Port Llyfni – Signalling Notes – Port Dinllaen Lines

by tynewydd » Tue Jun 3, 2014 5:07 am

Next up for discussion is Pont Llyni Junction (PLJ) – the junction station between Port Dinllaen (PD) and Caernavon for the Nantlle branch.

The track plan is here

I think it is fairly conventional design – with a couple of exceptions.

1) It has a single facing point in each direction leading to Pl3 and the marshaling and goods yards.
2) Because of the presence of a pair of high-level viaduct/bridges that encapsulate the station, there are some co-acting signals. This viaduct is use of modeling license, carrying the single line from PD to Tan-Y-Graig, but being nowhere near the correct route.
3) Both Pl2 and 3 are both signaled for bi-directional working to allow fast trains to overtake.Pl3 is also down relief while Pl2 plays that role for Up.

The signaling is part of the CLC/GCR Irish Extension as outlined for PD – and was equipped as that box is with pneumatic controls. The actual signaling is LMS.

Special types of working include branch slip coaches dropped by down express trains and added back onto up expresses and the adding of a milk tank to Nantlle trains. This is why there are calling on signals for Platform 2/3.

Anyway, please let me know if any thoughts occur or I have made some errors in judgement.

Adam

by Danny252 » Tue Jun 3, 2014 11:15 am

1) It has a single facing point in each direction leading to Pl3 and the marshaling and goods yards.

Not sure why that’s unconventional – it’s what I’d expect, to be honest, as it’s the minimum amount of facing points required!

3) Both Pl2 and 3 are both signaled for bi-directional working to allow fast trains to overtake. Pl3 is also down relief while Pl2 plays that role for Up.

My instinct is that P2 would only be signalled in the down, with P3 used for any overtaking use in either direction – however, the layout is all controlled by one box, so it’s doable.

For the branch homes, I would’ve expected a calling on arm to allow an engine to set back onto stock in the platform, even if this occurred rarely in actual operations (as per your brief). Depending on how you work the slip coach, it could also be used in shunting that to/from a branch train.

I also think there should be one on the Up Home bracket for P1, to allow the attachment of the return slip coach – I don’t think it would be acceptable to shunt it over the crossover via the discs, given that it would be carrying passengers from Nantle (after being detached from a branch train), but I could be wrong. However, perhaps you intend to do it in P2, hence the calling on for that platform?

I’m not too sure of the use of calling on arms at the right hand end of the platforms.

Double yellow discs – can someone confirm that this would be the practise? I’ve never seen stacked discs when a yellow disc was involved, but it almost certainly happened somewhere (cue new thread discussing examples?).

by jc92 » Tue Jun 3, 2014 12:10 pm

I’ve only seen yellow stacked discs once (can’t remember the location). if there is more than one route from the headshunt (eg a yard with a headshunt, and then a left hand exit onto a running line, and a right hand exit onto a running line). in this case, if both discs are on, movements can run into the neck/HS. if one or the other is off, it indicates the signalled route (as per normal red band discs).

the other situation i can think of is where the crossover from the exit signal could lead onto more than one running line (eg up line and down line) and indication of this would be preferable.

by Danny252 » Tue Jun 3, 2014 12:48 pm

To make it clear (which I didn’t before!), I was wondering whether it would be yellow/yellow, or top yellow and bottom red. Sounds like it may well have been the former, which makes more sense, else you’d have to argue that the top yellow gives you permission to pass the bottom red at danger.

by tynewydd » Tue Jun 3, 2014 2:18 pm

Danny252 wrote: For the branch homes, I would’ve expected a calling on arm to allow an engine to set back onto stock in the platform, even if this occurred rarely in actual operations (as per your brief). Depending on how you work the slip coach, it could also be used in shunting that to/from a branch train.

I was certainly missing a calling on arm for the Bay Platform on the branch home bracket – so I added one. Is that the only place you meant?

Danny252 wrote: I also think there should be one on the Up Home bracket for P1, to allow the attachment of the return slip coach – I don’t think it would be acceptable to shunt it over the crossover via the discs, given that it would be carrying passengers from Nantle (after being detached from a branch train), but I could be wrong. However, perhaps you intend to do it in P2, hence the calling on for that platform?

Well, yes, I did intend that to be a platform 2 operation, but now thinking about it, that is a long way round for an express, isn’t it? So I added one to the Up home bracket home for Pl1.

Danny252 wrote: I’m not too sure of the use of calling on arms at the right hand end of the platforms.

Actually I intended those for shunting out onto the Up as it is “wrong direction” working. That’s incorrect, I take it? I now realize that I subsequently added the two stacked ground discs that control the exit onto Down and the Up from the road as well, so that’s all certainly odd. I have removed both the subsidiaries on the RHS platform starters and those stacked discs and substituted a home (for the Down) with a bracket arm holding a miniature arm (for moves to the Up). As you pointed out, that could include a passenger coach being moved to Platform 1.

For Platform 2 moves shunt moves on the down, the Pl2 starter will have to be cleared. which is OK as there is an advanced starter. The alternative would be no AS, shunt signal + shunt limit and lots of “Shunt Ahead” bells, I think. Right now there will be “Blocking Back Outside Home” for all shunt moves to the Up and “Block Back Inside Home” for goods moves from Down to Up via the trailing crossover into the Brewery, which I could convert into only needing BBIH just for the coach working to Up by adding an outer home on the Up? The same applies to the Branch as well?

Edit: Actually, thinking about it, the shunt moves to the Up for a coach to be added to an express would be done under cover of the TOS not having been sent yet to the previous box (in this case Trefor), so no BBOH in that case.

Danny252 wrote: Double yellow discs – can someone confirm that this would be the practise? I’ve never seen stacked discs when a yellow disc was involved, but it almost certainly happened somewhere (cue new thread discussing examples?).

Also – what is the alternative you would suggest? In this case the route indications are to the Branch and (via the crossover) to the Up both for departing goods trains or shunting locos to the branch only. Those could have arms on a bracket or just a vertical arrangement of miniature arms, of course, but how would the shunting line be indicated – by a third arm that is nearly always pulled off, perhaps?

