Advice sought for lever sequence for Dartmouth

[Paul Townsend]

I have reached a point in building Dartmouth of 1870s where I must allocate numbers to levers in a frame for the station area. Although the frame does not yet exist I want the numbers to be fixed now so they are used in the CBUS configuration and JMRI.

The model is based on the Acts of 1860s that authorised the Broad Guage Totnes & Dartmouth Rly. and the standard gauge South Hams Rly. between Plymouth and Dartmouth. A financial slump in the period stopped either being built back then.

I attach the track plan as a non-scale diagram ( created as .pdf in ANYRAIL before converting into JMRI)
This shows the numbered turnouts and the numbered signals. The turnout numbers are set in stone but signals can be renumbered as convenient to match a logical lever numbering sequence. This diagram represents about half of the model. The remainder connects at the right hand end and represents rural lines, MPD , carriage sidings and a traverser fiddle yard.
The 4 tracks at that end are from top to bottom:
T&D Up, T&D Down, SHR single track so both ways, a goods shed and buffers.

FWIW the layout is controlled by Merg CBUS and has 2 mimic panels with push buttons and leds and will be automated by JMRI later.. I will be adding lever frame (s) for the station area but have yet to fix the signal box position. Indeed it may end up as a small box augmented by ground frame(s).
In this era, signalling was developing from scratch and did not conform to any modern standard. I will eventually add rudimentary partial interlocking. N.B. Totnes area on the SDR/GWR was used as a trial area for this emerging technology in mid 1870s. That is my excuse for my model of the Totnes and Dartmouth Rly to also have some safety features.

The colours on the diagram can be ignored, they identify electric block sections for CBUS detection and JMRI monitoring. Extra information is shown for Capstans, infra red spot detectors etc which should also be ignored as they are not relevant to signal lever frame(s).
There are 8 signals and 9 turnouts on this part of the model which at any later modernisation might have been controlled from a signal box, but 3 of these turnouts are goods only rated an probablywould have stayed as ground levers for ever, although for the model it is best if they are controlled from the main lever frame(s).
The positions of the signals on the diagram is an approxiamation to the real position on the layout.
e.g. Sig 1 will be further towards the right near the platform end.
Sig 16 is shown as 16/17 but 17 will probably be a fixed distant

The actual signals will be a mixture of conventional semaphores, Disc & crossbar, and a couple of slotted post semaphores. All turnouts are, and signals will be, driven by servos.

I am somewhat ignorant about signalling techniques in any peripod and so ask for advice.

How would the 16 or so levers be sequenced for the turnouts and signals shown?

Dartmouth-16_Stn.pdf

[Hardwicke]

Are the two distants for boxes off scene as they look like they are for trains leaving the station ? If so they will not be on your box and you will not need to actually numbered them as such.
However you will need two protecting distants for incoming trains and these need numbering even if not modelled. It all depends if you interlock track and signals and run to prototype rules. You will need corresponding advance starters off scene and inner homes to protect incoming trains.
As far as I understand it.

[Hardwicke]

Incoming signals have the lowest numbers with their associated points numbered as close to them as necessary. Outbound traffic has the higher numbers.

[Hardwicke]

You need signals for points 0,1,2,3,5 and 6 to protect movements on the running line. Possibly ground signals. 0 could be a simple lever with the crew giving directions or even spring loaded one way only.

[grovenor-2685]

Hi Paul,
Your period is clearly when the signalling principles were embryonic. Hence without any specific prototype info its difficult to be specific. Numbering pretty well settled down to having the signals and points as near as possible in the order of the levers in the frame. To follow this you must first know which side of the line your frame is on and then whether the frame is at the front or back of the box. Similarly if you are distributing the operation over several ground frames then each will have its own numbering so you have some decisions to make before getting to numbering.
Your point numbers, assuming they are the TO# nos, don't fit with lever numbers as a frame does not have a lever 0 and crossovers would most likely only have 1 lever each. Apart from that the numbers do seem to assume a frame on the down side with the frame at the front of the box. In this case the most common set up would be for the Up direction running signals to start with 1 for the distant the 2 for home and 3 for starter. The down running signals would start from the other end of the frame in reverse order. Any shunt sugnals fitted in close to the points they protect.

