miniature railway operations

[stantheman]

Hi everyone. I have been talking to a Canadian friend recently about club track signalling, and the various methods of control.
One idea that has given me food for thought followed his idea of using a PLC device as an interface between track switches/detection and the colour light signal. Just wondered if there were any other clubs here in the UK that have gone along this way. many years ago I constructed a system using miniature relays, obviously a saving in wiring would perhaps come about using a PLC, not sure of the applications issues though. Any thoughts would be appreciated.
Stan.

[jgb7573]

Hi Stan,

Sorry I can't help you with views on PLC, but our club uses miniature relays to control lights, semaphores and point locking. The biggest maintenance items are the track switches which are expeosed to rain sleet snow and steam oil, not to mention the occasional size 12 boot and the wrong kind of leaves. We occasionallly get relay problems (maybe once every couple of years) but the system basically works once it's been fired up. We are beginning to see some persistent problems with two signals which we think is down to some dodgy cable. So we're planning on replacing that. But it's done over 10 years so not bad really.

Ickenham is not far from Ashford if you want to come and have a look.
Good luck,
John

[mutley]

Whats a PLC device?

[Myford Matt]

Programmable Logic Controller

[stevep]

In my fairly limited experience, the biggest problem with track signaling is detecting the presence of the train.

When I set up the signalling system at the Basingstoke club, we inserted isolated sections that the wheels of the loco (and carriages) bridged. However, it was steel track, and when there was a good layer of rust on the track, the contact was not reliable.

After I left the club, I heard that a relay-based system as installed, and there was some talk of "shocks" caused by induction currents in the relays (?)

The original system at Andover club relied on a microswicth mounted under the track, which was allowed to flex at that point. The passage of any heavy load would operate the switch.

Cheers.....Steve

[JohnP]

When we resignalled the Bromsgrove track a few years ago, I did give some thought to using a microcontroller based system, or possibly a PLC, on the basis that a PLC has all the interfacing built in. In the end, we opted for a relay based system. The basic arrangement is that each signal has its own relay box, containing 3 relays. Each signal can be either a 3 aspect colour light or a home and distant semaphore on one post, and the three relays are as follows - one energises when the track section approaching the signal (in rear of) is occupied, one is energised for the home, and one for the distant. Additional contacts cover the colour light function.
The system is approach operated and fully automatic, but the track is elevated, so no pointwork.
Some things we found out..
12V is better than 24V for track circuit purposes - you can get enough leakage to simulate a train being in the section with 24V, but we never had problems with the previous 12V system.
Using a "distributed" system cuts down on the number of wires running around the system.

From experience, I prefer track circuit based systems to treadle based ones - partly this is because we can be an undisciplined crowd, and I have seen systems where someone overshoots a red signal, operates a treadle to register his prescence in the next section, then backs up so he can see the signal. The system is waiting for his train to clear the next section, so stays at red, and our driver gets stuck, effectively wating for himself! If you do use a treadle or similar system you need to think carefully about this kind of effect.

A description of the signals is on www.bromsgrovesme.myby.co.uk , although the circuits aren't! If you are interested, send me a PM.

Reagrds,

JohnP

[houstonceng]

The signalling system at Welling and District (W&DMES) is currently based on a microprocessor and solid state drivers feeding three aspect signalling with options for semaphores and reverse running.

Train detection is by insulated "pads" “scarfed” into the side of the all-steel welded (elevated) track. Not only are the drivers (electronic not human) getting obsolescent, but the bulbs in the signals are as well. The wiring is, also, getting a bit past it’s “sell-by date”.

Every now and again, one of the drivers is “deaded” when a bulb shorts or someone “works on a signal” with the power on. We're now in danger of running out of spares. Consequently, redesign is being discussed.

