Automatic Signalling Using Relays

This article first appeared in the January/February 1989 issue of AMRA's 'Journal'.
By Steve Chapman.

Building a model Railway to me has always meant incorporating all of the things that I know about that would reasonably be found in the equivalent situation on the twelve inches to the foot version. One of the most obvious of these which is not essential to the operation of a model railway but which can add greatly to the realism of a layout is operating signals.

How can signals be incorporated into a layout in such a way as to operate in a realistic manner while requiring a minimum of extra effort on the part of the layout operator?

The answer to this question that I have found (among the many possible solutions) can best be explained by considering the situation one step at a time.

Step one.

Signals come in a number of different varieties; lower quadrant semaphore, upper quadrant semaphore, two aspect colour light, three aspect colour light, and four aspect colour light; to name several common types. The first consideration after deciding which of these types of signal that you are using is how to make the signal operate ie how to get the signal arm to move for semaphore signals, or how to get the right colour displayed for colour light.

The operation of any signal can be handled using a relay. In the case of a semaphore signal the signal arm can be mechanically linked to the relay so that when the relay is on the signal arm is pulled off and when the relay is off the signal arm returns to danger. A two aspect colour light can be wired using one set of contacts on a relay with the red light wired to the terminal which gets the power when the relay is off and the green light gets the power when the relay is on.

Semaphore signal relay circuit diagram

Three and four aspect colour lights are slightly more complicated because the aspect which should be lit up is dependent on the next one or two signals in advance of the approaching train. The way to handle these cases is to cross connect the signals. The first set of contacts still controls when the red light is on but instead of the green light being on when the relay is off, power is fed to a second set of contacts on the signal relay which is next in advance of the approaching train. When this next relay is on not only will its associated signal show red but (provided that the previous relay is not also on and therefore making its signal show red) the amber aspect of the previous signal is also lit via the second set of contacts. For three aspect signals the green light is wired to the off terminal of this second set of contacts on the next relay in advance while for four aspect this terminal must be wired through a third set of contacts on the relay next in advance of it with the on terminal on this set of contacts being wired to the double amber aspect (with the top amber hooked to the bottom via a diode so that the top one doesn't light up when a single amber is selected) and the off terminal gets wired to the green.

This much work has enabled us to get the signals operating but so far the signals all need to be operated via on/off switches and the trains can still operate without paying any attention to the signals.

Step two.

The easiest way to get the trains to pay attention to the signals is to force the train to stop whenever a signal at danger is reached.

If a short section of track next to the signal (long enough to hold all of the power pickups on your train) is isolated when the signal is at danger then the train will stop when it reaches this section without requiring any intervention by the operator beyond setting the signal.

To isolate a section of track like this requires only that the signal relay have available another set of contacts so that the power to this section of track can be wired to operate in conjunction with the signal.

Use of yet another set of contacts on the signal relay and a resistor can enable the preceding section of track to be wired so that the power to the section is reduced when the signal is at danger thus slowing the train down. To do this wire the preceding section of track the same as for the isolating section and connect the resistor across the on/off terminals of this set of contacts.

A more complicated approach (beyond the scope of this article) is to use an inertia type circuit as used in some controllers to slow the train down gradually when it enters the section preceding the signal.

The trains will now obey the signals but the signal still needs to be operated manually.

Three Aspect signal relay circuit diagram

Step three.

To get a train which has just passed a signal to automatically turn the previous signal to danger.

To automatically set the previous signal requires a modification to the part of the circuit that supplies the power to operate the signal relay. If we keep the switch that we had on the panel so as to allow us to override the signal and turn the signal to danger whenever we wish (useful for trains which are to stop at eg a station even though the section in advance is clear) then we need another relay for the automatic control (let's call this the automatic relay). If manual override is not required then the signal relay power supply can be rewired to operate as shown for the automatic relay in the diagram.

