What's in the Box?

[mpeffers]

Hi all,

 

In 2019, I purchased a 3'x1' laser cut baseboard thinking I could use it to quickly get a layout up and running, trying out some new techniques and gaining some experience on the way without overthinking the concept too much. Naturally (and even with the extra time at home in the past year and a bit), things haven't panned out that way and there's certainly more to do with the layout than has been done. I'm also probably starting to overthink some aspects too...

 

Current case in point is I am close to completing my scratchbuilt signalbox and have reached the point where I am starting to consider what should be going inside, which has lead to a desire to have all the correct equipment and all the levers coloured correctly and in order (despite the fact the box faces in towards the backscene and the interior will be mostly obscured). After a while on the fence about this, my current thinking is that it might be a good exercise to learn more about the dark art of signalling, even if the ultimate visual impact is visible.

 

I have seen members very graciously produce signalling diagrams on request on here. In this instance, I've tried to have a go myself and am open to feedback on the bits I've got wrong. With the exception of lever numbering, which I've left off for now. I've tried to include any required context, but let me know if I've missed anything off.

 

The layout depicts Fossebridge station, roughly from the siding point (just off-scene) to the level crossing (marked as a red X in the diagrams below). This is my devised chronology for the site, including the line back to the branch junction at Chedworth:

 

Circa 1877 - The Railway Arrives

Those familiar with the MSWJR may recognise this section - I'm working in a slightly alternative universe where the Cheltenham Extension Railway (and Andoversford - Lansdown Junc.) was largely built by the East Gloucestershire Railway (as was originally intended) and operated by that company as an independent concern (which they almost certainly didn't intend). As such, the railway has reached Chedworth a little earlier although through running between Oxford and Cheltenham won't commence until 1881. Not much to see her signalling wise.

 

Circa 1882 - Northleach Branch Connected

Smarting from their exclusions from the original EGR route, the folk of Northleach and Burford formed their own railway company to get their settlements on the rail map. The first priority was to connect onto the EGR, with a short branches connecting Burford to Witney and Northleach to Chedworth. In the event of the EGR proving a profitable success (possibly getting sucked in to the EGR's promotional material), the two branches could be connected to form a more appealing through route (sidelining Fairford and Lechlade).

The Northleach Branch was just over five miles in length, with an intermediate station at Fossebridge. Although the EGR at Chedworth and Northleach are at similar elevations, the land drops steeply away inbetween. To avoid the worst of the earthworks and a large viaduct, the branch followed the EGR route south for around half a mile as two parallel single tracks before divergining and desceding about 30m at 1 in 100 into the Coln valley. The EGR made what I assume to be almost minimum Board of Trade provision for the junction - a signal box and loop - and the Northleach branch itself was initially worked as 'one engine in steam' with no signalling other than one signal to protect the junction.

 

Circa 1890 - Northleach Branch gets Signalled

After several years of through running on a shoestring, the EGR started to consider making alterations and improvements. One of these was to provide signalling on the Northleach branch (although by no means busy, I assume a trip freight could occupy a 5.5mi branch operating on one engine in steam for a significant chunk of time and their timings were likely to be more irregular than other traffic). Fossebridge gained a signal box and goods loop, with the crossover set back inside the platform so as not to impact the crossing.

 

Circa 1895 - EGR Mainline is Doubled

With continued punctuality issues and grumblings from the MSWJR and GWR (with running powers over the EGR from Cirencester and Andoversford to Cheltenham, respectively), the EGR agreed to double the line between Cirencester and Cheltenham. Chedworth gains a second platform, but the twin single track arrangment is swept away and the Northleach branch junction is moved to the point at which the two routes diverge (now Chedworth Junction). To economise, the Chedworth Signal Box is moved south to the junction, with Chedworth's (rarely used) siding lifted.

To simplify the Chedworth signaller's workload, the branch is also double-tracked between the junction and Fossebridge. A short second platform is added at Fossebridge, although it is only used when trains are required to cross (I assume they will be timetabled not to do so, but the operational flexibility was considered worthwhile) as the existing platform is longer and has the waiting room. The main loop now extends over the crossing, but the former goods loop cross-over is retained (I'll concede this seems unlikely - best excuse I can offer is that it supports attaching wagons from the siding to the rear of mixed trains, which would likely have been used at times in the branch's history).

