Are junction indicators/theatre boxes fail safe?

[PULF]

I’ve had a thought - can junction indicators and theatre boxes especially show a routing, say M for main, for example but the point-work actually be set elsewhere or are they actually fail safe, in that they 100% only show where the route is definitely set? Should drivers take these routings as gospel or should they always visually check to be sure?

Thanks

 

[snowball]

Others will know far more about this than me but I think the junction indicator lights up about a second before the signal clears. I've both observed this in cab videos and seen it mentioned in the RAIB Peterborough report.

 

[TSG]

There are a couple of ways of making these indicators work. With the older ones, the design considered the layout of the lamps, such that it should not have been possible to have a different valid character seem to appear and still get lamp proving (as alluded to above, the indicator is proved alight before the main aspect can be cleared. Enough lamps out to change the apparent character and this won't happen).

More modern designs use a single lamp/LED attached to an array of optical fibres that end as 'dots' on the indicator, so they'd usually all be lit or all be out. You would need to do serious mechanical damage to the indicator for the intended character to suddenly look different. Degradation should be gradual enough for maintenance to see bits of the indicator failing before a driver can misread it

Oh, and if you mean could the wrong one get lit, the interlocking drives it in much the same way as it drives the main aspect red or green, so it is fail safe

 

[PULF]

There are a couple of ways of making these indicators work. With the older ones, the design considered the layout of the lamps, such that it should not have been possible to have a different valid character seem to appear and still get lamp proving (as alluded to above, the indicator is proved alight before the main aspect can be cleared. Enough lamps out to change the apparent character and this won't happen).

More modern designs use a single lamp/LED attached to an array of optical fibres that end as 'dots' on the indicator, so they'd usually all be lit or all be out. You would need to do serious mechanical damage to the indicator for the intended character to suddenly look different. Degradation should be gradual enough for maintenance to see bits of the indicator failing before a driver can misread it

Thank you, very informative! So other than the unlikely event of LEDs going out, the letter stated on the box would always correlate to the route set on points? So effectively making it 100% fail safe!

Others will know far more about this than me but I think the junction indicator lights up about a second before the signal clears. I've both observed this in cab videos and seen it mentioned in the RAIB Peterborough report.

I think you’re right with that one

 

[Railsigns]

More modern designs use a single lamp/LED attached to an array of optical fibres that end as 'dots' on the indicator, so they'd usually all be lit or all be out.

On the LED-type indicators, each 'dot' is a separate LED, so there are no fibres involved. Individual LEDs can and do occasionally fail, as the attached photo illustrates.

 

[Annetts key]

I’ve had a thought - can junction indicators and theatre boxes especially show a routing, say M for main, for example but the point-work actually be set elsewhere or are they actually fail safe, in that they 100% only show where the route is definitely set? Should drivers take these routings as gospel or should they always visually check to be sure?

Others will know far more about this than me but I think the junction indicator lights up about a second before the signal clears. I've both observed this in cab videos and seen it mentioned in the RAIB Peterborough report.

It’s the interlocking that both commands all the points to move, detects that the all points have completed the required movement and that they are now detected in the required position.

It’s also the interlocking that determines what signals aspects are lit, including route and junction indicators.

As long as everything is being maintained correctly, has been installed correctly and the wiring and system design (including the interlocking) is correct, only the correct route or junction indication will be displayed. Most (not all older designs) junction indicators are required to be proved lit. This also applies to some route indicators. This is the reason that the junction or route indicator illustrates before the signal steps up to show a proceed aspect. The slight delay is the time for the proving system to detect that the junction or route indicator is consuming power (hence should be lit).

So, yes, it’s all as fail safe as it’s possible to get.

 

[TSG]

On the LED-type indicators, each 'dot' is a separate LED, so there are no fibres involved. Individual LEDs can and do occasionally fail, as the attached photo illustrates.

