A few years ago, I started working on a project to restore and re-imagine my childhood model railway layout. But to be honest, it didn’t get very far. When the COVID-19 lockdown hit, I ended up burning out, trying to tackle everything all at once. In my last post about it in 2022, I said I’d try to finish it if I found the motivation, but at that point I was in the middle of packing up the models and the layout board I’d built. I kind of thought that was it; that I’d given up. But, as it turns out, I’m still that little kid at heart when it comes to model trains—and here I am, looking at the project again.

Now that everything’s cleared away and the layout board I built is tucked into storage, I’ve had some time to rethink my ideas. Looking back, I think I know what really led me to give up last time, and it comes down to a few key things:

1. Lack of realism: I ended up with a mishmash of model trains from different eras and places that wouldn’t ever realistically be seen together. Many of these models are ones I’ve had since childhood, and I think nostalgia played a big role in wanting to run them all—even if they didn’t really belong together.
2. Taking on too much: I had far too many vehicles in my collection and wanted to run them all at once. This wasn’t helped by the fact that I’d chosen OO scale (1:76.2), which, in hindsight, didn’t leave much room to showcase everything without it all looking crowded and chaotic.
3. The layout got stale: Because of space limits, the whole layout ended up being basically just a double oval—nothing really exciting or dynamic.
4. It was just too big: The layout board itself was massive—almost 2 meters squared. It took up a huge chunk of space, which could’ve been used for other things.

It’s been nice to take a step back and reflect on all of this. Feels like I’m in a better place to figure out where to go next.

When I was a kid, I’d go to model railway shows with my dad, who probably got tired of taking me after a while! I was always fascinated by the smaller-scale layouts. N scale setups were my absolute favourite—they’re about half the size of the OO trains I have and can fit a lot more detail in a much smaller space (UK N scale is 1:148, which is just over half the size of OO). Now, as an adult (well, mostly) and looking back on those shows, I’ve realised that the best way forward might be to give up on running my OO scale locomotives and switch to N scale. If everything goes well, I’ll end up with a layout that takes up way less room in my house. If not, at least I’ll have an excuse to break out the power tools and cut up the board I built! I’m also planning to sell some of my OO scale trains to help fund the change, so I’ll be diving back into the wonderful world of eBay again soon!

The plan going forward

So, what’s the plan this time around to avoid the burnout I’ve hit before?

Well, I’m starting fresh—with a brand-new scale, no leftover mistakes from past layouts, and a better plan from the get-go. I’ve decided to switch to N scale, and I’ve also settled on a location and era: the modern-day GWR Didcot Parkway Station (UK era 11). It’s a practical choice—not only does it ground the layout in realism, but it’s also right down the road from me. If I need inspiration (or an excuse for a little field trip), I can just pop over and take a look in person.

Now, I get that the modern railway scene doesn’t have the same nostalgic glamour as the steam era, but that doesn’t mean it’s any less fun to model. In some ways, it’s actually more manageable—diesel and electric locomotives are simpler, with fewer moving parts compared to their steam-era counterparts—a.k.a. less parts to stop working. Simpler doesn’t mean worse though, as many models of modern locomotives have much more details on them than some of my OO scale steam ones. Plus, most new models come ready for Digital Command Control (DCC), which means I could eventually have independent control over multiple trains, lights, and sound effects. I’m not planning to dive into DCC right away, but knowing it’s an option for the future is a nice bonus.

And this brings me to another passion of mine: software engineering. I’ve got a few Raspberry Pi machines lying around, and I’ve already tinkered with getting them to interact with the world. Why not bring that into the layout? I’m thinking small automation tricks—street lights that turn on at real-life sunset, or real-time station announcements triggered when a train pulls into the platform. It’s a rabbit hole of possibilities, and honestly, that’s part of what makes this new attempt so exciting.

Starting off

Kato Precision Railroad Models, a Japanese manufacturer, has released a GWR Class 800 Intercity Express Train in GWR livery—and I know these pass through Didcot Parkway station as I regularly travel on them. The plain GWR livery one is no longer sold, but a depiction of the named Class 800 “Michael Bond and Paddington Bear” is available with an oval of N gauge track. The Class 800 is partially manufactured in Japan, so I figured that a Japanese model company would have the best representation of this train.

I also know that a Class 66 diesel locomotive “66017” in DB Cargo livery works around the Didcot Parkway station, and there’s evidence that it has been stationed there back in 2018. Revolution Trains are currently taking pre-orders for N scale Class 66 models in several liveries, where you can choose the running number. Needless to say, having a bespoke model of the train that has actually been in the station I’m modelling was too good of an opportunity to pass up.

Limiting myself to these two should help me stay focused and avoid getting overwhelmed. They’re a solid representation of both passenger and freight traffic through my chosen location, which makes them a great starting point. Now, the challenge is working out the best layout I can create within the board space I have, using a small selection of track pieces.

Beyond the trains themselves, I’m also starting to explore how I can integrate some Raspberry Pi magic into the layout using the hardware I already have tucked away in my desk drawers. There are so many small but exciting possibilities, and I’ve been taking inspiration from videos online. As I refine my plans, I’ll be sure to share updates here!

