Train-tech signal automation

[nhdesigns]

I've got some of the train-tech 4 aspect signal heads that I intend to fit onto a scratchbuilt signal gantry and I just need some help in a way of controlling them. Basically the two signals will be covering both a fast line and a slower line the merges back into one single down track line via a point. I'm looking at having both signals operated by means of block detecion or similar, that would change the from a clear or single or double caution signal to danger and then return the signal back to clear following the reverse sequence of single yellow etc.

 

Basically I want to have both signals change to danger when a train passes on either the slow exit line or fast line. And possibly to change the fast line signal to danger when the point is set for the slow line/exit and then change to danger when a train passes it. 

 

I don't know if that would work in real life, let alone in model form. But would anyone be able to help me out please?

 

What would be the best signal controller is best for automatic signal control?

 

How many signal controllers would I need for both signals?

 

I should also mention that my layout is:

 

DCC using Hornby Elite and Railmaster

 

My layout is a contious run with hidden sididngs

 

One line diverging into a slow and fast line

 

Two lines merging/joining back into a single line

 

I'd really appreciate any help and advice since this is something new to me.

 

I can provide photos if it will help

[Michael Hodgson]

In real life the signal protecting a converging junction must be at Danger before the point can be changed to the other route - it isn't changed to that aspect as a side-effect of operation of the point.  Rather the point is locked whenever a signal is clear for any conflicting movement.

 

The signal would however go back to Danger when the track ahead of it becomes occupied.  Signals at junctions are usually not automatic - that is, they normally remain at red until the signalman actively clears them for a train from one line or the other.   This is to avoid delay, because if a signal has to be restored to Danger in order to set up a route for the other line, it is common to have to wait for a timer to elapse (maybe 2 to 4 minutes) in case a train might be approaching that signal, as it would need to be given time either to run clear or to make an emergency stop short of the junction.  And the rules only allow you to put signals back in front of a train in the event of an emergency - doing so to give the route to some other train can be guaranteed to generate a justified complaint from the driver.

 

I am over-simplifying, but it is usually only on plain track that signals automatically step up from R to Y then YY then G after the passage of a train.

[nhdesigns]

17 hours ago, Michael Hodgson said:

In real life the signal protecting a converging junction must be at Danger before the point can be changed to the other route - it isn't changed to that aspect as a side-effect of operation of the point.  Rather the point is locked whenever a signal is clear for any conflicting movement.

 

The signal would however go back to Danger when the track ahead of it becomes occupied.  Signals at junctions are usually not automatic - that is, they normally remain at red until the signalman actively clears them for a train from one line or the other.   This is to avoid delay, because if a signal has to be restored to Danger in order to set up a route for the other line, it is common to have to wait for a timer to elapse (maybe 2 to 4 minutes) in case a train might be approaching that signal, as it would need to be given time either to run clear or to make an emergency stop short of the junction.  And the rules only allow you to put signals back in front of a train in the event of an emergency - doing so to give the route to some other train can be guaranteed to generate a justified complaint from the driver.

 

I am over-simplifying, but it is usually only on plain track that signals automatically step up from R to Y then YY then G after the passage of a train.

 

That does make more sense to me in a way. I'm rather new to using four aspect signaling and aspects such as route control. I've only really used two aspect signals in the past that would change to danger or green, depending on which line the point is set to.

 

I should think it would be okay to have one block detector close to the signal controlling the fast line and possibly another one controlling the signal for the slow line. Unless that wouldn't be effective or possible. As I say I'm new to all of this and don't want to be pressing buttons numerous times to change a signal when a train passes, since it would've already passed by the time it's changed to danger or similar! :-)

[Michael Hodgson]

To clear the signal, you only want to set it to "a proceed aspect" (= single yellow or better), which should be allowed as long as all relevant points are in the right position and the section immediately ahead is unoccupied.  Whether it should be showing YY or G is something your wiring or control logic ought to be able work out for itself, based on the aspect the next signal is showing, and that signal in turn is looking at the section in advance using the same criteria, all the way down the line.

