The reversing loop: how it works and how to avoid a short circuit
Benjamin | TrackDrawer.com · published 29 October 2026
A reversing loop is a piece of track that leaves a turnout, swings round in a large curve and comes back to the same turnout, so your train carries on in the opposite direction without any shunting. On two-rail DC that loop causes a short circuit, because at the end of the loop the left rail runs into the right rail. The fix: isolate a section of the loop at both ends in both rails, make that section longer than your longest train, and switch its polarity with a toggle switch or an automatic reversing module.
How a reversing loop works
Picture a single-track line that ends in a turnout. Your train runs into the turnout, follows the loop through roughly a full circle and comes back into the same turnout from the other side. It then heads back down the line it came from, with the locomotive still at the front. That is the whole point: no running round, no reversing the train, it just keeps going.
You find reversing loops at the end of a line, in a hidden fiddle yard under the layout, or as the two ends of a point-to-point line. A dogbone looks like two reversing loops, but electrically it is different: more on that below. A wye (three turnouts in a triangle), and any connection that joins a track to itself in the opposite direction, behaves electrically like a reversing loop as well.
Why two-rail shorts out
On two-rail DC (Trix C track, Roco, Piko, Fleischmann, Minitrix, Kato and most other systems) each rail carries one pole. Call the rail on the left in the direction of travel positive and the one on the right negative. Now follow the left rail round the loop. Because the train leaves the loop facing the other way, that rail ends up joined to the rail that was on the right at the start. Positive meets negative: short circuit. Your controller or command station cuts out as soon as a wheelset bridges the junction, or the moment you close the loop if there is no gap in it.
The same thing happens on a wye and with two crossovers that join the inner circuit of a double-track oval to the outer circuit the wrong way round. Once you understand the principle, you can spot the trap on the track plan before you lay any track.
The fix in three steps
- Isolate the loop. Fit insulated rail joiners in both rails at both ends of the reversing section. Gapping one rail is not enough: the other rail will still create the short. The section then has no electrical connection to the rest of the layout and gets its own feed.
- Make the isolated section longer than your longest train. As long as the train stands across a gap, its wheels or pickups bridge it. Lit coaches, a second locomotive or a pickup shoe join exactly the two sections you want to keep apart. Only when the whole train is inside the loop may the polarity flip.
- Switch the polarity. On analogue DC you use a double-pole changeover switch: once the train is completely inside the loop, you reverse the polarity of the rest of the layout and change the direction on the controller with it. The loop keeps its polarity, so the train runs on smoothly, and when it leaves the main line matches again. On digital (DCC, for example) it is simpler: a reversing module feeds the isolated section, detects the moment a wheel bridges the gap and flips the section in a fraction of a second. The decoder in the locomotive does not care about polarity, so the train simply keeps going.
Mind the turnout too: the train has to find it set correctly when it enters and when it leaves from the same side. A spring-loaded turnout, or a switching track that throws the turnout once the train has passed, makes automatic running possible.
Why Märklin three-rail has no problem
On Märklin C, K and M track the traction current runs through the centre conductor (the studs down the middle of the track). Both outer rails together form the return and are electrically joined. Follow the left rail round the loop and it ends up on the right rail, but that one is at the same potential: nothing happens. On three-rail you lay a reversing loop without gaps and without a module.
Two caveats. Trix C track has the same shape as Märklin C track, but it is two-rail: there you do need the gaps. And Märklin Z (mini-club) is also two-rail DC, despite the brand name. If you use the outer rails for occupancy detection with contact tracks on three-rail, the traction current stays safe, but plan the detectors so that a train in the loop does not accidentally link two blocks.
How much room a reversing loop needs
The loop itself is a curve of a little over 180 degrees, so it is at least as wide as the outside of a circle: twice the radius plus the width of the track bed. Add the turnout and its approach, and a reversing loop is always longer than it is wide. The table gives the width for the smallest radius of a few systems in our catalogue, with and without a 5 cm margin at each edge of the baseboard.
| System | Smallest radius | Track bed width | Loop width (2 x R + bed) | With 2 x 5 cm margin |
|---|---|---|---|---|
| Märklin C track R1 | 360 mm | 40 mm | 760 mm | 86 cm |
| Roco geoLine R2 | 358 mm | 38.5 mm | 754.5 mm | 85.5 cm |
| Piko A track R1 | 360 mm | 43 mm | 763 mm | 86.5 cm |
| Minitrix R1 | 194.6 mm | 16 mm | 405.2 mm | 50.5 cm |
| Kato Unitrack N R216 | 216 mm | 25 mm | 457 mm | 55.5 cm |
| Märklin Z R1 | 145 mm | 16 mm | 306 mm | 40.5 cm |
That is the minimum. A larger radius runs more smoothly and takes longer coaches, and you usually want some straight track before and after the turnout as well. With generous curves an H0 reversing loop is soon a metre wide, which is why it often sits at the end of a baseboard or disappears into a fiddle yard below the scenery.
Dogbone, figure eight and reversing loop: the difference
A dogbone has two turning curves, but the train always circulates in the same sense: electrically it is an ordinary continuous run without a short. Only when you join the two parallel tracks with a crossover do you create a reversing connection, and then you have to isolate again. A figure eight with a crossing or a bridge changes the train's turning direction every lap, but electrically it is a plain loop too. The test is always the same: does the left rail end up on the right rail anywhere?
Reversing loops in TrackDrawer
The Auto-designer lays reversing loops itself: the "Line with two reversing loops" variant is a single-track line with a loop at each end, built from the turnouts, matching curves and curved track in your stock. In every plan for a two-rail system TrackDrawer checks the polarity: if the left rail meets the right rail anywhere, you get a warning naming the two pieces between which you have to gap both rails, with the advice to add a reversing module. On three-rail that warning does not appear, because there is nothing to isolate.
Want to see how a reversing loop fits on your baseboard? Enter your track and let the Auto-designer lay out a plan, or browse examples in the Railfolio first.
