Train Speeds in Model Railways

The club has a couple of new train speed measurement gadgets — one permanent and the other portable. Why is measuring the speed of your model trains important? And what speed should they be running at anyway? Read more to find out.

By scaling the linear dimensions of a ‘thing’ in the real world down, we can build a representative scale model of that same ‘thing’ in miniature. The more accurately and consistently we do it in the three linear dimensions (length, width, and height), the more accurate, representative, and pleasing the resulting scale model is. The model railway hobby applies this principal to build accurate representations of real locomotives and trains that we can enjoy indoors.

But what about time? Sometimes called the fourth dimension, how should time be scaled when building a scale model of a railway? For static models like a building or a diorama, this isn’t a question that comes up. Nothing is moving so nothing changes with time.

But when modelling moving trains, it suddenly becomes relevant. The position of the model train changes over time. How quickly should it move through the scene? If we are aiming for a highly detailed and prototypically accurate model, thinking about this certainly deserves some attention!

The ‘natural’ approach is to simply not scale time at all. If the prototype runs with a maximum speed limit of 120 km/h (such as on the Perth suburban network), then an N Scale (1:160) model should run no faster than 750 m/h or 21 cm/s. 21 centimetres is only a little over half a foot and one second can seem like an eternity to a bored child or teenager! That probably explains why we are all guilty of running model trains much faster than we should if we are trying to be prototypically accurate.

Of course, modelling naturally means having to make some compromises. It is not possible to model everything 100% accurately – corners must be cut and adjustments made to accommodate real-world constraints. In HO Scale (1:87), modelling the original 1881 railway line between Perth and Fremantle would need to be 218 metres long! Obviously, that is not practical (at least not without some extraordinary effort!) so compromises must be made.

Typically, ‘scene compression’ is employed; the interesting station area and surrounding yards and sidings are accurately modelled, but the long stretches of tracks between stations are drastically shortened – sometimes to the point where when the back of the train is leaving the origin station, the loco at the front of the train is already arriving at the destination! Which means if we are already naturally running trains faster than we should, then how are we supposed to operate the model railway to any sort of prototypical timetable?!

There are many techniques that can be employed to alleviate some of the problems. Hidden sections between stations can be used where the train can rest out of site for a while before emerging in to view again to arrive at the destination – but how do you best achieve that transition on a prototype railway that has no tunnels or other scenery to ‘hide the train’?

Fast clocks’ can be used effectively to scale time, so the operator only has a few minutes in real world time to stay on track with a model’s timetable stretching over many hours. Personally experience shows this works quite well if the model runs smoothly, but any sort of problems like a sticky loco or derailment can suddenly introduce a lot of stress! You could argue that is accurately modelling real-life I guess, but running a model railway is supposed to be an enjoyable pastime, not an ulcer-inducer!

In any case no matter what method is chosen, the first thing to do is to get some better sense of how fast our models are running – to tame the inner child’s desire to go faster and faster!

The use of electronics in model railways goes back almost as far as the use of electricity. As soon as we stopped having to wind up clockwork mechanisms and instead started electrifying the rails, a whole new world of possibilities opened up.

Well over a decade ago, the club purchased a relatively advanced piece electronic gadgetry (for the time) that could accurately measure the scale speed of a moving model train by timing how long it took for the train to block a couple of light sensors embedded in the track. It was supplied in kit form with a pre-programmed micro-controller (PIC 16F886) and had to be assembled with a soldering iron.

DIP switches allow it to work with different scales and choose whether the display should be in miles per hour or kilometres per hour. Although it worked, it didn’t really work very well. The light sensors depended on ambient light only and proved to be very susceptible to changing light levels by something as innocuous as someone walking past nearby!

It was recently reinstalled on our new HO layout after sitting unused in a box for a few years since the old layout was dismantled and found to no longer work at all. Unfortunately, the original instructions are long gone and the lack of anything readily identifiable meant we no longer had any idea where we bought it from to try and find out anything about it! Faced with the prospect of difficult reverse engineering we figured there had to be a better way.

The explosion of the Internet in the years since has had many dramatic effects. Fortunately, not just Nigerian princes desperately trying to give away billions of dollars, but also a wealth of knowledge and resources about how to do just about anything – together with the ability to buy a huge range of parts to build stuff from all around the world easier than it used to be to get to your local hobby store.

One of our members (Garry) took it upon himself to find a better solution – which he did. He found a retired electronics engineer in the Netherlands who extensively documents on his web site everything he does in his numerous hobbies – one of which was a project called “A € 6 model train speed measurement device” [€ 6 is approximately A$ 10.]:‑e-6-model-train-speed-measurement-device-part-1-the-hardware/

Garry bought all the parts (from AliExpress as Rudy recommended), assembled them as described in detail, and programmed the Arduino board with the code Rudy made freely available. Before long the club had two new speed sensors – one permanently installed on the new HO layout and the other one assembled into a portable ‘tunnel’ that can be temporarily placed over a straight length of track to measure train speed on any layout, in any scale.

Not only is the display far more informative, the light sensors are more reliable, assembly was much simpler and total cost was far less.

There shall be no excuses for running a freight train on our layouts at 200+ km/h any longer!

Leave a Reply

Your email address will not be published. Required fields are marked *