GWR Signals and where they go?

[NorthHighlander]

This may not be the time or place BUT if that is the only problem then it is easily resolvable. 

 

1) take several photocopies for security - please put one copy in a fire safe it is a MOST PRECIOUS object and I am not bullshitting or fawning.  It is irreplaceable.

2) get the document scanned into text format - easily done by a good amateur or a professional service. It can be put into a format that you can easily work with very quickly.

3) tidy it up ONCE and then set about getting it published.  Face the fact that it can/will never be finished.  You will always find something, remember something else or learn about a new aspect.  Get issue 1 out and then worry about errata, revisions, extensions, volume 2 etc etc.

4) You CAN self publish and put it on the Internet - others will be better able to advise on that.

 

PLEASE, please don't leave it lying there.  I had to drag my father-in-law's memoirs off him and do something similar, but it was sadly too late for revisions then.

I would definitely echo what IMT has said.... it's not too difficult to read text, even written text, with OCR software. It's very advanced these days. And once digitised, keep one copy off site on a flash card, one copy in 'the cloud' on a site like Dropbox (which is useful for version control) and two copies on your own machine. And index as many words as pertinent, 100k words are about 60 pages of A4 so not easy to cross reference if (like me yesterday) you're trying to find a reference to 'slotted joints', for example. A pdf document is good for that, by the way.

[NorthHighlander]

The book still sits in type printed form with the discs it was on no longer available so it's re-write off the printed word = lots of work (100,000 words!).

 

Plus it would in any case need to incorporate everything (no - not all of it) I have learnt since I first wrote it.  With a subject like signalling getting hold of the basics is one, very important, thing but you can find yourself forever learning as you continue looking, even photos reveal fascinating things such as the exceedingly rarer GWR Backing Distant Signals, so rare there appear to have only been no more than two or three of them on the entire GWR network - unless, as usual, someone knows different.

 

Now the runaway catchpoint at Penhafod  (and I've now got the gradient the right way round thanks Tony).  So yes it has to be in advance of the Advanced Starter, No.28, and in view of what we are looking at I think it has to be double tongue and have an FPL but what I would suggest is basically the arrangement which existed at Abercynon only the other way round (at Abercynon the gradient fell, from the Merthyr direction, towards the station and after the line was singled a worked catch point became essential).  The handy thing about it from a modelling viewpoint, as far as I can recall, is that it did not have a crossing - the two switch rails ran into separate sand drags one of which was in the four foot and they went on for several hundred yards in its single line days.

 

The tunnel complicates the issue but equally I think doing the catch the way I have suggested might help as it could effectively be assumed to extend into the tunnel and you would not be diverting any potential runaways into the tunnel walls but merely into the long sand drag.  the gradient will of course impose shunting restrictions but fortunately the right way round as you would only be permitted to shunt anything at all with the engine on the lower end with suitable caveats about adequate brake power etc.  Most interesting will be its impact on the locking which oddly might result in some signals having to be passed at danger for some shunting movements if standard GWR locking rules are applied because signals 25, 26, and 27 should only be free if the catch point is closed and bolted by the FPL .

Thanks for this again The SM...! Grosvenor has suggested reassigning the southern half of the double slip to route to the engine spur as an alternative to the catch points. I think that your idea of local arrangements that the shunting movement has to have the loco at the gradient end of the approach to the advanced starter is the most practical. After all, the maximum length of train is 12 x 16T minerals (plus brake van) so I wouldn't have thought that brake force would be much of an issue, even with the humble Pannier! In practice, the only shunt movement I can envisage would be one or two wagons from Goods Loop North, the timetable would provide that any other shunt movement required from Goods Loop South would be into the NCB sidings.... aren't local arrangements wonderful!? In both goods loops the gradient is protected, north by a trap point, south by lying normally to the NCB sidings.

[Stephen Freeman]

This may not be the time or place BUT if that is the only problem then it is easily resolvable. 

 

1) take several photocopies for security - please put one copy in a fire safe it is a MOST PRECIOUS object and I am not bullshitting or fawning.  It is irreplaceable.

2) get the document scanned into text format - easily done by a good amateur or a professional service. It can be put into a format that you can easily work with very quickly.

3) tidy it up ONCE and then set about getting it published.  Face the fact that it can/will never be finished.  You will always find something, remember something else or learn about a new aspect.  Get issue 1 out and then worry about errata, revisions, extensions, volume 2 etc etc.

4) You CAN self publish and put it on the Internet - others will be better able to advise on that.

 

PLEASE, please don't leave it lying there.  I had to drag my father-in-law's memoirs off him and do something similar, but it was sadly too late for revisions then.

Mike, I think most scanners come with that sort of software bundled, they may not be perfect but it would be a good starting point.

