AFK signalling practice. A minefield of conflicting systems.

[ianathompson]

I do not usually post in the signalling section of the website but on this occasion I thought that I might as well describe my latest magnum opus.

I doubt that it will be of much interest to many regular readers as it follows continental practices, nominally at least, although there is also a large dose of freelancing.

 

 

Many spheres of influence fall within the remit of MaD, the AFK works.

In addition to the locomotive procurement and maintenance, and the commissioning of ferries, it is also responsible for the signalling equipment along the AFK.

Most NG lines, especially in Britain, did not have much in the way of signals.

The AFK follows the lead of the Roglasbanan and the Rhaetian railways, which, as quasi mainlines, possess full signalling.

It is easier to justify such provision in a continental environment than it is in Britain.

 

 

As has been noted elsewhere, AFK signalling practice is totally incredible, in the sense that it is not realistic.

Having a deep interest in the subject, the AFK has allowed me to deliberately incorporate elements of many countries procedures to produce a unique and unholy mixture.

Purists look away now!

Amongst other things the northern half of the line employs the archaic practice of green for “caution” and white for “clear”.

This practice ceased in Britain during the later nineteenth century when green was adopted to give a positive indication of “clear”.

The yellow for caution was adapted during the first decade of the twentieth century.

In France white remained as the “clear” aspect into the 1930s and was, as such, incorporated into colour light signals.

These often displayed so many lights in a proceed aspect display that French cheminots referred to them as “totalisers”, named for the displays used at contemporary greyhound stadiums.

The AFK colour light aspects at Fenditavalat and Lacono use such multi-light displays, although the aspects are ‘home brewed’.

 

 

The AFK colour light installation at Fenditavalat only comprises one the model.

White for “proceed” is also displayed in the semaphore installations at Ospicio and Ithilarak as does Caladonno in Narnia.

 They also use “green” for caution. Urteno, despite its importance, has lagged behind these locations both in signalling and scenic development.

It was now time to install the first signal of the Urteno colour light scheme.

After some consideration it was decided that Urteno should use searchlight signals, partly in homage to the Italian system although there would definitely be no reds whatsoever in the proceed aspects.

 

 

Initial attempts focussed upon the multicoloured LEDs available.

Certain of these can produce three colours although I understand that there are/used to be problems with the clarity of the three colours.

These are available in red/green/yellow and red/blue/white combinations, none of which suited my requirements.

After some deliberation, and consultation with a fellow signalman, I decided to employ mechanical searchlights for the Urteno scheme.

Whereas most prototypical searchlight aspects are projected by shining a light through an electrically activated coloured vane in the base of the signal there were certain examples where the vane was effectively a spectacle plate minus the arm.

I was aware that this practice was used in tunnel signals, where obviously the arm could not be seen, but I was surprised to discover that it was used in places where a normal semaphore could have been expected.

Perhaps someone more knowledgeable than my self could enlighten us as to why this was so.

 

In one final twist of complexity the Urteno installation, points and signals, is intended to be an electro-pneumatic system with the relevant equipment waiting to be modelled. Following this long pre-amble we can now move on to the prototype signal.

 

 

For many years the site adjacent to the station entry was occupied by this small equipment store, for use in replenishing the Pullman left in the adjacent carriage siding at the various times of the year when it runs.

Unfortuantely placing the signal to the right of the line, as is standard AFK practice, would place it in the middle of the public footpath that approaches the bridge.

The signal would have to go where the hut stood.

 

 

The hut was removed, no great hardship as it was loose in any case and the short, unelectrified, siding extension also went.

In one of the rare moments of foresight employed on the layout a feed and a return to the LED supply had been incorporated into the baseboard twenty odd years ago.

It was just a matter of discovering which was which hence the very temporary connections.

Needless to say the initial test of the signal on the bench had revealed that the two LEDs were wired opposite to one another so that one would light and the other would not.

 

 

A short time later and the signal was nailed down, lit, the operating rods connected and ready for testing.

In the manner of things AFK it is a crude rendition although it will receive a little more attention before it is considered complete.

Here it shows red for “stop”. The white LEDs shine through plastic sheet and the red is redder than the camera shows.

This was achieved by doubling up the plastic sheet to give clearer colours.

The red vane and the black vane are in the central positions although they do not automatically revert to this position “when the current is switched off”.

