Junction signal operation at a station

[G.M.R.]

I am trying to automate signalling on my layout and am currently working with a junction through station, my question is in regard to the operation of the signals on departure towards the junction. The station and junction have two running lines, up and down.

 

Currently I have things set up so that the main line through route is automatically signalled. This means that an approaching non-stop train will have a green aspect as it passes through the station whether the route set is straight or off the junction road. In the latter case, the feather will of course also be showing.

 

However, if a train is stopped in the station, things look different. If the train is departing straight on, then again green will be showing if the line ahead is clear, but if the train is to be sent along the junction route I am not sure what the procedure should be.

 

Because the junction crosses the down line, the junction route is normally set for straight on. This means that whilst the train is waiting in the station, the driver would see a green but no feather. When the route gets set for the junction, the feather would illuminate.

 

That sounds wrong. Would the signalman override the green for the straight ahead route and keep the station signal at red until he was ready to clear the junction route when green would show at the same time as the feather?

 

If so, my automation will have to include something clever to simulate the signalman's override.

 

I've tried to find documentation on this question, but no luck so far, so any help gratefully received.

[Del]

From what I recall of my S&T training many years ago, the proceed aspect and the route indication would be given at the same time. I'm pretty sure if a driver was to be routed over the junction, but had a green for the mainline and no feather, he would question the route with the signaller.

[TheSignalEngineer]

Era and location would have a bearing on the controls provided as would relative traffic levels on the routes from the signals concerned. Could you provide more information and a plan as this would help people making comments.

[SwissRailPassion]

In principle yes you will have to change the signal from red

The signalman would accept the train as far as the line is clear. So if the line was clear for straight ahead then a proceed aspect will show. The same applies to the junction, if the line is clear the route can be set and the signal will be switched to show proceed over the junction. That's the basic situation. However if the speed of the junction is restricted, which is very common, then the signal will show red until the train is close and has slowed. At a station the normal working would be to give the route when the train has almost stopped or stopped or is ready to depart.

So your signal should show red clearing to feather and proceed when the train arrives if the route is clear.

[Edwin_m]

The signalman may clear the signal before the train arrives, particularly if it is a minor station with little risk of the train being late departing, and there are no trains approaching that might conflict with it.  As mentioned above, if the diverging route had a lower speed then some form of approach control would apply and this might result in the signal remaining at red until the train was in the platform.  Details depend on era and on the permitted speeds over each route. 

 

For a more major station where there are driver changes or other sources of delay, there could be a Train Ready to Start button, operated by platform staff, and the signalman would normally not clear the signal until this was done. 

 

If the diverging route was needed but the signalman was not able to set it immediately the signal would remain red until the route was set and the conditions satisfied to clear the signal, when the feather would actually light a fraction before the change of aspect.  You would never see it changing from one route to another in front of a train, as the interlocking would force it to remain red for (usually) 2min in this situation. 

[beast66606]

The signalman may clear the signal before the train arrives, particularly if it is a minor station with little risk of the train being late departing, and there are no trains approaching that might conflict with it.  As mentioned above, if the diverging route had a lower speed then some form of approach control would apply and this might result in the signal remaining at red until the train was in the platform.  Details depend on era and on the permitted speeds over each route. 

 

 

Not necessarily, I've worked in at least one box where the lever controlling the colour light junction signal was locked until the berth was occupied.

 

Geoff - as well as Erics (SE) comments, do you envisage a PSB controlling your signals or local signal boxes

[G.M.R.]

Thank you for all the very helpful replies.

 

Here is the diagram for the entire junction, wonderfully drawn up for me several years ago by a member here. It is still in an old layout thread of mine in the archive.

 

 

The signals in question are numbered 36 and 37, but I've changed these two to coloured lights. The premise is that the era is mid fifties "ish" in an area which could see both Eastern and Midland practices and traffic. I'm have a mixture now of semaphore and colour lights, indicating a time of change.

 

Not necessarily, I've worked in at least one box where the lever controlling the colour light junction signal was locked until the berth was occupied.

 

Geoff - as well as Erics (SE) comments, do you envisage a PSB controlling your signals or local signal boxes

All still local boxes. You can see the box in the above diagram. It's the South Box ( actually it is the other side of the one on the layout ).

 

I have nearly finished signals 54 to 62, which are all semaphore and shown in my "new signals for my layout" topic within this sub forum.

[G.M.R.]

Incidentally, although several years have passed, I have now got signals 2 to 9 in place on a gantry (again some semaphore, some colour light - as originally suggested by beast!); signal 27 as a semaphore is also in place with a second goods arm for a siding which has appeared off to its right; 20 to 23 are in place as coloured lights and feathers; and as well as the signals the question was about, all the other south box signals are in progress.