[imt]

Mike, I think most scanners come with that sort of software bundled, they may not be perfect but it would be a good starting point.

 

Please, I don't want to divert this into a technical discussion on OCR.  Can I just say that if going that way you need a good quality printer/scanner with a sheet feeder and some idea of what kind of output would be best for your editing needs.  Doing this yourself by hand one sheet at a time will be time consuming and error prone. Best of all get a professional to do it for you - it'll take no time and be done properly.  Like all things, why learn about a new technology which you have no further need for when pressed, get a specialist to do it.

[NorthHighlander]

I think that, thanks to The Stationmaster and Grovenor (Keith), we have now bottomed out all the practical implementations for a P4 model, with all its errors and imperfections. I think the final questions to be resolved were the use of an Annett's key box to release the engine release ground frames at the buffer stop ends of the platforms and the need for a catch point in advance of the advanced starter. The removal of the trap point on platform one and the addition of a ground signal for the NCB sidings results in two spare levers. If we decide that there ought to be a catch point for completeness (even though it won't actually feature on the layout, so it is there to pull to release the advanced starter) then I would use one of them rather than reorganising the lever frame yet again! However, by diverting the double slip at the platform ends to the engine spur, the goods loop south set to normal to the NCB sidings and the goods loop north protected by trap points, the only risk of a runaway down the 1 in 50 bank is for a loose shunting movement towards the advanced starter. I conclude that this would be handled by local arrangements, and in this case, I daresay it would be acceptable to handsignal any shunting movement to the advanced starter, with the proviso that the locomotive must be at the tunnel end of the train being shunted. In any case, there is a sand trap at the foot of the bank (well off the layout).

For completeness and reference for other modellers, here is the signalbox diagram going forward and the pulling sequence and locking table.

We've said it before but it bears repeating constantly, without the help of the Stationmaster, Grovenor and others here, this would not be the accurately signalled model that it will become. My sincere and grateful thanks to you all. You can come and have a shift on the lever frame!

 

TH

[NorthHighlander]

..sorry, I'm having problems adding the files. I'll try again later..

[Stephen Freeman]

Please, I don't want to divert this into a technical discussion on OCR.  Can I just say that if going that way you need a good quality printer/scanner with a sheet feeder and some idea of what kind of output would be best for your editing needs.  Doing this yourself by hand one sheet at a time will be time consuming and error prone. Best of all get a professional to do it for you - it'll take no time and be done properly.  Like all things, why learn about a new technology which you have no further need for when pressed, get a specialist to do it.

That's fair comment, I never was very impressed with the OCR software that came with my scanner but as the scanner is of some age now I had hoped that things might have improved, clearly not. (I uninstalled the OCR software as I didn't/don't have a use for it).

[The Stationmaster]

Tony a couple of things (having had a look at the thread elsewhere in P4 land) -

 

1. Unless signalbox operation of the points is added there is no point in having an extra signal reading out of the colliery and to do so would add two levers of course plus relocating disc No.7.  It depends on two things (apart from having enough levers available) - thus

A. Is there some sort of sighting (of signals) or gradient difficulty coming out of the colliery, AND

B. Will shunting to/from the NCB Sdgs be taking place at the same time as movements are likely to arrive from the colliery?

 

At most I would expect the line from the colliery to have a STOP board (proper GWR pattern obviously) and more likely there'd be nothing as the whole lot will be under the control of one Shunter so no need for a signal of any sort.

 

I'm not entirely clear about your 'period modelled' which from comments elsewhere seems to be quite late and well after the final major cull of South Wales Valleys passenger services c.1962/63.  The only ones which retained complete steam operation that late were the ones slated for early closure (such as the Vale of Neath) with all the Cardiff and Newport Valleys fully dieselised albeit with occasional steam worked replacement of the odd diagram.  By 1971 - when I went down there workwise the passenger services on the various Valleys were at their lowest point.  Nothing of course to say that you can't avoid various closures but don't forget they really began in the 1950s with various minor branches going early.  For another thread of course - wherever you might happen to have one (and there's quite a rich seam of South Wales knowledge hiding away on RMweb should you wish to tap into it).

 

However on another tack entirely one advantage of going late is that - assuming it was all still there - the facing point lock(ing) bars at Penhafod would have been replaced by track circuits;  I can't recall anywhere I knew on the Valleys (which still had signalling) that had anything other than track circuits by the early 1970s.  

[bécasse]

 

However on another tack entirely one advantage of going late is that - assuming it was all still there - the facing point lock(ing) bars at Penhafod would have been replaced by track circuits;  I can't recall anywhere I knew on the Valleys (which still had signalling) that had anything other than track circuits by the early 1970s.  