 

 

The wiring has been connected above the baseboard, which might seem barbaric.

The model stands on the middle level of three layout levels, however, so placing wiring under the board interferes with the level below which is Relforka at this point.

The single white indicates “clear, take the RT”.

The RT is the rektatrakoj in AFK terminology and means main route.

Yes I am aware that the vane has not fully cleared but this was not apparent from the normal viewing angle.

 

 

Green over white indicates “caution take PFT”.

The PFT is the passing loop and the train will diverge from the main route at the points just visible in the background.

The operating parts of the signals, and the heads, were constructed in plastic with a view to testing their efficiency and then creating metal replacements.

Laziness has won out. I am no master blacksmith when it comes to working metal so they will remain plastic unless they disintegrate.

 

 

I am not greatly happy with the projection of the pins holding the signal into the ‘ground’ so they might receive some attention in the near future.

Green over green indicates “caution, take KPFT”, this being the goods loop.

I use this term loosely as the locking permits passenger trains to enter the loop but gradually, across the years it has evolved to deal mainly with goods trains.

 

 

Now we come to the joker in the pack.

Rather like the “Ruhe” position of a Bavarian semaphore red over yellow indicates that the points are released from the locking and can be thrown by hand for shunting purposes.

Unfortunately I have introduced an inconsistency here in that the Bavarian example at Ithilarak incorporates the correct blue for “Ruhe”.

My excuse is that yellow in pre WWII French installations was often a subsidiary aspect before the violet was adopted.

 

This is the limit of photographs allowed in one upload.

Those interested enough to have read so far will have to go to post 2 to see the conclusion.

It is not as long as this I can assure you!

 

Ian T

Edited August 30, 2022 by ianathompson
typo

[ianathompson]

Continuing on from the first post this is the conclusion showing the final photograph.

 

 

The final view of the signal shows the tidied up arrangements.

Yes I did say tidied up!

The AFK eschews sophistication for simplicity.

Some location cases will eventually cover the wiring connections and a simple colour chart will be provided for the push/pull levers.

Both buffer stops, such as they are, have been set back from the siding ends. NG stock overhangs the running gear by a sufficient amount to allow this dodge to grab a little extra length for the sidings.

 

I hope that somebody finds something of interest here.

 

AFK signalling practice is reviewed in more detail here.

 

If you just want a visit to the layout and an overall look at the project try here.

 

Ian T

[5BarVT]

I like that, it appeals to my sense of different but still consistent.

Paul.

[uax6]

They have turned out nicely, and give you plenty of colours to play with.... Are you going to add hoods to them?

 

Andy G

[The Stationmaster]

If you're into 'different' signalling what does this one mean - showing green in the upper lamp unit and red in the lower lamp unit.  The illuminated letter 'A ' means - I hope obviously - that it is working as an automatic.

 

Edited August 31, 2022 by The Stationmaster

[ianathompson]

5 hours ago, uax6 said:

They have turned out nicely, and give you plenty of colours to play with.... Are you going to add hoods to them?

 

Andy.

 

The short answer is no there will not be any hoods.

I think that hoods will restrict the sightlines and make them harder to see.

The LED signals have got hoods, as you are aware, but the light is (incorrectly) displayed from beyond the body of the signal housing.

This makes them easier to see from wider angles whereas the prototype lights were very "directed" beams.

 

The only problem now is that there are potentially another eight of the things to knock up and I still have not found the LEDs that you gave me.

I would lose my head if it came off!

That reminds me I still owe you for the sockets from last time.

Don't let me forget next time I see you!

 

Also on the horizon are some home brewed electro-mechanical shunting signals.

It would have been nice to make some Italian "marmoset" signals.

Unlike the originals, which shine in one direction only, mine would need to simultaneously shine in two directions.

I have not been able to work this out using a LED.

A bulb would allow it but would melt the plastic.

I am not going down the metal road and the bulbs would require another power supply.

 

I cannot find the Urteno LED supply, despite this supposedly being behind the station building.

The current signal is hooked up to stuff installed around twenty years ago. 

I can see, being preamatic, that this will end up powering all the station signals except for those out on "the hill".

 

I have some vague ideas in mind for the shunt signals.

Time will tell.

The penalties of freelancing with regard to signalling!

