Interlocking Turnouts and Signals

This article first appeared in the May/June and July/August 1990 issues of AMRA's 'Journal'.
By Steve J Chapman.
Go to Part Two

Part one - why and what.

It is very rare to find a prototype railway that does not use signals to control the movement of trains. With Model Railways signals appear much less often. There is a very good reason for this.

On the prototype the signals are there for the purpose of controlling train movements. Very basically, the line is broken up into a number of sections called blocks. One or more signals are placed to control the entry of trains into each block so as to reduce the chance of two trains attempting to occupy the same section of track (and hence colliding) to the lowest possible level.

With a Model Railway we also have the track broken up into blocks and allow only one train into each block. There is however a major difference between the way entry of trains into blocks is controlled on a Model Railway. Usually we can see where each train on a Model Railway is. When a Model Railway is large enough or arranged in such a way that we cannot see all of the trains we can still tell whether a block is clear to receive our train because it is necessary to link that block to the controller that is driving our train before the train can enter the block. If the block is occupied then usually it will be connected to another controller and this will be visible on the control panel.

Only when we have a block (or a number of blocks) which are not visible from the control position and on which it is desired to hold trains and not just run them do we need to consider some form of signalling. Even then the indicator usually consists of a block occupancy light on the control panel rather than a physical signal on the layout itself.

The need for signals which exists on the prototype is totally absent on a Model Railway. This explains why so many unsignalled layouts exist.


The reason for including signals on a Model Railway is not the same as the reason for their use on the prototype. Signals appear on Model Railways because the person building a Model Railway wants to create a miniature copy of the real thing. Depending on the degree of accuracy that is desired (with regard to signalling) a number of alternatives are available.

If physical accuracy is all that is required then non-operating signals set in the appropriate locations will be all that is required. Signals are fixed in the clear position (or possibly the warning position for distant signals) so as to avoid having trains pass a signal that is at danger. The signals are then effectively ignored when the layout is operated.

An appearance of operational accuracy can be obtained if signals which can be operated from the control panel are installed. When a train passes a signal it can be changed to show the danger aspect and once the block is clear again the signal can be returned to a clear indication ready for the next train. The problem with this is that the trains are not dependent on the signals and the layout can still be ignored when the layout is run. Also it is possible to give clear indications on signals when a clear road is not set.

The prototype uses interlocking between turnouts, signals, etc to ensure that signals can only be cleared when the road ahead has been set and to ensure that the set road is not changed while the signal is clear. This function is performed by the signal box or (as the Americans call it) interlocking tower. If this interlocking is duplicated on the model then greater realism can be obtained.


Before we can apply interlocking to our layouts we first of all need to determine what to interlock and in what way. In order to understand what sorts of interlocking are required let's first look at a number of simplified situations involving a small number of turnouts and signals.

Let us first consider the situation with a turnout preceded by a signal (figure one). The signal can be cleared to allow trains to use either the main line or the branch. There doesn't seem to be any relationship between the signal and the turnout that requires interlocking. In fact there is a need for interlocking when this situation occurs on the prototype.

one turnout and one signal

Let's consider what can happen if there is no interlocking and assuming that trains obey the signal. A train may be passing the signal which is clear in order to travel down either the main line or the branch (it doesn't matter which). While the train is travelling over the turnout the signalman decides to change the so that the next train can take the other road. The result would be a derailment.

The signal therefore needs to be interlocked with the turnout in such a way that the turnout is locked in position whenever the signal is cleared. The signal must then be returned to danger before the turnout can be reset to the other road. In order that we know what we are referring to when we look at how we can achieve the various types of interlocking lets call this (for the purposes of this article) a type ONE interlock.


If we replace the signal with a junction signal (figure two) then the situation is very different. What form of interlocking would be required in order that the signal can only show a clear aspect for the road that is set?

one turnout and a junction signal

There are in fact quite a number of interlocks which should exist in this situation. The two parts of the junction signal can be treated as separate signals. Each of these signals needs to be interlocked with the turnout.

The main signal needs to be interlocked so that the signal can only be cleared when the line is set for the main. This ensures that a train can not be given a clear setting on the main line unless the main line is actually set (obviously). The turnout also needs to be interlocked with the main signal so that when the main line signal is clear that the turnout is locked in position.

