Understanding Signal Box Diagrams

[Halton Boy]

Hello everyone

This is really interesting. I am learning so much information. I am really grateful to you all.

I have produced a revised plan below:

So what is the point of all this?

Well two things, one is to learn how to read signal box diagrams. 

The other is to produce a model of a branch line station that no one else has modelled. That may not be possible.

If I can understand how the real station operated, then I can reproduce the station within the limits of a model.

No FPL levers and hand operated point levers on the end of the frame.

But it is important that I know what to include and what to leave out.

Thank you again for your help.

I assume that lever 1, when locked, locks levers 2 , 7 and 10 but not levers 3 and 8.

[Halton Boy]

Hello everyone

I am trying to find a signal box diagram for Coleford branch station on the Seven and Wye railway. Not the GWR station.

I have found some photo's and will try and build the station in Anyrail.

[Michael Hodgson]

Last time I was in Coleford, the GWR box was extant and on display at a fairly little-known museum

https://www.picturesofengland.com/

 

I assume you mean this box though ...

https://www.flickr.com/photos/trains-travel/44518484085/in/photostream/

Diagram available from the SRS...

https://www.s-r-s.org.uk/html/gwx/S2528.htm

 

 

 

[TheSignalEngineer]

On 22/01/2023 at 17:28, Jeremy C said:

C appears to be a token instrument. There is something odd about the shape, though, which I don't recognise but I am sure someone will explain.

Do you mean that it looks as if it has been hit by a train or the letter "K" inside it which denotes that it is a Key Token machine

 

10 hours ago, Halton Boy said:

What would the token machine in the station masters office be used for?

Was it the token for single line running?

The token machine in the SM's office would be an Auxiliary machine. It would enable the SM to withdraw a token instead of the signalman taking it. This means it can be given to the driver in the platform befor the train starts away rather than the signalman coming do to do it. More common at locations where the train doesn't have to pass the box such as the Blaenau branch at Llandudno Junction. Also used when the train draws up to a branch starter off the main line to await the previous train clearing the single line. I remember putting in an arrangement like that at Madeley Junction so that MGR trains to Ironbridge could draw into the loop on the branch to clear the main line. When the single line was clear the signalman would release the token for the driver to take it.

I think one of the furthest from the box to an Axilliary token instrument must have been after the Cotswold singling when the token for the single line at Norton Junction could be given to the driver at Worcester Shrub Hill about three miles away.

There were other instances of long distances but configured as "No Signalman" instruments where there was just a ground frame at the end of a branch and it was possible to to shut a train inside then let another train come through.

[Halton Boy]

Hello Michael

I mean this station:

 

I have only found a few very old pictures of the station. None of the pictures show the right hand end of the station past the water tower.

Here is a map I have drawn. The station is Coleford on the SWSB.

Here is an early attempt at the station in Anyrail:

This plan has the main elements but the tracks off to the right cannot be correct, and the station was on a curve.

[TheSignalEngineer]

The ground frame points would be 2a and 2b, as 1 is the release lever, which is locked normal by the signal box.

Originally Ground Frames were usually released by a key which had to be taken out of a lock on the lever frame in the signalbox. Removal of the key locked all signals going over the ground frame points. The key was then put into an Annett's Lock on lever 1 at the ground frame to release it from the normal position. The key could only be removed from the Annett's Lock with the lever 1 normal.  As the key was captive with lever 1 reverse that lever could also control the points. Trains could only be signalled over the ground frame points by the signalman when the key had been returned to its lock on his lever frame.

The lock on the ground frame lever could also be configured for operation by a key token or key fixed to a train staff for operating intermediate sidings on a single line.

 

Later it was decided that remote ground frames needed to be released without having to walk all the way to and from the box for the key. A mechanical device called a Midway Lock was devised. A lever in the signalbox pulled a rod which operated a lock half way to the ground frame. Then Lever 1 at the ground frame could be pulled as it was linked to the Midway Lock by a second rod and that action released points lever 2. (I must be one of the few remaining people to have worked a Midway Lock)

 

As ground frames became more remote a controller on the signalbox lever energised a relay circuit which enabled an electric lock on lever 1 at the ground frame to be released. 

 

The reason for having separate release and points levers is to prevent the signalman inadvertently taking to release back during a shunt move which requires the points normal. The release lever can however by used for an FPL, where reversing it withdraws the bolt from the lock stretcher. This can be done if the FPLis not required to be locked during any of the ground frame moves.