 

That is a very interesting point (sorry!) that Mike has made there, and it set me wondering whether that was actually true generally across BR. I can certainly remember plenty of facing point locking bars in the early to mid 1960s, but with closures, rationalisation and modernisation, the majority would have disappeared anyway by the 1970s, and I can't remember any other location where they hadn't been replaced by track circuits. Doubtless there were still a few around (away from the heritage railways) but I reckon that anyone modelling a fictional location from 1970 onwards (and perhaps at any time in the "blue" era) could safely ignore the need to provide them (and to allow the space they necessarily took up).

[NorthHighlander]

Tony a couple of things (having had a look at the thread elsewhere in P4 land) -

 

1. Unless signalbox operation of the points is added there is no point in having an extra signal reading out of the colliery and to do so would add two levers of course plus relocating disc No.7.  It depends on two things (apart from having enough levers available) - thus

A. Is there some sort of sighting (of signals) or gradient difficulty coming out of the colliery, AND

B. Will shunting to/from the NCB Sdgs be taking place at the same time as movements are likely to arrive from the colliery?

 

At most I would expect the line from the colliery to have a STOP board (proper GWR pattern obviously) and more likely there'd be nothing as the whole lot will be under the control of one Shunter so no need for a signal of any sort.

 

I'm not entirely clear about your 'period modelled' which from comments elsewhere seems to be quite late and well after the final major cull of South Wales Valleys passenger services c.1962/63.  The only ones which retained complete steam operation that late were the ones slated for early closure (such as the Vale of Neath) with all the Cardiff and Newport Valleys fully dieselised albeit with occasional steam worked replacement of the odd diagram.  By 1971 - when I went down there workwise the passenger services on the various Valleys were at their lowest point.  Nothing of course to say that you can't avoid various closures but don't forget they really began in the 1950s with various minor branches going early.  For another thread of course - wherever you might happen to have one (and there's quite a rich seam of South Wales knowledge hiding away on RMweb should you wish to tap into it).

 

However on another tack entirely one advantage of going late is that - assuming it was all still there - the facing point lock(ing) bars at Penhafod would have been replaced by track circuits;  I can't recall anywhere I knew on the Valleys (which still had signalling) that had anything other than track circuits by the early 1970s.  

Thanks for the further update, SM..... one point to clarify: the layout is based on the late 1950s, before the great cull and after the introduction of DMUs to the Valleys. The confusion about '1978' is because that's when the layout was started!

Re the questions about the colliery: no, there are no sighting difficulties or gradients, there is a straight and clear run from the sidings to Goods Loop South. However, there is a potential for shunting movements independent of the sidings. The NCB sidings are operated by NCB locos which have running powers within station limits. However, I agree that signal 7 is probably superfluous and can be replaced by a stop board. I assume that the authority to proceed in that case is given by the signalman? It is possible for a conflicting propelling movement from Goods Loop South to the NCB sidings at the same time that an NCB brings a train out of the colliery, but I would expect that, because the sidings are owned by the NCB, any such propelling movement would have to be authorised, as you said, by the NCB yard shunter.

I shall investigate track circuit locking and whether, for modelling purposes, it would be more attractive to attempt locking bars on the double slips. Because all the trackwork is dumb and the train axles and wheelsets can be solid, we could have working track circuits.

I'll update the track diagram and locking table before posting here for reference.... thanks again

 

TH

[NorthHighlander]

For completeness and to summarise the comments made in previous replies, here are the 'final' (bound to be some errors subsequently discovered!) signalbox diagram and pulling sequence and locking table documents.

(document deleted, see below)

 

If you are interested in how I drew the signalbox diagram, I used paint.net. I referred to an image of Worcester Goods found on the web from a Google search on GWR Signal Box Diagrams. This gives you a variety of drawing eras. I wanted an old drawing.

Create a new drawing on paint. net, I used the default page size.

Using the LINE tool, draw the track layout using a black line 2px wide (the default on my system). You can adjust the curve of each line by moving the points along the line. You COULD use the spline tool, but I find this to be too inflexible for my purposes.  You can zoom the image in and out to fit the screen.

When you have a rough layout to your satisfaction, click Add Layer. Name this layer something like 'Track layout'.

At the top right hand corner of the screen there is an icon for Layers, select it to bring up the Layers window. Both the layers are shown and should be ticked. BUT, and this is important, click on the Track layout layer (or whatever you've called it) to make this the working layer, otherwise anything you do from now onwards will overwrite the rough layout you made.