 

Tommo

[ianathompson]

7 hours ago, The Stationmaster said:

If you're into 'different' signalling what does this one mean - showing green in the upper lamp unit and red in the lower lamp unit.  The illuminated letter 'A ' means - I hope obviously - that it is working as an automatic.

 

I did not claim to be an expert on unuusal signals, just that I liked them!

My first thought was that it might be an US style aspect such as "medium approach".

US signals seem to incorporate red into clear aspects.

 

The number plate seems to indicate that it is a Colchester PSB signal.

Since I retired I no longer have access to Anglia region CCF so I cannot confirm this.

A poke through published photos of Colchester panel seems to suggest that there is currently no signal 141.

I cannot see what the notice below the four aspect head says.

The poisition of the number plate might suggest that the two aspect head is an independent signal.

My best guess then would be that it is a junction signal with a two aspect head over the four aspect head.

This would suggest that it was "stacked" like some sempahore signals and dodds.

The green would apply to the left hand route, such as the branch line from a bay, and that the four aspect signal applied to a train leaving the bay to enter the mainline.

If this was a branch bay starter the signal could be left on automatic until a train needed to enter the mainline.

 

The only reason for placing the signals in this way that I can think of wouuld be sighting constraints.

Normally the two aspect head would have been to the left of the four aspect head in  an old fashioned array.

As the modern practice would be to use a feather I can only surmise that it is an anachronistic survivor from older times.

 

Now go on and prove me wrong!

 

Ian T

Edited August 31, 2022 by ianathompson
typo

[Grovenor]

Try https://www.nswrail.net/library/signalling.php

Unfortunately the yellows are hard to make out.

[ianathompson]

56 minutes ago, Grovenor said:

Try https://www.nswrail.net/library/signalling.php

Unfortunately the yellows are hard to make out.

 

Thanks for that.

I am still a bit concerned that the "Low speed " indication allows a driver to pass a signal displaying two reds.

 

Ian T

[5BarVT]

1 hour ago, Grovenor said:

Try https://www.nswrail.net/library/signalling.php

Unfortunately the yellows are hard to make out.

Hadn’t gone far enough south!  I’d got as far as probably not HK or Singapore and with a nameplate like that it had to be somewhere with ex-Pat British influence.

Paul.

[Grovenor]

1 hour ago, ianathompson said:

 

Thanks for that.

I am still a bit concerned that the "Low speed " indication allows a driver to pass a signal displaying two reds.

 

Ian T

Its no different to our standard here, we use a subsidiary to pass reds, one red here is the same as two reds there. Low speed to pass two reds is the same in the USA as well.

[ianathompson]

On 31/08/2022 at 22:19, Grovenor said:

Its no different to our standard here, we use a subsidiary to pass reds, one red here is the same as two reds there. Low speed to pass two reds is the same in the USA as well.

 

Yes. But the authority to pass the red comes from a distinct subsidiary signal.

The authority to pass the red in US practice, and NSW practice by the look of it, is given in a main aspect.

 

Okay there is more than one way to skin a cat but I find it a little unnerving.

 

A quick look at Italian signallling, which went down this same road, shows that red over green can authorise proceed at 100 kph when shown with two white bars (thanks Wikipedia).

I believe that the LMS Mirfield "speed signalling" system attempted something similar but was never adopted as a norm.

 

Ian T

Edited September 1, 2022 by ianathompson
typo

[St Enodoc]

20 hours ago, The Stationmaster said:

If you're into 'different' signalling what does this one mean - showing green in the upper lamp unit and red in the lower lamp unit.  The illuminated letter 'A ' means - I hope obviously - that it is working as an automatic.

 

This is an NSW controlled signal that is operating in automatic mode (shown by the illuminated A). It is displaying a CAUTION indication (PROCEED. Next signal may be at STOP). It is at the East end of Circular Quay station on the City Outer track of the City Circle line in Sydney NSW, Australia.

 

The plate below the aspects reads: WHEN "A" LIGHT IS OUT THIS SIGNAL MUST NOT BE PASSED AT STOP WITHOUT AUTHORITY FROM SIGNALLER.

 

See page 8 of this document:

 

https://railsafe.org.au/__data/assets/pdf_file/0006/39759/City-Circle-Central-to-Central-via-Circular-Quay.PDF

[St Enodoc]

6 minutes ago, ianathompson said:

The authority to pass the red in US practice, and NSW practice by the look of it, is given in a main aspect.

