Model railway signal box interior advice

NSW would put ground frames in for infrequently used points, or points too far from the box when another box wasn't justified.

So that makes me wonder what kinds of distances are imagined here. There could be some very heavy pulls in both directions, and likely the points on the hollow track in the sidings would be thrown by hand by a shunter or someone on the ground.

That level crossing is also interesting, I would have thought the derails were better placed to the right of the crossing to better protect road/pedestrian traffic against a runaway off the sidings.

Also are the points motorised or are there going to be FPLs for every facing move on the main line? That would increase the number of levers.

May I just check: is the layout based on Australian practice?

Lever layouts

Number of levers
This could be dramatically affected by the lever spacing, which varied very widely. The Midland Railway's 'tumbler' frame had levers at 6" centres, necessary to fit the mechanism in. Later, a 'tappet' variant was developed, again initially with 6" centres, but the REC/LMS/LMR developments of that reduced them to 4.5" centres. If I recall rightly the GWR's 5-bar vertical tappet frame had levers down to 4.25" centres, though I think older models (the 'double-twist' for example) were a bit wider. The Springside kit is definitely GWR (though I've never measured the lever centres accurately enough to tell what the spacing is). The Peco example looks like an uneasy amalgam of GWR and McKenzie & Holland (the latter particularly in the level crossing wheel). The LNWR used 6" centres, presumably because it was impossible to get levers any closer with the stirrup catch handle; their later habit of signalling everything and (usually) operating a crossover from 2 levers not 1 led to them having some very large boxes.

Some companies left plenty of room at the ends of their frames. The GNR was a notable example; in their case it often meant new frames could be installed controlling extended layouts without having to build a new box. Others were more parsimonious.

Some (the Midland and the LNWR for example) produced boxes and frames in standard sizes. The MR's were prefabricated at Derby and assembled like a big garden shed, using absolutely standard-sized 'flakes' delivered on a special wagon.

In the UK if there's a signalbox present then all the routes including the dock siding that lead to the mainline would normally be controlled from there

flabberdacks in post 159369 said:

NSW would put ground frames in for infrequently used points, or points too far from the box when another box wasn't justified.

The yard between arrival home signals is probably about 350-400 yards long.

flabberdacks in post 159369 said:

Also are the points motorised or are there going to be FPLs for every facing move on the main line? That would increase the number of levers.

When were motorised points introduced at Scottish mainline crossing stations (featuring about thirty to sixty trains per day)?

Andy174 in post 159372 said:

In the UK if there's a signalbox present then all the routes including the dock siding that lead to the mainline would normally be controlled from there

Would the dock siding be signalled directly, or worked via Annett's Key?

Further to that, would normal working here involve shunting out onto the single line to get from one road to another? That would probably necessitate what NSW would call a starting signal. Something a reasonable distance down the single line that can be held at stop to control shunting movements but cleared to authorise a train to depart the area towards the next interlocking.

Lever layouts
So far as setting levers out was concerned, in UK practice points/FPLs/etc tended to be clustered towards the middle of the frame, usually interspersed with shunting signals. Signals were at the ends of the frame and - so far as sensible - arranged so they could be pulled in a left-to-right or right-to-left sequence (then having to go back for the distant), including where there were junctions. An example of setting that out for a junction is visible for the up fast - up slow route at St Albans South https://signalbox.org/~SBdiagram.php/?id=%20150 shown as it was in 1970. (That was impossible to do, of course, on a 4-track layout where there were 'both-ways' crossovers.)

I've just found an interesting layout at Leamside (https://signalbox.org/~SBdiagram.php/?id=%201519). There's 55 levers, of which 35 are spares. The through moves, with no points movements, are 3,4,2, on the Up and 53, 52, 51, 55 on the Down. But if a train is reversing Up to Down, the move out of the Up platform is points 10, 8 and then signals 16, 51, which will be a long walk!

Your service levels are at the upper end of that seen on Scottish single lines, except maybe wartime north of Perth.

A single bidirectional platform is quite unusual though, but 'preferred' vs 'crossing' arrangements still exist - eg at Dunkeld and Dalwhinnie (HR) while WHL stations were mostly island platforms.