Adam

by Danny252 » Tue Jun 3, 2014 3:53 pm

I was certainly missing a calling on arm for the Bay Platform on the branch home bracket – so I added one. Is that the only place you meant?

Yep, that’s the one.

Actually I intended those for shunting out onto the Up as it is “wrong direction” working. That’s incorrect, I take it?

Yes, as a subsiduary has a specific meaning (calling on, shunt ahead, or warning, none of which are applicable here) – the miniature arm you replaced it with is more suitable. You will also need an LOS on the Up line to indicate where the short arm gives authority to – otherwise, the engine driver could chuff wrong road all the way to Port Dinllaen, as he’s not seen any signal or LOS telling him to stop!

Right now there will be “Blocking Back Outside Home” for all shunt moves to the Up and “Block Back Inside Home” for goods moves from Down to Up via the trailing crossover into the Brewery, which I could convert into only needing BBIH just for the coach working to Up by adding an outer home on the Up? The same applies to the Branch as well?
Edit: Actually, thinking about it, the shunt moves to the Up for a coach to be added to an express would be done under cover of the TOS not having been sent yet to the previous box (in this case Trefor), so no BBOH in that case.

Shunting a train into the section before giving TOS for a different train seems a bit iffy. If nothing else, I’d want confirmation from the guard that his train is complete before shunting out onto the Up Main. Otherwise, there’s a risk that the slip coach shunt could be hit by a runaway part of the arriving train, following some way behind the first part.

Otherwise, your use of BBO/BBI seems fairly sensible. Depending on the company’s regulations, shunts onto the branch may have been performed under token/tablet/staff released for shunting purposes, to protect the shunt move – that would also depend somewhat on the instruments used, as I know some early instruments weren’t able to return an authority to the same instrument it was released from.

tynewydd wrote:Danny252 wrote:Double yellow discs – can someone confirm that this would be the practise? I’ve never seen stacked discs when a yellow disc was involved, but it almost certainly happened somewhere (cue new thread discussing examples?).
Also – what is the alternative you would suggest? In this case the route indications are to the Branch and (via the crossover) to the Up both for departing goods trains or shunting locos to the branch only. Those could have arms on a bracket or just a vertical arrangement of miniature arms, of course, but how would the shunting line be indicated – by a third arm that is nearly always pulled off, perhaps?

See my later clarifying post – I had meant to ask if it would prototypically be two stacked yellows, or instead a yellow and a red, purely as stacked yellows looked a bit odd. I have no issues with the actual arrangement and intention!

by John Hinson » Tue Jun 3, 2014 5:38 pm

Danny252 wrote:See my later clarifying post – I had meant to ask if it would prototypically be two stacked yellows, or instead a yellow and a red, purely as stacked yellows looked a bit odd. I have no issues with the actual arrangement and intention!

http://shop.studio433.co.uk/index.php?r … ct_id=3824

John

by kbarber » Tue Jun 3, 2014 5:59 pm

As the signalling seems to be LP pneumatic as used by both the LSWR and the GCR, and installed over a wide area (thus akin to the LSWR Woking – Basingstoke scheme), I suspect that rather than Absolute Block working and consequent BBI/BBO and shunt ahead signals, there would be long stretches of automatic signalling (home-over-distant or even the new-fangled three-position arms), with PL box having a control on a rear signal sufficiently far out to protect any shunts. So far as shunting ahead is concerned, I don’t know what the regulations were on the LSWR main line but I suspect, at this early date, the automatic sections (apart from one or two on the approach and the first in advance) weren’t indicated in any box (I would welcome confirmation). That would replicate American practice of the day, from which this system was derived. If there were un-indicated auto sections between signalboxes, there would seem to be no need for any bell signals for shunting ahead – the advance box would never need to know what you were doing in a section he had no indication for. (Of course a UK operator might have felt it needed to be done anyway, however superfluous; again further information would be welcome.)

None of this applies to the Nantlle Branch, of course, where standard single line instructions of some flavour would apply. I think it’s a bit unlikely that anyone would go so far as to install any kind of tokenless or track circuited system to even a single section up there

by davidwoodcock » Tue Jun 3, 2014 6:51 pm

Can anyone confirm whether or not the CLC used stacked discs, I would have assumed that they didn’t until the joint lines got subsumed into the LMR and then (surely) only where layouts were altered?

The LSWR LP pneumatic signalling areas originally used rather distinctive miniature arm shunt signals (and a yellow version certainly existed from c1925 because there is an HC Casserley photo showing one), but I am not sure what the GC used. Presumably the CLC would have followed the GC.

by John Hinson » Wed Jun 4, 2014 6:00 am

davidwoodcock wrote:Can anyone confirm whether or not the CLC used stacked discs, I would have assumed that they didn’t until the joint lines got subsumed into the LMR and then (surely) only where layouts were altered?

Probably not – but the first post says that the signalling is LMS and the LMS certainly did.

John

by tynewydd » Wed Jun 4, 2014 1:54 pm

kbarber wrote:As the signalling seems to be LP pneumatic as used by both the LSWR and the GCR, and installed over a wide area (thus akin to the LSWR Woking – Basingstoke scheme), I suspect that rather than Absolute Block working and consequent BBI/BBO and shunt ahead signals, there would be long stretches of automatic signalling (home-over-distant or even the new-fangled three-position arms), with PL box having a control on a rear signal sufficiently far out to protect any shunts. So far as shunting ahead is concerned, I don’t know what the regulations were on the LSWR main line but I suspect, at this early date, the automatic sections (apart from one or two on the approach and the first in advance) weren’t indicated in any box (I would welcome confirmation). That would replicate American practice of the day, from which this system was derived. If there were unindicated auto sections between signalboxes, there would seem to be no need for any bell signals for shunting ahead – the advance box would never need to know what you were doing in a section he had no indication for. (Of course a UK operator might have felt it needed to be done anyway, however superfluous; again further information would be welcome.)