 

During WW1 - Fossebridge Becomes a Terminus

The contrivance continues. This change is essentially driven by the fact I only have one fiddleyard, but the best I can do is suggest that a major earthworks failure during WW1 resulted in the closure of the line beyond Fossebridge, with the wartime economy and general unimportance of the Northleach traffic not supporting reinstatement. Fossebridge was retained as a terminus, acting as a railhead for Northleach (with road transport used to bridge the remaining 3.5mi) and the nearest convenient point at which trains could be turned without also reducing the level of service at Chedworth.

 

Signalling Diagram

So this is my stab at signalling the 1895 configuration. The signalbox was based on the MSWJR boxes at Withington and Foss Cross, which were provided by Duttons of Worcester. As Fossebridge Signal Box would be installed at a similar time in a similar part of the country, I'm assuming the same suppliers would have been used. If practices varied between companies, the EGR's nearest 'ally' is likely to be the Midland Railway (the GWR were heavily against; LNWR likely uninterested) so I'd expect it would lean that way operationally.

 

Trains from Chedworth towards Northleach are straightforward - all pass through the original platform line; signals are distant > home > starter. I'm assuming no advanced starter is required as the middle crossover can facilitate shunting of the siding without passing the starter (or crossing the level crossing).

Trains from Northleach towards Chedworth can be routed two ways: 'wrong line' through the original platform then over the leftmost crossover or over the rightmost crossover and through the new, shorter platform. To signal these moves, the signals are: distant > 'junction' home (to indicate route) > starter (one for each platform) > advanced starter.

 

I have assumed fixed distant signals in both directions due to the level crossing (manually worked, so not interlocked with the signalling?) and as all trains will have to slow for a token exchange. The three blue lines denote FPLs. Assuming the cross-overs were worked by a single lever, I make this to be 14 levers (7 signals, 4 points, 3 FPLs). This seems good because it's the same size as Withington's frame (so the box is the right size); but also doesn't seem like an unreasonable number to have been installed when the box was new in 1890 (with a few spares) which I guess would be critical to any credibility for the middle cross-over being retained.

 

Outstanding Questions

• Other than levers, what else needs to go in the box? Currently I'm thinking stove, chair, table, token machine for Fossebridge - Northleach single line and 2x bells for communication with Chedworth Junction and Northleach. Anything else? Storage? Would telephones be common in fairly remote boxes by the late 1910s/early 1920s?
• I've not accounted for a trap point on the siding at the moment. Were they always required? Although the 1 in 100 will ease off significantly before the station area, it in reality would probably still decend from left to right as the branch continued towards the Coln so would this adverse gradient (say 1 in 400?) be sufficient to guard against wagons rolling out of the siding and fouling the running line?
• Similarly, I've not accounted for any catch points protecting the station/level crossing from anything running away down the 1 in 100 gradient. Is this reasonable? The railways were a relatively mature system by the final track alterations in the mid-1890s, by which point uncontrolled runaways should have been relatively rare...?

 

I've rambled a lot. I hope it made some sort of sense and thanks to anyone who read through the whole lot, as well as for any pointers/comments/questions its raises.

 

Cheers,

Matthew

Edited July 28, 2021 by mpeffers
Errant drawing removed

[mpeffers]

Layout shots included from the phone:
 

 

[Stephen Freeman]

Springside do Signal box interior kits, probably best to ask them which one would be best as they do 2 different sizes.

[Nick C]

12 hours ago, mpeffers said:

Signalling Diagram

So this is my stab at signalling the 1895 configuration. The signalbox was based on the MSWJR boxes at Withington and Foss Cross, which were provided by Duttons of Worcester. As Fossebridge Signal Box would be installed at a similar time in a similar part of the country, I'm assuming the same suppliers would have been used. If practices varied between companies, the EGR's nearest 'ally' is likely to be the Midland Railway (the GWR were heavily against; LNWR likely uninterested) so I'd expect it would lean that way operationally.

 

Trains from Chedworth towards Northleach are straightforward - all pass through the original platform line; signals are distant > home > starter. I'm assuming no advanced starter is required as the middle crossover can facilitate shunting of the siding without passing the starter (or crossing the level crossing).