 

 

[Jeremy Cumberland]

19 minutes ago, TheSignalEngineer said:

Do you mean that it looks as if it has been hit by a train or the letter "K" inside it which denotes that it is a Key Token machine

The D shape outer.

 

28 minutes ago, TheSignalEngineer said:

The token machine in the SM's office would be an Auxiliary machine.

Need it have been? I thought the practice of token instruments being in the station building rather than the signalbox existed into the 1980s in parts of Scotland.

 

A lucky flash of memory and a quick check finds that this is still the case at Barrhill: https://www.southwestscotlandcrp.org.uk/stations/barrhill-station

 

2 hours ago, Halton Boy said:

I assume that lever 1, when locked, locks levers 2 , 7 and 10 but not levers 3 and 8.

1 will lock 7, and I might expect it to lock 2, but the absence of an FPL suggests that isn't necessary. I can't really see the need for it to lock anything else. @TheSignalEngineer will know, of course.

[Halton Boy]

Hello Michael

I have found a photo of Milkwall station and it has double track, but I do not know if it extends to Coleford.

[TheSignalEngineer]

Try this site which shows the layout in 1902, and you can see how the Severn and Wye was connected to the GWR line at Coleford.

https://maps.nls.uk/geo/explore/#zoom=16.7&lat=51.79195&lon=-2.61584&layers=168&b=1

 

Move down a bit and you come to Milkwall. It was just a loop with a connection to a mineral railway off it.*

https://maps.nls.uk/geo/explore/#zoom=17.4&lat=51.78047&lon=-2.60207&layers=168&b=1

Edited January 23, 2023 by TheSignalEngineer
* Sling branch?

[Halton Boy]

Hello everyone

"As ground frames became more remote, a controller on the signalbox lever energised a relay circuit which enabled an electric lock on lever 1 at the ground frame to be released. "

I am now sure that is the case at Penicuik station.

Can you tell me which signals would be locked when the ground frame was in use.

I have been told that stations in Scotland had the token instrument in the station masters office, not in the signal box.

If there was a single line token would that mean that only one engine would be on the line at anyone time.

It seams odd that there are signals on posts but no ground signals.

If it was one engine in steam then it seams like overkill with all the signals.

The more I think about this station the less I know. 

 

[TheSignalEngineer]

Another shot showing the trackwork in 1958 

 

 

Edited January 24, 2023 by TheSignalEngineer

[Halton Boy]

Thank you very much for the information.

I thought Coleford would be a small branch line station, but it is quite big with a long distance between the signal box and the platform.

My plan is not correct. 

It would need a lot of space to fit the whole thing in.

I looked everywhere for maps and photos.

Thank you so much.

[TheSignalEngineer]

7 minutes ago, Halton Boy said:

"As ground frames became more remote, a controller on the signalbox lever energised a relay circuit which enabled an electric lock on lever 1 at the ground frame to be released. "

I am now sure that is the case at Penicuik station.

Can you tell me which signals would be locked when the ground frame was in use.

Any signal worked from the box that could send a train over the ground frame tracks.

For convenience some ground frames had their own signals for some of the moves they controlled.

10 minutes ago, Halton Boy said:

I have been told that stations in Scotland had the token instrument in the station masters office, not in the signal box.

This is sometimes the case. I remember some stations where the 'box' was no more than a frame open to the elements and operated by a porter/signalman. The instruments were then inside the station building. The Bletchley - Bedfored line which is double track still had some of its signalling controlled by frames out on the platform to the early 2000s. 

I also did instances where the token was in a cabinet on the platform and remotely released for the driver by the signalman, and one in colour light signalling where it was released by a route being set and the train standing at the signal.  In that instance the signal back onto the main line couldn't be cleared to let the train off the branch unless the driver had put the token back in the machine when he got back to stop him from going off with it still in the cab.

24 minutes ago, Halton Boy said:

If there was a single line token would that mean that only one engine would be on the line at anyone time.

Only one engine (or two coupled) could be on the running line between loops or sidings. At some sidings an intermediate instrument was proided where a train could shunt inside then get the token out again when the section was clear. 

27 minutes ago, Halton Boy said:

It seams odd that there are signals on posts but no ground signals.

If it was one engine in steam then it seams like overkill with all the signals.