Work out which of the tracks are siding, running lines and goods only lines. Thanks to the StationMaster, I would have got these completely wrong in my first version. Start with the running lines, so using the colour picker tool, select a grey line that approximates to the grey on the (Worcester Goods) image. Now overdraw the visible track line for what would be the running line. I can't tell you how long to draw the line, that depends on your track layout, but typically a straight line attached to a curved line would be two lines, dragging the points on the second line to the required curve. It's easier to do that to explain, but for a curve you draw a straight line to the extents of the curve and then drag the inner points to the required curve. If you get it wrong you can UNDO the line (Edit-Undo). NOW, important, once you have got the line in the right position DON'T click Enter. With the points on the line highlighted, go to the menu bar and change the line width. I selected 20px. You now have the track shown as a respectably thick line (which would have been water coloured with a paint brush at Reading). When you've finished the running lines, click Enter and now change the colour to a light blue for the sidings (if you do this with the previous line still selected, that will change to blue - much swearing ensues). Continue until you're ready to do the goods lines, with the colour now a sort of watery orange. I haven't found a way of getting the various colour densities that resulted from the Reading designer over painting each line, only because life is too short! 

When you have finished the track layout toyour satisfaction, you can add the track boundaries, selecting not the black colour but, say, 60% black colour. CHANGE THE LINE WIDTH to 2px.... Now draw the edges of each track to the outside of the thick coloured lines. Points and crossings would continue the thin lines through the thick coloured lines.

Now add a layer. Call is something line 'Signals' and again select the layer as the working layer. SInce it is likely that you will draw several signals, I draw them once and then select the drawing, copy and paste for as many similar signals as needed. If you have a bracket signal, say, just select the top part of the post and arm and copy that, adding the bracket to it accordingly. It is easier to switch off the other layers when you do the copy and paste and switch on the Track layout layer to select and move the signal to the required place. You can select the signal and there is a small curved line with two arrows which, if you move the mouse over it, enable you to rotate the selection as required. DON'T put the signal and point level numbers on this layer. Add another layer, call it e.g. 'Lever numbers' and, selecting it as the working layer with all other layers (except the very first default layer) visible, add the numbers for the signals. The default Windows fonts aren't suitable, and I found and downloaded a close enough font called 'Bryant'. I use 26pt for the signal numbers in red.

I wanted an old drawing where the point numbers are shown over the heel of the point in a yellow, grey edged circle. I created a new layer for the point number circles and another layer for the point numbers themselves. The reason for having point and signal numbers on their own layers is for the inevitable changes you'll be making to the diagram as you get the expertise from the StationMaster and Grovenor  and others...

I place the station platform and other fixed items such as the signal box on the Track layout layer because they are unlikely to change.

Remember to SAVE the document frequently and do backups of each version (either locally or on something like Dropbox which has its own version control). You'll save the document as a .pdn file for further editing, but if you SAVE AS and select .png or .jpg (I prefer the former), the layers will all be flattened to form a single document. You still have the .pdn file but DON'T flatten the layers in that file.

That's it, really...  any questions, just ask..

 

Tony Hagon

Edited June 16, 2017 by NorthHighlander

[NorthHighlander]

...sorry, uploaded the wrong version of the Pulling Sequence and Locking Table doc, here is the correct version, with acknowledgements to Keith for his help in creating the locking table.

Penhafod Pulling Sequence and locking table.pdf

[St Enodoc]

For completeness and to summarise the comments made in previous replies, here are the 'final' (bound to be some errors subsequently discovered!) signalbox diagram and pulling sequence and locking table documents.

Penhafod signalbox diagram_filled_v3.2.png

(document deleted, see below)

 

If you are interested in how I drew the signalbox diagram, I used paint.net. I referred to an image of Worcester Goods found on the web from a Google search on GWR Signal Box Diagrams. This gives you a variety of drawing eras. I wanted an old drawing.

Create a new drawing on paint. net, I used the default page size.

Using the LINE tool, draw the track layout using a black line 2px wide (the default on my system). You can adjust the curve of each line by moving the points along the line. You COULD use the spline tool, but I find this to be too inflexible for my purposes.  You can zoom the image in and out to fit the screen.

When you have a rough layout to your satisfaction, click Add Layer. Name this layer something like 'Track layout'.

At the top right hand corner of the screen there is an icon for Layers, select it to bring up the Layers window. Both the layers are shown and should be ticked. BUT, and this is important, click on the Track layout layer (or whatever you've called it) to make this the working layer, otherwise anything you do from now onwards will overwrite the rough layout you made.

Work out which of the tracks are siding, running lines and goods only lines. Thanks to the StationMaster, I would have got these completely wrong in my first version. Start with the running lines, so using the colour picker tool, select a grey line that approximates to the grey on the (Worcester Goods) image. Now overdraw the visible track line for what would be the running line. I can't tell you how long to draw the line, that depends on your track layout, but typically a straight line attached to a curved line would be two lines, dragging the points on the second line to the required curve. It's easier to do that to explain, but for a curve you draw a straight line to the extents of the curve and then drag the inner points to the required curve. If you get it wrong you can UNDO the line (Edit-Undo). NOW, important, once you have got the line in the right position DON'T click Enter. With the points on the line highlighted, go to the menu bar and change the line width. I selected 20px. You now have the track shown as a respectably thick line (which would have been water coloured with a paint brush at Reading). When you've finished the running lines, click Enter and now change the colour to a light blue for the sidings (if you do this with the previous line still selected, that will change to blue - much swearing ensues). Continue until you're ready to do the goods lines, with the colour now a sort of watery orange. I haven't found a way of getting the various colour densities that resulted from the Reading designer over painting each line, only because life is too short! 