Trains from Northleach towards Chedworth can be routed two ways: 'wrong line' through the original platform then over the leftmost crossover or over the rightmost crossover and through the new, shorter platform. To signal these moves, the signals are: distant > 'junction' home (to indicate route) > starter (one for each platform) > advanced starter.

 

I have assumed fixed distant signals in both directions due to the level crossing (manually worked, so not interlocked with the signalling?) and as all trains will have to slow for a token exchange. The three blue lines denote FPLs. Assuming the cross-overs were worked by a single lever, I make this to be 14 levers (7 signals, 4 points, 3 FPLs). This seems good because it's the same size as Withington's frame (so the box is the right size); but also doesn't seem like an unreasonable number to have been installed when the box was new in 1890 (with a few spares) which I guess would be critical to any credibility for the middle cross-over being retained.

 

Outstanding Questions

• Other than levers, what else needs to go in the box? Currently I'm thinking stove, chair, table, token machine for Fossebridge - Northleach single line and 2x bells for communication with Chedworth Junction and Northleach. Anything else? Storage? Would telephones be common in fairly remote boxes by the late 1910s/early 1920s?
• I've not accounted for a trap point on the siding at the moment. Were they always required? Although the 1 in 100 will ease off significantly before the station area, it in reality would probably still decend from left to right as the branch continued towards the Coln so would this adverse gradient (say 1 in 400?) be sufficient to guard against wagons rolling out of the siding and fouling the running line?
• Similarly, I've not accounted for any catch points protecting the station/level crossing from anything running away down the 1 in 100 gradient. Is this reasonable? The railways were a relatively mature system by the final track alterations in the mid-1890s, by which point uncontrolled runaways should have been relatively rare...?

 

I've rambled a lot. I hope it made some sort of sense and thanks to anyone who read through the whole lot, as well as for any pointers/comments/questions its raises.

 

Cheers,

Matthew

 

Do you actually need to be able to run Chedworth-bound trains through the original platform? Not doing so saves 2FPLs and 2 signals, which is fairly significant. I think bi-directional signalling at small rural stations was quite rare.

 

The crossing would be interlocked - there would be a gate lock controlled by a lever in the frame, possibly also a second locking the pedestrian wickets. I think the starter should be before the crossing, so that the crossing is protected when a Northleach-bound train is in the station.

 

You will need a trap on the siding, I think that was a BoT requirement? 

 

In terms of box equipment, above the frame will be a block shelf holding the block instruments, and the diagram will be above that. There will be a desk to hold the train register, usually with storage underneath and a lamp above. There would have been a telephone - at that stage I think a 'code phone' providing an omnibus circuit amongst several boxes. A stove to keep the box warm in winter, and storage - lockers for the signalmen and a cupboard to hold cleaning equipment etc. Often also a bookshelf for rule books, working timetables etc. Regarding token machines, you'll need to decide what the company would have used - I think Tyers key tokens were quite unusual outside of the GWR at that stage, so more likely either a tablet or the more traditional staff-and-ticket (which many lines kept right up until closure)

[5BarVT]

I can see you’ve done a lot of thinking about this!  Good stab at a back story.

Trap for the siding would be a requirement irrespective of gradient.  Add a dummy for model purposes, no extra lever as worked with the siding point.  Catch points also a requirement (I think) for 1:100, but won’t be linked to the signalling.  To allow shunting, it would be beyond the Advanced Starting signal so off your layout.

You will need extra levers for shunting discs for the siding and crossovers, as you appreciate, retaining the crossover fouls things up a bit but you need it for your reduced layout.

How would you feel about a modified back story that didn’t extend the loop across the level crossing and kept it as goods only?  It might mean a bit of alteration to the goods platform ramp and signalbox but could give a better overall lever allocation.

Paul.

[bécasse]

43 minutes ago, Nick C said:

 

Do you actually need to be able to run Chedworth-bound trains through the original platform? Not doing so saves 2FPLs and 2 signals, which is fairly significant. I think bi-directional signalling at small rural stations was quite rare.

