problem exists in chair, not in computer.
(I was using the wrong settings in Cura – the clue is in the word “draft”).

Cura has a number of different print profiles – from “Extra coarse” through “draft” and “normal” to “extra fine”. The difference is mostly in the layer height – the vertical thickness of each layer laid down by the 3D printer. I printed off the same “thing” multiple times with different profiles to compare the results.

Before we get to the results, here is some context. The “thing” I printed is a tower support for the amazing Gravitrax marble run from Ravensburger.

We never have enough tower supports, so this seemed like a great opportunity to try out the printer, and create some more opportunities for amazing marble runs whilst also trying out different slicer settings. Of course, someone has already designed the bits.

The default profiles in Cura are mostly the same settings : 20% infill, 200 degrees temperature, 0.8mm wall thickness, 60mm/s print speed. I chose to use supports but not raft or brim. The main difference in each test is the Layer Height – i.e. how thin each layer deposited is. Thicker layer heights equate to poorer quality but faster prints. Hence the use of the word “draft”.

The print is actually done upside down on the printer, with the “bottom” lugs facing upwards. Supports are used so that the cutout in the “top” of the print has a smooth surface.

COARSE: Layer height 0.4mm. Cura estimated a print time of 21 minutes, using 283cm of filament. Actual print time was 25 minutes. This is the result (in white) next to a real (ABS injection moulded) tower support (in grey)

“Coarse” 3D printed Gravitrax Tower piece (left) next to genuine injection moulded part.

On this “coarse” print, the supports remained on the bed when I removed the model. There are some imperfections, but these are to be expected on the coarse settings.

DRAFT: Layer height 0.2mm. Cura estimated a print time of 41 minutes, using 292cm of filament. Actual print time was 47 minutes. This is the result (in white) next to a real (ABS injection moulded) tower support (in grey).

“Draft” 3D printed Gravitrax Tower piece (left) next to genuine injection moulded part.

NORMAL: Layer height 0.15mm and an infill density of 10% instead of 20%. Cura estimated a print time of 47 minutes, using 294cm of filament. Actual print time was 50 minutes. This is the result (in white).

“Normal” 3D printed Gravitrax Tower piece (left) next to genuine injection moulded part.

FINE: Layer height 0.1mm. Cura estimated a print time of 81 minutes, using 294cm of filament. Actual print time was 93 minutes. This is the result (in white).

“Fine” 3D printed Gravitrax Tower piece (left) next to genuine injection moulded part.

Comparison: Most of my problems can be resolved by using the “normal” profile in Cura instead of Draft. On these prints. I can’t see a lot of difference in quality between Normal and Fine.

“Coarse” (left), “Draft”, “Normal” and “Fine” (right).


  • Look up the word “Draft” in the dictionary.
  • Understand how the slicing software works
  • Running a test like this (same “thing” different settings using the default profiles of the slicer) is a great way to understand what’s going on.

Posted in 3D printer | Tagged , , | Comments Off on A 3D Printer diary – Slicer settings

Running on air

I am building a steam locomotive from a kit. Last week, I tried steaming up for the first time, and it was a very exciting success. The instructions discuss correcting the timing, and then running in. To do the timing, it is suggested that one runs the loco on compressed air instead of steam, because steam makes everything hot, and fingers can’t adjust things very well if they are hot.

Everyone I’ve asked about running on air has bought a compressor. They recommend all sorts: quiet ones, small ones, large ones, compressors for air brushes or for commercial use. I don’t want to buy one, because I will use it rarely, and I don’t have anywhere to store it.

A friend of mine does a lot of his own motorcycle maintenance. He has added pressure gauges inside his tyres and somehow wired that up to a custom dashboard (link to his Facebook page). To test the pressure gauges, he used a fizzy drinks bottle. And that got me thinking.

running on air

The working pressure of the steam loco is about 40 to 60 psi (lbs per square inch). A fizzy drinks bottle can apparently withstand about 170 psi. I drilled a hole in the cap to take a Schrader tyre valve (same as on a bicycle or a car). I bought a spare foot-pump hose from Halfords for about £5, which fits snugly on the end of the superheater pipe. A foot-pump is used to pressurise the bottle.

It only runs for a few seconds, but that should be enough to check if the pistons and valve gear are getting stuck anywhere, and hence to try out the tuning. For a proper test, it needs to be run on steam again.

Here is a photo of the whole setup.

Posted in 16mm, Garden Railways, Trains | Tagged , , , | Comments Off on Running on air

First run of my loco

I am building a live steam loco. It’s a kit from Roundhouse Engineering, and it’s relatively straightforward to build. Last weekend I lit the burner for the first time, raised steam and tested to see if it would run.

Roundhouse “Billy” kit steam up 25 October 2020

The black vertical tank at the back is the gas tank, and contains gas – basically the same stuff as Camping Gaz. There is a wheel to control how much gas flows out of the tank into the burner, which runs in a tube in the middle of the boiler.

The boiler (the large horizontal copper tube) is mostly filled with water, and has another smaller tube within it containing the burner (so that the burner doesn’t get wet). The burner heats up the boiler, heating up the water, producing steam. The steam gathers at the top of the boiler, building up pressure.

The regulator controls how much of that steam is sent to the pistons (at the front, via an internal pipe that you can’t see). The reversing lever moves the mechanisms at the side – “valve gear” – to determine if the loco goes forwards or backwards.

The safety valve correctly operates when there is too much pressure in the boiler, preventing the boiler from exploding. In the second run, the loco moves. The motion is jerky because the valve gear it hasn’t been properly “tuned”, and the loco hasn’t been run in yet. Running in requires running the loco on the track, or raised up on blocks, for several hours.

Posted in 16mm, Garden Railways, Trains | Comments Off on First run of my loco