When you have finished the track layout toyour satisfaction, you can add the track boundaries, selecting not the black colour but, say, 60% black colour. CHANGE THE LINE WIDTH to 2px.... Now draw the edges of each track to the outside of the thick coloured lines. Points and crossings would continue the thin lines through the thick coloured lines.

Now add a layer. Call is something line 'Signals' and again select the layer as the working layer. SInce it is likely that you will draw several signals, I draw them once and then select the drawing, copy and paste for as many similar signals as needed. If you have a bracket signal, say, just select the top part of the post and arm and copy that, adding the bracket to it accordingly. It is easier to switch off the other layers when you do the copy and paste and switch on the Track layout layer to select and move the signal to the required place. You can select the signal and there is a small curved line with two arrows which, if you move the mouse over it, enable you to rotate the selection as required. DON'T put the signal and point level numbers on this layer. Add another layer, call it e.g. 'Lever numbers' and, selecting it as the working layer with all other layers (except the very first default layer) visible, add the numbers for the signals. The default Windows fonts aren't suitable, and I found and downloaded a close enough font called 'Bryant'. I use 26pt for the signal numbers in red.

I wanted an old drawing where the point numbers are shown over the heel of the point in a yellow, grey edged circle. I created a new layer for the point number circles and another layer for the point numbers themselves. The reason for having point and signal numbers on their own layers is for the inevitable changes you'll be making to the diagram as you get the expertise from the StationMaster and Grovenor  and others...

I place the station platform and other fixed items such as the signal box on the Track layout layer because they are unlikely to change.

Remember to SAVE the document frequently and do backups of each version (either locally or on something like Dropbox which has its own version control). You'll save the document as a .pdn file for further editing, but if you SAVE AS and select .png or .jpg (I prefer the former), the layers will all be flattened to form a single document. You still have the .pdn file but DON'T flatten the layers in that file.

That's it, really...  any questions, just ask..

 

Tony Hagon

Tony, that looks superb and thanks for the comprehensive description. i will have a go myself.

[NorthHighlander]

Grovenor reminds me that I didn't take up his kind offer to review the locking now that the east end of the double slip is set normal to the engine spur. I have asked Keith if he would kindly review my corrections so I'll be interested to see if I got it right... here is the revised version (3.1.1):

Penhafod Pulling Sequence and locking table.pdf

[NorthHighlander]

I've just read my entry earlier today and one sentence stands out...

"Thanks to the StationMaster, I would have got these completely wrong in my first version."

I didn't of course mean that... what I meant to say was 'I got the colours completely wrong in my first version and thanks to the StationMaster for correcting me'. Sorry, StationMaster!!

Edited June 16, 2017 by NorthHighlander

[NorthHighlander]

True to his word, Grovenor has checked my workings on the pulling sequence and locking table (and many errors found!) so here is the amended document with Keith's amendments, for which I am very grateful... (fun will start when the dog table is created!).

Penhafod Pulling Sequence and locking table.pdf (at version 3.1.2)

 

I finished off the trap point from the engine spur today, but I remember that because I was building the Brighton layout (now on ice because of a difficulty finding a suitable turntable and point motors) I had lent my copy of GWR Switch and Crossing Practices to a pal... can anyone point me at an image of the rodding arrangement for a single bladed trap, please?

 

Regards

Tony

[NorthHighlander]

.....sorry, please ignore last request, I have found the thread on RMWeb!

http://www.rmweb.co.uk/community/index.php?/topic/122322-gwr-type-b-single-tongue-catch-point/

[The Stationmaster]

True to his word, Grovenor has checked my workings on the pulling sequence and locking table (and many errors found!) so here is the amended document with Keith's amendments, for which I am very grateful... (fun will start when the dog table is created!).

Penhafod Pulling Sequence and locking table.pdf (at version 3.1.2)

 

I finished off the trap point from the engine spur today, but I remember that because I was building the Brighton layout (now on ice because of a difficulty finding a suitable turntable and point motors) I had lent my copy of GWR Switch and Crossing Practices to a pal... can anyone point me at an image of the rodding arrangement for a single bladed trap, please?