 

 

By the time that the station became a terminus it was a Board of Trade requirement that two platforms were provided at termini of double-track branches - so both platforms would have had to be retained and signalled even if turn back facilities were provided from the arrival platform.

[Nick C]

1 minute ago, bécasse said:

By the time that the station became a terminus it was a Board of Trade requirement that two platforms were provided at termini of double-track branches - so both platforms would have had to be retained and signalled even if turn back facilities were provided from the arrival platform.

Interesting - presumably then you'd also need a facing crossover to allow for arriving trains to get into the second platform?

[bécasse]

3 minutes ago, Nick C said:

Interesting - presumably then you'd also need a facing crossover to allow for arriving trains to get into the second platform?

No. The requirement was merely to provide two platforms, they didn't both have to accept arrivals. 

 

I would foresee two possibilities at the time of conversion to a terminus. Firstly, if the box was tight on levers, almost no alterations would have been made, so the train would arrive, be run round using the loop, and then shunted to the other platform ready for departure. Alternatively, turn back facilities would have been provided from the arrival platform which would have required a new fpl on the crossover (which was facing to a train departing from the arrival platform) and a running signal to cover the move - thus requiring two new levers.

 

My suspicion is that the former is more likely to have been the solution at the time and that any later (and probable) "economies" would have singled the whole branch. Retaining the former would certainly make operation of the model more interesting and your historical background isn't that far-fetched.

[The Stationmaster]

15 minutes ago, Nick C said:

Interesting - presumably then you'd also need a facing crossover to allow for arriving trains to get into the second platform?

No, the arriving train could be shunted via a trailing crossover and you would in any case need two crossovers (or one crossover and the double to single line connection) in order to run round.  Having a facing crossover for an arriving  trains in such situations at small stations would be very unusual.

Edited July 29, 2021 by The Stationmaster

[mpeffers]

Thanks all for your contributions. Hugely informative.

 

7 hours ago, Nick C said:

Do you actually need to be able to run Chedworth-bound trains through the original platform? Not doing so saves 2FPLs and 2 signals, which is fairly significant. I think bi-directional signalling at small rural stations was quite rare.

Even as I was sketching out my diagrams, it felt tenuous. The justification in my mind is that the original platform is around twice as long as the new one and had all the amenities, although the station building/waiting room being adjacent to the crossing probably makes the difference in distance negligable. The new platform has 112' of usable length, which conveniently does just about accommodate the doors of my adopted non-bogie branch set (luggage brake - third - comp. - brake third) if the driver nails his stopping point. Strengtheners would break this, and trailing loads would risk fouling the crossing. Maybe this is sufficient though?

 

8 hours ago, 5BarVT said:

How would you feel about a modified back story that didn’t extend the loop across the level crossing and kept it as goods only?

It sometimes crosses my mind, if only because a set of crossing gates to cover twin track at PECO's 00 gauge track spacing would need to be wide. I've not mocked them up yet but I sometimes wonder if they will look a bit silly. It would essentially kaibosh ever 'reopening' the Northleach side if I ordered another fiddle yard and opened up a hole in the right-hand end (the stopping point for trains in either direction would end up being in the fiddleyard) but, to be honest, 3 scenic feet seems like quite a poor return on a 9' long layout so I probably won't bother anyway.

 

7 hours ago, Nick C said:

The crossing would be interlocked - there would be a gate lock controlled by a lever in the frame, possibly also a second locking the pedestrian wickets. I think the starter should be before the crossing, so that the crossing is protected when a Northleach-bound train is in the station.

Interesting to know - even in quiet backwaters? I think I read somewhere that the red disk on the crossing gates was to be interpreted as a de facto stop signal by engine crew, so assumed that the positioning of the starter could sit either side. A quick Google survey of preserved stations I can think of with adjacent gated crossings suggests that you're correct though. This is important as it would be on scene!

 

7 hours ago, 5BarVT said:

Trap for the siding would be a requirement irrespective of gradient.  Add a dummy for model purposes, no extra lever as worked with the siding point.  Catch points also a requirement (I think) for 1:100, but won’t be linked to the signalling.  To allow shunting, it would be beyond the Advanced Starting signal so off your layout.