The interlocking between the main line signal and the turnout therefore needs to be such that reversing either the signal (clearing it) or the turnout (setting it for the branch) should lock the one not reversed in the normal position. Lets call this a type TWO interlock.

The interlocking for the branch signal is slightly different because the branch signal should only be able to be cleared when the turnout is set for the branch (ie reversed) and the turnout should be locked in the reverse position when the branch signal is cleared. Lets call this a type THREE interlock.


The same interlocking applies if the signals are placed on the other side of the turnout as in figure three. The main line signal requires a type TWO interlock and the branch signal a type THREE.

one turnout and two signals


In figure four we have both sets of signals. The interlocking that is required between each signal and the turnout remains the same but we now have the additional requirement that the two main line signals be interlocked so that only one can show a clear aspect at a time and similarly for the two branch signals. This interlocking is in fact the same type of interlocking as exists between the main line signals and the turnout (type TWO).

one turnout and all three signals


Let us now consider a double track junction as in figure five. The signals need to be interlocked with their appropriate turnout in the same way that we determined above. The signals for the two directions do not need to be interlocked because they now apply to different tracks. However we now require the two turnouts to be interlocked so that both paths through the diamond crossing cannot both be set at the same time. We need interlocking so that having turnout A set for the main line locks turnout B also set for the main line and having turnout B set for the branch locks turnout A also set for the branch. This interlocking is identical to that required between each turnout and its branch line signal (type THREE).

double junction


The signalling for a passing station on a single track is shown in figure six. We will assume for the moment that the train will always take the left hand road through the station. Again as for the double junction we have four signals and two turnouts. Once again the signals need to be interlocked with the turnouts in the same way as we have already discussed. The signals labelled MA and MB need to be interlocked with the corresponding turnout (A or B) in the same way as the main line signals were interlocked for the junctions (type TWO). Similarly signals BA and BB need interlocking with turnouts A and B (respectively) in the same way as the interlocking of the branch signals at the junction (type THREE).

passing loop at a station
Graham Plowman has provided further info on interlocking this configuration.

To obtain the corresponding situation for right hand running just hold the diagram up to a mirror. Everything else remains the same.


Figure six introduced the interlocking of signals which will be passed one after the other by a train travelling down the line but which are all close enough together that they all belong to the one station (or block). The interlocking described between signals MA and BB will apply to any two signals which will be passed one after the other by a train except for the situation leading up to a junction. Let us consider this now (figure seven).

signalling in advance of a junction

The interlocking between the junction signal (C) and the turnout has already been determined when we looked at figure two. So what type of interlocking is required if the junction is preceded by a warning (distant) signal (A).

Signal A will depend on the aspects displayed by both the main and branch signals at the junction. Only if one or the other of these two signals shows a clear aspect can signal A also be cleared. Conversely when signal A is cleared then whichever of the two parts of the junction signal shows a clear aspect must be locked in that position. This is a more complicated situation than those we have discussed previously but is still achievable. Lets call this interlock a type FOUR.


The example situations given will probably make up at least part of the track work of your layout. Wherever these situations occur the appropriate types of interlocking can be applied. The rest of the layout can then be dealt with using similar principles - examine the situation to see what combinations of settings could result in an accident (assuming that all signals are obeyed). The appropriate types of interlocking can then be applied.

The individual simple situations for which interlocking has been determined can then be combined together and any additional interlocking between the sections can then be worked out. In this way the entire prototype style of interlocking requirements for the layout can be worked out (except for the facing point locking bars and only a signalling enthusiast would bother including these).

To make things easier to keep track of when working out the interlocking we can use a diagram similar to the ones that the prototype use. For each signalbox they draw up a track plan showing all of the turnouts and signals. The signals are considered to normally be at danger. To record the normal setting for turnouts the two lines are drawn so that they do not quite touch and only the normal road is shown as a continuous path. The various turnouts and signals are then numbered and can be referred to by these numbers when determining the interlocking. Each point tiebar and each signal arm should be numbered separately unless there are two which are always intended to operate together from the same lever or switch.

Once the entire interlocking requirements have been worked out you can then determine just how much of the interlocking that you intend to actually apply to the layout (since it is only a model some or even all of the interlocking can be left out without causing a major disaster). I suggest that if you intend to incorporate any interlocking at all that you work out the whole thing first before deciding how much you actually wish to apply - you might find that with a little extra effort that quite a lot more of the interlocking can be easily included and improve the realism of your layout quite substantially.

go to top

Part two - how.