Fashions changed over the years and some lines operated under Light Railway provisions had only one signal for two or three routes from the same place.

If token worked then there could still be another train within station limits. Signals were provided for box worked points so that the signalman didn't have to keep going to the window or even outside to give a hand signal to the train crew. Detection on the points also ensured that they were properly closed before the signal could be cleared.

[TheSignalEngineer]

3 hours ago, Halton Boy said:

I assume that lever 1, when locked, locks levers 2 , 7 and 10 but not levers 3 and 8.

56 minutes ago, Jeremy C said:

1 will lock 7, and I might expect it to lock 2, but the absence of an FPL suggests that isn't necessary. I can't really see the need for it to lock anything else. @TheSignalEngineer will know, of course.

Sometimes it will come down to what moves can potentially take place at the same time, distances involved and visibility. 

It would be advisable for locking between 1 and 2 as there is nothing stop a freight arriving and being broken up there and a wagon getting pushed over the points then derailed. 

[Dagworth]

1 hour ago, Halton Boy said:

Thank you very much for the information.

I thought Coleford would be a small branch line station, but it is quite big with a long distance between the signal box and the platform.

My plan is not correct. 

It would need a lot of space to fit the whole thing in.

I looked everywhere for maps and photos.

Thank you so much.

Does this help?

 

https://maps.nls.uk/geo/explore/side-by-side/swipe/#zoom=17.8&lat=51.7918&lon=-2.6151&layers=206&right=osm

[Michael Hodgson]

I don't see any signals in those photos.

 

I would have thought it would have been worked under One Engine in Steam regulations so you would timetable freight to run in a long gap when the passenger train was elsewhere (only one or two services a day I believe) - or maybe even run a mixed train.  I don't know that it needed a signal box, I would have thought it sufficient to use a train staff normally kept at Coleford Junction signalbox, points connecting with the platform line (including Milkwall Sidings for the iron mine, Futterhill's Siding and the Sling Branch) all being locked by Annetts key attached to the train staff.  The hut in the background in the photo of Milkwall probably houses a ground frame.

 

From the way the map above is drawn, I imagine the connecton between the two companies at Coleford would have been operated as an exchange siding and only used for goods wagons.

[5BarVT]

11 hours ago, Halton Boy said:

I have been told that stations in Scotland had the token instrument in the station masters office, not in the signal box.

Highland Railway practice.  This was the arrangement at Nairn until late 90s.  There was a ‘signalbox’ at each end of the station which controlled the points and signals at that end.  In the station building were the token instruments (only one at Nairn since Inverness resignalling in late 80s) and slots on the two starting signals.

Paul.

[The Stationmaster]

20 hours ago, Jeremy C said:

I was looking for a picture like that one!

 

@Halton Boy are you intending to have working interlocking between levers?

For a model, you may as well use lever 1 for the run round points and trap. The prototype only has a separate ground frame to avoid the need for signals and so the person operating the levers can see what the engine is doing (and vice versa, in the absence of signals).

 

This leaves the four hand-worked points which, if you want a prototypical feel, can be controlled entirely separately (individual levers in line with the actual points, perhaps). Michael offers a number of good suggestions.

 

Personally, I wouldn't be inclined to include the loop siding point on 1 (or GF1) as this restricts how you might use the loop. You would not, for example, be able to bring a goods train into the loop siding and run round it if the back of the train was foul of the loop siding points.

 

Yes, and the whole idea seems bizarre. In a busy goods yard a ground frame might control points some distance away, and have signals too, so the person operating the ground frame can let the driver know the route has been set, but you wouldn't get this at a branch line terminus.

 

The four hand-worked points at this station would have been operated by levers immediately adjacent to each point.

 

Something of a tangle developing here I think.  Grovenor has already made the point that the freight would arrive at the platform and run round there.  That happened at the overwhelming majority of branch termini for two very simple reasons -

 

1. The platform line usually - as in this case - gave by far the longest length available for a train which was being run round which could be far more critical for freight traffic than the passenger train.  Plus of course if a train arrived in the loop as it is shown in this case it would be running towards an open trap point on a relatively short siding, not a sensible way to do the job.

 

And of course the signalling and Signalling Regulations were arranged in that basis at, again, the overwhelming majority of places.

