The Signalling Project I

The CPMR actually consists of two quite separate railways, and the signalling requirements of each are different. The raised track, being basically an endless single line that can have up to five trains running on it at any time, lends itself to a fully automated block system of control. The ground level track, on the other hand, with its various loops and sidings, calls for a semi-automatic system controlled from a signal cabin. Each track has only a very basic signalling system at present, and of course there is no signal cabin as such. For many years now, there has been talk of an upgrade, but not much has happened on the ground. We have now decided that the only way to make it happen, is to get on and do it!

The problem with the ground level track is that we don't, at this stage, know what the final layout will be. Any work carried out now could turn out to be a waste of time and effort, if, for example, the station is re-located, the steaming bays are moved or the line is extended out into the park. Therefore, any changes to the ground level signalling will be quite minimal at this stage, and will be dictated purely by safety requirements.

The raised track, on the other hand, is not likely to alter much, if at all. We have therefore decided to make a start by re-signalling the raised track, and the experience gained from this will stand us in good stead when the time comes to tackle the ground level system. The main considerations for the system were drawn up as follows:

1) It must be fully automatic, and not require the presence of a signalman.

2) It must use simple logic, such that someone with minimal knowledge of the system can trace faults using a circuit diagram.

3) It must use proprietary equipment wherever possible, to minimise the construction time and to ensure future interchangeability of parts.

4) It must be weatherproof and vandal resistant, and able to remain in situ throughout the summer season.

5) It must not require any special setting-up before a running session.

6) It must require minimal maintenance.

With these objectives in mind, we decided at the outset to use 3-aspect colour light signals, and to control these by relays, rather than using any form of microprocessor control. This last decision probably reflects the backgrounds of the signalling group, but we do feel that those who inherit the railway from us will have a better chance of understanding how it works! However, the method of train detection posed a bigger problem. Track circuiting was considered, but was dismissed as being potentially unreliable. There are of course many miniature railways that successfully employ track circuiting, but these mostly use aluminium rail, whereas ours is steel. We did not feel that a good contact could be maintained with a rusty rail, particularly so for the 3½" rail which sees infrequent use. Add to this the difficulty of making insulated joints, the likelihood of current leakage, and the fact that "Loctite" is used widely for securing wheels to axles these days, and it will be seen that track circuiting is probably a non-starter.

We therefore looked at other methods of train detection, and considered optical, inductive and electro-mechanical systems. Optical devices, such as PIR detectors, were eliminated because these could be treiggered by a bird, a squirrel or even a falling leaf. Inductive proximity switches were given serious consideration, but we soon found that the detection range of most of those available was inadequate. Those that did have an adequate range were prohibitively expensive. We are therefore planning to use treadle switches, as used by several other clubs, including Cardiff. A suitable ready-made unit (actually an industrial process control switch) has been identified.

Having decided how the system will work, we set out to produce a signal plan and schematic wiring diagram. A voltage of 12V DC was decided on, as this will give the best availability of components. Suitable locations for the signals were identified and marked out using pegs driven into the ground. The track was divided up into nine sections, seven of these being long and of roughly equal length, with the remaining two sections being in the station area. With a maximum of five trains in operation, there will always be four empty sections and so a log-jam should not occur. Each section will display a green light if the next section ahead is unoccupied, a yellow light if the next section ahead is occupied, and a red light if the section itself is occupied, as per full size practice. There will be times, of course, when a whole journey is made on "yellows", but this is no different to the full size railway! Operation of the traverser or swing bridge will cause a red light to be displayed, with a yellow in the preceding section. Each section will be able to hold only one train, except for the departure platform, which will display only a yellow light when occupied, enabling trains to "close up" ready for loading. The schematic diagram has evolved to cover all of the "What if" situations we can think of, but it probably isn't finalised yet!
[This diagram was published in 2002. A more up to date version is available here. To view this page requires the Acrobat Reader which is available free from Adobe.]

As for the actual signal heads, three possibilities presented themselves:

1) Fabricate our own, to the same design as the unit at the rear of the ground level carriage shed. In fact, five more of these currently exist in an unfinished state, but we need a minimum of nine heads just for the raised track. The five may eventually be finished and find use on the ground level circuit.

2) Cast more of the aluminium units used elsewhere on the system. We do not have a pattern for these, but an actual head could be "borrowed" and modified to make its own pattern. However, we have found it difficult to get suitable lenses for these heads.

3) Purchase ready-made proprietary units. One of our members collected a leaflet from Parkside Electronics at the Harrogate Exhibition, and as this obviously merited further investigation, a 3-aspect signal was obtained from them on loan. A trial was arranged at Cutteslowe Park, and we were sufficiently impressed to order a batch of twelve of these units. They comprise a moulded polycarbonate body with 2.2W bulbs, reflectors and prismatic lenses, and are incredibly bright. Design of the signal posts is now taking place.

The whole system will work automatically, with no action required by anyone, except for one scenario when an entire train is taken off the system and put into the carriage shed. When this happens, the signals will remain at danger, as there will be no train to restore the system as it continues on its way. To overcome this minor hiccup, we propose to site a reset button just inside the door of the carriage shed.

This is an exciting project that will not only enhance the safety of the raised track, but will give a "real railway" feel to driving, and will add to the interest for both drivers and public alike. The signalling group presently comprises Chris Kelland, Tony Vereker and myself. There is obviously a lot of work to be done and the more people who can help, the sooner we will see the fruits of our labours. Come and join us on any Saturday afternoon.