[Paul Townsend]

Are the two distants for boxes off scene as they look like they are for trains leaving the station ? If so they will not be on your box and you will not need to actually numbered them as such.
However you will need two protecting distants for incoming trains and these need numbering even if not modelled. It all depends if you interlock track and signals and run to prototype rules.

Thats helpful, TVM.

However few rules applied in this peripod!

You will need corresponding advance starters off scene and inner homes to protect incoming trains.
As far as I understand it.

That is good for modern practice but may not have applied in 1870s !

[Paul Townsend]

Incoming signals have the lowest numbers with their associated points numbered as close to them as necessary. Outbound traffic has the higher numbers.

Clear and simple answer, ta.

[Paul Townsend]

You need signals for points 0,1,2,3,5 and 6 to protect movements on the running line. Possibly ground signals. 0 could be a simple lever with the crew giving directions or even spring loaded one way only.

These turnouts would likely have been controlled by ground frame and/or capstans in the real world. These devices threw the switches and had a sort of flag on top that functioned as a signal. However in the model I may have the sig box lever frame do the biz

[Paul Townsend]

Hi Paul,
Your period is clearly when the signalling principles were embryonic. Hence without any specific prototype info its difficult to be specific.

Prototype info is like hen's teeth. SRS info that I have seen does not go back to prehistory!

Numbering pretty well settled down to having the signals and points as near as possible in the order of the levers in the frame. To follow this you must first know which side of the line your frame is on and then whether the frame is at the front or back of the box. Similarly if you are distributing the operation over several ground frames then each will have its own numbering so you have some decisions to make before getting to numbering.
Your point numbers, assuming they are the TO# nos, don't fit with lever numbers as a frame does not have a lever 0 and crossovers would most likely only have 1 lever each. Apart from that the numbers do seem to assume a frame on the down side with the frame at the front of the box. In this case the most common set up would be for the Up direction running signals to start with 1 for the distant the 2 for home and 3 for starter. The dowcatchline that n running signals would start from the other end of the frame in reverse order. Any shunt sugnals fitted in close to the points they protect.

All good useful advice so TVM.

Because I adhere to the principle:
there are 10 types of people in this world, those who understand binary notation, and those who don't.
(acknowledgements to Duncan from another place)
I started with Turnout 0 and am a bit stuck with that....too many things to change.
Do I regret that decision....Yes!

[grovenor-2685]

Incoming signals have the lowest numbers with their associated points numbered as close to them as necessary. Outbound traffic has the higher numbers.

Clear and simple answer, ta.

However, not common for through stations, possible for a terminus.
Popular arrangement, depending on frame orientation, Up running signals are in sequence from one end, Down running signals in sequence from the other end, points and shunt signals in the middle. Alternative that was briefly fashionable, turn it inside out, Up and Down running signals in the middle on the basis they are the most used and that will minimise walking, points and shunts on the ends.
Unless you introduce a look up table to decouple your point numbers from your lever numbers you are pretty well stuck with having the point numbers at one end.

[martin goodall]

As Keith has pointed out, period practice in the Pleistocene era may have differed from later norms.

Subject to this, two points occur to me. First, it depends which way round your Signal Box will face as to whether, at a terminus, No.1 will be the outermost Down signal or the innermost Up [Starting] signal. Second, someone mentioned incoming Distant signals. At a terminus like Dartmouth these would be Fixed Distants, i.e. always at Danger, and nailed to the signal post, so would not have numbers in the Signal Box lever frame.

In GWR practice, it was usual for running signals for one direction to be at one end of the box, and for running signals in the other direction to be at the other end of the Box. (As I mentioned above, it depends which way round the Box faces as to which way these signals will be arranged.) In the middle of the frame, points, FPLs (where needed) and ground signals would have been set out roughly in geographical order, subject to avoiding any 'Pull Betweens' (which might necessitate a slight departure from strictly geographical order).

