Linear Motor Tracks

martink Jan 9, 2022

  1. martink

    martink TrainBoard Member

    I would like to recruit a small group of experienced modellers who would be interested in trying out my linear motor track system. I only build a new layout every year or two, but it would be great to see what more people could achieve. Using a linear motor track system offers some unique possibilities for complete layouts in the tiny scales, and for animations and scenic features in the normal ones.

    The basic skill requirements are decent levels of ability in general modelling and soldering. If you can build and wire a conventional layout, and build a simple electronic kit, then that should be enough.

    You may have seen some of the posts or videos of my linear motor powered T Gauge models. Up until now, intellectual property issues and my agreement with the patent owner have meant that I could only make and use this track for my own projects (with a very limited exception a couple of years ago). As of the start of 2022, the situation has changed and these restrictions have disappeared, although there are still some regulatory issues that would obstruct selling commercially.

    The general plan would be to have interested modellers spend a month or so sketching out ideas, exchanging suggestions, playing with track plans, etc., and only at the end of that phase decide whether or not to commit and go forward. I would then have a suitable batch of track manufactured for the combined set of projects, while also sourcing the components for the controller kits. Payment would not be required until after this point when all the bits and pieces were ready to ship. Ongoing help and support would be provided via this forum, and hopefully each of you would be willing to document your progress in a suitable forum topic (warts and all).

    This linear motor system was originally designed for T Gauge, but can handle trains from 1:720 (two thirds of T) up to Zn30, and even industrial narrow gauge in N. For roads, the range is 1:480 up to cars only in N. Canal boats, trams, etc. are also viable.

    The main advantages of a linear motor compared to conventional propulsion are: very reliable running, no track or wheel cleaning, excellent low speeds, long trains, steep gradients, and easy automation. The main drawbacks are: no wheels (models slide along the track), flat (paper) track surface, slightly jerky or wobbly motion, and over-wide double-track spacing.

    Each piece of track is a small printed circuit board, which get soldered together into a complete layout, hence the requirement for decent soldering skills. Despite of their unusual nature, these track pieces form a complete sectional track system with the usual pros and cons of such a system. To keep things simple, no turnouts will be offered at this time. I would also prefer to limit things to rail only, but if enough people are interested then the 2-lane road pieces could be included as well. The road and rail tracks are compatible, so you can build a road layout using rail track, as long as the wider curves and single lane are acceptable.

    The absence of turnouts means that your track plans will have to be either basic end-to-end or basic continuous run (preferably more than just a simple oval!). The controller can power up to 8m of track, and has built-in automation that is more than adequate for sequences like an end-to-end shuttle or twice-around-and-stop-at-the-station, all with smooth acceleration and deceleration. You can build your own trains or I can include some of my low-resolution 3D printed models (unpainted versions of those seen in the videos).

    Prices in USD are expected to be $70 for the controller kit, $35/meter for track, $7/vehicle for trains or cars, $20 shipping from Australia, using PayPal, with a planned maximum of 10 participants.

    If you are interested, I would suggest starting by watching (or re-watching!) the videos from one or more of my T Gauge layout topics here, followed by a selection of the other videos from the YouTube channel to get a better idea of the full range of possibilities.

    To experiment with track plans, you can use any common track planning software (such as AnyRail) starting with several lengths of T gauge flex track to create the various pieces of sectional track, then copy and paste as needed. The rail straights are 96mm, 48mm and 36mm long; curves are 122mm radius at 45 and 22.5 degrees and 140mm radius at 45 degrees. Double track spacing is 18mm. If the ends don't quite join up, you can fudge it as long as the ends are perfectly parallel and within 5mm or so. For roads, start with HO or S flex track, and straights are 108mm and 36mm, with curves of 54mm radius at 90 and 45 degrees and 244mm radius at 11.25 degrees. There is also a T piece which is equivalent to a 108mm straight overlaid with two 54mm 90 degree curves. Road designs must line up and join properly - fudging is not possible. You can mix and match road and rail, and the 2-lane road track uses the same 18mm spacing as rail double track.

    This technology has quite a few subtle differences from conventional model rail, so please feel free to ask questions. Then let us see what happens.

    SLSF Freak, Mike VE2TRV and gmorider like this.
  2. MegaBlackJoe

    MegaBlackJoe TrainBoard Member

    Could you please clarify why do turnouts create problems? How does decoupling work? Does the track itself makes this weird noise or the controller? Is it possible to put some plastic layer between the engine and the track? For example to 3d print plastic ties and rails? Thank you.
  3. martink

    martink TrainBoard Member


    All of the vehicles are self-propelled, and what appear to be couplings are actually just alignment pins to reduce sway and wobble, particularly for shorter UK models. Coupling and uncoupling involves parking a train across a section break, then turning the track power or one side on or off.

