LED Signals

[ComUtoR]

If the signal isn't needed (can be covered in snow without problem), why have signals in the first place?

Because collisions are frowned upon.

There would no doubt be someone who would chance the junction if there was nothing to stop them. Or those who would always want to be the first into the platform.

 

[ac6000cw]

In fact, in the USA you can see whole gantries of junction signals all dark until a train approaches. This is often in conjunction with arrays of solar battery chargers mounted alongside. On single lines the signals for both directions will switch on. This has the advantage over conventional signal lighting that staff on the ground get a warning that an oncoming train is getting close.

...and very handy for visiting train-watchers like me. Seeing the signals light up does generate a nice feeling of anticipation, even though a train might be miles away at that point.

I think LED signals in the US do still have hoods quite often - snow and ice build up is a bigger problem over there, with much harsher winter conditions in some areas. I don't know if they have heaters in some signal heads. A long-handled broom in the back of the track inspectors Hy-Rail truck (or on a loco) doubtless comes in useful at times too...

 

[Taunton]

...and very handy for visiting train-watchers like me. Seeing the signals light up does generate a nice feeling of anticipation.

I wonder if that is actually part of the intention; in the US, Automatic Block Signals on single lines, a quite separate way of working from anything in Britain, only controls the safety separation of trains rather than their organisation, which continued to be done by wayside operators handing up train orders (nowadays all by radio). ABS signals are normally green in both directions, unless there is a train ahead, and at US traffic frequency can remain so for hours, or even days. The first sign there is something approaching, detected by track circuits, is the signal for the opposite direction changes back from green to red. This can be useful to a crew instructed to pass an oncoming train miles from anywhere, who get down and hand-throw the points into a passing siding, then enter and wait, seeing the signal on the main track ahead at green. As it goes through yellow to red they know the oncoming is getting close (yes, nowadays the crews radio to each other as well).

I as always reminded of this when I saw what was done in the Mound Tunnels at the west end of Edinburgh Waverley in the 1970s, with the indicators for the four bidirectional tunnel tracks W to Z that lit up whenever a movement was set

 

[Philip Phlopp]

If the signal isn't needed (can be covered in snow without problem), why have signals in the first place?

Indeed. Much better putting them in the cabs of trains and letting Hitachi/Ansaldo or Siemens worry about making the damn things work, instead of sending the S&T division out on manoeuvres.

Flashing lights have no place at the side of a railway line, falling over and getting in the way of trains, or staying upright and getting in the way of electrification masts. They're suitable for discotheques and nothing else.

On a serious note, signalling has always been susceptible to snow and ice, semaphore signals could have wires and signal heads weighed down by snow and ice, so fail to show a correct aspect, it blocks up colour light heads even when they're filament type bulbs, especially bits of Scotland where you get a wind chill factor in addition to snow, so the bulb doesn't heat up enough to melt all the snow and overcome the wind chill factor.

Which is why there are a large number of rule book requirements to follow when running in heavy snow, which can be boiled down to - run slower, check your brakes more often, and expect to come across something blocking the line - a tree, a power line, a snow drift or a broken down train.

It is possible, btw, with LED bulbs, to fit a heating circuit - but heating LEDs kills the light intensity, unlike in filaments where hotter equals brighter, so you want to avoid that where possible.

 

[Taunton]

snow and ice ... it blocks up colour light heads even when they're filament type bulbs, especially bits of Scotland where you get a wind chill factor in addition to snow, so the bulb doesn't heat up enough to melt all the snow and overcome the wind chill factor.

I've often wondered why, on conventional 4-aspect signals in Britain, they are mounted vertically with a hood over each, which does indeed make them susceptible to snow/icicle buildup. Common in the US on multi-lens signals (they are not all searchlight) is a single large hood right over all the lenses. These used to have parallel sides but more recently are tapered outwards, known as "Darth Vader" hoods.

http://www.railroadsignals.us/signals/color/index.htm

 

[edwin_m]

Compared with four separate hoods, a single deeper hood of the same length will be less effective at shading the lenses from the sun, especially the bottom (red) lens.

 

[ac6000cw]

Compared with four separate hoods, a single deeper hood of the same length will be less effective at shading the lenses from the sun, especially the bottom (red) lens.

I suspect it's like most things in engineering - a compromise, between minimising snow build-up problems (single large hood) and sun shading (a hood per lens). As light sources have got brighter - first halogen lamps and now LEDs - I assume the balance has shifted towards minimising the snow problems.

 

[edwin_m]

I suspect it's like most things in engineering - a compromise, between minimising snow build-up problems (single large hood) and sun shading (a hood per lens). As light sources have got brighter - first halogen lamps and now LEDs - I assume the balance has shifted towards minimising the snow problems.

All new LED signal heads now have at least one of: no hood at all, only one lens or two lenses widely separated. All these are less prone to snow buildup than a classic multi-lens signal head. I'd guess the ones most vulnerable are where the drop-in LED replacements have been fitted to an existing signal head.