In such a system, how does the box get informed/requested to accept a train with its class/direction early enough to not cause delay to an express but not too early on a goods (and so inhibit local shunting)? Would there be some form of train describer (and how was it set) or were there still block bells from the box in the rear?

by Mike Hodgson » Wed Jun 4, 2014 3:21 pm

I think a train describer is unlikely in North Wales for the scenario being modelled.

Sometimes a special bell code might be used for a train conveying a slip portion. It was important to identify such trains correctly as if a coach was slipped prematurely, there was a risk that train out of section might be given when the main train passed and the slip guard was sitting in the middle of nowhere wondering how late his shift was going to finish today. There was also a risk that the guard might attempt to slip but the mechanism might fail and the coach would be overcarried.

Special tail lamps might be carried to distinguish the end of the the various slips portions (some trains could convey slips for more than one destination).

Box instructions sometimes specified that trains of certain classes were to be offered on immediately whilst others must wait for TES etc. The journey time for slow goods vs express passenger would be known, so the signalman in advance should know whether he was all right to continue shunting for a few minutes (assuming there were outer homes etc to allow this). He also has his WTT (which in practice he knows) and special train notices so he knows when the crack trains are imminent and when he should be OK to block the station with shunting. The pick up goods will be timetabled to allow this. And there would be telephone/telegraph traffic to keep him advised of late running of important trains.

by tynewydd » Wed Jun 4, 2014 7:03 pm

So I now realize from all this discussion that I should probably have included the signaling at Trefor junction as being inside the scope of the Pont Llyni box rather than having it’s own box – which would make perfect sense as there are no stops in between. That way the Cricceth box can offer goods trains (maybe even using permissive block if we assume Goods Only) to PLJ. Port Dinllaen would be the down box, and the next box up the line would be Caernavon on the Up. By having an automatic section between PD and Trefor and another between Trefor and PLJ, we can have increase traffic density during those boat arrival and departure rushes.

Would there need to be something special about the Block Instruments for PLJ and PD when with working automatic blocks? After all, one assumes that PLJ could accept a train from PD or Cricceth even if a subsequent section (like the station) was occupied as long as the initial section was clear to its own clearing point. But you would still want the signalman to be aware of the occupancy of the subsequent sections (like the station) to avoid clearing signals too soon. I guess I can see how this would work if the LC and TOS was all about the initial block and then rigorous track circuiting took care of the subsequent section signal interlocks, but what about where there is a junction and so a subsequent (automated) block could be impeded by a merging train from a different direction? Does that imply that the junction has to be at the boundary between boxes? So PD would be offering to PLJ and the acceptance is really at Trefor?

Adam

PS I’ll add the Trefor section(s) to the picture later today.

by Pete2320 » Wed Jun 4, 2014 10:43 pm

I’ve looked through this discussion and it seems to be resolving itself nicely but I’ll comment on one or two points I’ve picked up on.

Trevor Junction
I think this would have been provided with its own signalbox even if the LP system was in use here as until 192x there was still a limit on the distance that power operated points could be operated from a box. I think this distance would be (well) exceeded because Pont Lyni box is at the opposite end of the station and TJs down distant is not under PLs advance starter (with possible repeat under at least the P2 platform starter) suggesting TJ is more than braking distance from PL. However, there is nothing to stop the box having been abolished subsequently but my gut feeling is that this would not have happened if TJ is as far away as I suspect. I also have a suspicion that the true LP system was not efficient over long distances but of course we can reasonably assume that these boxes were converted to EP operation as were all the GCR and L&SWR boxes.

Automatic Signalling
As suggested, this might well have been a possibility certainly for the longer “block (sic)” sections. Outer homes would be essential, at least in the sense that the last signal before the inner home would need to be controlled. No LC, TOL or LB would be used, trains only being descried. Not sure how this would be done, a bell is probably simplest but it would be reasonable to assume that the same system as used on the L&SWR would be used- whatever that was!

The West End
I don’t think that route from P3 to the Limit of Shunt would have been provided unless clearing the Down Main was a priority. I don’t see any objection to shunting the loaded Slip Coach onto the back of an Up express via the down main and then the “Brewery” crossover on the authority of disc signals. This may not be permitted by modern signalling principles but we are not talking a modern installation here. As it is the said Slip coach has to get to the LOS on the authority of a miniature arm which has the same meaning as a disc. However if that LOS was provided, given its’ use I feel certain that there would have been an outer home. But is this going to conflict with Trevor Junction?

Pete

by StevieG » Wed Jun 4, 2014 11:49 pm

John Hinson wrote:Danny252 wrote:See my later clarifying post – I had meant to ask if it would prototypically be two stacked yellows, or instead a yellow and a red, purely as stacked yellows looked a bit odd. I have no issues with the actual arrangement and intention!
http://shop.studio433.co.uk/index.php?r … ct_id=3824
John

Not necessarily appropriate for Pont Llyni, but as stacked yellow discs have been a reality (as have 2-up stacked ‘yellow’ ground position-light shunts, for that matter), hypothetically, in general terms, it would appear that a single yellow disc with a route indicator, still passable when ‘On’ to a headshunt or such, could have been a possibility, though hardly an ideal solution.

But has there ever been a real example of this?

by davidwoodcock » Thu Jun 5, 2014 8:12 am

I still think that stacked shunts wouldn’t have been used. The LP pneumatic shunt signals used by the GCR (and LSWR) were of an integrated design where the LP operating piston formed part of the post and the lamp was mounted on top. I don’t see how a stacked version could have existed, although I admit that twins, mounted side by side or fore and (taller) aft, might have been a possibility. In the latter case, it would seem perfectly reasonable for both arms (for these signals had miniature arms, sometimes of rubber, rather than discs) to have been painted yellow post-1925.

The Southern seems to have replaced these signals with (non-standard) discs where boxes were later converted to EP operation and possibly elsewhere as well, but I don’t know what happened on the ex-GCR installations.

by John Hinson » Thu Jun 5, 2014 8:30 am

davidwoodcock wrote:. . . but I don’t know what happened on the ex-GCR installations.
http://shop.studio433.co.uk/index.php?r … ct_id=2612

Looks like a modified LNER shunt signal to me, but I can’t speak for the detail.