I think that's the VMS type. The Signal House and Dorman (I believe) types, effectively replaced the halogen light source with a LED one, but kept the fibre optic bundle. I suspect that's partly because it makes ensuring an unambiguous indication simpler

 

[GardenRail]

Our Theatre for platform 7 at Sheffield hasn't been working for months. The signal still clears for that route, but we have to verbally advise the drivers where they are going. It's been going on that long, the drivers are telling us what we want to tell them

 

[John Webb]

I think that's the VMS type. The Signal House and Dorman (I believe) types, effectively replaced the halogen light source with a LED one, but kept the fibre optic bundle. I suspect that's partly because it makes ensuring an unambiguous indication simpler

The Dorman Theatre unit we have on display at St Albans South, dating from around 2010-2015, is individual LEDs, not fibre-optics. The label in the back states the normal current consumption, which we understand was monitored in the Location Cabinet from which the indicator was fed to check that sufficient LEDs are lit to give a clear indication of the set route before the main aspect is cleared.

 

[Signal Head]

I think that's the VMS type. The Signal House and Dorman (I believe) types, effectively replaced the halogen light source with a LED one, but kept the fibre optic bundle. I suspect that's partly because it makes ensuring an unambiguous indication simpler

The drive from the interlocking will always be fail safe. Proving, where provided by the interlocking, will also be fail safe.

The possible failure mode(s) may vary according to the method of construction-

Dorman type - each LED is used only once, this means there is a limitation on how many indications can be shown in a single box, as there is with (others) fibre-optic type before them, because each fibre end takes up a position on the matrix plate at the front, like an individual LED.
For a Standard LED indicator (ie a large one replacing a traditional 'theatre'), I believe the limit is 6 indications. For the Miniature indicator, replacing a 'Stencil', I think it's 4, but it varies depending on how many characters each indication has - eg 2 x 3 char, 3 x 2chr.
I'm not aware of Dorman ever having made fibre optic indicators.

Signal House make fibre optic indicators, I believe they are the only manufacturer to do so now. Originally they used a 55W halogen lamp but now offer an LED alternative (but still with fibre bundles).

VMS (now Mallatite) - original LED indicators were similar in principle to Dorman I think, ie chains of LEDs specific to each indication - but some years ago they introduced a 'Mk 2', which provides for up to 20 indications in a Standard. The only way I can see this working (ie physically fitting in the case), is that the LEDs are used in multiple indications, much as the lamps were in the old 7x7 Multi-lamp Route Indicators (Theatres). Whereas the logic for an MLRI was done with relays in the location case or relay room, I suspect the Mk 2 has some internal electronic or processor-based logic to sort out which LEDs to light from a given (single) input, so that would presumably need to have an appropriate SIL rating in order for the indicator, as a unit, to be considered 'fail safe'

 

[Tim M]

A small tale from Edinburgh in 1976/1977 during the Edinburgh and East of Scotland Resignalling. A couple of signals on the approach to Waverley from the Princess Street Gardens has multi-lamp route indicators including displays for platforms 18 and 19. Platform 19 is a through line, 18 a short bay. A corruption to the ‘18’ with a couple of lamps failed could show ‘19’ leading drivers to under brake thinking they had a route into the through platform, but finding a buffer stop in the way.

This problem was only realised after much of the design and wiring had been completed! I was asked to design a solution that required additional lamp proving relays in the adjacent Apparatus Room, loss of critical lamps ensuring the relevant signals staying at red. However (1)!

Due to sighting problems with the Mound Tunnel, the signals concerned were ground mounted at the exit of the tunnel with 7*7 matrix Westinghouse route indicators PLUS signal off repeaters and pairs of 7*5 matrix GEC-General Signal (main contractor) route indicators on a gantry just before the tunnel entrance. However (2)!

The Westinghouse unit lamps were direct fed from the Apparatus Room, the GEC equivalents had lamp configurations selected by relays in the indicators themselves with separate power feeds. Following selection and bench testing of suitable relays and quite a lot of modification to the wiring, a solution was implemented and rolled out for other similar situations around Waverley.

Such issues wouldn’t happen with modern indicators.

 

[ComUtoR]

It's been going on that long, the drivers are telling us what we want to tell them

My ever growing list of "funny things the Signaller says" (would make a superb thread)

"Driver, can you stop at the next signal please; I have no idea what its showing. It's showing as blank on my panel"
"Showing, single yellow with position 4 but it looks like three lights aren't lit"
"Ahh, that makes sense"
"Driver, can you let me know if you go into the correct platform please"

 

[John Webb]

The Dorman Theatre unit we have on display at St Albans South, dating from around 2010-2015, is individual LEDs, not fibre-optics. The label in the back states the normal current consumption, which we understand was monitored in the Location Cabinet from which the indicator was fed to check that sufficient LEDs are lit to give a clear indication of the set route before the main aspect is cleared.