 

Regards

Tony

 

And the first one I looked at is wrong!  It doesn't reflect the requirements of the Block Regulations so 

 

Branch to Goods Loop North would be (unless Special Instructions stated otherwise) - Route set to an unoccupied passenger platform in order to accept the train (the 'box Instructions might however say points set towards the relevant loop in order to accept a freight or mineral etc train and all facing points locked.

 

So as a  minimum (no Special Instruction) in order to accept a train it would be 10 standing normal, 11 reversed and 13 reversed whichever way 12 stands . Far simpler with a Special Instruction so it would be 10 and 11 standing reverse in order to accept the train, train to be brought almost to a stand at 1 before it is lowered (In view of the tunnel there might be something to indicate the presence of the train, and then train to be brought almost to a stand at 2 before it is lowered.   There might be an exemption from Rule 39(a) in respect of No 1 signal in view of the gradient rising towards it but in any event 2 would have to be kept at danger until an approaching train had almost come to a stand at it.

 

Branch to Platform 1 is correct apart from the order in which the signals are cleared - they should by your layout date be cleared in successive order and there might well be sequential locking in place by then in order to enforce that.

 

Branch to Goods Loop South we could well assume an Instruction exists to allow trains to be accepted with the route set towards the loop so the order would be as follows -

11, 12, 1314, 18, 19 all standing reverse in order to accept a train.  No 1 only to be cleared as for a train heading for the other loop, 4 not to be lowered until the train has been brought almost to a stand at it.

 

Important features are that -

1. All facing points must be bolted before a train which will pass over them can be accepted from the 'box in rear (and the points must remain set that way and bolted until the train has arrived in its destination line or has been brought to a stand at the Inner Home signal.

 

2. It was well established on the Western by the 1950s (and earlier in most places) that stop signals should be cleared in correct order starting with the rearmost and working forward to the most advance (and then the Distant).  This made logical sense when considering Rule 39(a) which required signals to be lowered in the correct order and one at a time - after a train had been brought almost to a stand - in order to allow train forward to the next stop signal.

 

3.  You need to remember that clearing Signal No.1 will lock points 10 & 12 in whatever position they might standing and should also lock points 19 with 12 standing reverse.

[NorthHighlander]

And the first one I looked at is wrong!  It doesn't reflect the requirements of the Block Regulations so 

 

Branch to Goods Loop North would be (unless Special Instructions stated otherwise) - Route set to an unoccupied passenger platform in order to accept the train (the 'box Instructions might however say points set towards the relevant loop in order to accept a freight or mineral etc train and all facing points locked.

 

So as a  minimum (no Special Instruction) in order to accept a train it would be 10 standing normal, 11 reversed and 13 reversed whichever way 12 stands . Far simpler with a Special Instruction so it would be 10 and 11 standing reverse in order to accept the train, train to be brought almost to a stand at 1 before it is lowered (In view of the tunnel there might be something to indicate the presence of the train, and then train to be brought almost to a stand at 2 before it is lowered.   There might be an exemption from Rule 39(a) in respect of No 1 signal in view of the gradient rising towards it but in any event 2 would have to be kept at danger until an approaching train had almost come to a stand at it.

 

Branch to Platform 1 is correct apart from the order in which the signals are cleared - they should by your layout date be cleared in successive order and there might well be sequential locking in place by then in order to enforce that.

 

Branch to Goods Loop South we could well assume an Instruction exists to allow trains to be accepted with the route set towards the loop so the order would be as follows -

11, 12, 1314, 18, 19 all standing reverse in order to accept a train.  No 1 only to be cleared as for a train heading for the other loop, 4 not to be lowered until the train has been brought almost to a stand at it.

 

Important features are that -

1. All facing points must be bolted before a train which will pass over them can be accepted from the 'box in rear (and the points must remain set that way and bolted until the train has arrived in its destination line or has been brought to a stand at the Inner Home signal.

 

2. It was well established on the Western by the 1950s (and earlier in most places) that stop signals should be cleared in correct order starting with the rearmost and working forward to the most advance (and then the Distant).  This made logical sense when considering Rule 39(a) which required signals to be lowered in the correct order and one at a time - after a train had been brought almost to a stand - in order to allow train forward to the next stop signal.

 

3.  You need to remember that clearing Signal No.1 will lock points 10 & 12 in whatever position they might standing and should also lock points 19 with 12 standing reverse.

Hello Mike....

Many thanks for this additional info... in fact, this assists the creation of the layout as an exhibition model. In reality, the fiddle yard has a 'bridge' between the traverser and the tunnel mouth which can be wired to emulate a treadle for a train standing at the outer home. As mentioned before, signal 1 actually exists and is the indication for a train in the fiddle yard to proceed up the branch (remembering that the signalmen and drivers are separate people on this layout!).

