Three-Section signals for DC control


3sig5.jpg Philip Willis
Pictures by the author

An earlier version of this article appeared in the MERG Bulletin. I am pleased to acknowledge their help with that version, some of which will be reflected in this version

I HAD LONG planned to have working semaphore signals on my layout. I built it without signals but once it was working satisfactorily, I started to think of how to go about adding them. I had used MERG CBUS from the start, to control points and section relays, so adding a few servo-controlled signals was not a problem; but I wanted the signals to be more than decorative elements.

What I describe below includes the heart of the system, which can be implemented without needing CBUS or any other digital control. A version with relays and switches only will still give the operational benefits of three-section signals. Adding digital control of some kind can give the additional benefit of automatically setting the road ahead of the changed signal to be safe, meaning the operator is freed from the responsibilities of real signallers. Simply put, it gives a form of automatic interlocking which is nevertheless hidden from the operator. I won’t describe that here but will stick to the basics.

I like simplicity of operation, so I didn’t want a full simulation of a signal box. As I have West Wilts GOG visitors who like to put their trains on the track and run them for themselves, ease of use matters for them too.

For safe operation I wanted the signals to control the track power, as a minimum to prevent a loco running past a signal that should have halted it. This is known as ‘Signal Passed at Danger’ (SPAD) protection but, as I will describe, more can be achieved.

The Three-section Signal

First, a few words on terminology. In the prototype, a signal is ‘on’ when it is set to danger and ‘off’ when it is set to proceed. However, as we control model trains through power to the track, it too can be on or off. To avoid confusion I will use the terms ‘Stop’ and ‘Go’ for signals, with apologies to the many modellers who prefer prototype signalling terminology. I will also use the word ‘section’ to mean a length of track which can be electrically isolated from other sections. These do not play quite the same role as prototypical ‘blocks’. I will sometimes refer to a ‘signal section’, meaning a short section adjacent to a signal. The signal section is electrically isolated when the semaphore is set to Stop and energized with the signal set to Go. I will use the term ‘intermediate section’ for a section between two signal sections, powered from either signal.

The basis of all my signal wiring is the arrangement in Figure 1 which, in addition to the signal, shows track sections and the relay contacts that power them. The signal sits next to an electrically isolated length of track, B, which is powered only when the signal is set to Go. This is a signal section. There is another section A on the approach to the signal and a third section C beyond the signal. These too are each powered from different contacts on the same relay (or switch) when the signal is at Go.

Figure 1 – Basic signal section

Figure 2 – Home and starter signals

Using a relay to isolate a section of track near a signal is not a new idea. Cyril Freezer describes it in one of his books [C. J. Freezer, Model Railway Signalling, Patrick Stephens Ltd, 1991, reprinted 1992]. What he describes is a one-section signal: only power to B in Figure 1 is controlled by the signal; the two adjacent sections A and C are always live to the controller. He was though very clear that, although turning the signal section power off prevented SPADs, the real benefit was isolating a loco to let the same controller move other locos. As will be seen, greater operating flexibility will come from extending this one-section concept to three sections of track, A, B and C in Figure 1. But an immediately obvious problem is that sections A and C must be additionally controlled from somewhere other than this one signal, or nothing could move, in front of or behind a loco held at the stop signal.

Suppose we have a Home signal and a Starter signal, Figure 2, one on the approach to the station and one at the departure end respectively. As shown, we can arrange for intermediate section C to be powered when either or both signals are set to Go. Then, for example, when the Starter shows Stop but Home shows Go, we can bring a train in from the left, drive it through A and B and stop it in anywhere in C before the Starter. Now suppose we protect the rear by setting the Home to Stop but release the train by setting the Starter to Go. The train can move into D and onward to E, from wherever it stopped in C.

If these are the only one or two signals on the track, we can dispense with the two outermost relay contacts controlling A and E and wire those sections directly to the controller. We can bring a train in from the left as far as the Home. If the Home is set to Stop, the signal’s section prevents a SPAD. If the Home is set to Go, the train can proceed as far as the Starter. If the Starter is set to Stop, its signal section prevents a SPAD. So far, so useful but routine.

Now however, something else is possible. If we deliberately park the loco in the Starter’s signal section D, with the Starter at Stop, the loco is isolated for as long as that Stop remains. In turn this means we can bring a second loco in from the rear through A and B, on the same controller, and use it within C to remove or add stock to the train in front on D. We are beginning to see some operational benefits, beyond simple SPAD control.

Using Two Controllers

Figure 3 – Branch line signalling

Let’s look at a slightly more complicated example, to see why that intermediate section, C in Figure 2, offers more than we have seen so far. Figure 3 represents a branch line feeding into a main line. We have two controllers, red and blue, needing access to the shared intermediate section, but not at the same time.

The diagram illustrates how the blue controller can feed power to the intermediate section of the main line. That section will be adjacent to the platform. This permits the branch loco to reach the Starter, while the main line loco is isolated at its Home. Once it has waited for its passengers to board or disembark, the branch train can move forward on the main line when the Starter is set to Go. Alternatively, if there is sufficient room on the main track, a train can be held on the mainline by the Starter. If the loco is isolated in the Starter’s signal section, with a couple of coaches behind it, then a branch line loco can be brought behind using the blue controller, to remove the coaches and take them onto the branch.

The method readily extends to more controllers, each powering its own part of the track. The photograph shows the main panel for my four-controller layout, with green, yellow, red and blue controllers. The red/green LEDs show the state of a signal. The yellow LEDs show the route selected by a point. At this level it really does help to have some intermediate digital system doing the route setting in response to your changing a signal, especially a programmable one because you will find circumstances which you hadn’t anticipated.

I use CBUS to ensure mutually exclusive access to the shared intermediate sections. Setting either approach signal to Go should set the other to Stop. Their respective relays also need to be set or cleared in the same fashion. It is still possible to set both signals to Stop and open all relays to remove power from all tracks. It is also arranged so that all signals can be set to go around the two main circuits, for those days when you just want the trains to run. Other ways of achieving safe operation may be possible, such as a diode matrix, or indeed careful operators!

Four power controllers

.

Further examples of using three-section signals to control track power may be found in the Guild Knowledge Base Wiki, in the Signalling section.