Signalling help on my simple 8x4

[RJS1977]

 

 

What is required is a means to permit moves from siding or yard to headshunt.   In the days of rotating ground signals one stacked above the other would have been employed to differentiate the route offered on the basis of top = left, centre = ahead, bottom = right.  Using colour light ground dolls it is possible to give a route indication by means of miniature feather (an angled white line displayed in a small square pane) or alpha-numerically but such things are too complex for a small layout.  We can assume that if the point to the headshunt is controlled by the signalbox, fully interlocked and set to that route then by definition there is no access to the main lines and any move may be authorised by the yard-master.

 

 

IIRC A position light with a yellow light rather than a red one (or a ground signal with a yellow line rather than a red one) means that the driver only has to stop at the signal (if at 'danger') if the points are set to a certain route (in this case out on to the main line). So provided the points are set to the headshunt, as much shunting can go on as necessary until the point is switched to the main line, then the train can't leave until the position light clears.

[Jaggzuk]

Some really useful comments and advise here Chris and as a fellow 8x4er I have followed this topic with interest.

 

The angle I have applied for the signal planning for my layout has been to try and not see the complete layout as a one single entity.  But rather to consider each part/area as from something bigger.  The use of the tunnels to give the illusion that trains are in fact going off scene to somewhere else helped.  As such I have not worry too much about signals been to close to one another.

 

I have also considered the placing of signals at locations where I think they will add to a certain view point, a place where I like to watch trains going by and I hope to take photos from in the future. Of course when viewing an 8x4 from above it will look over signalled or unrealistic, but close up is a totally different thing.

 

The big difficulty I found was that it is very hard, on a 8x4 multi loop layout, to decide which lines are uni-direction and therefore only to be signalled in one direction or which lines are bi-directional .  All my lines can be run in both directions, so deciding which way is normal was very difficult.

 

I shall look forward to your chosen plan.

 

Have you considered which brand of coloured light signal you might get?

[RJS1977]

Also - if you want the signals giving access to the tunnel to change to red as a train passes, and back to green after a given time interval, MERG do an inexpensive circuit kit for that, though you do need to be a member to purchase them. 

[28XX]

A very long time ago.

Wish I'd stayed. 20/20 hindsight a wonderful thing.

[Chris64B]

Hi again all,

 

Here's v3 of my diagram - nearly there hopefully!

• Signal added to control inner main line access to yard/parcels siding - I might be able to fit this one on the layout rather than have it off-scene.
   
• Signal added to control outer main line access to yard/parcels siding and branch line via the two feathers. This will be off-scene so not installed, only added to the diagram for completeness.
   
• Revised location for the position lights and one with yellow light instead of red for headshunt access. 'Stop and await instructions' board added.
   
• Trap point description corrected.
   
• I've also added some estimated locations for AWS inductors following some advice I've read on this site that they are usually c.200yds before a signal and not generally used on two aspect signals. Please shout if I've got this wrong!

 

 

......Have you considered which brand of coloured light signal you might get?

 

Yeah been looking into it - I'm thinking probably Traintronics and/or CR signals? They seem to be a good balance between detail and cost.

 

Also - if you want the signals giving access to the tunnel to change to red as a train passes, and back to green after a given time interval, MERG do an inexpensive circuit kit for that, though you do need to be a member to purchase them. 

 

Thanks, that'd be quite smart, will look into it!

 

Cheers,

Chris

[TheSignalEngineer]

Hi again all,

 

Here's v3 of my diagram - nearly there hopefully!

• Signal added to control inner main line access to yard/parcels siding - I might be able to fit this one on the layout rather than have it off-scene.

   

• Signal added to control outer main line access to yard/parcels siding and branch line via the two feathers. This will be off-scene so not installed, only added to the diagram for completeness.

   

• Revised location for the position lights and one with yellow light instead of red for headshunt access. 'Stop and await instructions' board added.

   

• Trap point description corrected.

   

• I've also added some estimated locations for AWS inductors following some advice I've read on this site that they are usually c.200yds before a signal and not generally used on two aspect signals. Please shout if I've got this wrong!

Layout_Signals_v3.jpg

 

 

 

Yeah been looking into it - I'm thinking probably Traintronics and/or CR signals? They seem to be a good balance between detail and cost.

 

 

Thanks, that'd be quite smart, will look into it!

 

Cheers,

Chris

Looking much closer to getting there. A few comments to consider.