 

There are quite a lot of signal driver modules (printed circuit boards) on the market that are intended to control this for you.  Some of them will do this linking one signal to the next.   Some other firms produce modules that are simple timers that cycle through 4 possible aspects.

 

One thing many modellers do is have a signal at the scenic break before re-entering the fiddle yard, the train being deemed to be proceeding unchecked on its way to the rest of the world - these signals might actually be driven by a timer - so after it changes to red behind a train, a suitable period elapses and it goes back to Y, and steps up again after a further period.  

[nhdesigns]

20 hours ago, Michael Hodgson said:

To clear the signal, you only want to set it to "a proceed aspect" (= single yellow or better), which should be allowed as long as all relevant points are in the right position and the section immediately ahead is unoccupied.  Whether it should be showing YY or G is something your wiring or control logic ought to be able work out for itself, based on the aspect the next signal is showing, and that signal in turn is looking at the section in advance using the same criteria, all the way down the line.

 

There are quite a lot of signal driver modules (printed circuit boards) on the market that are intended to control this for you.  Some of them will do this linking one signal to the next.   Some other firms produce modules that are simple timers that cycle through 4 possible aspects.

 

One thing many modellers do is have a signal at the scenic break before re-entering the fiddle yard, the train being deemed to be proceeding unchecked on its way to the rest of the world - these signals might actually be driven by a timer - so after it changes to red behind a train, a suitable period elapses and it goes back to Y, and steps up again after a further period.  

 

That makes sense to me as I only have a modest size scenic section and don't have enough space to have multiple signals and signal blocks realistically. I'm only having one signal with a feather for the up line prior to it splitting into a slow and fast line and showing the route set.

 

I'm also having just two four aspect signals for my down slow and fast lines before it merges back into a singe down line, which will go onto the hidden storage loops.

 

I'd be best going for something with a sensor that will set the signal(s) to danger and then go through a timed sequence.

 

Do you have any possible suggestions for what would be best to use? Also would I need a seperate sort of sensor for switching a feather?

 

I've seen a few around from BLOCKsignalling, heathcote electronics, and mimic points but just seem confused by them.

[MarshLane]

With signalling there really are two options, totally prototypical or 'good enough'.  If you want to be totally prototypical, then there is a lot involved, as Michael has explained.  Yes the option is way over the top for a model railway, but some of us have an desire/ambition/want to replicate things as much as we can. In the case of signalling, that involves route setting, route locking, full interlocking, accurate signalling etc...

 

For the 'good enough' option, it can be anything from a two-aspect or semaphore signal on an 'on-off' switch through to any part of the above.  The advantage with model railways is it is all personal preference, its your railway, so it works how you want it to.  Some will say not having accurate is wrong, but what is the point of getting it all spot on if it confuses you and/or makes the layout complicated to operate - the enjoyment has gone, and at the end of the day, enjoyment is what we are all about.

 

Reading your posts, it seems like you want the signals to 'appear' like they are working correct, without necessarily needing to go to all the hassle in the background.

 

Something like the MAS Sequence from Heathcote Electronics would be an easy option - see http://www.heathcote-electronics.co.uk/mas_sequencer.html - they are about £17.90 - or £20 if you want a route indicator to be controlled as well - according to the website.

 

In essence this connects to a three or four aspect signal, but has no connections anywhere else, minimal sensors, no track circuiting etc.  Basically the board mounts under your baseboard, and there are two holes through the baseboard to an infra-red transmitted/receiver, which is positioned between the track, probably just beyond (5-10cms) the signal.

 

When the train passes over the infra-red beam, the sequencer electronics put the signal to red.  Once a set time has passed, it changes to yellow, then after the same time period to double yellow and again after the same period to green.  If at anytime during that period the train passes over the sensor again, it simply reverts back to red and starts again.

 

That is the simplest way to achieve what you want.  The board has the ability to connect a switch to the RR terminal (a simple on-off or SPDT switch should be fine). When on, it defaults the signal to red.  Switch it off, and the sequencer goes back to controlling the aspect of the signal.