No. In NSW two reds with no subsidiary aspect means Stop.

 

https://railsafe.org.au/__data/assets/pdf_file/0020/31763/NSG-606-Responding-to-signals-and-signs-V5.1.pdf

Edited September 1, 2022 by St Enodoc
link added

[The Stationmaster]

St Enodoc - who is of course 'a local' - is spot on right down to correctly identifying the location.

 

However the most interesting point has been missed.  The upper, 2 aspect, signal head, is the stop signal, the lower, multi aspect head, is a distant for signals in advance.  As these colour lights date from the time when red was still used in distant signals they show a red aspect when the distant is a caution.  So a Driver is not passing a red aspect which means stop but a red in that position means the next stop signal in advance is at danger.

 

In the full aspect sequence the distant will change to yellow when the next stop signal in advance is at green but the one beyond that is at danger.  When the second stop signal in advance goes to green the distant in this signal will go to green.

 

The next aspect change is when a train passes the signal and both signal heads will return to red - i.e. stop signal at danger. distant signal at caution.  Depending on track circuit and overlap occupation at the next signal in advance a miniature green will appear in the bottom lens indicating that the signal may be passed at danger at very limited speed  and run at caution etc.  The next change of aspect will extinguish the miniature green and the stop signal will clear to green with the distant still at caution - as in the photo.

 

The signal number indicates Circle (the City circle), Outer rail.

 

The system allows running at very close headways and fits well (the miniature green) with the very short (18 second) station dwell times at most stations on the City Circle.  in connection with the work I was doing (on timetabling reliability and risks) for Cityrail - the then operator of Sydney suburban services -  I did several samples of stopwatch recording at Circular Quay on my way to work in the morning peak and the dwell times were very rarely exceeded there while headways were normally exactly maintained.

[ianathompson]

Just a slight update on this subject.

 

 

Despite my stated intention not to add hoods I have now done so.

I suppose that the signals look better for it.

They might also look better for a white ring so that also is under consideration.

Unfortunately, as the prototype installation, it went in as raw styrene so that any defects could be addressed.

Much to my surprise it functiioned at the first time of asking.

 

The problem is, as anticipated, that the hoods tend to obscure the aspects.

This one is version MK IV and has been pared back to the bare minimum.

The previous versions meant sticking the head down to track level, as per a scale sized driver, to see them.

This does not seem to be a very clever idea given the proximity of the delicate lifting bridge.

 

 

As has been stated the signal stands on the middle level of the layout so that trains run in scenes above and below this one.

The station at Urteno is  therefore fairly dark without its lights turned on.

The photo only gives an approximation because the lenses of modern cameras tend to automatically componsate for dark conditions.

The photo has therefore been fiddled with.

One unexpected problem has been the amount of light being thrown out behind the LEDs.

As much black paint as possible has been poked into the heads containing the LEDs to cut this down.

The photo is a a premodification shot.

This has partially solved the issue ,or at least made it less significant when the lights are on.

The other problem is the light thrown back from the coloured vanes.

Not quite sure what can be done about this altough I have a few thoughts.

 

Once again following a prototype would have eliminated these problems.

 

Ian T

[uax6]

The UK mechanical searchlights just used a normal semaphore spectacle plate, with a high intensity Adlake electric lamp behind with circular stepped lense. I'm guessing that the lense cut out a lot of the back light, but I can't now remember how wide the Adlake lamp was, and how it blocked the rear of the lense to reflected light.

(I can't find a rear shot of the one at Quorn at the minute, I think Stationmaster Mike has one).

 

It might be that you might need to paint a bit of your lenses to reduce their size so that only the bit that you can see through the front hole is not black, or, which might be easier, cut another black circle the size of the lenses, and stick that on the front of the led bit, so the lenses are inside two discs.

 

As for the bleeding from the rear of the leds, smear them with a bit of filler (miliput would be ideal) and then paint.

 

I have to say I prefer my searchlights with long hoods, the Westinghouse one in the gardern here, has one about 24" long....

 

Andy G

 

[Michael Hodgson]

38 minutes ago, uax6 said:

The UK mechanical searchlights just used a normal semaphore spectacle plate, with a high intensity Adlake electric lamp behind with circular stepped lense. I'm guessing that the lense cut out a lot of the back light, but I can't now remember how wide the Adlake lamp was, and how it blocked the rear of the lense to reflected light.