In the first part of this article we have discussed why turnouts and signals are interlocked and have looked into what interlocking is required between individual turnouts and signals. Now we need to look into how to apply the interlocking on an actual layout.

In the situations that we looked at we found a need for four different types of interlock that we needed in order to meet our requirements. All of the interlocking that is needed should be able to be built up by combining these four types. So all that we need to do now is to determine a mechanism whereby each of these different interlocks can be achieved and a method for combining all of these interlocks together so as to interlock the whole layout (or at least as much of it as you decide to do).


There are two basic ways of providing interlocking. There is the physical interlock via a mechanical link which duplicates in model form the interlocking system used with the old mechanical signal boxes and there are various electrical methods which duplicate the functionality of the more modern power signal boxes. The actual method used may not duplicate exactly the way that the signal box carries out the interlocking but the effective results are the same.

Let us begin by considering the way in which mechanical interlocking might be applied. This form of interlocking can be applied when using any of the various mechanical means of operating the signals and turnouts on your layout.

If we assume that we are using a miniature lever frame to work the signals and turnouts then instead of connecting the wire or rod directly to the lever we attach the lever to one end of a bar and the wire or rod to the other end of the bar. You don't necessarily need a lever frame as there are other ways of providing a means to operate the system but the miniature lever frame duplicates the signal box more closely so let's stay with it for the moment. The exact method of operation doesn't affect the operation provided that all of the controls are centrally located. On the prototype these controls are centrally located in the signal box.

So we have our lever frame with a bar attached to the base of each lever. Each of these bars is held in a frame so that it can only move forwards and backwards and does not have any sideways or up and down movement. Another set of bars sit in the frame underneath the bars which are attached to the frame. These bars are held so that they are free to move from side to side only. (see figure eight).

an interlocking frame

Notches are cut into the sides of the top set of bars and pieces are attached to the top of the lower bars to slide into these notches. The pair of notches and the raised section shown in figure eight make up a type three interlock.

All of the bars are shown in the "normal" position. The interlock shown operates as follows - bar A is locked in place by the projection which is fitted into the notch, if bar B is reversed then the projection is free to slide into the notch on bar B which has now been brought into alignment, reversing bar A will force the projection sideways into bar B thus locking bar B in the reversed position until bar A has been returned to normal.

The other interlocks can be set up in a similar way.


To make it easier to discuss how to set up each of the different interlocks we can simplify the diagram and use a two-dimensional representation of the mechanical interlocking. Each of these two dimensional views can be physically created using the method that we have already discussed.

The other methods of interlocking are electrical and here the circuits fall readily into two categories. Basic electrical circuits using relays and/or switches attached to point-motors is one alternative.

Another alternative is using electronic components to build logic circuits or to use a computer to provide the interlocking. This alternative will not be dealt with at this time.


Let us consider the mechanical and electrical means of providing each of the four different types of interlock that we have decided (in part one) that we require.

In figure nine A we have a two dimensional representation of what we earlier called a type one interlock. Lever A is free to move as long as lever B is in the "normal" position. The turnout that is operated by this lever will then be locked in place when lever two is reversed. This locking takes place regardless of whether lever A is in the normal or reversed position.

type one interlocking
type one interlocking

Figure nine B shows an equivalent electrical circuit using single solenoid (or ordinary) relays. A relay being off corresponds to a lever being "normal" and on corresponds to reversed. These relays may then be mechanically or electrically linked (via additional contacts) to the various parts of the layout to which the interlocking belongs.

The way in which the relay circuit works is that the set of contacts on relay A is used to stop relay B from switching relay A off if relay A is on. If relay A is off then the first set of contacts on relay B (the left most) will lock relay A off if relay A is already off. The other set of contacts on relay B will lock relay A on by connecting it directly to the supply provided of course that it is already on.

Figure nine C shows a corresponding circuit using double solenoid relays (a point motor is basically a double solenoid relay). In this case the circuit is simpler because there is no need for the part of circuit that holds a relay on. The set of contacts on relay B cut the power to relay A when the reverse position is selected thus locking it in place.

What to do if you have one part of the circuit as a single solenoid relay and the other as a point motor? Simply take the parts of the circuit from one diagram that relate to the relay and the parts of the circuit from the other that relate to the point motor. The circuits relating to the two halves of the circuit are shown without any connection between the two so that you will be able to do this without too much difficulty.