[TheSignalEngineer]

In such a location with all signals and points worked from the frame automatic working would not be provided. It would not be possible to clear the signal for the lower speed route until the train occupied the berth track circuit. 

 

Assuming that the main line is worked under Absolute Block Regulations, signals through the section would be cleared sequentially after a line clear is obtained from the next box. The South Box Home signal lever would be returned to the Normal position and the signal proved at Red before another Line Clear could be given to the North Box.

 

Eric

[G.M.R.]

Thanks Eric. My aim is to try to "automate" the signalman wherever possible, leaving me to just drive the trains. I can simulate what you are saying should happen by using the occupancy detector I have in the platform road, together with a single switch or dcc command to the automation electronics when I want a stopping train to leave the station.

[TheSignalEngineer]

I've had a look at the layout revived thread, should be no problem in automatically working the string of colour lights. They should clear in sequence, stepping up the aspect as the next one clears, then revert to red as the train passes.

 

When a lower speed route, calling-on or shunt signal is set the aspect would not clear until the berth track circuit is occupied.

[G.M.R.]

Thanks again for looking into my layout in detail Eric. As you say, the sequencing of the string of colour lights works well for the main circuit, however, I'm still not completely sure when the signal at the station exit would change to red aspect ready for a diverging train.

 

If a main line train has passed through the station, the aspects will go through the sequence of red, yellow and green, as the train continues into the distance. So now if a train approaches from the same direction ready to stop and then take the lower speed route, I now understand that the signal needs to return to red until the train is in the berth and the diverging route set.

 

I guess that I will need to simulate a signalman operation returning the aspect to red before the second train arrives, even though the through route is clear? Maybe just after the train has passed the North box?

[Edwin_m]

In this situation a signalman would have to reverse and normalise the lever for each individual train.  This would only be done when he knew the next train approaching would be taking that route - they wouldn't leave the signal displaying a proceed aspect on the offchance that the next train was taking the main route, then potentially have to replace it in the face of a train if it turned out to be headed for the branch. 

 

For automation purposes I suggest the junction signal should stay at red until a train (taking the main line) had passed the last green signal before it.  The signal would then clear for the main line allowing the intermediate signal(s) to step up, so that the main line train would not see any adverse aspects.  This assumes of course that no conflicting route is set. 

 

For the diverging route I agree the junction signal would not clear until the train was closely approaching it. 

[G.M.R.]

Thanks Edwin, that works well except that the preceding signal also has a yellow aspect which is currently programmed to show when the signal being discussed shows red. Maybe go back just a little and use a train entering North box's control being the trigger, so that south box would set the last of north's signals to green at the same time as showing green at the end of the station.

 

If holding at red due to the train stopping or preparing to diverge, then the last North signal would correctly show a yellow aspect.

[G.M.R.]

I've thought of a supplementary question in regard to the home/distant combined semaphores. For trains approaching on the down line, the signals are semaphore. The south box signalman will have cleared his home signal arm, but then what is the timing for the distant on the same doll? Has North already pulled his lever, so the south box distant goes automatically, or does the south box signalman also have to pull a lever?

 

At the moment, my electronics have both arms moving together, but I could delay the distant arm behind the home arm, if that is more realistic. Or am I getting carried away here?!

[G.M.R.]

Here is my post from 8 yrs ago which yielded the signalling diagram I'm using.

 

http://www.rmweb.co.uk/forum/viewtopic.php?f=5&t=16931&sid=544ff0a6a541fb42c6bfec54d7f4ea99&start=25

 

Just to show that the contributions at the time were all taken on board, and that I've stayed with it. Making these signals work as closely to prototype as reasonable is one aspect of the model that really interests me.

[TheSignalEngineer]

If the North signals were already Off right through the section the North signalman would pull his Distant lever. The Distant lever would work a 'slot' weight bar. The Distant arm is pulled Off by a weighted bar dropping when both the South Home signal and the North Distant weight bar are operated. If the North weight bar is already operated when the South Home lever is pulled both arms would move at the same time. When the South signalman replaces the Home levr both signals will return to the horizontal posituion together.

 

Ther's a picture here, albeit lower quadrants operated by push rods to the arms.

http://janfordsworld.blogspot.co.uk/2013/05/railway-signalling-in-britain-part-3.html

[micknich2003]

UQ Signal Slot.

 

[G.M.R.]

Thanks Eric. The penny has dropped!

[G.M.R.]

Thanks Mick. I have the picture in front of me as I purchased your excellent book a number of years ago. It has been my guide along the way. The order of pulling the two levers was what I was missing. Now that Eric has made it clear that the North box (B in your diagram) would pull his lever first, I can see that it is okay for both home and distant signals to move simultaneously. I haven't tried to model the slot mechanism and just rely on a motor directly operating the distant arm.