(I can't find a rear shot of the one at Quorn at the minute, I think Stationmaster Mike has one).

 

 

 

Mechanical searchlight at Quorn

The signal is a starter back towards Loughboro from the wrong platform via the main crossover, so it isn't used very often.  Backlight is a bit irrelevant here, although it has the usual semaphore backlight blinder. The face of the signal rather than the rear is visible from the box, and the overbridge osbtructs any view of the rear anyway.

Edited September 2, 2022 by Michael Hodgson

[LNERGE]

I have a working mechanical searchlight at home and all the bits to make another that will most probably be at the MNR. The bits came from the last one in use at Littlebury 

 

 

[Michael Hodgson]

You're too modest - it was you that put that signal at Quorn !

[ianathompson]

14 hours ago, Michael Hodgson said:

Mechanical searchlight at Quorn

 

Thanks for posting these photos in this thread.

I was pointed in their direction by Andy (uax6) and used them as the basis for the model.

To an extent I have solved the problem by placing large black styrene "blockers " around the light.

The next ones in this particular scheme, whenever progress is made, will have these added before they are built not after!

 

Ian T

[Michael Hodgson]

1 hour ago, ianathompson said:

 

To an extent I have solved the problem by placing large black styrene "blockers " around the light.

 

That's essentially all the prototype is - it's a semaphore signal with a large metal plate and a hood bolted onto the front of the post   It is fitted with a lamp powerful enough to be seen by day, unlike the usual oil lamps, shining through the spectacle in the usual way, so there's no need to fit the enamelled arm to the spectacle casting.  The usual signal arm operating rod can been seen running down the front of the post, through timbers in the platform where it is worked by a galvanised wire like all the other semaphores. 

 

A more conventional  searchlight signal isn't really all that different either.  It's electro-mechanical in that a powerful light shines through a (3-position) spectacle plate, but that is fitted within a signal head mechanism containing solenoids to deflect one way or the other from its centre (fail safe) position.  The big advantage is that now you don't need mechanical wires & pulleys, just an electric wire from a relay cabinet.  In the UK if you needed a double yellow aspect, you just fitted another lamp above it, with a fixed yellow lens fed through relay circutiry.  American signalling systems tended to need two of the mechanisms.

 

And of course, a multiple lamp colour light has the advantage that it is all-electric, there are no moving parts to fail, though to be fair searchlight mechanisms mounted at the top of the post were remarkably reliable.  But nowadays we're going back to what looks like a searchlight but with different coloured LEDs within built into one lamp.  These need less space than a 4-lamp assembly are are easier to site so that all aspects can be seen among OHLE.

 

 

 

I think the colour distortion problem you are experiencing is because you are viewing an LED through a coloured translucent sheet.  Unfortunately this doesn't work as well in terms of changing the colour as it does with a filament bulb (or an oil lamp).

 

It is preferable to use an LED of the right colour (ie one that emits light at the right wavelength).  Various combinations of bi-colour LEDs are available, Red/Green being the one most relevant here.  These are simply two LEDs in one component and wired such that the Red lights for one polarity of DC, Green for the other.  If fed fast AC, both light up, in opposite AC phases.  The human eye doesn't notice this rapid flicker but tends to see the two combined as yellow. 

 

 

 

 

[5BarVT]

10 hours ago, Michael Hodgson said:

though to be fair searchlight mechanisms mounted at the top of the post were remarkably reliable.

But they did need to be maintained - cycled through Reading Works (WR ones) to keep them in spec.  Thus significantly more whole life cost than all electric heads.

Paul.

[The Stationmaster]

On 03/09/2022 at 21:51, 5BarVT said:

But they did need to be maintained - cycled through Reading Works (WR ones) to keep them in spec.  Thus significantly more whole life cost than all electric heads.

Paul.

And they could be prone to 'bobbing' - ie.e changing from one aspect to another and back again and even going through all three aspects depending on where they started.  Not an enjoyable sight, or effect, when the signal did that in front of an approaching milk train which might well have finished up delivering cream, or even butter, to Wood Lane.

[5BarVT]

When commuting to school for a year on the North Tyne Loop (before the metro) I used stand at South Gosforth watching searchlights stepping up.  And the green did bob back to yellow as the signal in advance changed aspect.  I think sometimes even the one after (depended on when my train arrived!).

Paul.