Dewsbury Change to LED signal heads

[Crossover]

Does anyone know when the heads were changed in the Dewsbury area? First saw them a few weeks ago and swore they hadn't been like that last time I was there. Also noticed today from the adjacent Huddersfield Road that the change has included the first one out Westbound, which I can only assume is a semi auto as it is most often at yellow (where the signal protecting the junction is presumably held at red)?

 

[Tomnick]

That signal to the west is, if I'm not mistaken, Batley's last one - I'm not sure whether it's an auto signal or not, but it's very unlikely to be a semi-auto signal. The next one is Healey Mills' first signal (controlled) - although it doesn't protect the junction, it'll normally be at danger (until a route is set from it), so it does still explain why the signal in rear is normally showing a yellow aspect.

 

[LNW-GW Joint]

Aren't they the ones due to change to improve speeds through Ravensthorpe?
Flashing yellows to avoid the approach control etc.

 

[Llama]

Flashing yellows were put in a couple of years ago, there are three sets of flashing yellow signals, on the Down line approaching Mirfield East Jn for the turnout to the Down L&Y for Wakefield, on the Down line approaching Dewsbury for the turnout to the Down Passenger Loop and on the Up line approaching Mirfield East Jn for the turnout to the Up Slow.

The Dorman LED heads at the Batley end of Dewsbury were changed recently, although quite when I am not sure.

HM686 signal after Dewsbury in the Up is an automatic signal, however is still another two signals from Thornhill LNW Jn after Ravensthorpe - the approach to Thornhill LNW Jn is subject to 'double blocking' whereby HM684 signal (the signal before the junction signal) may be held at danger until an approaching train is proven by occupation of track circuits to be travelling at a low enough speed to be permitted to approach HM682 signal at danger, HM682 being the signal protecting Thornhill LNW Jn.

 

[Tomnick]

Apologies, I mistook that auto signal - 686 - for one of Batley's from a photo of the panel. Thanks for the correction! The double blocking (actually sounds more like a reduced overlap with the delayed yellow) also explains why 684 is a controlled signal and normally at danger. It's still definitely not a semi automatic signal though .

 

[Sapphire Blue]

Does any of this explain why the stopper out of Leeds is regularly shown as being up to 7 mins late approaching Dewsbury, giving unwary passengers the idea they have time to nip to the loo or go for a very swift half, only for the train to trundle in smack on time?

 

[Crossover]

Thanks to you all for the insight - very interesting, and glad I am not going mad on the Dorman heads being new! Wonder if the stop boards were done at the same time...

It will be HM686 signal that I see from Huddersfield Road (there is no immediate starting signal at Dewsbury on the Up) - has had me wondering for a while as it has been showing yellow even on Christmas Day when I have driven by! That one looks to have recently gone over to a Dorman head too - it was noticing the new head on HM686 that prompted the thread.

Having since found this thread, I would suggest I have got confused in terms between auto/semi-auto signals - from what has been said above, HM686 being auto makes sense as it automatically moves up aspect depending on the status of those, and the block occupation, ahead of it.

Regarding the double blocking, could that be because, from what I recall, HM682 is close to the edge of the Up platform at Ravensthorpe with a relatively short distance until the junction? Of course, in the case of that particular junction and its layout, a SPAD could put a train into the direct path of an oncoming service bound for Healey Mills as a short section of the track there is bi-di, it only being 3 track and Up towards Mirfield and Down towards Healey Mills traversing the same track at the junction?
--- old post above --- --- new post below ---

Does any of this explain why the stopper out of Leeds is regularly shown as being up to 7 mins late approaching Dewsbury, giving unwary passengers the idea they have time to nip to the loo or go for a very swift half, only for the train to trundle in smack on time?

It does also happen to the TPE at times as well. Maybe that is more related to a non-report at a previous timing point?

 

[Tomnick]

It's a common misconception - as I think you've now established, a controlled signal with an auto working facility isn't a semi-automatic signal! The latter, as the linked thread explains, are usually found protecting ground frames, level crossings and so on, on sections of otherwise plain line. The signals protecting some of the ECML level crossings between Peterborough and Doncaster, for example, follow the same principle - as far as the controlling signal box is concerned, they operate as auto signals, but only actually clear when the crossing keeper pulls off their 'slot' on the signal. I can't remember whether they're actually played as semi-autos, though I'm pretty sure that they're shown as such on the panel.

Broadly speaking, a delayed yellow on the signal in rear is used where there's insufficient room for a full overlap (up to 200yds, depending on linespeed). In this case, that might be something to do with the position of Ravensthorpe station, or there might be other factors to consider. It's hopefully clear, though, that by bringing the train nearly to a stand at the signal in rear, it'll approach the protecting signal for the junction at a lower speed, and therefore a shorter overlap is acceptable. Some installations have the ability to select either a reduced overlap (with delayed yellow in rear) or a full overlap, but that doesn't seem to be the case here. Double blocking is slightly different, and more to do with mitigating against a high SPAD risk at a given junction - most of the examples that I can think of have been applied retrospectively, probably in the light of experience.

 

[Railsigns]

I had a quick look at the signalling plan for the area, and the situation is this: The signal protecting the junction (HM682) has a short overlap (correctly referred to as a "restricted overlap" in this scenario) and, accordingly, the signal in rear (HM684) is subject to approach release when HM682 is at red. HM686 is a fully automatic signal. There is no double blocking.