John

by davidwoodcock » Thu Jun 5, 2014 9:43 am

John Hinson wrote:Looks like a modified LNER shunt signal to me, but I can’t speak for the detail.

I tend to agree, it is certainly very different from the original LP installations, which may not have been suitable for conversion to EP operation, and different again from those installed by the SR which I think were a Westinghouse offering (but quite different to the SR mechanical Westinghouse shunt signals).

I presume that the SR went in for a mass conversion only because the originals were unsuitable for EP operation as, at that period, it was distinctly careful with the pennies. What I don’t know is whether any of the originals remained long term on the SR in places like Salisbury West where LP operation was retained. I presume that Chris knows.davidwoodcockRest-day relief

by kbarber » Thu Jun 5, 2014 10:05 am

tynewydd wrote: So I now realize from all this discussion that I should probably have included the signaling at Trefor junction as being inside the scope of the Pont Llyni box rather than having it’s own box – which would make perfect sense as there are no stops in between. That way the Cricceth box can offer goods trains (maybe even using permissive block if we assume Goods Only) to PLJ. Port Dinllaen would be the down box, and the next box up the line would be Caernavon on the Up. By having an automatic section between PD and Trefor and another between Trefor and PLJ, we can have increase traffic density during those boat arrival and departure rushes.

Would there need to be something special about the Block Instruments for PLJ and PD when with working automatic blocks? After all, one assumes that PLJ could accept a train from PD or Cricceth even if a subsequent section (like the station) was occupied as long as the initial section was clear to its own clearing point. But you would still want the signalman to be aware of the occupancy of the subsequent sections (like the station) to avoid clearing signals too soon. I guess I can see how this would work if the LC and TOS was all about the initial block and then rigorous track circuiting took care of the subsequent section signal interlocks, but what about where there is a junction and so a subsequent (automated) block could be impeded by a merging train from a different direction? Does that imply that the junction has to be at the boundary between boxes? So PD would be offering to PLJ and the acceptance is really at Trefor?

Adam

PS I’ll add the Trefor section(s) to the picture later today.

I concur with Pete2320 that Trevor Junction would have had a separate box. At the supposed date of the installation, even local power operation was still quite advanced and schemes for control of remote locations were still a way off (multiplexing technology that would minimise the number of wires/pipes from the controlling to the remote location even more so). Likewise every box would almost certainly have control of an additional automatic signal in rear of their control area, as Pete says a de facto outer home, ensuring the requisite clearing point/overlap to protect any conflicting move. That in itself would ensure nothing was impeded by a merging train – the rear signal would be interlocked with the junction points, as well as the signal at the junction itself.

As to the method of signalling, my fantasy was that the LP system was installed throughout from Aynho to PD (replacing existing mechanical signalling from Shrewsbury through N Wales to PD). That would have involved automatic signalling throughout with signalboxes being provided only at stations and junctions where something other than straight through working was required (and even many of those being capable of switching out, with automatic working maintained). There would have been no block signalling as such, apart from the section from PD to (my supposed) PD Yard (mentioned in my PD post of 13th May), but automatic signalling with trains described from box to box by bell (or even, given Bound’s interest in US practice, by a box-to-box phone). Each signal would clear automatically when the line was clear to the termination point of the overlap beyond the next signal. The other place you might have found a section of AB was PLJ to Trevor junction, but that would depend how far apart these locations were. I have an idea the LSWR auto signals were spaced about 1000 yards apart (confirmation anyone?) and I’d expect the installation here to be similar (a way of handling the intensely concentrated peaks of traffic when the boat arrives – my partner tells of three expresses leaving Holyhead pretty much block-and-block in the small hours during the 1960s/’70s). So if there’s much more than a mile between PLJ and Trevor I’d anticipate at least one set of ‘open country’ signals between the two, albeit if there was only one set they’d actually be controlled as the outer homes referred to above. If the distance was insufficient for that, Absolute Block working (or slots on the other box’s starter) would be necessary to ensure protection for conflicting moves (and to handle block backs, shunts into forward section, etc). In this case working would most certainly be by block bell, even if the ‘slotted starters’ option was chosen, as it would involve block messages (messages regarding line occupation affecting the other box) rather than simply descriptions of trains (whose line occupation would be controlled entirely by track circuits operating the signals). Whether Bound would have been so radical as to choose slotting over traditional block working I somehow doubt (AB was retained at Mirfield) so, subject to a short-enough distance, you could have proper AB working there. The other end of PLJ, of course, would undoubtedly be automatic sections from Caernarv(f)on. (I’ve just looked at your latest Pont Llyni files and realised there is at least one set of open country signals between PLJ and Trevor, so most of what I’ve said about AB is redundant.)

As to the signals themselves, I think it’s almost certain the signals through PLJ would give three indications, whether by twin arms (home-over-distant as per the LSWR scheme) or by using 3-position signals (which the GCR tried out around Keadby and could well have adopted more widely in a scheme like this). If the signal spacing could be worked appropriately, each main signal would be at about the right distance to take its place in the 3-aspect sequence. At PLJ as you’re creating it, that almost certainly isn’t the case so my money would be on each main line signal having 3 indications but the ‘distant ‘ indication being slotted through so that both home and starter cleared from caution to green simultaneously (preserving braking distance). Quite likely to be distants on loop- and bay- to main starters as well (but not starters to the branch). Whether there would be splitting distants for the routes to P3 is a moot point; the scheme is early enough for them and they would certainly help clear the main lines more quickly when trains are following closely.

If you really want another AB section, you maybe need to postulate some big installation close to PLJ on the Caernarfon side; maybe a major slate quarry (or even a find of lead ore or somesuch), sufficient to require 3-shift working of the railway facilities. The difficulty with that, given your total scheme, is that local trip workings have to leave PLJ and disappear out of the system. (Or is everything on the Caernarfon side part of the fiddle yard? In that case, although you’d need enough FY capacity to handle those local moves you’d have complete freedom to imagine what you like at PL North.)