 

1. The signal reading off the incline would need to be a three-aspect, showing the same as the two platforms.
2. The signal in the lower platform road would most likely have an AWS, although this was not universal in loops wher trains were usually brought to a stand, Sometines in such cases there was an AWS beyond the signal before reaching the main line. 
3. The signal at the left hand end on the inner circuit should have a subsidiary aspect to the yard and parcels bay rather than a feather. Likewise the the off-scene signal would also have a subsidiary aspect, but in this case may still have two feathers to avoid confusuion as to where a through-running train is routed.
4. I would not put a Stop board in the position shown as it would lead to delays by blocking the running line for longer during a shunt. Instead I would make the subsidiary aspect on the signal reading in also controlled by a shunter's acceptance switch in the yard. 
5. You could probably then get away with one PLGS reading out of the sidings, positioned between the yard/parcels bay points and the crossover to the main line.
6. There is no provision for attaching/detaching on the inner platform. How do you intend to get traffic from the anticlockwise trains into the yard. Is the a run-round in the sidings?

[phil-b259]

I've also added some estimated locations for AWS inductors following some advice I've read on this site that they are usually c.200yds before a signal and not generally used on two aspect signals. Please shout if I've got this wrong!

 

 

AWS is provided at all signals capable of displaying a 'caution' indication to drivers. So that means it will be fitted to 4 aspect, 3 aspect and 2 aspect yellow / green colour lights and distant semaphore signal arms. It must NOT be used at 2 aspect red / green colour light signals or stand alone 'stop' semaphore signals as neither is capable of showing 'caution'

 

Also while 200 yards is the standard distance, where line speeds are low or in situations where a bi-directional platform has a single installation relating to both signals then departure from the standard distance is acceptable.

 

AWS is fundamentally a system designed around the needs of a mechanically signalled railway and is the modern equivalence of the ATC system pioneered by the GWR. Its implementation on 4 aspect colour light signalling was long considered controversial by the Southern Region as there was a very real danger that a driver being confronted with a series of signals all showing double yellow would get so used to cancelling the AWS, that they would continue to do so when they came across a single yellow and ultimately a red signal. This is what led to the 1989 Purley rail crash which showed that something better was needed - yet it took a further two decades before TPWS closed the major flaw in the system by preventing the driver from overiding the brakes at red signals (which AWS was quite happy to let them do)

[Gwiwer]

 

 

a driver being confronted with a series of signals all showing double yellow would get so used to cancelling the AWS, that they would continue to do so when they came across a single yellow and ultimately a red signal. This is what led to the 1989 Purley rail crash

 

And was probably a contributory factor in the 1978 crash at Patcham (SR again, between Hassocks and Preston Park) where the train in rear had not braked at all after passing a single yellow and only applied brakes (far too late) when the stationary train ahead was sighted.  In this case the signal protecting the stationary train was alleged to be out - showing no aspect at all - but the driver's lack of action was attributed to the common practice of "riding yellows" and not braking until a red was sighted.  It was reported at the time that this was common practice on the Brighton line (and probably others) because drivers were keen to keep moving when possible and it was quite normal to run on yellows for much of the trip between Brighton and London due to high line occupancy.  It still is.  The driver of the train in rear was killed and unable therefore to offer his own version of events.

[phil-b259]

Also I note from your diagram that you seem to have a two aspect head (coming off the incline) leading to the same place as a three aspect head (on the end of the platform).
This is not acceptable on the prototype,but you can of course have different heads leading to different locations (See the signals ringed in red on the attachment*) but not to the same place.

 

You also lack a shunt signal coming out from the headshunt - have a look at the extract of the plan shown bellow and you will note that both shunt necks are fitted with shunt signals as it is physically possible to place something in them clear of the points leading out onto the main line.

 

 

* If set along the mainline T277 4 aspect signal operates as a 4 aspect signal. If the route is set into the branch T277 operates as a 3 aspect one, so it matches the sequence given by T281

[Chris64B]

More invaluable stuff, thanks chaps!

 

 

Looking much closer to getting there. A few comments to consider.

