Let’s see what is needed, assuming there are two locos on one track, with only one power controller. Here is what I came up with.

1. We need to isolate behind loco 1, in case loco 2 is active there.
2. We need to isolate in front of loco 1, in case loco 2 is active there.This agrees with Freezer’s approach, assuming the two gaps are near a signal.

Where do we need to apply power, when the signal tells us the loco can go?

1. We must power between the track breaks (the signal section), in case the loco is there or needs to pass through there.
2. We must power the section to the rear of the signal section, in case the loco is there, waiting for the signal to change.
3. We must power the section in front of the signal section, because that is where the loco will go.

In Figure 1, only the signal section is switched. What is needed in practice can be seen in Figure 2. The key difference to the first version is that I use a relay with three poles (in practice, there is much more choice, in both price and size, of four-pole relays.) One pole powers the signal section, one powers the section to the rear and one powers the section in front. It would also be possible to use four-pole switches: three poles for the track and one to activate the signal. Figure 2 All sections switched

Figure 3 Signal to signal movement

This arrangement comes into its own when you have two successive signals, such as a Home signal on the approach to a station and a Starter signal for departing the station. Figure 3 shows how this works. The loco at the Home signal is isolated, even when the Starter turns to Go. When it does, the loco in the station can still move forward, even if it halted before the Starter’s section break. This is important because we don’t want to find the platform loco can’t depart, just because neither signal can supply it with electricity. Equally, when the loco arrived, that central section of track needed to have power via the Home signal, even though the Starter is against it. In other words, the track between the two signal sections must be electrically-feedable from either end. Hence, there is one pole potentially feeding it from the Home and another potentially feeding it from the Starter.

Figure 4 Shunting a coach to/from the train loco

This is where it becomes more fun to operate: other manoeuvres are now possible. Figure 4 shows a simple example of shunting stock, to couple to a stationary train engine. The reverse is clearly possible too: the detachment of one or more coaches from the tail.

Figure 5 Two controllers

Depending on your track plan, the approach extends to having multiple controllers. If we add a point in that central section (Figure 5), then a branch line train can feed into the station using its own controller. The main line Home and the Starter will be set to danger, so there will be no power in that central section from the main line controller. There could even be a train safely waiting on the main line Home signal. The branch train is arriving within station limits, so the driver is still within their cleared route. As soon as the point is once more set to the main line, there will be no longer be a feed into the central section from the branch line controller. In practice this all feels very natural. After all, if the branch train is to continue on the main, then it will need a driver cleared for the main. In other words, it must be driven onwards from the red main line controller, not the blue branch controller.

The one clear difference between what can be done and what happens on the prototype can also be seen by looking again at Figure 3. If the Starter is against, then the incoming loco is only powered from the rear of the station, by the welcoming Home signal. If you follow prototype practice and set the Home to danger as soon as the train has passed, then the loco will stop. You need to wait until the loco has reached the intended position and then protect it from the rear, by setting the Home against further arrival. With the relatively short distances that many of us have to use, this is scarcely an issue. By the time the train has passed, it is close to the Starter signal anyway.

For those of you with long runs, perhaps in a garden railway, then you might prefer having the power permanently connected on those long runs. This doesn’t matter, as long as the permanently connected sections of track are not used to hold a loco when others are intending to move using the same controller.

This can be taken further. If you have an Outer Home, Starter and Advanced Starter signals, the same electrical arrangements can be used. In conjunction with the signals this is very natural to use. However, if you just want to let the trains run then, you can set all signals to green. Just don’t tell the Railway Inspectorate.

If there are two operators it may be desirable that each controls the signals permitting a train to approach them, as well as using their own power control. Figure 6 shows the arrangement to achieve this. There are two power controllers, red and blue. Here we assume each signal controls the isolation its own short section but the power to that section and the subsequent one is fed from the controller in advance. (Though not prototypical, it is convenient if the operator of the blue controller has control of Signal 1. That way they have complete control of the power to the sections of track that they can run over. However, this is not essential.) Figure 6 Operator to operator