Here's a photo of the above indicator while we trying it out:

It only shows the single 'R'.

 

[Signal Head]

My ever growing list of "funny things the Signaller says" (would make a superb thread)

"Driver, can you stop at the next signal please; I have no idea what its showing. It's showing as blank on my panel"
"Showing, single yellow with position 4 but it looks like three lights aren't lit"
"Ahh, that makes sense"
"Driver, can you let me know if you go into the correct platform please"

If three PLJI lamps are out, then there shouldn't be anything better than a Red in the main aspect.

When separate adjustable proving resistors were the norm, they were typically set so as to require four lamps in to clear the signal, relay characteristics would normally allow it to stay energised if one more failed (ie three lit), lose another and the proving drops out - signal reverts to Red.

Except on the SR of course, who only had three lamps to begin with.

 

[ComUtoR]

I've always been taught that 3/5 and 2/3 (unlit) should always be treated as broken and to call it in.

Even with theatre boxes I've always been told that 2 lights out are acceptable but 3 needs a call to confirm. I've seen many theatre boxes with lamps out but I've always been able to read the number correctly.

I've generally found indicators to be robust. Signals (main aspects) on the other hand...

 

[Tomnick]

Our Theatre for platform 7 at Sheffield hasn't been working for months. The signal still clears for that route, but we have to verbally advise the drivers where they are going. It's been going on that long, the drivers are telling us what we want to tell them

It's interesting to note that, on all the signals with route indicators around Sheffield station, the main aspect steps up to a proceed aspect *before* the route indication appears – the opposite way around to what you'd expect if the route indicator was lamp proved (in which case, of course, you wouldn't get the signal off into 7 at all!).

 

[MadMac]

A small tale from Edinburgh in 1976/1977 during the Edinburgh and East of Scotland Resignalling. A couple of signals on the approach to Waverley from the Princess Street Gardens has multi-lamp route indicators including displays for platforms 18 and 19. Platform 19 is a through line, 18 a short bay. A corruption to the ‘18’ with a couple of lamps failed could show ‘19’ leading drivers to under brake thinking they had a route into the through platform, but finding a buffer stop in the way.

This problem was only realised after much of the design and wiring had been completed! I was asked to design a solution that required additional lamp proving relays in the adjacent Apparatus Room, loss of critical lamps ensuring the relevant signals staying at red. However (1)!

Due to sighting problems with the Mound Tunnel, the signals concerned were ground mounted at the exit of the tunnel with 7*7 matrix Westinghouse route indicators PLUS signal off repeaters and pairs of 7*5 matrix GEC-General Signal (main contractor) route indicators on a gantry just before the tunnel entrance. However (2)!

The Westinghouse unit lamps were direct fed from the Apparatus Room, the GEC equivalents had lamp configurations selected by relays in the indicators themselves with separate power feeds. Following selection and bench testing of suitable relays and quite a lot of modification to the wiring, a solution was implemented and rolled out for other similar situations around Waverley.

Such issues wouldn’t happen with modern indicators.

Ah the indicators W, X, Y and Z. Memories…..

My ever growing list of "funny things the Signaller says" (would make a superb thread)

"Driver, can you stop at the next signal please; I have no idea what its showing. It's showing as blank on my panel"
"Showing, single yellow with position 4 but it looks like three lights aren't lit"
"Ahh, that makes sense"
"Driver, can you let me know if you go into the correct platform please"

Oh, the stories I could tell…..

 

[Signal Head]

It's interesting to note that, on all the signals with route indicators around Sheffield station, the main aspect steps up to a proceed aspect *before* the route indication appears – the opposite way around to what you'd expect if the route indicator was lamp proved (in which case, of course, you wouldn't get the signal off into 7 at all!).

Well, it's a relatively low speed layout, and I would expect no great variation in speeds between routes so, with historical principles, they wouldn't need to be proved - except that there are bays at both ends of the station, so all the routes aren't of equal 'value', therefore proving would be desirable at least.