 

1. The signal reading off the incline would need to be a three-aspect, showing the same as the two platforms.
2. The signal in the lower platform road would most likely have an AWS, although this was not universal in loops wher trains were usually brought to a stand, Sometines in such cases there was an AWS beyond the signal before reaching the main line. 
3. The signal at the left hand end on the inner circuit should have a subsidiary aspect to the yard and parcels bay rather than a feather. Likewise the the off-scene signal would also have a subsidiary aspect, but in this case may still have two feathers to avoid confusuion as to where a through-running train is routed.
4. I would not put a Stop board in the position shown as it would lead to delays by blocking the running line for longer during a shunt. Instead I would make the subsidiary aspect on the signal reading in also controlled by a shunter's acceptance switch in the yard. 
5. You could probably then get away with one PLGS reading out of the sidings, positioned between the yard/parcels bay points and the crossover to the main line.
6. There is no provision for attaching/detaching on the inner platform. How do you intend to get traffic from the anticlockwise trains into the yard. Is the a run-round in the sidings?

 

 

Thanks for looking through it again...

1.  OK have changed it to a three aspect as you describe.
    
2. Thanks for the info, I realised I'd put it on the wrong line by mistake, should have been the one above (loop platform) so now fixed.
    
3. Yes, I was just beginning to wonder that myself having been looking on the CR Signals website where they sell a 'position light signal' described as "position light signal showing two white lights at 45 degrees, used as an addition to the main aspect signal in order to 'call on' rail traffic at yards and stations."
   
   There's no picture of it on the website unfortunately so is that the right thing? I assume it's very much like a normal ground position light signal but mounted up on the post of the main aspect signal? Do they have the red light like a GPLS or not?
    
4. Good point re blocking the main line. Will remove the stop board.
    
5. Would this be a red/white or yellow/white (as per the earlier suggestion to allow access to the headshunt if points are set)? Or perhaps the yellow/white would be where the stop board is currently shown and the red/white after that but before the main line? Or stacked on top of each other, yellow/white at the bottom and red/white above?
    
6. Yes, that issue was mentioned before - there isn't a run-round in the sidings so trains running anti-clockwise on the inner line emerge from the tunnel and have two choices, either run to the end of the inner platform and have their coaches/wagons shunted away from there or go directly into the yard themselves (which does mean the loco is trapped until the shunter comes to the rescue to remove the wagons).
   
   If it was the former, would some extra shunt signalling be needed at the inner platform left hand end then?

 

Also I note from your diagram that you seem to have a two aspect head (coming off the incline) leading to the same place as a three aspect head (on the end of the platform). This is not acceptable on the prototype,but you can of course have different heads leading to different locations (See the signals ringed in red on the attachment*) but not to the same place.

 

You also lack a shunt signal coming out from the headshunt - have a look at the extract of the plan shown bellow and you will note that both shunt necks are fitted with shunt signals as it is physically possible to place something in them clear of the points leading out onto the main line.

 

Thanks Phil, as per The Signal Engineer's comment, I've changed the two aspect on the incline to a three aspect.

 

Re the headshunt, thanks for the diagram and yes I see what you mean. My headshunt isn't very long but will fit a Class 08 plus 2-3 wagons or a larger loco on it's own so I've added a ground signal as suggested facing that way - would it be red/white or yellow/white?

 

Here's v4 - still not sure the shunt signals are correct, sorry, struggling with them a bit! If anyone can sketch out how they should look that would be great, particularly if more are needed near the platform for point 6 above!

 

 

Cheers, Chris

[phil-b259]

Re shunt signals (note that this is based on Southern region practice)

 

I would also suggest you have another look at the platform signals - Generally where you have crossovers, etc you tend to find either fixed red aspects, shunt signals or limit of shunt boards provided for flexibility.

 

Some examples are included

 

 

[St. Simon]

 It must NOT be used at 2 aspect red / green colour light signals or stand alone 'stop' semaphore signals as neither is capable of showing 'caution'

 

Interesting, Table 1 in the last issue of GE/RT/8035, although superseded, states this:

 

 

 

It doesn't say that 2 aspect stop signals are exempt from fitment at all, I have seen (and drawn) several scheme plans that have AWS magnets for 2 aspect stops, in fact Modular signalling requires all signals to have AWS and TPWS.

 

Of course, this doesn't mean that it has always been the case.

 

Simon

[phil-b259]

 

It doesn't say that 2 aspect stop signals are exempt from fitment at all, I have seen (and drawn) several scheme plans that have AWS magnets for 2 aspect stops, in fact Modular signalling requires all signals to have AWS and TPWS.

 

Of course, this doesn't mean that it has always been the case.