The bridge is actually a cassette that enables a loco to be replaced from the other end of the train so the presence of the cassette can short the two aluminium angles to indicate in the signalbox.

It will also be possible for visitors to see the train leaving the tunnel and approaching the inner homes so bringing the train almost to a halt before proceeding to the respective road. Justifies slow running, but the severe curve at the top of the gradient will require a maximum speed limit (I had planned to install a '10' board in both directions).

I'm struggling a bit to understand why the route would have to be set for an unoccupied platform in order to bring a train up for the Goods Loops, why not the unoccupied respective Goods Loop? Also does this mean the signalman can't accept a train if both platforms are occupied?

I like the idea of Special Instructions, after all, aren't we trying to provide as close fidelity to the prototype as we can??

Excellent stuff, many thanks again..

 

TH

[The Stationmaster]

Tony,

 

Firstly no need for the '10' signs - distinctly unusual on the WR for the 1950s - Drivers were expected to know the road and that meant knowing the speeds in particular places.

 

1950s so you would be working (believe it or not) to GWR 1936 Block Regulations although the numbering of later reissues suggests they were probably reissued a couple of times after 1948 as there seem to be two different BR document numbers, the latest I have being BR30176 but it is undated (as it was actually an unamended reprint of the 1936 book) although the original book had been amended on a number of occasions after 1936.  But to get back to the point the GWR acceptance at a terminus station on a line worked under ETT Regulations required the line where the train would run to be clear to the point at which the train would normally stop and all facing points set for that line.  Normally this would mean that the line would be set into the station platform as that was inevitably be where the train would normally stop - whatever sort of train they happened to be.

 

Equally of course you can bend the meaning to include a properly signalled good loop/arrival line although they were not at all common at most single line branch termini apart from those which dealt regularly with lots of mineral, i.e. coal, traffic.  So perfectly acceptable, and no doubt confirmed by the Signalbox Special Instructions ('footnotes' in GWR speak) for a train of the right type to be accepted with the route set towards such a line or in your case either of your Goods Loop lines.  But we now run into the Rule Book because in the case of any such train it is very unlikely the service timetable would nominate a particular line for it to arrive on and in any event local circumstances might require it to arrive on the other one even if it was timetabled to a particular line and such trains could in any case, and to some potential advantage, actually arrive at either of the passenger platform lines.  So no clearing the Inner Home Signal in advance and the Signalman can only clear it once he considers the Driver has sufficiently slowed his train ready to stop at that signal and is thus travelling equally slowly to run into any terminal line and safely stop without running beyond the stop blocks.

 

the point many modellers miss - as they gaily run branch freight trips head on into branch terminus yards and sidings is that the line (almost inevitably the platform line(s) which has a trailing engine release crossover connection will also accommodate the longest train which can be runround and that shunting actually commences once the train engine has runround  (N.B.  note re pedantry of language - the correct term is 'runround', not 'runaround' (which in proper usage means something completely different). In some parts of Scotland the process of running round was referred to a 'rounding a train' so if you hear an old Scottish railwayman talking about 'rounding a train' he means that it is being run round.

[XN593]

 

I hope it is ok to reopen this thread after a break of 3 years but I'm hoping for some advice in signalling my GWR Flight of Fancy?

 

Like many that have gone before, there has been a lack of proper forethought and the layout is what it is. Installing two traps after the layout was ballasted was bad enough and that was just a straight replacement.

 

It is an O gauge layout, based around the early 30's. I tried from the onset to make operating it interesting but with only 14 feet to play with there had to be compromises, hence the use of two double slips. As it is, the loop has just enough room for a loco to run round two coaches or to hold 6 wagons and a guards van. Slightly more if I use the main platform and reverse into the cattle dock/headshunt but then the length of the fiddle yard becomes the limiting factor.

 

Another confession, I do like signals and interlocking. I have just completed a 25 lever frame using the Howard Bolton/Scalefour designed frets and have the associated interlocking ready. But I don't want to start on the tappets until I know that the signalling is more or less ok. And then of course there is the locking table to do!

 

Can you have too many signals?

One of my aims is to include as many different types of wooden post signals as could be deemed appropriate. For example, a Bracket Signal instead of two starters (same size arms and same height dolls), a 3 way Route Indicator for the Home Signal (Bay, Main & Loop) and an Advanced Starter above a Shunt Signal. All are kits bought from either MSE or Scale Signal Supplies with Modelu accessories.

 

But I have issues!

1. Signalling the exit from the loop and the goods yard is giving me a problem, along with the same route in the opposite direction. I would like to install a Goods Line ringed arm signal (or should that be a siding signal) but with limited clearances around the points in the throat, I think the only place available would be just off the loop, between 17A and 17B rather than on the loop between 9A and 17A.