A final thought. At PD I postulated a Midland signalbox into which the EP frame was built. How about, at PLJ, an original B&I box with the EP frame in the back? Of course the B&I might have used one of the major signalling contractors, hence a good excuse for a Saxby & Farmer or perhaps a McKenzie & Holland box in a place where none such existed in real life. Or maybe the B&I had their own design, with numerous similarities to others’ but some unique features of their own – in particular, it would be helpful if the B&I’s design was wider than the average so there had been room to install the LP frame without interfering with operation of the original mechanical frame.

The more I see of your scheme Adam, the more I like it – a might-have-been that really ought to have been. Hope my ramblings will be useful.

Keith

by Pete2320 » Thu Jun 5, 2014 7:53 pm

davidwoodcock wrote:John Hinson wrote:Looks like a modified LNER shunt signal to me, but I can’t speak for the detail.

I’m sure the signal in John’s picture is a converted standard LNER disc but I’m a little (but only a little) surprised it was at Ashton. My memories of around Guide Bridge is that standard LMS discs were used with a horizontally mounted air cylinder in front of the signal driving the usual balance lever. Across the Pennines in the Wath area a similar arrangement was used but with LNER dollies whereas the Ashton signal seems to be a more sophisticated conversion.
The Southern does indeed seem to have used the standard Westinghouse EP shunting signal although I’m a bit surprised that they worked on the much lower air pressure. This all rather suggests that the standard LP signal was not suitable for conversion to EP but I can’t see any reason why. However, if we are going to discuss this further I think it ought to be in a new thread.
BTW Salisbury West certainly was converted to EP working, despite George Priors’ writings. East had a different conversion which I think was discussed on this Forum.

Pete

by Pete2320 » Thu Jun 5, 2014 8:25 pm

kbarber wrote:As to the signals themselves, I think it’s almost certain the signals through PLJ would give three indications, whether by twin arms (home-over-distant as per the LSWR scheme) or by using 3-position signals (which the GCR tried out around Keadby and could well have adopted more widely in a scheme like this).

I think I’d go with the twin arm option. Not least, afaia all three position signals in this country were electrically operated. If we stick with the same thinking as previously, perhaps BPRS supplied “their” all electric system but this would have been post 1912ish. But the three position signals would be an interesting feature on a model and would promote discussion (ok probably not, sadly). However, I don’t think the signals at PLJ would all be three indication. I would suggest only those acting as the distants (including outer distants if applicable) for the first automatic signal in advance of the station, in fact exactly those that would have distant arms if the “auto” was the next boxes home or an IB.

by davidwoodcock » Thu Jun 5, 2014 8:30 pm

Pete2320 wrote:My memories of around Guide Bridge is that standard LMS discs were used with a horizontally mounted air cylinder in front of the signal driving the usual balance lever.

Am I right in thinking that that would have been a pretty difficult conversion on stacked discs?

However, I now wonder whether there may not have been a pair of stacked EP discs at one time at Clapham Junction on the SR – a railway on which, of course, stacked discs were a great rarity.

Finally, looking at Warburton’s A Pictorial Record of LMS Signals for a good photo of a pair of stacked discs, I noticed that Plate 52, an official photograph, does in fact show a pair of yellow stacked discs. They are of the 1941-type so it is just possible that they are an LMR installation (although the adjacent track is bullhead).

by StevieG » Thu Jun 5, 2014 10:01 pm

Pete2320 wrote: ” …. But the three position signals would be an interesting feature on a model and would promote discussion (ok probably not, sadly).

… Pete… Or a few comments just might be forthcoming, in the vein of “Huh: That’s wrong. No UK signals ever did that!”

by MRFS » Thu Jun 5, 2014 10:09 pm

Why ‘Llyni’ – not being able to contribute much to the GC power erudition, but as the language of heaven is my first tongue can I say that it looks peculiar?

by tynewydd » Thu Jun 5, 2014 10:17 pm

kbarber wrote:I concur with Pete2320 that Trevor Junction would have had a separate box. At the supposed date of the installation, even local power operation was still quite advanced and schemes for control of remote locations were still a way off (multiplexing technology that would minimise the number of wires/pipes from the controlling to the remote location even more so). Likewise every box would almost certainly have control of an additional automatic signal in rear of their control area, as Pete says a de facto outer home, ensuring the requisite clearing point/overlap to protect any conflicting move. That in itself would ensure nothing was impeded by a merging train – the rear signal would be interlocked with the junction points, as well as the signal at the junction itself.

But suppose, I thought, that the removal of the Trefor box was a later (say 1920s) development. Could the LMS have decided that this was an area where experimenting with speed signaling could have helped (between PD and PLJ)? I mocked up such a scheme for the PD->Trefor->PLJ stretch in here

The modelling motivation is that running the Trefor box would be a relatively boring but high-stress job. (I know that’s also probably prototypical for many intermediate boxes, but still!). It is possible that using 21st century technology the manually run model box could be automated, however. I have the code for an “automated signalman” who uses block bells almost completed – my idea was to use it as my own “Automatic Crispin” for the FY, able to run the Caernavon, Cricceth and Pwhelli boxes. But, at a pinch, it could run a box in vision as well. The 21st century advance would be that it actually recognizes bell sequences and responds appropriately rather than having an unvaryingly fixed repotoire as the electro-mechanical original one did.

kbarber wrote: As to the method of signaling, <snip> but automatic signalling with trains described from box to box by bell (or even, given Bound’s interest in US practice, by a box-to-box phone). Each signal would clear automatically when the line was clear to the termination point of the overlap beyond the next signal.

So basically there would be no response to the description bells other than setting up the exit signal to the automated section? And that would be proceeded by enough other signals to ensure that the driver would have time to stop and that the entrance signal would be circuited so it couldn’t be pulled off if there was a train standing waiting at the next signal. I think I got it now. The aide-memoire that the BI represent are replaced by track circuits and visual indicators.
Must be fun when the circuits or indicators break! It would take a long time to get back to a safe working condition.

kbarber wrote:I’d expect the installation here to be similar (a way of handling the intensely concentrated peaks of traffic when the boat arrives – my partner tells of three expresses leaving Holyhead pretty much block-and-block in the small hours during the 1960s/’70s).