 

Simon

 

Indeed - and its a point worth remembering (and I include mysef when I say that), the railway environment changes over time. In your position as a designer you will naturally be working to the latest standards where as us in the maintenance department work in an environment which frequently was designed to a different, older set of standards. It is also something modellers need to consider as using the latest design rules may lead to mistakes if they are attempting to replicate an earlier time period

 

As for AWS, a change in the standards to fit it to all colour light signals makes sense - however the change is relatively recent because there are plenty of late 80s / early 90s resignalling schemes which featured two aspect red / green signals which did not feature AWS. The change in fitment criteria also indicates a subtle but distinct change in emphasis over why AWS is provided in the first place. Instead of being an aid to help drivers identify a cautionary signal in fog, etc it now effectively is a way of identifying whether a signal is at Green or not

[Chris64B]

Re shunt signals (note that this is based on Southern region practice)

 

I would also suggest you have another look at the platform signals - Generally where you have crossovers, etc you tend to find either fixed red aspects, shunt signals or limit of shunt boards provided for flexibility.

 

Some examples are included....

 

Thanks for the sketch Phil!

 

I was planning to have the headshunt and parcels bay points controlled by the signalbox (motor) and everything else further into the yard by hand on the basis that access to the parcel bay would be fairly frequent so likely to be 'automated' if you will, rather than manual like the rest of the yard.

 

So I take it that would mean going with your 'alternative option' - could/should the bottom left signal be a white/yellow to allow moves into the headshunt when the points are set that way as mentioned by others earlier in the thread? Or is that what you've allowed for even with red/white signals in your starred description?

 

Saying that, it could be a lot cheaper to only need two ground signals rather than five if I want to have them working!

 

Re the additional signals around the platforms for LOS etc, I'm wary of it all being too complex and cluttered - it's only 8x4' after all and I'm not sure I have the space really and the curves are pretty tight so ground signals around them might look a bit odd. I guess I'm sort of working on the principle that the crossovers are out on the main line further from the platform rather than part of the station area infrastructure if that makes sense?

 

Cheers, Chris

[The Stationmaster]

SN150, installed 1959 in connection with Swindon resignalling was a 2 aspect red/green signal and had AWS from new.

[TheSignalEngineer]

SN150, installed 1959 in connection with Swindon resignalling was a 2 aspect red/green signal and had AWS from new.

At that time the standard was that a two-aspect red/green signal in a run of multiple aspect signals should be fitted with AWS.

[TheSignalEngineer]

Re shunt signals (note that this is based on Southern region practice)

 

I would also suggest you have another look at the platform signals - Generally where you have crossovers, etc you tend to find either fixed red aspects, shunt signals or limit of shunt boards provided for flexibility.

 

Some examples are included

One ground signal at the exit points to the main line and one in the head shunt would be the most likely in the 1980s. If the parcels siding points were motor worked then I would only provide three signals, omitting the two between the exit points and the parcels siding points.

For attaching and detaching on the anti- clockwise platform all I would provide on-scene is one additional ground signal at the left-hand end of the platform to get back behind the signal reading into the yard.

[phil-b259]

SN150, installed 1959 in connection with Swindon resignalling was a 2 aspect red/green signal and had AWS from new.

  

At that time the standard was that a two-aspect red/green signal in a run of multiple aspect signals should be fitted with AWS.

Interesting - However I can show you a 1980s two aspect based scheme where AWS was not provided on red / green signals.

 

Another case of regional variation possibly?

[Chris64B]

One ground signal at the exit points to the main line and one in the head shunt would be the most likely in the 1980s. If the parcels siding points were motor worked then I would only provide three signals, omitting the two between the exit points and the parcels siding points.

For attaching and detaching on the anti- clockwise platform all I would provide on-scene is one additional ground signal at the left-hand end of the platform to get back behind the signal reading into the yard.

 

Hopefully I've interpreted that correctly in v5 below (but maybe not?! I think I'm going signal blind - fair play to those of you working on this stuff for real!)

 

 

What I'm a bit confused about is that compared to Phil's suggestion below (in his alternative option box), doesn't the version above leave some of the route options in both directions unsignalled? Perhaps that's ok in a yard situation where a shunter would be in control of movements?

 

 

Thanks for your time and patience all, I'm learning loads!

[phil-b259]

Hopefully I've interpreted that correctly in v5 below (but maybe not?! I think I'm going signal blind - fair play to those of you working on this stuff for real!)