 

If you have two or more signals on the same line in the scenic area, you can put a second MAS sequencer under the second one, and with a single wire feedback information, so that the two work together.  When the train passes the second unit and it goes to red, it will trigger the first unit to go to yellow.  When the timer on the second unit expires and the signal goes to yellow, the first will go double yellow etc..  But there is no need to do that if you just have one signal on each line.

 

Its not the only option, but as your just getting into things, and the simplicity of your track layout - you should be able to achieve it with that.  You could also use a couple of those units automatically to set the signals for you on the 'down line' depending on how you want it to work.

 

Hope that helps.

 

Richie

[dave55uk]

On 01/11/2021 at 00:11, MarshLane said:

you can put a second MAS sequencer under the second one, and with a single wire feedback information, so that the two work together.  When the train passes the second unit and it goes to red, it will trigger the first unit to go to yellow.  When the timer on the second unit expires and the signal goes to yellow, the first will go double yellow etc..

Although better than nothing, this bugs me.

 

As soon as the loco passes over the IR sensor, the signal goes red (that's ok) and the signal behind it goes to yellow (NOT ok).

IRL the second signal would not change to yellow until the whole train had passed the signal (at a minimum) or had cleared the associated track circuits.

[MarshLane]

8 hours ago, dave55uk said:

Although better than nothing, this bugs me.

 

As soon as the loco passes over the IR sensor, the signal goes red (that's ok) and the signal behind it goes to yellow (NOT ok).

IRL the second signal would not change to yellow until the whole train had passed the signal (at a minimum) or had cleared the associated track circuits.

 

On the real real railway, yes you are right. This comes back down to the exactly like the prototype, and 'good enough'.  To get the exact operation requires additional work, some form of track circuiting (detection, reed switches or more infra-red) and more boards.

 

The OP said:

On 31/10/2021 at 13:45, nhdesigns said:

I only have a modest size scenic section and don't have enough space to have multiple signals and signal blocks realistically.

 

So I was working on the principal of keeping it simple and trying to achieved what he wanted in a 'good enough' setting.

 

Edited November 3, 2021 by MarshLane

[nhdesigns]

On 01/11/2021 at 00:11, MarshLane said:

With signalling there really are two options, totally prototypical or 'good enough'.  If you want to be totally prototypical, then there is a lot involved, as Michael has explained.  Yes the option is way over the top for a model railway, but some of us have an desire/ambition/want to replicate things as much as we can. In the case of signalling, that involves route setting, route locking, full interlocking, accurate signalling etc...

 

For the 'good enough' option, it can be anything from a two-aspect or semaphore signal on an 'on-off' switch through to any part of the above.  The advantage with model railways is it is all personal preference, its your railway, so it works how you want it to.  Some will say not having accurate is wrong, but what is the point of getting it all spot on if it confuses you and/or makes the layout complicated to operate - the enjoyment has gone, and at the end of the day, enjoyment is what we are all about.

 

Reading your posts, it seems like you want the signals to 'appear' like they are working correct, without necessarily needing to go to all the hassle in the background.

 

Something like the MAS Sequence from Heathcote Electronics would be an easy option - see http://www.heathcote-electronics.co.uk/mas_sequencer.html - they are about £17.90 - or £20 if you want a route indicator to be controlled as well - according to the website.

 

In essence this connects to a three or four aspect signal, but has no connections anywhere else, minimal sensors, no track circuiting etc.  Basically the board mounts under your baseboard, and there are two holes through the baseboard to an infra-red transmitted/receiver, which is positioned between the track, probably just beyond (5-10cms) the signal.

 

When the train passes over the infra-red beam, the sequencer electronics put the signal to red.  Once a set time has passed, it changes to yellow, then after the same time period to double yellow and again after the same period to green.  If at anytime during that period the train passes over the sensor again, it simply reverts back to red and starts again.

 

That is the simplest way to achieve what you want.  The board has the ability to connect a switch to the RR terminal (a simple on-off or SPDT switch should be fine). When on, it defaults the signal to red.  Switch it off, and the sequencer goes back to controlling the aspect of the signal.