    For a more familiar analogy, think of it all as a conventional DC layout with one controller plus on/off sections, where every loco and carriage is powered with speed-matched mechanisms, and all the Kaydee coupler springs have fallen out. The capabilities and limitations are very similar.


    The combination of track and train forms an electric motor with a lot of very loose rattly parts. The track is the stator and the trains are the rotors. The longer the train, the more pieces there are to buzz and rattle. The controller uses something similar to PWM to get fine speed control and low speed running, and these frequencies fall in the audible range. As a train accelerates, the frequency keeps changing but for practical reasons it has to be kept within a certain range, so it sounds like notching up and changing gears.

    The trains naturally move in 1mm steps, based on the coil spacing. This works fine at high speeds, and if the controller is set up this way the trains are almost silent. However, low speed running is terrible, so I use a modified PWM technique to reduce the effective step size to 0.25mm. This gives much better running, but is noisy, very much like low-frequency PWM with conventional trains.

    Plastic Track:

    I have tried this using 2-color 3D printing to make proper (but thin) track with ties/sleepers and raised rails, and it does look quite good. The train magnets are arranged to protrude slightly below the model, and slide along between the rails. Unfortunately, the magnetic field strength falls off very quickly with any sort of gap between the track and magnets, so even 0.5mm is problematic. Further, while it works well enough on straight track, the rigid wheelbase (magnetbase?) and the need to hunt slightly left or right so that magnets retain perfect alignment with the coils means the track gauge on the curves has to be widened very noticeably. Alas, not viable.

    Any sort of smooth running surface improves things compared to running on the raw track, which is very much like a cobblestone road. A better appearance, smoother running and less noise. I have tried a variety of materials, such as thin styrene sheet painted or printed on the underside. In the end, the best I have found is self-adhesive label paper from an office supply store, with suitable track or road designs printed with an inkjet or laser printer, then coated with several layers of matte spray varnish. The canal used the same paper, but painted and with gloss varnish.


    One of the main complications with any linear motor system is the electrics. The goal is to have a certain current flowing through the coils. Too low, and the trains come off the track; too high, and there will be power consumption and overheating problems. For this application, 0.3-0.4A is about right. For a simple layout, that all just boils down to spacing the power feeds correctly: 0.8m-1.2m (8-12 pieces of track) apart for 12V. Nice and easy.

    Turnouts consist of two very short sections of track, only one of which can be powered at a time. In addition, there will usually be one or more stop sections (sidings, loops, etc) associated with each turnout, and these rarely line up with the normal power feed spacing. This means that several of these short sections need to be chained together into correct length power sections. Then some of these sections need to be dynamically turned on and off to switch the turnout and stop sections, but the 0.8-1.2m rule still has to apply at all times. Dummy load resistors often have to be added to make things add up. To further complicate things, the curved track on the turnouts is nearly 1mm below the straight one, significantly further away from the train magnets and so has to be driven at nearly double the normal current to compensate. Designing the wiring and controls for even a simple switching layout becomes something of a logic puzzle.

    A simple power section needs a single 3-wire cable. Each subsection needs two cables, an input and an output to connect to the next part of the section. Each turnout needs three 3-wire cables on the surrounding tracks to interrupt the normal sections, then three more to actually control the turnout sections. All cables are 3-wire, and that unfortunately means two DPDT relays per switching element. So, lots of cables and relays. And connectors. My new Penzance layout has 9 turnouts, 19 other sections, 2 improved controllers, 12 relay boards with 6 relays each, and something like 65 3-wire cables underneath the board. Plus custom software to control it all.

    I use a highly modular control system to keep things manageable, with standard relay boards that plug into and daisy chain from the controller. All 3-wire cables have connectors so the wiring is really just a big plug-board.

    The standard controller using standard software can handle up to 6 toggle switch inputs for points and sections, and drive up to 3 relay boards, enough for a modest layout with a loop and a couple of sidings, say 2-4 turnouts. The automation features include all this, with the turnout and stop section operation included in the playback, so sequences involving passing two trains, switching a few cars on or off, etc. are easy enough to do.

    Modest layouts with a handful of turnouts really aren't as scary as all of this makes it sound, but the wiring design approach is very different to a conventional model railway, and I do not want to cope with them at this time. If I did so in future, it would probably start with a few standard designs with variants, etc.

    Hope this helps.
    MegaBlackJoe and Mike VE2TRV like this.

Share This Page