Right. Exactly.

kbarber wrote:As to the signals themselves, I think it’s almost certain the signals through PLJ would give three indications, whether by twin arms (home-over-distant as per the LSWR scheme) or by using 3-position signals (which the GCR tried out around Keadby and could well have adopted more widely in a scheme like this).

I added a home and distant “traditional but automated” variant here

kbarber wrote:If the signal spacing could be worked appropriately, each main signal would be at about the right distance to take its place in the 3-aspect sequence. At PLJ as you’re creating it, that almost certainly isn’t the case so my money would be on each main line signal having 3 indications but the ‘distant ‘ indication being slotted through so that both home and starter cleared from caution to green simultaneously (preserving braking distance). Quite likely to be distants on loop- and bay- to main starters as well (but not starters to the branch). Whether there would be splitting distants for the routes to P3 is a moot point; the scheme is early enough for them and they would certainly help clear the main lines more quickly when trains are following closely.

kbarber wrote:If you really want another AB section, you maybe need to postulate some big installation close to PLJ on the Caernarfon side; maybe a major slate quarry (or even a find of lead ore or somesuch), sufficient to require 3-shift working of the railway facilities. The difficulty with that, given your total scheme, is that local trip workings have to leave PLJ and disappear out of the system. (Or is everything on the Caernarfon side part of the fiddle yard? In that case, although you’d need enough FY capacity to handle those local moves you’d have complete freedom to imagine what you like at PL North.)

Happy to do things right. Yes, everything Caernavon-wards is hidden (staging plus reversing loop). Looking at the map the next likely major freight site would be an interchange with the WHR at Dinas where the NWIR route reverts to the Caernarvonshire Railway route about 4 miles away. I think that’s a little far way to insist on AB but maybe not.

For reference Dinas Junction -> PLJ 4 miles. PD->TJ 8 miles. TJ->PLJ 5 miles. TJ->Criccieth 10 miles. PLJ->N 5 miles. PD->TYG 3 (hard) miles. TYG->Pwhelli 5 miles. In the model… well, let’s just say “Your Mileage May Vary”

kbarber wrote:A final thought. At PD I postulated a Midland signalbox into which the EP frame was built. How about, at PLJ, an original B&I box with the EP frame in the back? Of course the B&I might have used one of the major signalling contractors, hence a good excuse for a Saxby & Farmer or perhaps a McKenzie & Holland box in a place where none such existed in real life. Or maybe the B&I had their own design, with numerous similarities to others’ but some unique features of their own – in particular, it would be helpful if the B&I’s design was wider than the average so there had been room to install the LP frame without interfering with operation of the original mechanical frame.

Do you mean for installation or that there would have been a need for both an LP (?L?-Pneumatic) and traditional frames to coexist longer than just during install?

kbarber wrote:The more I see of your scheme Adam, the more I like it – a might-have-been that really ought to have been. Hope my ramblings will be useful.

Very, thanks, Keith!

by kbarber » Fri Jun 6, 2014 8:39 am

tynewydd wrote:kbarber wrote:I concur with Pete2320 that Trevor Junction would have had a separate box. At the supposed date of the installation, even local power operation was still quite advanced and schemes for control of remote locations were still a way off (multiplexing technology that would minimise the number of wires/pipes from the controlling to the remote location even more so). Likewise every box would almost certainly have control of an additional automatic signal in rear of their control area, as Pete says a de facto outer home, ensuring the requisite clearing point/overlap to protect any conflicting move. That in itself would ensure nothing was impeded by a merging train – the rear signal would be interlocked with the junction points, as well as the signal at the junction itself.

But suppose, I thought, that the removal of the Trefor box was a later (say 1920s) development. Could the LMS have decided that this was an area where experimenting with speed signaling could have helped (between PD and PLJ)? I mocked up such a scheme for the PD->Trefor->PLJ stretch in here

Certainly possible. We’re now looking at a conversion of this section of the 1908 installation from LP pneumatic to some form of electric or electro-pneumatic before the system was 20 years old, on a joint railway (thus needing the agreement of both companies – always a hindrance to significant change). And I previously suggested the MR was responsible for structures and the GCR for signalling, whereas this would be the reverse (but I think I’m right in saying there was one joint line where responsibility swapped back & forth at intervals, so maybe the LMS was in charge of signalling in the ’20s). Either way, I think I’d say possible but pushing your luck a little.

tynewydd wrote:The modelling motivation is that running the Trefor box would be a relatively boring but high-stress job. (I know that’s also probably prototypical for many intermediate boxes, but still!). It is possible that using 21st century technology the manually run model box could be automated, however. I have the code for an “automated signalman” who uses block bells almost completed – my idea was to use it as my own “Automatic Crispin” for the FY, able to run the Caernavon, Cricceth and Pwhelli boxes. But, at a pinch, it could run a box in vision as well. The 21st century advance would be that it actually recognizes bell sequences and responds appropriately rather than having an unvaryingly fixed repotoire as the electro-mechanical original one did.

Sounds a good idea, from the point of view of operating the model. As Trevor would undoubtedly have auto sections all round, given the distances you set out, there would be no bell signals to handle other than train descriptions. The LMS was fond of routing codes in its bell signals so you could easily imagine the A codes (say) being used for trains to the Criccieth branch and the Automatic Adam being programmed to set the road accordingly on that basis; no need to differentiate on the up, provided the Trevor module can be persuaded to regulate properly rather than simply send ‘first-come-first-served’ or (equally problematic) ‘booked-route-booked-order’.

tynewydd wrote:kbarber wrote: As to the method of signaling, <snip> but automatic signalling with trains described from box to box by bell (or even, given Bound’s interest in US practice, by a box-to-box phone). Each signal would clear automatically when the line was clear to the termination point of the overlap beyond the next signal.