One final point - It was usual to leave a few spaces in the frame to allow for possible enlargement or re-arrangement of the station layout in the future. These were spaces, NOT spare levers. The practice of having white-painted spare levers in the frame came much later when signalling layouts started to be simplified, and it was easier to leave the disconnected levers in the frame and simply to paint them white to show that they were no longer in use.

[Tim V]

I am reminded that the Ffestiniog Railway was signalled at the same time as your model is set. Have a look at https://www.festipedia.org.uk/wiki/FR_signals - it was modern in 1860s. You may have to re-think your signalling plan!

[bécasse]

I disagree, quite strongly, with Martin Goodall's view that the distant would have been fixed. Not just because in general distants started to be fixed quite late in signalling history, but also because, in the early 1870s at least, the distant signal would have looked just like a home signal without even a fishtail to differentiate it - I think that they started to appear about 1873 although the arms were still painted red for most of the next half-century. And how did a driver know that it was only a distant signal that could be passed when "on", that was what route knowledge was for.

The early 1870s were definitely still in prehistory so far as signalling on many railways was concerned and that only started to change with the 1873 Act which gave the BoT more powers to insist that new works were carried out in accordance with their "advice", retrospective conformity had to wait until the 1889 Act (and even that gave Railway Companies a fair amount of time to bring their railways up to scratch).

There was no requirement to provide shunting signals at all in the early 1870s (and when they were required it was only for movements on to running lines), although some coacting point indicators might have been provided.

Looking at early 25" OS maps for the local area, it would seem that the only signal initially provided at the Brixham Road terminus (later Churston) was a single home, quite possibly a disc.

[Hardwicke]

I've had a look through "Victorian & Edwardian Railways from old photographs" and pre 1870's there wasn't much signalling at stations. Nothing for crossovers or sidings and just Bar and disc signals, double arm semaphores or early single arm semaphores. So I'd suggest you need signals to protect your incoming lines and outbound only. Signals 8, 16,1 & 4, plus one for trains leaving on the SHR.

[Winander]

In this case the most common set up would be for the Up direction running signals to start with 1 for the distant the 2 for home and 3 for starter. The down running signals would start from the other end of the frame in reverse order.

Incoming signals have the lowest numbers with their associated points numbered as close to them as necessary. Outbound traffic has the higher numbers.

I was interested in what position levers would be in a box, but didn't dare ask. In my mind the above two statements appear to contradict each other. Isn't the up direction towards the capital (usually), therefore, outbound?

[Hardwicke]

I've checked the Ashburton line and the Up is towards Totnes where the Up runs to the Down.
Confused, you will be.
Up is to London, the railway headquarters or the more important location on the line if not towards the aforementioned.

[Paul Townsend]

I am reminded that the Ffestiniog Railway was signalled at the same time as your model is set. Have a look at https://www.festipedia.org.uk/wiki/FR_signals - it was modern in 1860s. You may have to re-think your signalling plan!

Iconoclast!

However I rather like the double arm sigs and the 3 position ones. I may introduce some of each.....easy scratch builds above ground but the 3 posn one will be a challenge for servo operation!

[Paul Townsend]

Many helpful replies here.
I can now adopt a scheme which is logical and historically justifiable.
I will come back with a draft plan soon.

[Neil Smith]

Isn't the up direction towards the capital (usually), therefore, outbound?

Well not always Richard and even on the same bit of track. The Furness Railway Working Timetables show that it changed is mind at least twice over which way was "up" on its branches. So for Lake Side (to give the FR spelling), trains heading to the terminus were "up" in Victorian times, then "down" before the Great War, then back to "up" again before Grouping!

And in my experience levers were numbered from the left of the frame and so they lowest number depended simply on whether the frame was at the front or back of the box and which side of the track the box was built.

All the best

Neil

[Alan Turner]

Many helpful replies here.
I can now adopt a scheme which is logical and historically justifiable.
I will come back with a draft plan soon.

You may care to look at: http://www.gwsbristol.org/hsignals.html

About half way down.