Ok, that makes sense. I have some C&L chairs still somewhere so I should be able to knock something up. I think I was trying to save a lever in this exercise, but as you point out, it doesn't matter either way... If not linked to the signalling, does that mean the catch point would be sprung and on the 'uphill' line only?

 

8 hours ago, 5BarVT said:

You will need extra levers for shunting discs for the siding and crossovers

Shunt signals have proven the limit of my understanding of signalling so far. I guess I was hoping with a loop and one siding, there were sufficiently few available movements that they might not be required? Going back to Withington, the 1902 signalling diagram has a ground signal at each end of the goods loop but they don't have a corresponding lever number. I believe they were operated by the crossover lever and just showed which way the points were set (siding or mains). Not sure how common this arrangement was, though? Obviously saves levers, although Withington had three spares so could have accommodated them.

 

7 hours ago, bécasse said:

I would foresee two possibilities at the time of conversion to a terminus. Firstly, if the box was tight on levers, almost no alterations would have been made, so the train would arrive, be run round using the loop, and then shunted to the other platform ready for departure. Alternatively, turn back facilities would have been provided from the arrival platform which would have required a new fpl on the crossover (which was facing to a train departing from the arrival platform) and a running signal to cover the move - thus requiring two new levers.

There might have been a spare lever or two (or levers for the mothballed section of the line beyond the level crossing could be repurposed), although I guess these changes would require alterations to the interlocking even if the levers were available and maybe that would be sufficiently onerous to put the company off? It would have been operationally interesting to use the new platform a departures platform and shuffle the stock around. Sadly my fiddleyard won't accept it (currently one-third siding point, two-thirds sector plate).

 

Plenty to consider here. I'll get a pen and paper out when I get the chance and try and craft this knowledge into some definitive decisions. I do appreciate all the contributions here - aware that I'm trying to overlap some prototypical theory onto a mildly nonsensical scenario to justify what I've already created...

 

A Slight Aside

Following a train of thought I had, could a passenger train on a single line branch pass a goods train to the right (i.e. 'wrong line') if the goods train was in a loop not signalled for passenger traffic? Or did all opposing direction passes have to occur to the left? I envision a single morning and evening passenger train continuing through to Cheltenham (it'll vary the stock a bit, but maybe some of the rural folk would take to commuting), but the rest of the passenger service will be provided by a single set, shuttling between Andoversford and Fossebridge/Northleach so it's unlikely both platforms would ever be occupied at the same time. I'm just wondering what operational constraints downgrading the loop to goods only (either with it continuing over the crossing or not) would impose if a Chedworth-bound passenger train met a Northleach-bound goods.

[mpeffers]

Also, this is the box:

I thought I had a picture of it to include last night, but I did not. Doesn’t add anything to the scenario (maybe a maximum frame size?), it’s just I only recently managed to cut plasticard square and I’m feeling unduly proud of myself.

[5BarVT]

55 minutes ago, mpeffers said:

I think I read somewhere that the red disk on the crossing gates was to be interpreted as a de facto stop signal by engine crew,

Correct.  Probably still are some on Network Rail, certainly were 5 years ago, but they only had a level crossing.  As soon as you put points in and signals for the points, then the gates will be interlocked so that if the signal is cleared the gates must be open to rail too.

57 minutes ago, mpeffers said:

A Slight Aside

Following a train of thought I had, could a passenger train on a single line branch pass a goods train to the right (i.e. 'wrong line') if the goods train was in a loop not signalled for passenger traffic?

Most definitely Yes!  But by the time the line is doubled that is no longer an issue as the goods would wait on the the double line until the passenger has left the platform.  Goods could still be shunting while the passenger runs through in the other direction.

 

59 minutes ago, mpeffers said:

Going back to Withington, the 1902 signalling diagram has a ground signal at each end of the goods loop but they don't have a corresponding lever number. I believe they were operated by the crossover lever and just showed which way the points were set (siding or mains). Not sure how common this arrangement was, though?

More common that long ago.  May well have been upgraded with the alterations when the line was doubled (but you’re in charge of the back story and Helston kept one until 1958).

 

So, if it works for you, how about:

Single line with passenger platform and goods ‘loop’ with siding off.  After doubling, you would need a separate trap protecting the passenger crossover but still only three point levers as the siding would be hand points.  Two FPL, one for the goods xover and one for the passenger.