So basically there would be no response to the description bells other than setting up the exit signal to the automated section? And that would be proceeded by enough other signals to ensure that the driver would have time to stop and that the entrance signal would be circuited so it couldn’t be pulled off if there was a train standing waiting at the next signal. I think I got it now. The aide-memoire that the BI represent are replaced by track circuits and visual indicators.
Must be fun when the circuits or indicators break! It would take a long time to get back to a safe working condition.

The sequence would be that Call Attention would be sent and acknowledged, then the description of the train (the same code that’s used for offering under Absolute Block) would be sent and acknowledged (by repetition). As it’s for information only, the receiving box doesn’t have the option to not acknowledge.

Yes, you’ve got the idea of the intermediate signals and their operation. In later practice, track circuits would be indicated on an illuminated diagram. 1908 might be considered a little early for that but, given the number of TCs that would need indicating, attempting to do so with galvanometer-type indicators on the block shelf would give problems both of space for the installation and of reading the indications sufficiently clearly and rapidly. Therefore I would think you had an early illuminated diagram in those boxes.

In later UK main line practice, all track circuits would be indicated for trains approaching a box, starting with the first TC in advance of the previous box’s starter. For trains leaving the control area, TCs would be shown to the terminating point of the overlap beyond the first automatic signal. Signals themselves would not be indicated – their aspect could be inferred from track circuit occupation. Every auto signal would have a phone to the box that supervised it. In case of failure (a track circuit left showing occupied or a signal unlit or incorrectly at danger) the drill would be for the driver to phone the signalman, who could authorise the failed signal to be passed at danger. It can get rather pressured under those circumstances, but traffic will keep moving (albeit with delays).

I don’t know whether that practice was followed in the LSWR installation. US practice (followed by London Underground) was to indicate track circuits for approaching trains only for one or two signal sections back from the first controlled signal, and it’s quite possible the LSWR did the same. In that case, automatic signals would be designated as such in some way and drivers would have authority to pass them at danger having stopped for a set time (one or two minutes), thereafter proceeding cautiously until they’d passed at least two further signals in the clear position. No telephone would be provided at the signal. The driver would then be required to report the circumstances at the next signalbox.

tynewydd wrote:kbarber wrote:If the signal spacing could be worked appropriately, each main signal would be at about the right distance to take its place in the 3-aspect sequence. At PLJ as you’re creating it, that almost certainly isn’t the case so my money would be on each main line signal having 3 indications but the ‘distant ‘ indication being slotted through so that both home and starter cleared from caution to green simultaneously (preserving braking distance). Quite likely to be distants on loop- and bay- to main starters as well (but not starters to the branch). Whether there would be splitting distants for the routes to P3 is a moot point; the scheme is early enough for them and they would certainly help clear the main lines more quickly when trains are following closely.

Ok I’ll look at that. I assume that equivalent color signal “splitting distant” would not be used in speed signaling based on the notes on the page describing them – only the main line would have the ability to set high-speed, so the preceeding signal(s) would not clear to high-speed if the Pl2/Pl3 directions were set.
You wouldn’t get a clearance to high speed for P2/P3 but a positive indication of a clear route at the junction signal is what would help clear the main – the driver can regulate his speed so he reaches the junction at the maximum turnout speed (and obviously passes the protecting signal at a speed dictated by his braking curve). The alternative is that, passing a distant at caution, the driver must slow down sufficiently to be able to stop at the protecting signal; the whole approach is therefore considerably slower and the main line is blocked for rather longer, possibly 3 or 4 minutes more in the most adverse situation, with consequences for the achievable headway.

tynewydd wrote:kbarber wrote:If you really want another AB section, you maybe need to postulate some big installation close to PLJ on the Caernarfon side; maybe a major slate quarry (or even a find of lead ore or somesuch), sufficient to require 3-shift working of the railway facilities. The difficulty with that, given your total scheme, is that local trip workings have to leave PLJ and disappear out of the system. (Or is everything on the Caernarfon side part of the fiddle yard? In that case, although you’d need enough FY capacity to handle those local moves you’d have complete freedom to imagine what you like at PL North.)

tynewydd wrote: kbarber wrote:A final thought. At PD I postulated a Midland signalbox into which the EP frame was built. How about, at PLJ, an original B&I box with the EP frame in the back? Of course the B&I might have used one of the major signalling contractors, hence a good excuse for a Saxby & Farmer or perhaps a McKenzie & Holland box in a place where none such existed in real life. Or maybe the B&I had their own design, with numerous similarities to others’ but some unique features of their own – in particular, it would be helpful if the B&I’s design was wider than the average so there had been room to install the LP frame without interfering with operation of the original mechanical frame.

Do you mean for installation or that there would have been a need for both an LP (?L?-Pneumatic) and traditional frames to coexist longer than just during install?
Frames would only need to co-exist during the installation period. But the box would need to be wide enough to accommodate both and still be worked properly (unless it was possible to completely shut the Nantlle branch during installation and work with the box ‘switched out’). If there wasn’t sufficient width in the box, a new box would have to be built to accommodate the new frame. Remember these power frames were quite wide and could make a normal-width box feel surprisingly cramped, even without having to work a mechanical frame behind them during installation & testing.

kbarberRest-day relief

by kbarber » Fri Jun 6, 2014 4:20 pm

Pete2320 wrote:kbarber wrote:As to the signals themselves, I think it’s almost certain the signals through PLJ would give three indications, whether by twin arms (home-over-distant as per the LSWR scheme) or by using 3-position signals (which the GCR tried out around Keadby and could well have adopted more widely in a scheme like this).
<snip>

However, I don’t think the signals at PLJ would all be three indication. I would suggest only those acting as the distants (including outer distants if applicable) for the first automatic signal in advance of the station, in fact exactly those that would have distant arms if the “auto” was the next boxes home or an IB.