Also see bottom for more information.

regards

Alan

[martin goodall]

I disagree, quite strongly, with Martin Goodall's view that the distant would have been fixed. ............. The early 1870s were definitely still in prehistory so far as signalling on many railways was concerned

I did qualify my remarks with a caveat that early practices may well have differed from the later signalling practices with which we are all more familiar.

In a Bristol area group meeting yesterday evening, I facetiously suggested to Paul that many of his signalling requirements might perhaps be met by having a strategically placed Company Servant who could raise a green flag at the appropriate moment.

[Paul Townsend]

Thanks to Alan for the link.

The Frome D&C is useful but the piccy of the capstan is even better.
The Broad Gauge Society has etched and cast parts for both of these but to see them in colour is a bonus.

Who should I approach to get permission to reproduce those two pics for dissemination within BGS?.
GWS Secretery perhaps?

[Paul Townsend]

I disagree, quite strongly, with Martin Goodall's view that the distant would have been fixed. ............. The early 1870s were definitely still in prehistory so far as signalling on many railways was concerned

I did qualify my remarks with a caveat that early practices may well have differed from the later signalling practices with which we are all more familiar.

In a Bristol area group meeting yesterday evening, I facetiously suggested to Paul that many of his signalling requirements might perhaps be met by having a strategically placed Company Servant who could raise a green flag at the appropriate moment.

I assume Becasse is French so Jingoism to the fore; what can a Frenchman possibly know about early GWR practise?
OOer IKB was half French, I almost forgot!

I may well add a bobby somewhere. BGS commissioned an early bobby from Modelu. The scan being of a BGS member who likes dressing in ancient uniforms. Saw off an arm and add a wire to a servo.......

[Paul Townsend]

I now have a cunning plan for 16 levers + 2, 4 or 5 spares. The S4 Soc lever frame is in multiples of 5 ISTR.
The number sequence follows the recomendations from experts above. Thanks again chaps.

I decided to bite the bullet of renumbering the turnouts, complicated, and the signals, easy. This enables me to get rid of TO 0 and arrange that all sigs and T/Os have numbers that correspond to the lever no. It was a challenge to avoid conflict or duplicate numbers in the other half of the layout.

It will be a pain to have to go through all my records for CBUS and JMRI ( paper and PC files) to implement the changes but in the long run will be worth it.
Fortunately the temporary and too small station mimic panel needs a rebuild anyway so revised fascia legend is a doddle.

I will show the revised numbers on the Anyrail diagram and the lever table soon for further comments.

I have a scratch built 21 lever frame in fairly heavy brass which is more robust then the Society etched one, and cheaper too ! I think it cost me £30 many years ago. It has full mechanical interlocking for some other station which can be fully disabled initially and rejigged later for Dartmouth when time permits. It doesn't have microswitches yet but looks practical to add them...essential for Dartmouth's CBUS.

The tracks run north from Dartmouth towards Totnes. ( RHS of diagram ). The SHR turns left under the T&DR passing below Olton Bridge as mentioned in my workshop blogs. That enables it to turn West towards South Hams and Plymouth.

The "signal box" location is at right hand end of the island platform so facing East (preferred) or West ( to be decided) onto the UP departure line or the dual purpose lower platform face.
The taper of the platform is much wider than the Any Rail diagram shows.
The model lever frame will be at the bottom of the ANYRAIL diagram, so in the River Dart and facing West. The numbering function ascent will be chosen as recommended above, to accomodate East/West facings .

The sig box will be on stilts so still allowing use of the whole platform length. Thus provision is made to add a locking frame below if the Directors so decide, possibly reducing platform access for passengers.

[bécasse]

I assume Becasse is French so Jingoism to the fore; what can a Frenchman possibly know about early GWR practise?

Actually Bécasse is a Belgian, having successfully escaped from the country where he was born and where lived for the first 65 years of his life.

Having been one of the small team half-a-century ago that produced one of the first P4 layouts to hit the exhibition circuit, he also appreciates the importance part that historical research plays in creating a believable layout. That first P4 layout was an accurate depiction of a prototype in a particular year, fortunately nowhere near as far back as the 1870s, an experience which created a useful discipline for anything one did later.