There’s a lot to think about, but I feel like the pieces are finally coming together. I’ll try to post regular updates on this blog and hopefully, in the coming months, I’ll have something interesting to show. Stay tuned!

Credit to ChatGPT for helping refine my writing and make it less technical.

I started a project at the beginning of the COVID-19 lockdown, which was to restore and reimagine a 25-year-old model railway. While I ran out of steam over the course of the next few years, I did make a small amount of progress which I documented in another place on this site. I’ve decided to move that here now that I’m doing some much needed clean-up around here – it helps reduce the amount of duplication there is anyway!

That said, both the build and this website have been neglected for so long that most of it isn’t as relevant as it used to be. However, it probably makes more sense to have the updates all in one central location and it means the other posts aren’t as hidden as they were.

Perhaps moving forward I’ll be able to regain the motivation to finish the project.

The Class 09 shunter I rebuilt has had a few more tweaks added to it now it’s working properly again.

First off, I’ve put some weight into the locomotive — more weight means there is less likely to be any wheel spin when pulling larger loads. Since the model is mostly plastic, it doesn’t weigh much at all. Originally, it had a solid lump of lead placed inside it to weigh it down but this doesn’t exist in my model — and besides, lead is toxic. I picked up a pack of iron car wheel weights and placed a bunch inside. There’s now an additional 45g of weight just above the wheels.

The second change is that I’ve put in window glazing which gives a more realistic appearance to the cab.

I’ve mentioned a few times that I’ve had to service a locomotive in order to get it working. I also have several draft posts for each locomotive I’ve serviced but each of these posts reads pretty much the same with only the photos and names differing. Therefore I think it probably makes sense to write a single post outlining the way I do the servicing, as opposed to duplicating a post structure for each separate locomotive.

The logical first step to giving a locomotive a full service is to remove the decorative bits that can get damaged easily. This includes things like the body shells and any detailing parts present near wheels. Many of the locomotives have their body shells connected to their chassis by a few screws so removing them is relatively straightforward. This also lets me give these shells a quick clean so they look better when reattached.

Once the shells are removed, I can clean down electrical contacts and metallic parts with rubbing alcohol or similar. I have a small bottle of isopropanol alcohol (IPA) that I put in a bottle cap and use with a cotton bud to remove any caked on dirt. This removes any possible sources of electrical isolation for things such as motors and lights, and mainly involves spinning wheels underneath the cotton buds or wiping down larger electrical contacts. Sometimes this is all that is needed to improve the performance of a locomotive.

It’s recommended to do a more thorough clean of motors and gearboxes at least once a year if the model has been run quite significantly. Since many of these locomotives have not been run in quite some time, I thought it would be good to try and do this for them all anyway.

The older locomotives tend to have the ringfield-style motors (a small version seen above, partially opened) which are easy to take apart. The most important thing to do in here is to clean the brushes and commutator as these are the components that route electricity in the motor. Since they are moving parts that physically touch, they do build up grime. So far I’ve only gone as far as dipping the brushes in the IPA bottle cap and rubbing down the commutator with the cotton bud, but it cleans them up enough to get them going pretty steadily again.

A couple of the locomotives have had damaged gearing or loose parts. These have all been fixed with small amounts of epoxy glue and hopefully they will hold for the foreseeable future.

Before reassembling the locomotive, I will oil all the motion mechanisms. I do this by putting a small amount of oil in a container and using a paper clip to directly apply drops on important parts. The paperclip is metallic so the oil doesn’t seep into it like it would a toothpick for example, and it’s precise enough that I can apply just a drop as opposed to a massive blob. I’ve also used other stainless steel pins in the past where a paperclip didn’t work — the most memorable of these was the bent leg of a burnt-out LED, which was surprisingly efficient!

These techniques also work for non-locomotives. Old coaches with slightly seized wheels can be oiled the same way and their wheels can also be cleaned up as they can spread dirt around a layout. I’ve also got an old turntable with an ageing mechanism which I cleaned up somewhat. I’ve cleaned up the majority of the running fleet for the Guelder Valley layout and they mostly seem stable now asides from Cadbury Castle which struggles going forward and needs further work. I’ll have to take a closer look at this locomotive in the future.

In one of my earlier posts, I mentioned that I had one more damaged locomotive that I had yet to reveal. Truth be told, I wasn’t entirely sure I wanted to touch this one — it wasn’t in the best state and definitely not suited for the era that Guelder Valley is set in — but I’m genuinely glad I have now. While the repairs I’ve done on the other locomotives so far have been freeing up wheels, reattaching gears or fixing cosmetics, this one required quite a significant rebuild: let me introduce you to the little shunter that could: a British Rail Class 08/09.

When I was a younger nerdlet than I am now, my father took me to a model railway show. I can’t remember what year it was, or which country, but I seem to recall getting three significant pieces from the show: the locomotive shed, the Eurostar and rescuing what could only be described as a sacrificial model — that is to say, parts had been taken from it to repair other models. The person who gave it to me probably thought I was crazy when I took it off his hands but back then I thought I could repair it. However it soon became apparent that I didn’t have the skills. It sat in a box in several pieces as I grew up and forgot about it.