One of the newer, young “wiz-kids” (nick-named “Micron”) is suggesting a bus based system with three wires around the track and PLC in each signal head. I’m a bit of an “old fashioned” Electronic Engineer and prefer to use a more rugged system that more than one club member can diagnose & repair. As we used to say when I designed military communications equipment – “it has to be repairable in the field with a bit of wet string and an AVO”. I'm also a little worried about reliability since I own a car fitted with "CAN-Bis" and that is often causing odd "electrocal faults". Consequently, I prefer relays in this environment.

[mutley]

A bus system is fine until the bit of wire gets cut buy the strimmer or set on fire by the first person without an ashpan. When either happens I guess the whole system goes down. Is the designer of this new system also prepared to spend the time maintaining it/developing it as he will probably end up being the only person who understands how it works.
I know of a club that uses a system similar to this on a ground level track that suffered a lightning stike. That took out half the signal heads. They now have an earth stake in the ground with the track connected to it. Similarly the two guys who designed it now spend more time tweaking/maintaining it than using it.
Im with houstonceng keep it simple enough that anyone can maintain it and rugged anough to take the abuse the elements will throw at it.

[Boadicea]

Some good points here. JohnP makes excellent points that appear to be based on experience. I agree the only foolproof system is a track circuit system with electrically isolated sections of track – like the big boys do. I think anyone who has done this deserves great credit and probably has the best system. Any other system has drawbacks as JohnP describes when someone overruns a red light etc.
However, with multi-gauge track and points, track circuit systems get complicated. Other systems tend to be simpler and cheaper. As houstonceng suggests, it is easy to overcomplicate things, especially if you have one “expert”.
Why not accept a simple system operated by, say, microswitches attached to a small length of track and acting like a hand-on system, accept its shortcomings and plan for when the odd things happen. Remember we are not running at 125mph here. A hand-on system running with a train entering section, go to red, release the red when going into the next section will work fine under normal conditions – work out what to do when you have a SPAD. When it happens (very occasionally), the driver knows when he has done it and can initiate the process.
I would design it with the minimum of hardware out in the field and bring each cable back to a cosy central point. This way you will have minimal weather problems and minimal problems caused by vandalism etc. Most cables are extremely resilient to water ingress and if you minimise the number of plugs/sockets (preferably zero) out in the field, this can be an extremely reliable system.
Back at the central point, with inputs from switches and outputs to lights, you have a very flexible system. Interface all of these via relays (ref. houstonceng) which are robust and you will be relatively immune from any external nasties. You could start with a manual system, upgrade to a solid-state electronic logic system or to a computer system. It would allow itself to be upgraded centrally and be much more flexible and future-proof than a hand-on, series cable system. You could even try out different detectors and indicators out in the field at will. Keep it simple and rugged (houstonceng) is good advice. Any problems with a sliced cable etc. can be overridden at the central point without affecting other parts of the network and a sign to proceed with caution erected.
There you are – in my humble opinion of course!

[Tony K]

You might like to look at what our mates did at Oxford, building and installing the automatic signalling system for the Cuttleslowe Park Miniature Railway between 2002 and 2006, just click on Cutteslowe Park .....
www.cosme.org.uk/

[Myford Matt]

Nice work at Oxford.

Keeping it simple, robust and easy to maintain must be the best way to go.

[stantheman]

Hi, at last.

Sorry not to have responded to any of the previous messages, must admit since coming home from Canada everything seems to be running at twice the speed, except my own energy levels.

I did at one time use treadle switches mounted against the main running rail, operated by the wheel flanges. A locking system was used with the relays so that no matter how many wheels went over it only took one to register. I suppose in all honesty relays are perhaps better because most people can look and see if one is not clicking, whip it out and replace it if that is what is needed. One of my main issues at the moment is trying to remember how I wired each of the relays so that the next section unlocked the previous, I do have a sheet of paper with the details on somewhere, who knows!!

Yes it would be nice to see other ideas working and as John says, Ickenham is not far from home, perhaps will pay a visit, will warn beforehand.

I will keep thinking and hopefully before long will decide which course of action to take, thanks for all the ideas.

Stan.