To see how this works, first notice that the red controller must be used to drive the incoming loco into the dead section at Signal 1, as Figure 6 indicates. When the blue operator is ready to take that loco, they set Signal 1 to permit the loco to proceed, as in Figure 7. This energises the section at Signal 1 and also the section between the two signals. The loco can now pass Signal 1, which can immediately be returned to its ‘stop’ setting. Figure 7 Figure 8

The loco now proceeds until it reaches the dead section centred on Signal 2, as in Figure 8. Setting signal 2 to permit the loco to pas energises its signal section and the loco proceeds. Once again the signal can be returned to its ‘stop’ setting as soon as the loco has cleared the track break. Figure 9

Finally, Signal 2 can be used to release the loco for its onward travel, as in Figure 9. This method readily extends to three or more sequential operators and has some of the merits of classical signalling. On larger layouts, suitably positioned operators will not need video cameras to know what is going on elsewhere.

As shown here, this is using only the signal sections; they are used as isolating sections when the signal is against. However, as the remarks at the end indicate, it is possible to combine this with multiple-switched sections wherever more than one loco may be in use on the same controller

If you wish for full prototypical operation then none of this is for you. You will expect your drivers to observe the signals and your signallers will need to communicate, perhaps by bell codes, with the signal box in the rear and the signal box in front.

If you operate in DCC, then the same comments apply.

If you work DC and wish to make it relatively safe for visitors to run on your rails, or you wish to have the operational benefits of having more than one loco on the track, then the examples given are readily adaptable to your own track plan. Whether to use multipole switches or multipole relays is perhaps your first decision. If you use an appropriate form of intermediate electronic control, you can ensure that signals and any associated points are always set in safe combinations. This becomes a form of route setting, presented to the user as using the signal to send the train onwards, with electrical sections and points being automatically set for the intended route. In practice, this is very easy, even for the casual user.

The above examples also indicate that, in some instances, it may not be necessary to have three switched sections associated with each signal. There is, within this approach, an easy way of deciding. Suppose we classify the section connections around a signal as P, for permanently powered, or S for switched power controlled by a signal. So a signal which switches the section to the rear, the signal section and the section in advance would be classed (S,S,S). A signal in which the rear section was permanently powered and the signal and advance section was switched would be (P,S,S).

Now suppose we have a station with an outer home, a home, a platform, a starter and an advanced starter. With this method the two interior signals, the home and the starter, would need to be (S,S,S). The home signal, being the first met by an incoming loco, could be (P,S,S). The advanced starter, being the last met, could be (S,S,P). This permits a loco to stop anywhere alongside the platform, without necessarily having to drive to the starter’s dead section. However, if the loco is indeed isolated at the starter signal, a second loco can be brought in behind, as described earlier.

The general rule is make the first home signal (P,S,S) and the last starter signal (S,S,P). Any signals in-between are (S,S,S). So a station with just a starter, or perhaps a home and a starter, will not need (S,S,S). This ensures a loco can always be driven to the first home signal, even if it is set to stop. It also ensures that the last starter can be returned to stop, once a loco has gone beyond it. In plain English, switch all sections within station limits.

These days you should be able to find most of what you want via the web. The only caution is to check that the article is appropriate to your railway, country and era. The general principles of semaphore signalling in the UK were relatively stable throughout the 20th century. Here are a few books about prototype railway signalling.

Model Railway Signalling C J Freezer (Patrick Stephens Ltd 1991, reprinted 1992) This was primarily written to help the modellers of its time. It has advice on how to create your own signals from scratch, whether semaphore or lights. The relevant section is quite a short one, towards the end.

Signalling in the Age of Steam, Michael A. Vanns (Ian Allan, 1995) This gives a good overview of the details, such as the different kinds of block instruments, the interiors of signal boxes, the evolution of signalling and the block system. If you plan to operate your railway in the fashion of the prototype or you enjoy the history, then it has much to offer you.

The First Principles of Railway Signalling, C.B. Byles (London Railway Gazette, 2nd edition 1918) The author was widely experienced on railways and it was intended as a book of tuition. It is accordingly very clear about the “why” as well as the “what”. The pencil drawings are a delight in themselves. Mine is an original in hard covers and was a lucky find in a second-hand bookshop. There are also facsimiles, a 2010 paperback and a 2019 hard cover, which you may be able to get.