Figure ten A shows the mechanical linkage for a type two interlock. Either lever is free to move provided that the other lever is in the normal position. As soon as one lever is reversed the other lever is locked in the normal position.

type two interlocking
type two interlocking

The electrical equivalent circuits for single and double solenoid relays are shown in figures ten B and ten C respectively. In this case the wiring for the two types of relays is almost identical. The set of contacts on each relay is arranged to lock the other relay in the off position (normal) when that one is on (reversed).


The mechanical linkage for a type three interlock is shown in figure eleven A. Lever A is locked in the normal position by the locking bar until lever B is reversed. The locking bar is then free to move into the notch in lever B thus enabling lever A to be reversed. Once this has been done the locking bar then holds lever B in the reverse position until after lever A has been returned to normal.

type three interlocking
type three interlocking

The electrical interlocking is fairly straightforward and is identical to the type two interlock in so far as the wiring of relay A is concerned. The differences to the wiring for relay B can be seen in figures eleven B (for single solenoid relays) and figure eleven C (for double solenoid relays or point motors).


The situation with the type four interlock is more complicated than the three types that we have discussed so far. The main reason for this is that we now have three things being interlocked together. Our type four interlock is concerned with the way that the first two (A and B) are interlocked with the third (C). Additional interlocking (probably a type two interlock) will be required between A and B. This additional interlocking will not be shown in the diagram.

Figure twelve A shows the mechanical linkage. In this case we have three separate locking bars which are connected together with a bar which is pivoted on the central locking bar (which locks lever C) . This pivot bar has slots in which a projection from the top of each of the other two bars may slide as the locking bars move to the left.

type four interlocking

With all of the levers in the normal position both the locking bar for lever A and the locking bar for lever B are forced to the right. The locking bar on lever C is also forced to the right via the pivoted connection to the other two bars and hence lever C is locked in the normal position. If either lever A or lever B is reversed then the corresponding locking bar becomes free to move into the deeper than normal slot. The locking bar on lever C is now also free to move (by a shorter distance) because of the way that it is connected to the other two bars. Lever C may now also be reversed which will lock whichever of the other two levers that was reversed in that position until lever C is returned to normal. Note that unless additional interlocking is supplied between levers A and B is supplied that the other of these that has not yet been reversed is still free to move and that if it is reversed then either but not both of levers A and B will then be free to be returned to normal without having to return lever C first.

Figure twelve B shows the same interlock using single solenoid relays. Relay C requires that either relay A or B be on before it can be turned on. When relay C is on then whichever of relays A and B are on will be locked on. The circuit differs from the mechanical linkage because in this case both A and B will be locked on. As specified before a separate interlock is required between relays A and B.

The interlock can also be done using double solenoid relays as in figure twelve C. In this case relays A and B will be locked in the reverse position (on) by having the solenoid that returns them to normal cut off when relay C is reversed. Similarly the relay C solenoid for reverse is cut off if both relay A and B are normal. The effect on relays A and B is the same as with single solenoid relays.


Mechanical interlocking can be easily built up to include all of the different interlocks that you require on your model railway by setting up a large enough frame with enough locking bars to provide the various locking that is required. To most readily determine the layout of such bars and where the corresponding notches need to go the entire interlocking for the layout should be drawn up on paper first. The simplified arrangements that I have used in my descriptions of how the interlocking works can be readily combined together and shuffled around on paper until the preferred arrangement involving a minimum number of bars is obtained. This interlocking diagram can then be kept as a record of what interlocking exists between the various turnouts and signals on the layout. The bars can be numbered from the track plan that you drew earlier (in part one) to identify which bars operate which turnouts and signals.

Electrical interlocking can be worked out by combining the circuits for each of the interlocks required. A circuit diagram can then be used to record the information that is required for wiring and fault finding purposes. In this situation it may still be an advantage to draw up an interlocking diagram the same as would be required for a mechanical system of interlocking.


Interlocking is not hard to do when you break it down into individual interlock requirements and then build it back up to include the whole layout. Adding interlocking to a layout adds a whole new dimension to the operational interest of the layout and gives those who like building layouts something else to design and build. Why don't you think about interlocking for your layout?

go to top
Copyright Stephen Chapman