Signals would be (at doubling, assuming Down from Junction , Up to junction):

Down - fixed distant; home (1/4 mile from first xover so train can be accepted whilst another approaching on single line); inner home protecting first xover (and second too); starting on end of platform at level xing.

Up - fixed distant; junction home reading to platform and goods yard; starting on platform; goods starting at trap points; possibly Advanced starting beyond the crossover.

Shunt signal to come back through the passenger xover; shunt to come out of yard towards Dn starting.

A lever to lock the gates across the road (lever reverse allows the gates to be opened). (Make the gates hand operated so you don’t need space in the box for a wheel.)

16 levers on 4” centres, less than 6’.

Then after closure of the single line, a disc signal (or two) could be put at the toes of the run round points still operated by the levers for the junction signal.

Colours: points black, FPLs blue, signals red, gate lever brown.

Paul.

[The Stationmaster]

I agree fully with Paul's ((5BarVT) post immediately above with one possible exception and that is in respect of the distant signals because they might be worked distants and not fixed at caution.  In the  pre 20th century days distants fixed at a caution seem to have been relatively uncommon and even in the 20th the GWR was rather unusual in having them at crossing stations on single lines.

 

Incidentally operationally there is little real need for an Advanced Starting Signal at a single line crossing station - they were no more than a belt, braces, and more, precaution and could really be only justified in special circumstances.  You could also avoid the Outer Home on the double line approach by the use of Special Instructions (probably by using a Single Line Crossing station Clearing Point instead ofa double line Clearing Point) and there were definitely places without them on relatively busy stretches of railway where lines changed from double to single.

[mpeffers]

22 hours ago, 5BarVT said:

More common that long ago.  May well have been upgraded with the alterations when the line was doubled (but you’re in charge of the back story and Helston kept one until 1958).

The S-R-S also has the 1943 diagram for Withington (I knew I'd seen it somewhere), with the point indicators still present. This was when the loop was extended to accommodate wartime traffic and I'm not aware of any subsequent changes, so they were likely there too until the box closed in 1956. All 14 levers are in use by this point, so this figures. I'm not certain if they were retained from installation, or removed when the line was singled in 1928 and reinstated in 1943.

 

I have had a go at drawing up the following, with some amendments suggested by The Stationmaster (although just noticed I have left the Outer Home in). Hopefully correctly interpreted.

22 hours ago, 5BarVT said:

Single line with passenger platform and goods ‘loop’ with siding off.  After doubling, you would need a separate trap protecting the passenger crossover but still only three point levers as the siding would be hand points.  Two FPL, one for the goods xover and one for the passenger.

Signals would be (at doubling, assuming Down from Junction , Up to junction):

Down - fixed distant; home (1/4 mile from first xover so train can be accepted whilst another approaching on single line); inner home protecting first xover (and second too); starting on end of platform at level xing.

Up - fixed distant; junction home reading to platform and goods yard; starting on platform; goods starting at trap points; possibly Advanced starting beyond the crossover.

Shunt signal to come back through the passenger xover; shunt to come out of yard towards Dn starting.

A lever to lock the gates across the road (lever reverse allows the gates to be opened). (Make the gates hand operated so you don’t need space in the box for a wheel.)

 

1895

 

1920

 

Catch and trap points added, along with ground signals (hopefully in the correct places). I've also converted the line up to the box into a second siding - possibly mostly because I'll want to use it as such on the model but also in the scenario, Fossebridge will also handle Northleach's goods do some improvised additional capacity might have been welcomed. On a similar note, I feel like I want to keep the platform/dock for the visual effect. Leaning towards placing cattle pens at the Chedworth end (I assume they were positioned in accessible parts of yards for swift onward transit). Milk seems like the other plausible regional traffic that could be loaded from a platform (perhaps a wagon or two left overnight to attach to the morning Cheltenham train?).

 

The signal box is probably poorly located for this scenario, and would be better placed on the platform. This arrangement maybe hints at future intent (or at least provision) to extend the double track through to Northleach though. I think I'm happy with what we've proposed?

 

 

 

Edited July 30, 2021 by mpeffers
Formatting