These would be single box setups as bidirectional was never popular with the regulators.

Anothersignalman in post 159373 said:

The yard between arrival home signals is probably about 350-400 yards long.

UK had rules about distances from lever to points, which did slowly increase. This resulted in longer loops needing two lever frames, although not necessarily staffed as full signal boxes.

You might find the Oban line website interesting as it contains fairly detailed, if Caledonian, layouts and boxes. Callander, Balquhidder, Killin Jn and Connel are the most relevant to you being the most complex through stations and generally better signalled.

Any siding onto a passenger line had to be signalled and locked, but the extent of in-yard signalling varied.

Anothersignalman in post 159373 said:

When were motorised points introduced at Scottish mainline crossing stations (featuring about thirty to sixty trains per day)?

Very few ever did, as many did not have mains power. Some still do not have motor points - eg Stanley.

Anothersignalman in post 159373 said:

Andy174 in post 159372 said:

In the UK if there's a signalbox present then all the routes including the dock siding that lead to the mainline would normally be controlled from there

Would the dock siding be signalled directly, or worked via Annett's Key?

Direct locked and signalled.
The '1935 modernised' Killin Jn layout on Oban-Line shows electric locks on some of the lever frame functions.
Balquhidder West shows a dock siding exactly as your layout.

bill_gensheet in post 159375 said:

Any siding onto a passenger line had to be signalled and locked, but the extent of in-yard signalling varied.

The dock platform isn't for passenger trains (or railmotors with passengers in them), it's for storage of a spare passenger carriage that could be used for the Mixed or to boost a train for holiday traffic, or to unload horseboxes.

bill_gensheet in post 159375 said:

Your service levels are at the upper end of that seen on Scottish single lines, except maybe wartime north of Perth.

Good to know. This, combined with the diagrams you referenced through the rest of your post, seems to indicate that any such station would have long-since earned a larger lever frame than 20-25 levers, and therefore the proposed building probably isn't suitable? Maybe I should ask if we can kitbash two kits together; 12x windows across the front should allow for up to 50 levers.

bill_gensheet in post 159375 said:

A single bidirectional platform is quite unusual though, but 'preferred' vs 'crossing' arrangements still exist - eg at Dunkeld and Dalwhinnie (HR) while WHL stations were mostly island platforms.

We're sticking with a single side platform due to the local terrain, though I could imagine that if a suburban electric service ever came through (and the yard was still necessary), the dock platform may well be extended around the back with a shotcrete retaining wall.

bill_gensheet in post 159375 said:

You might find the Oban line website interesting as it contains fairly detailed, if Caledonian, layouts and boxes. Callander, Balquhidder, Killin Jn and Connel are the most relevant to you being the most complex through stations and generally better signalled.

Would any of these have had similar traffic numbers, 30-60 trains per day?

bill_gensheet in post 159375 said:

Anothersignalman in post 159373 said:

Andy174 in post 159372 said:

In the UK if there's a signalbox present then all the routes including the dock siding that lead to the mainline would normally be controlled from there

Would the dock siding be signalled directly, or worked via Annett's Key?

Direct locked and signalled.
The '1935 modernised' Killin Jn layout on Oban-Line shows electric locks on some of the lever frame functions.
Balquhidder West shows a dock siding exactly as your layout.

Were plunger locks a thing in the UK?

Very few ever did, as many did not have mains power. Some still do not have motor points - eg Stanley.

Below an absolute minimum setup which does keep to 24 levers, graphics from copying bits from the oban-line site.
It has fixed distants, not well suited to a main line with non-stop trains, and has the mineral siding worked by ground frame with electric lock. (ideas from Killin Jn).
The shunt discs each end have a route display worked off the point lie (as Oban).

Anothersignalman in post 159381 said:

bill_gensheet in post 159375 said:

Any siding onto a passenger line had to be signalled and locked, but the extent of in-yard signalling varied.

The dock platform isn't for passenger trains (or railmotors with passengers in them), it's for storage of a spare passenger carriage that could be used for the Mixed or to boost a train for holiday traffic, or to unload horseboxes.