Pete

I think I’d disagree here Pete. What we’re looking at is effectively 3-aspect mas (Multiple Aspect Signalling) with signals spaced at approximately braking distance to get the best possible throughput in open country. With the scale of resignalling we’d be talking about here, I anticipate stations would be worked in to that general scheme so far as possible. It would make most sense if the home in each direction was at braking distance from the preceding auto or, more accurately, from the preceding evenly spaced signal that had been given a control to act as an outer home. That means there would need to be, in turn, a distant under the home at braking distance from the next auto; any starting and advanced starting signals between would also have distants.

I’ve done some jiggery-pokery with one of the PLJ diagrams – I hope you don’t mind, Adam – and the result is here.

D287 is an auto (I assume a mileage-based numbering system would’ve been used and that sounds about right from Marylebone) and 1 is the ‘outer home’ which is located at full braking distance from 2. The distant beneath 2 reads to the auto beyond the country end of the station (presumably numbered D288, or perhaps D289 depending where the mile post is) and so do the distants beneath 4 and 6. 3 doesn’t need a distant as it would be assumed to read towards a signal at danger and the driver would already have reduced speed for the turnout, hence able to pull up within the length of the platform. (At best, I’d anticipate a fixed distant there. Equally, the splitting distant under 1 would be taken as reading to 5 at danger. This would be a matter of drivers’ route knowledge.)

Therefore, on receiving a caution at 2 off, a driver would know he was braking to stop at D288; the distant under 4 would be a reminder and prevent a wrong assumption that the road was now clear past D288. When D288 clears, the distants beneath 2 and 4 will clear simultaneously (if those signals are off). The distant beneath 6 also refers to D288 but there’s no need for one under 5 as the train is approaching under a caution aspect and, in any case, speed through the turnout is low enough that braking distance to D288 won’t be an issue.

Of course similar arrangements would apply in the up direction also.

Incidentally, it might be that 1 could be allowed to work automatically when the London end crossover isn’t being used (likewise at the other end of course). It seems most likely that this would happen by simply leaving that lever (slide) in the ‘reverse’ position. Given that we’re talking about a BPRS frame with dynamic indication, that would lead to the slide putting itself part-way back when 1 returns to danger then putting itself back fully to the reverse position when the overlap beyond 2 (or 3 if the points have been reversed in the meantime) clears; from my one visit to such a frame I recall that this happens with rather a loud thump and it’s very peculiar to watch the slides moving in & out with no human intervention.

by Pete2320 » Sat Jun 7, 2014 12:34 am

davidwoodcock wrote:Pete2320 wrote:My memories of around Guide Bridge is that standard LMS discs were used with a horizontally mounted air cylinder in front of the signal driving the usual balance lever.
Am I right in thinking that that would have been a pretty difficult conversion on stacked discs?

It would have involved a degree of ingenuity, probably the air cylinders one behind the other, slightly off line. Actually I can’t remember any stacked discs at Guide Bridge- which isn’t to say there weren’t any.

davidwoodcock also wrote:However, I now wonder whether there may not have been a pair of stacked EP discs at one time at Clapham Junction on the SR – a railway on which, of course, stacked discs were a great rarity.

Westinghouse certainly produced a two arm version of their standard EP shunt signal, however this is irrelevant as EP was never used at Clapham Jct. The LSWR boxes were originally LP and then “electric”. I think BPRS produced a two arm shunting signal for both LP and electric although the latter is unlikely at CJ. I’m not sure if the Westinghouse electric discs were intended as stackable but it would be easy to mount one above another.

Pete.

by Pete2320 » Sat Jun 7, 2014 1:22 am

tynewydd wrote:But suppose, I thought, that the removal of the Trefor box was a later (say 1920s) development. Could the LMS have decided that this was an area where experimenting with speed signaling could have helped (between PD and PLJ)? I mocked up such a scheme for the PD->Trefor->PLJ stretch in here

What you have drawn is fine as early MAS but isn’t really speed signalling. A “splitting” speed signal would have a top signal for the main (fast) route and below that another signal for medium speed routes, ie all other running lines. Below that again would be a miniature signal for routes to sidings etc. In practice what was provided at Mirfield for a splitting signal was three searchlight heads one above the other with a fixed red marker light below these with a miniature yellow (originally green?) where there were routes to sidings. The top head displayed R/Y/G for the main route, the next one down yellow only to give fourth and fifth aspects for the main route, the third one down R/Y/G for medium speed routes (ie to all lower speed running lines) and below that again a two aspect head showing a red “marker” light (extinguished only if the top head was at green) or a miniature yellow for subsidiary routes or routes to sidings. This resulted in some funny looking aspects. From the top, “Stop” was red, black, red, red, “advanced preliminary caution” for the main line was green, yellow, red, black. Combinations of heads were provided as required, but main line signals were always five aspect and the red marker light was always provided even where the miniature yellow was not needed. In fact Bound “bottled out” at Mirfield and true speed signalling as above was only used between parallel running lines, geographical junctions being provided with separate side by side heads.

Pete

by tynewydd » Sat Jun 7, 2014 6:04 am

MRFS wrote:Why ‘Llyni’ – not being able to contribute much to the GC power erudition, but as the language of heaven is my first tongue can I say that it looks peculiar?

You can not only say it – you might even be right!

Well, it is clearly a spelling error on my part

– but which one? It clearly should be Llynfi (place of the lake) as the Afon Llynfi is nearby but a Victorian might have spelt it Llynvi – as in The Llynvi and Ogmore Railway (originally Duffryn, Llynvi and Portcawl Railway) a part of the GWR. But as I already made several mistakes with Dinllaen previously – I guess in London that would be pronounced “Din-lay-en” we’ll try for the correct spelling, shall we?

Let’s just hope that a “man” will appear soon with a virtual pot of paint and clear the matter up without any “trafferth”. During my many summer holidays in Anglesey, if you wanted a man to come to do a job the initial response was always what sounded like am-hos-ib but was “amhosibl” so I guess the linguistic misspelling/appropriation cuts both ways!

<edit>
Many thanks to the inspector, John, for doing the deed!
</edit>

Diolch yn fawr,
Adam