Signalling is a particular issue. When I was young it was commonplace to see layouts where the signalling was just ignored, either not there at all or non-working dummies. Now the average British layout tends to be festooned with far more, and far more modern, signalling than the imagined prototype would ever have been. I suspect that the preserved railways, which tend to be over-signalled (albeit for good reason), have a lot to answer for, together with an unfounded assumption that what can be seen on the ground today was much the same as it was fifty-plus years ago when steam finished. Incidentally that P4 layout had no signals, they had been removed by the Southern Railway as an economy measure, but it did have the original signal-box and frame, by the time depicted by the model unlocked by the tablet and just used to work the points (and we did install accurate point0-rodding).

Ultra fine scale modellers, whether 2FS, P4 or S7, tend to be an exception to the generality even if their knowledge of signalling rarely quite matches their knowledge of thee finer points of locos and rolling stock. At least they usually know who to ask, although knowledge of the dark ages of railway signalling is inevitably a bit thin on the ground. Not only were none of us alive then but none of us ever had the chance to talk knowledgeably with people that were alive then, so the only knowledge available is from the careful interpretation of contemporary reports, maps, drawings and photographs.

I found it rather sad, therefore to find that Paul was proposing signalling arrangements that would only have become appropriate several years after the abolition of the broad gauge in 1892 as the GWR resignalled in accordance with the obligations imposed retrospectively by the 1889 Act, the earlier 1873 Act having only effectively imposed obligations in respect of new works, while before that "cowboys" ruled supreme. Some railways were at the forefront of signalling modernisation, almost leading the BoT on, the London & South Western Railway was one of them (quite possibly because of its close links with the military), the Great Western Railway, and its protegé in South Devon, wasn't.

The South Devon Railway had three branch lines in south Devon - to Kingswear, to Moretonhampstead and to Ashburton - and fortuitously the National Library of Scotland gives us the opportunity to look at all three of them in the broad gauge era, they are reproduced below. Most interesting is Brixham Road, later Churston, in the mid-1850s before the line was extended to Kingswear. 25" OS maps are not always quite as accurate in their depiction of railways as they might have been but what these maps show accords well with descriptions of the GWR approach to signalling in the southern part of its network and on the SDR in particular. The maps only show signal posts which tell us nothing of the type of signal but descriptions suggest that the disc and crossbar ruled supreme in this area until the broad gauge was abolished. The d & c signals were popular with the BoT because they gave clear stop (crossbar) and go (disc) indications. Some parts of the GWR especially narrow gauge lines started to use three aspect signals in the 1860s which could indicate stop (arm horizontal), caution (arm at 45°) and clear (arm hidden within post), stop and caution displaying a red light at night, clear a white light. These don't seem to have been used in Devon and certainly not on branches, and their use was strongly discouraged following the Abbott's Ripton accident on the GNR in the 1870s.

These maps show that Brixham Road and Ashburton possessed only a single home signal (which would have been the down line form of the disc & crossbar signal) while Moretonhampsteaad, at least by the mid-1880s, also had a starting signal (which would have been the up line form of disc & crossbar signal). None of the stations possessed signal boxes, which only came in the mid-1890s, and any interlocking would have been rudimentary, not dissimilar to the more modern point detectors on signal wires. Signals and points would have been worked, probably from their locations, by the station policeman (hence the use of the colloquial term "bobby" for railway signalmen).

So what does this mean for Paul's Dartmouth set in the 1870s. The layout is rather over-generous in its provision of double track, it's perhaps noteworthy that the real Kingswear across the river was never reached by more than its original single track, and that makes it necessary to provide a starting signal from the "departure platform" (separating arrivals and departures was commonplace, and not just on the GWR, at the period), otherwise home signals protect the station from arrivals on both the main line and the South Hams line - and that is it, no ground signals, no semaphore signals and no signal box.

signals-Dartmouth.jpg

map-BrixhamRoad.jpg

map-Moretonhampstead.jpg

map-Ashburton.jpg