With the rebuild of Guelder Valley put on hold due to the current crisis, I decided to take a closer look at the model. After all, many years had passed; I must have the skills to fix it by now… right?

Twenty-or-so years in storage had put a solid layer of dirt on the shell and oxidised the wheels somewhat — but what was there was at least in good shape, all things considered.

The most logical place to start would be to see if there were service sheets for the model. But what make was it? What locomotive class? I fired up my favourite search engine and typed in the little I knew about the model: “BR 0-6-0 diesel”. The first few results confirmed that this was either a Class 08 shunter or the similarly-designed Class 09. But which company made it?

Model locomotives usually have the company name printed on the bottom. Unfortunately, it appears that the bottom of this locomotive had been part of the sacrifices — there was nothing there to hold the wheels in or that I could use to work out who the manufacturer was. I went back to the search engine.

Turns out most of the major manufacturers who create models in this this scale have made a model Class 08. I ruled out the Bachmann products as they were far too detailed. Hornby have made quite a few models of the Class 08, but none of theirs seemed to match. Dapol make theirs in a different scale. I’d come up with nothing and was worried that perhaps I might not be able to do this “to spec” so to speak.

I wasn’t going to give up, though. Not sure where to start, I looked for the easiest things that could be done with minimal guidance. The most straight-forward was to reconnect the coupling rod to the wheels — in the photo above, you can see the rear wheels are not connected at all to one side. The process was actually quite straightforward: I took an old panel pin, clipped off the end using bolt cutters and glued it to the coupling rod connector.

This fix seemed good enough, but it wouldn’t stop the wheels falling out the bottom of the chassis. With renewed confidence, I pulled out my callipers and took a whole bunch of measurements of the bottom of the locomotive. With a sheet of A4 paper covered in numbers and scribbles, I opened up TinkerCAD, a 3D modelling program.

After a few hours of designing and refining, I put the design on my 3D printer and created the first draft of the bottom chassis. It fit pretty well — screw holes were correct and it didn’t obstruct anything. Unfortunately, it wasn’t very secure: the screw holes in the upper chassis are slightly recessed and since I wanted to keep the design free from the wheels, there was about half a millimetre of wiggle room. I went back to the drawing board.

It took me a few more attempts to get the size correct and movement-free. I’d noticed the upper chassis had a square hole near the back which I decided to use as an anchor, so I placed a raised wedge on on that part of the design. The first revision used rounded screw supports but was far too thin and ended up snapping when attached — this is the left one in the photo above. The second revision was thicker and used square screw supports to make it easier to print — pictured on the right of the photo. This proved to be a much better fit and I was planning to use this one until I realised I could cover a larger part of the axles by adding small lips to each side of the design where the wheels were. Using the second revision, I measured out the locations of the axles and went back to TinkerCAD.

This new revision had the added benefit of interlocking with grooves on the upper chassis to prevent any sort of wiggling, as well as covering the exposed areas of the axles and preventing any lubricant from dripping onto the track below while the locomotive was running.

At this point, I’d been looking at a spares website to get a digital blanking board for the Peppercorn A1 ‘Tornado’ when I noticed a spare Lima motor that looked exactly the same as the one on this model. I hadn’t considered that an Italian model railway company would have made British locomotives, but it turns out that this is a Lima Class 09 model! Unfortunately, Lima became part of Hornby many years ago and spares for the models they released before merging are hard to find. I was able to purchase some new buffers for the locomotive which I’ve applied, but couldn’t find many more spares.

Now that the wheels were back in the chassis and I knew the manufacturer, I took a look at the cosmetics on the model. One side of the cab had a missing ladder. This was pretty straightforward to design in TinkerCAD and I printed this out and applied it straight away, no revisions needed. I gave the model some detailing paints and looked to the final hurdle: getting the motor running.

With the wheels back in, the motor was able to pick up power from the rails again. Unfortunately, this simply made loud grinding sounds. Gearing between the motor and wheels was missing, as well as the TV suppressor. Fortunately, I have a set of capacitors on hand and was able to fashion a Hornby-style TV suppressor with a bit of soldering. The gear, however, was more difficult.

I’d read somewhere that the Lima gears that were used had either 19 or 20 teeth so I started from there and worked backwards. I measured every dimension possible — the motor gear dimensions, the size of the axles, the length of the gear teeth on the existing gears to name but a few. It took me several hours of poking things with callipers and pretending to do maths to end up with an absolute maximum and minimum diameter of gear that I would need. Fortunately for me, TinkerCAD provides a means for creating a gear from a number of required teeth, so I scaled that down to get a gear I was even remotely confident would print and fit. I didn’t expect much from the first draft, so I was actually quite surprised that it fit perfectly first try.

Of course, perfect fit means nothing if the motor isn’t back in the chassis, and I will admit that when reassembling the motor, I sent one of the washers pinging across the room by accident and had to crawl around with a magnet to find it again. With that excitement out of the way, I placed the body shell back on and prepared to witness either the greatest moment of this year… or the most disappointing.

Suffice to say, that’s quite the positive result!