 

2. The second reason is that traffic levels and train frequencies on most branch lines left a more than adequate gap in the passenger train service to run the freight trip and in any case ata small terminus like this it also needed that gap in order to be able to shunt the various sidings..

[Michael Hodgson]

1 hour ago, 5BarVT said:

This was the arrangement at Nairn until late 90s.

There was an article about it (in Modern Railways?) when the arrangements were modernised.

The lady signalman was famously provided with a bike to get from one end of the platform to other, as the sequence of operations required at the two boxes and at the station office for trains to cross meant a lot of moving from one place to another.  The system worked very efficiently when stations had plenty of staff, because the station master just had to send porters out to each of the boxes at the platform ends, but the clumsiness of the procedure became almost comical once they had single-manning of the station.

 

 

[The Stationmaster]

Interestingly the practice of having token/tablet etc instruments in the station building - officially in order that they would be operated by the Stationmaster - was severely criticised by Major Pringle, the HMRI who investigated the collision.  GWR records show that the Company took particular note of Major Pringle's recommendation that such instruments should be sited in the signal box and proceeded to relocate accordingly any machines which were not in the signal box.

 

Alas as a consequence of the Grouping the GWR suddenly found itself with considerably more instruments which were not in signal boxes - particularly on the former Cambrian Railway where the majority of instruments were in the station building.  So the pace of r4e-sting instruments suddenly slowed very considerably.

 

Incidentally talking about extreme examples the longest distance I can find between the controlling Signalman and a token instrument under his control is on one of  the WR's 'No Signalman Token (Remote)'  - NST(R) - installations where the signal box is almost 70 miles from its most remote token instrument.  I'm fairly sure the same route also boasts the longest token worked section - NST(R) in its case - on the national network at over 31 miles.  This was also previously the longest section on BR worked under Train Staff & Ticket (TST) Regulations

 

PS I think the former WR might also hold the record for the longest distance ( a few chains short of 94 miles) over which a token was over-carried instead ad of being handed over by the Drover to be replaced in an instrument.  It was probably the first, and hopefully only, time when a single line token went all the way to Paddington.

Edited January 24, 2023 by The Stationmaster

[Michael Hodgson]

27 minutes ago, The Stationmaster said:

Incidentally talking about extreme examples the longest distance I can find between the controlling Signalman and a token instrument under his control is on one of  the WR's 'No Signalman Token (Remote)'  - NST(R) - installations where the signal box is almost 70 miles from its most remote token instrument.  I'm fairly sure the same route also boasts the longest token worked section - NST(R) in its case - on the national network at over 31 miles.  This was also previously the longest section on BR worked under Train Staff & Ticket (TST) Regulations

The Esk Valley line is worked under NST(R) and WHitby is about half that distance from the controlling box at Nunthorpe.

There are instruments at Battersby (where trains must reverse), Glasidale and Whitby, and an intermediate instrument at Gromont for the NYMR connection.

[5BarVT]

7 hours ago, Michael Hodgson said:

The lady signalman was famously provided with a bike to get from one end of the platform to other,

Staying off topic (Sorry): the last place I saw a token exchanged on the move.  The driver was stationary and the signalman (then) cycled past on the bike collecting it on his arm as he went.

Paul.

[Michael Hodgson]

28 minutes ago, 5BarVT said:

Staying off topic (Sorry): the last place I saw a token exchanged on the move.  The driver was stationary and the signalman (then) cycled past on the bike collecting it on his arm as he went.

Paul.

Here's one that had just been collected a little further along the line.  December 2010.

[Halton Boy]

Thank you everyone this has been very interesting and I have learned a lot.

Georeferenced maps is my new friend.

These are the stations that I am looking at now:

Thornbury South Gloucester

Fairford

Burnham on Sea

Marlow Bucks

Cinderford

Nailsworth Midland branch South Gloucester

Dursley   Midland branch  South Gloucester

Chesham by Chesham park

Rickmansworth branch L&NWR

Watlington in Oxfordshire  South of Thame.

Stroud Midland branch

Once I have been through these I will take a look at Scotland's branch railways.

[RailWest]

On 23/01/2023 at 19:06, Michael Hodgson said:

Last time I was in Coleford, the GWR box was extant and on display at a fairly little-known museum

https://www.picturesofengland.com/

 

 

 

 

 

Err, not exactly....the SB seen at the entrance to the Museum is actually the structure from Cogload (near Taunton). Don't let the nameplate confuse you :-)