 

Layout_Signals_v5.jpg

 

What I'm a bit confused about is that compared to Phil's suggestion below (in his alternative option box), doesn't the version above leave some of the route options in both directions unsignalled? Perhaps that's ok in a yard situation where a shunter would be in control of movements?

 

Layout_Signals_v4_B.jpg

 

Thanks for your time and patience all, I'm learning loads!

Basically it amounts to this - if a set of points is worked from the signal box then routes across them need signals. Hence the two different options I have given.

 

 

With regard to the alternative option, what has been suggested is that the middle set of shunt signals (between the each set of points) can be omitted. This is fine, but means every shunt move from the parcels dock and back into the yard has to go right up into the headshunt (so the entire train is clear of both sets of points) before it can go back again into the yard / parcels dock.

 

By having the extra shunt signals added, a train making such a move doesn't need to go right up into the headshunt - it only has to clear the parcel dock / yard points and can remain sitting across the points leading out onto the main line while the points giving access to the yard / dock are reset. This could be useful if space is limited.

 

However as with all things signalling era, region, traffic density and the amount of money the railway authority wish to spend all have an impact in what the end layout looks like - so my suggestions are just that,

[Chris64B]

Thanks for the explanation, that does make sense when I visualise the moves on the diagrams.

 

On the layout the space is limited - the yard entrance, headshunt and parcels points all run directly in to one another with little or no plain track in between so how many signals can actually fit may be the clincher...

 

Off to the garage to have a think and a plan!

[phil-b259]

Thanks for the explanation, that does make sense when I visualise the moves on the diagrams.

 

Ultimately that's what signal designers do on the prototype. Yes they have lots of fancy tools to help them, but it all basically comes down to 'these are the moves we wish to do, now what signals are necessary to allow them to take place' approach.

 

So take some time and work out what you need. One technique would be to draw up your diagram on a large bit of paper then get hold of some games counters / buttons / pasta shells / scraps of paper to represent trains and try simulating the moves on the plan. If a train needs to move to a particular place add the appropriate signal (main aspect or a shunt) to the plan. If you find you have two routes you now need a route indicator etc

[Chris64B]

Now in the garage playing trains and shunting back and forth to see how it would work. I have a thought that picks up on your last point about route indication - if a train or loco is leaving the headshunt it has a choice to enter the parcel bay or the yard, how does the ground position light signal indicate which route is set?

[LongRail]

Date a lot of yards don't use signals given the size of the layout I would suggest using a shunt signal to,allow trains back onto the mainline assumingbeverything else is controlled by a shunter on the ground manually changing points.

[TheSignalEngineer]

Basically it amounts to this - if a set of points is worked from the signal box then routes across them need signals. Hence the two different options I have given.

 

 

With regard to the alternative option, what has been suggested is that the middle set of shunt signals (between the each set of points) can be omitted. This is fine, but means every shunt move from the parcels dock and back into the yard has to go right up into the headshunt (so the entire train is clear of both sets of points) before it can go back again into the yard / parcels dock.

 

By having the extra shunt signals added, a train making such a move doesn't need to go right up into the headshunt - it only has to clear the parcel dock / yard points and can remain sitting across the points leading out onto the main line while the points giving access to the yard / dock are reset. This could be useful if space is limited.

 

However as with all things signalling era, region, traffic density and the amount of money the railway authority wish to spend all have an impact in what the end layout looks like - so my suggestions are just that,

In reality if only two ground signals were provided in the sidings and both could be cleared at the same time for shunting, you would find that the train would stop and start wherever the shunter dictated by handsignals to the driver. If in a location out of sight of the signalman there would often be a local instruction as to what could be done with both signals off and that the signalman had to ensure that no shunt move was about to take place before putting the signals back to danger and setting a conflicting route across the exit points. 

As far as route indications are concerned at ground signals, around 1980 these were usually minimal, and were limited to differentiating between routes of a considerable difference in nature or length such as

1. Out to the main line or into the headshunt - indication NCK for the headshunt.
2. Electrified and non-electrified lines - e.g. in the days of the 1500v electrics over Woodhead, Piccadilly station had indications to differentiate between routes available to AC or DC electric trains.
3. Routes running wrong line eg to a limit of shunt - 'X' would be included such as XDM 

From the Headshunt to different sidings there would be no need for route indications if all moves were under control of the shunter. The shunter should instruct the driver as to the moves to take place before starting.