 

[swt_passenger]

In the general sense is it now fair to say that LED signal heads can appear anytime, anywhere, as part of routine maintenance? IIRC they have LED signal heads (or LED inserts for existing lamp positions) that appear to the rest of the signalling system to be electrically identical to the normal filament lamps they are replacing.

What I'm suggesting is that the sudden appearance of LEDs in any given area may not necessarily mean any alterations to the underlying signalling system have been made at the same time.

 

[edwin_m]

If anything it probably means that the signalling system has several years life left in it, as reduced maintenance over a longish period makes it worthwhile to replace the heads. This isn't universal though - for example the signal that fell over at Newbury last week had a new head but presumably would have been replaced before electrification (whenever that is).

 

[Railsigns]

for example the signal that fell over at Newbury last week

In addition to the one that fell down three weeks ago? Goodness.

 

[edwin_m]

In addition to the one that fell down three weeks ago? Goodness.

Was it really that long ago? Time flies when you're enjoying yourself...

 

[Marklund]

In the general sense is it now fair to say that LED signal heads can appear anytime, anywhere, as part of routine maintenance?

Yes, and no. Depends on workload.

IIRC they have LED signal heads (or LED inserts for existing lamp positions) that appear to the rest of the signalling system to be electrically identical to the normal filament lamps they are replacing.

Well, they're not quite electrically identical.
The LED inserts (or LED Light Engines as they're branded) are a pcb with 5 led chips mounted, forward facing, in a 3 pin holder, which fits in to the majority of existing SL35 heads.

However, due to the current loading of the lamp (drawing about 2 amps), and the current draw of LEDs being a fraction, it's not enough to energise the ECR (Lamp Proving Relay). So the pcb had a ballast resistor fitted to draw the required current to energise the ECR

However, as there is no auxiliary filament now, the EKR is removed, and a dummy unit is fitted.

They seem to have a 5 year planned life, compared with the now standard 8000 hour SL35, although I've had a few failures already.

SL35 LED Light Engines
They say 15 minutes for a 4 aspect signal conversion?!? Would love to see that!

 

[edwin_m]

I think the point is that the electrical interface to the signalling is identical, so they can be substituted for filament lamps without changing anything else.

 

[Marklund]

I think the point is that the electrical interface to the signalling is identical, so they can be substituted for filament lamps without changing anything else.

No, you still need to change the EKR.

 

[swt_passenger]

Depends on what your point of view is surely? Is the EKR part of the signal, or part of the 'signalling system'? If it's physically part of the signal, but is effectively an adaptor to decide what the rest of the signalling system electrically 'sees', then the change to LED (including the changed EKR) does make the type of illumination transparent to the controlling system?

 

[Tomnick]

The EKR is surely part of the signalling system, rather than merely part of the signal itself? That said, I remember some of the distant signals on our patch being converted to LEDs (not new heads, just fitting LED 'light engines' in the existing lamp holders) between trains. I seem to remember them being left indicating a first filament failure for some time, presumably because nothing was done in the relay room to accommodate the change.

 

[MarkyT]

The EKR is surely part of the signalling system, rather than merely part of the signal itself? That said, I remember some of the distant signals on our patch being converted to LEDs (not new heads, just fitting LED 'light engines' in the existing lamp holders) between trains. I seem to remember them being left indicating a first filament failure for some time, presumably because nothing was done in the relay room to accommodate the change.

Signal head EKRs are for 1st/2nd filament switching and indication. In a conventional filament head for each lamp unit there's a small plug in relay whose coil is wired in series with the lamp, the armature of which falls away if there's a 1st filament failure. That switches the supply to the 2nd filament and, using another voltage free contact breaking an alarm circuit to inform technicians its about time to think about changing the lamp. Usually in relay-based signalling groups of signals share a 'zone' alarm circuit. In SSI and related technologies an alarm channel exists for each signal.

With vastly more reliable LED heads and plug in replacement LED 'light engine' lamps for conventional heads, there is no need for a filament changeover mechanism; the unit works or it doesn't. That is not unsafe as lamp proving still functions to prove sufficient current is being drawn to assume the unit is alight. With a LED head there is simply no auxiliary contact to pick up for the individual or zone alarm circuit. With a light engine fitted conventional head, the series relay coil of the head EKR is bypassed by modified wiring, so is permanently in the 1st filament failed position. Unless the zone alarm circuit is also modified by some means, a permanent alarm will persist.

http://www.howells-railway.co.uk/ledlightengine.html

"Light engine" modules (as link above), essentialy bayonet socket replacements for a traditional SL35 or similar filament lamp are supplied in kits with a dummy EKR module that replaces the relay and bridges out the contact, without having to design a wiring change. For a new LED head a small designed wiring change is required to achieve the same effect.

 

[Tomnick]

Thanks, that makes sense and broadly confirms my limited understanding, although I didn't realise that the relay in question was actually located in the signal head itself. We were left with a first filament failure indicated constantly (for individual signals, so not a problem once acknowledged) and lamp proving continuing to operate as normal.

 

[edwin_m]

That could be another reason to replace all the signal heads in an area at the same time - if some but not all the ones on a filament proving circuit are replaced then filament proving won't work for the remaining ones.