 

2. Should I signal the routes from either platform to the cattle dock/headshunt or can the starters be used?

 

3. I understand that the GWR treated double slips as compound points using 3 levers. There is example on the Kidderminster signal box diagram which has helped me understand the principle but I'm struggling to work out the FPLs. Ignoring the route to the cattle dock, four facing point locks seems a bit excessive. Is that correct and which points should they be associated with? It could mean running short of levers so the goods yard compound will have to be a 2 lever one.

 

4. Hopefully, after all of the above, there may still be a lever left for a Disc Signal or two, including perhaps a white light . It would be nice to operate the layout from just one panel which may mean I might have to use point indicators for the loco release and headshunt trap instead of levers.

 

I know it's a lot to ask but does this Flight of Fancy have wings or should I stick with a single starter á la Ashburton?

 

Many thanks.

 

 

[RailWest]

Some random comments....

 

>>> I understand that the GWR treated double slips as compound points using 3 levers....

 

Sometimes, but not always IMHO and I do not see any need for it in your location.

 

>>> I think the only place available {for signal 5} would be just off the loop, between 17A and 17B rather than on the loop between 9A and 17A.

 

Neither. In must be in rear of the relevant 'traps'. You already have disc 6 in rear of 7. For the sidings exit it must be in rear of 17B.

 

I do not see the need for FPL 16.

 

23/24 should be in rear of 9A, but IMHO the Down Home should be in that position anyway.

 

You need shunt discs below 1 and 2 to read into the cattle dock siding.

[Grovenor]

And 18A & B would be hand points, not in the interlocking.

[XN593]

Many thanks to you both. I really thought I might get shot down in flames. Instead, it's all positive and easy to do. So I'm glad I posted. The only reason I thought I needed FPL 16 was once the loco has run round it backs the coaches over the release points, straddling them ready for the departure.

Kind regards

Peter

[The Stationmaster]

14 hours ago, RailWest said:

Some random comments....

 

>>> I understand that the GWR treated double slips as compound points using 3 levers....

 

Sometimes, but not always IMHO and I do not see any need for it in your location.

 

>>> I think the only place available {for signal 5} would be just off the loop, between 17A and 17B rather than on the loop between 9A and 17A.

 

Neither. In must be in rear of the relevant 'traps'. You already have disc 6 in rear of 7. For the sidings exit it must be in rear of 17B.

 

I do not see the need for FPL 16.

 

23/24 should be in rear of 9A, but IMHO the Down Home should be in that position anyway.

 

You need shunt discs below 1 and 2 to read into the cattle dock siding.

Generally agree with this.  

 

The important thing is to move signal  20/21/22 to the toe  of No.9 crossover and place disc 23 (or 24) adjacent to it there are only 4 routes from there so you only need a maximum of 4 'arms' to read to them so you don't need a double disc.  A farm crossing does not need to be protected by a fixed signal of any kind.

 

Incidentally I do wonder about having a route indicator on that Home Signal in the 1930s.  We would probably be looking at quite an old signal and that would be unlikely to have a route indicator at a small station especially as two of the routes are towards sidings however some examples did exist.

 

Having the lever frame in the back of the 'box would be rather unusual for the GWR in the 1930s but if you move it to the front it means renumbering quite a few signals so maybe there's back story to explain the frame being that way round?

 

FPL 16 would be very unusual in a GWR setting.  Your logic for putting it there was absolutely correct - it is just that the GWR's logic for not having an FPL there was different from yours!  As the train has to stop short of the crossover in order to release an engine it is obvious that it will not be passing over those points once the engine is on the other end and when the train starts away .   Excellent logic so no need to spend money on an FPL.  But no doubt there were numerous places where the train was pushed back once the engine was on the other end.

 

As far as the double slip etc is confirmed you have separately numbered the point ends in the only possible way which allows them to be worked properly - 14 has to be on its own lever because of the way 11 works.  But if you regard the bay as a passenger line being used by passenger trains then 11B needs an FPL. (your signals suggest that it is a passenger line).

 

The route towards the cattle dock should be provided with fixed signals and that gives you quite a choice of how you might tacl kle the job.  For the period you are modelling most likely would be a short (or even centre pivot) arm bracketed off the main signal.  See my photo of the signal at Bugle on Page 4 of this thread or my photo of the signal at Slough at the bottom left of the only page of this thread -

 

https://www.rmweb.co.uk/community/index.php?/topic/123710-gwr-points-and-signals-on-a-blt/

 

While the Slough signal shows an arm bracketed off a tubular steel post the arm and bracket fitting was in fact transferred to that post when the original timber post Branch Home Signal was renewed with a tubular steel post.

 

Another way of doing it is the arrangement that existed at Helston - visible in several photos in this link -

http://www.rosewarnejunction.me.uk/helston-station-history/