The automatic relay makes use of two sets of contacts (or one extra set on the signal relay if no manual override is required). One of these sets of contacts (the set not required if using only the signal relay) is wired into the power feed for the signal relay so that when the automatic relay is on the signal relay must be off (ie the signal is at danger). The signal is therefore changed to danger whenever the automatic relay is turned on (or the switch on the panel is open).

The second set of contacts is wired so as to supply power to the automatic relay so as to hold the relay on, this will enable the relay to be turned on by a single pulse of power.

A detector of some sort (eg a reed switch) is positioned near the track just passed the signal and set up so as to provide a pulse of power when the tail end of the train passes it (eg by having a magnet attached to the underside of the last coach or wagon). This pulse of power will momentarily turn the automatic relay on, the second set of contacts on the relay will take over supplying power to the relay after the pulse ends so as to hold the relay off, and the signal relay will therefore be switched off and will therefore change the signal to danger.

Step four.

To get the signal to clear again when the train has left the section.

With the circuit as described so far the signal will be set to danger and held this way forever. A way is required to tell the automatic relay that the train has left the section and that the signal no longer needs to show danger.

One way of doing this would be to use an additional set of contacts on the automatic relay wiring the relay power to the previous automatic relay so that when this automatic relay is switched on the previous automatic relay (together with the signal relay for this signal) is switched off. The problem with doing this is that the previous automatic relay is not only switched off, it is locked off so that, even if triggered to turn the signal to danger, no power can flow through the relay and the signal is not set. Using this method does not allow for two adjacent signals to both be set automatically to danger and therefore if realism in signal operation is to be maintained no train must be permitted to enter the section preceding a signal at danger and thus no train will be stopped by a signal. If a train is permitted to enter a section like this then the preceding signal will not be set and this train will not be guarded by a signal. This is not a satisfactory solution.

The solution to the problem is to provide a momentary break in the supply of power to the automatic relay in order to reset it but not locking it so that it can't be turned on.

One way to do this is to provide another track detector beside the first which is wired to the previous automatic relay. This track detector should be of a different type from the first one in that the first passes a momentary current when it detects the train, the second should momentarily interrupt the current when it detects the train.

An alternative is to only use the one detector but to use an extra relay (let's call it the reset relay). The power to the automatic relay is wired in such a way that the power is cut whenever the reset relay is turned on. The reset relay can be turned on whenever, the automatic relay of the next section in advance is turned on by the momentary pulse from the track detector.

The difference from using a set of contacts on the automatic relay however is that the reset relay only goes on momentarily and is not held on in the same way that the automatic relay is. Therefore unless two trains pass adjacent detectors at the same time (which shouldn't be able to happen since the rear train cannot pass its signal until the next section has been cleared by the train in advance) each signal is available for the track detector to set it to danger.

The circuit is now complete. As each train passes a signal it sets the signal to danger (and in the case of three and four aspect signals sets the preceding signals appropriately). A section of track next to the signal is isolated thus stopping another train from entering the same section and the track detection circuit for the previous section has been reset so that if a train enters that section the signal for that section responds appropriately. Any signal can be set to danger if required by overriding the automatic system using a switch on the panel thus allowing danger signals wherever appropriate when not required to protect trains.

Automatic relay circuit diagram

Other circuits are available which can do the same job as this circuit using solid state components (eg transistors, diodes etc) but for those railway modellers who are not electronics experts the functioning of this circuit should be much easier to follow. And for those railway modellers who can see a way to substitute solid state circuitry for part or all of the circuit I have described, congratulations, and perhaps you might be so kind as to send me a copy of your circuit modification.


One last point: I have actually built and tested the circuit as described above with the circuit also being interlocked with the various sets of points along the route (I will describe how to interlock in a later article). I used relays obtained cheaply on ex telephone circuit pc boards and carefully stripped the relays off the board. This method gave me a relatively large number of relays for a very low cost (about 40 cents each) and all I had to do was build a 24 volt power supply to run them off. So just because the circuit shown appears to require a large number of relays doesn't mean that the circuit will cost the earth to build and may in fact work out cheaper than other equivalent circuits.

go to top
Copyright Stephen Chapman