 

If you have two or more signals on the same line in the scenic area, you can put a second MAS sequencer under the second one, and with a single wire feedback information, so that the two work together.  When the train passes the second unit and it goes to red, it will trigger the first unit to go to yellow.  When the timer on the second unit expires and the signal goes to yellow, the first will go double yellow etc..  But there is no need to do that if you just have one signal on each line.

 

Its not the only option, but as your just getting into things, and the simplicity of your track layout - you should be able to achieve it with that.  You could also use a couple of those units automatically to set the signals for you on the 'down line' depending on how you want it to work.

 

Hope that helps.

 

Richie

 

Hi Richie;

 

That is a real big help indeed to me. I've had a good look at what Heathcote Electronics offer and what you suggested fits the bill. I've asked them for their advice as well as train-tech to see if they can be of help. Train-tech told me that the LEDs are common negative fitted to a PCB with solder pads for wires. I was able to clarify all of this with Heathcote electronics who said these signals will work since there's no complex fittings like the plug and play and auto signals. I've also been told that the boards have resistors fitted so there's no need for adding any as suggested in the intstructions, unless I would choose to lower the brightness by using the resistors supplied with the signal kits.

 

Given the posistions of the signals and the lines they control, I would think the IRDASC-4RI -2EW would be suitable for use on the single four aspect signal with a feather as I intend to have it work inline with the setting of the points from the main, up fast line, to the diverging, up slow, line. I've looked at it and it would be more likely to suit the setting of this signal to red whether a train uses the up slow or up fast.

 

I would think that the MAS Sequencer(s) would be as you said better to use for the signals being on a gantry or perhaps the IRDASC 4RI since the signals are going to be controlling the up fast and up slow prior to them converging  and would set the signals accordingly to which way the point is set.

 

 

[MarshLane]

On 05/11/2021 at 20:47, nhdesigns said:

That is a real big help indeed to me.

No problem, your welcome.

 

On 05/11/2021 at 20:47, nhdesigns said:

Given the posistions of the signals and the lines they control, I would think the IRDASC-4RI -2EW would be suitable for use on the single four aspect signal with a feather as I intend to have it work inline with the setting of the points from the main, up fast line, to the diverging, up slow, line. I've looked at it and it would be more likely to suit the setting of this signal to red whether a train uses the up slow or up fast.

 

I would think that the MAS Sequencer(s) would be as you said better to use for the signals being on a gantry or perhaps the IRDASC 4RI since the signals are going to be controlling the up fast and up slow prior to them converging  and would set the signals accordingly to which way the point is set.

 

No problem with doing things that way at all - just remember that the IRDASCs depend on other signals for their aspect.  So if you have just one single four aspect in that direction, there are no other signals to tell it what to do, so the MAS Sequencer would be the best bet if that is the case.  If there are other signals, then yes the IRDASCs are fine, and the MAS can control the signal aspect of the last signal the train passes, the others will then respond to what that does.

 

Richie

[Sir TophamHatt]

Take a look at:

 

 

 

This is automatic signal control - remember the trains will not obey the signal.

I spent around £50 odd on a whole bunch of the control boards (job lot from ebay, three mas sequencers, the rest "dummy" units - they still need the mas to tell them what to do).  They were from Heathcote Electronics and work pretty good in tandem.  Would recommend them as a simple way to add automatic signal operation but just take care on the sensors.  Either my room doesn't work or they're not quite sited right but sometimes the trains don't trigger the signals properly.

My layout isn't big enough for many but the next layout is likely to be bigger so have the same system for future needs.  They're a little expensive but then I don't know how much it would cost to make them (thinking time as well).

 

However, I didn't get anywhere when looking at adding a feathered version of the signals with a point.  Like you, I'd want the feather to light with a green light every time the point is thrown a certain way, and just red at all other times.  Red/green is easy as it's independent of the rest of the layout but the feather...  can't seem to work out how to attach it to the PM-2 seep point motors I have (although I think there's a way).  Does anyone know if turtle motors have a soldering section for signals?  I think the switch I have to throw the point (an off-on-off momentary switch) won't help here either?

I may have to look at introducing an accessory decoder that can switch both rather than relying on wires, although the latter is preferred.