The signal is to prevent whatever is in the siding from getting onto the main line, not for passenger movements. Remember the rules applied to these layouts were designed with horse/human powered shunting in mind.

Intermediate block signals: Red over yellow

Great efforts were made to avoid 'pull-throughs' (where a lever had to be pulled between two others that were already reversed).

If that logic holds true, then the Peco box would have room for a maximum of 25 levers; and I'm not certain that's enough to work this sort of track layout, at least without considering mechanical selectors on the shunt signals. Was that a common practice in the UK?

Off the top of my head, the local limit was about 400 yards for signal pulls, but less for points.

bill_gensheet in post 159375 said:

Any siding onto a passenger line had to be signalled and locked, but the extent of in-yard signalling varied.

The dock platform isn't for passenger trains (or railmotors with passengers in them), it's for storage of a spare passenger carriage that could be used for the Mixed or to boost a train for holiday traffic, or to unload horseboxes.

What a splendid discussion you've got going here, although I'm still a bit unsure whether you're going for an Australian layout, a Scottish layout, or are you just generally interested in either or both?

I notice for example you use the term derails and the symbol does look more like that of a derailer. To me a derailer is a device that goes on a rail and, uh, derails a train. They are not really that common in the UK I would say, we would tend to use trap points (which the GWR tended to call 'catch points', although a common convention is that 'catch points' are the ones you use to catch runaways down a gradient and 'trap points' are for trapping trains in sidings, providing flank protection, etc.)

Anothersignalman in post 159373 said:

If that logic holds true, then the Peco box would have room for a maximum of 25 levers; and I'm not certain that's enough to work this sort of track layout, at least without considering mechanical selectors on the shunt signals. Was that a common practice in the UK?

Mechanical selection was certainly done.

Peco kit SS91 contains parts for 2x four lever and 2x two-lever ground frames; they can probably be kitbashed to make a five-lever machine, but four should be fine: up and down signals, point, and lock lever. The latter would be released from the main frame, and when reversed, capture the key as well as unlocking the points. Then, with points reversed, either the up or down shunt signal could be worked.

I note that you've shown the down end points sequence as signal-point-lock-point-lock; that's opposite to how I'd have expected it, because with out-both-ways lockbars that means you need to pull the points, then the lock on the right, then the signal on the left instead of running right-to-left directly.

I gather that local practice was far more strict (after c.1910) about placement of shunting signals than in the UK, so we'd have to provide two signals from the derails down end 3 and 4 roads. As an example, this locking sketch from 1902 shows shunting ground disc 4 - https://drive.google.com/file/d/1Yfuxvk7AETALPze-OmpBmU0ixmtd9bKO/view - between home signals 2,3 and mainline roads 1 and 2. I think that sort of mid-section shunt signal was common for UK practice, but that signal was likely removed circa November 1912. So, is it acceptable in UK practice to have a single combined shunt signal further along like that, or is it that the turnout 3/4 roads becomes a hand-throw, which is why you added the catch? We don't have room for a catch or derail in that position, but we could probably interlace catches with the hand-throw turnout.

I think the 24th lever needs to be a signal, you've shown 12 red levers for 8 homes and five dwarfs. Unless the dwarfs from dock and up end 3 road can be worked from one lever?

Could the distant signals in each direction be automated, working based on slots from the relevant home signals similarly to how a home-over-distant on a single post would have been locked?

kbarber in post 159371 said:

Lever layouts
So far as setting levers out was concerned, in UK practice points/FPLs/etc tended to be clustered towards the middle of the frame, usually interspersed with shunting signals. Signals were at the ends of the frame and - so far as sensible - arranged so they could be pulled in a left-to-right or right-to-left sequence (then having to go back for the distant), including where there were junctions. An example of setting that out for a junction is visible for the up fast - up slow route at St Albans South https://signalbox.org/~SBdiagram.php/?id=%20150 shown as it was in 1970. (That was impossible to do, of course, on a 4-track layout where there were 'both-ways' crossovers.)

I read somewhere that there's an alternative - less common but still found - where the signal levers are all put in the middle and the points etc. at the ends. If most moves are just straight through or just one or two lever pulls, there's much less walking around required.

<snip>