Starter signal at OXF

[At_traction]

Hello, a blocks/signals question ..

While waiting for a train towards Reading at the Oxford station, I noticed that the starter signal for the (I guess it's called) up line seems to display the red aspect for a particularly long time and then change to green in a rapid succession. This happened with FGW and XC DMU services with "sufficient" acceleration and I suspected that the first block out of the station might be particularly long with the next two blocks being, in turn, relatively short, changing the starter signal from red to yellow and green.

Or rather, I never actually even saw the yellow(s) as I got tired of eyeing the signal (a limit to my anorakism I guess ), just that the green came after a long wait at the red. According to this photo at least my memory about the signal on the second, bypass up rail (four, in all, going through the station) being a four-aspecter with a feather is correct -- the signal being on the extreme left of the photo. I think that the platform signal had a four-aspect signal too, although I'm not certain if the signal in fact does change straight to green due to the length of the block. But I'm certain someone here knows the answer...

 

[Old Timer]

It is a controlled signal, that is, it is operated by the Signalman for each train, hence why it is staying at red after the train has passed beyond.

The Signalman is setting up the route for the train on each occasion as it approaches. The fact that the signal changes immediately to green is almost certainly because the signal ahead is an auto, or a controlled signal set at auto mode, hence why you are not seeing the signal step-up sequence.

 

[At_traction]

The Signalman is setting up the route for the train on each occasion as it approaches. The fact that the signal changes immediately to green is almost certainly because the signal ahead is an auto, or a controlled signal set at auto mode, hence why you are not seeing the signal step-up sequence.

That would indeed explain the longish period at red. And judging from the amount of time, the distance to the next signal from the station starter must well be such that any, say, bypassing train through the station would have time to stop within the block (whatever the speed limit there is)?

 

[Metroland]

It's not that, the next signals in either direction tend to get left on auto working (some controlled signals can be set to auto) to cut down workload, the routes only killed when there is something to cross over. So when the signals in the station are left at danger, and then cleared, the effect is they go straight to green because the next signals are at least yellow on the down or double yellow on the up.

If the next signals on the down/up are at red, the platform signals would still show yellow.

Depends on the style of working from the signaller, some prefer to have platform signals always cleared, long is there is a train about, others held at red so not to block in other possible workings.

 

[At_traction]

Hmm, now this is in fact getting more complicated...

Do I have to understand from above that there in fact, in case of four-aspect signals, would be three vacant blocks before the signal that triggers the release of starter signal (or any signal previously at danger)?

preceeding train - red ------ yellow ------ dbl.yellow ------ green - following train

Would the controlled section of railway normally have track magnets and at automatic operation work like a non-manual sequence of blocks with incremental change towards green, but only the third magnet after the red (in the OXF case the starter) triggering the signal and making it change at once from red to green?

At a loss.

 

[Railsigns]

In order for a controlled signal to display a green aspect, two fundamental conditions require to be satisfied:

1) The requisite number of blocks ahead of the signal must be proved clear

and

2) The signalman needs to have set the route from that signal (otherwise it stays at red).

 

[At_traction]

Thanks, so in a sense I had understood semi-correctly.

In fact, while pondering this dilemma I made a search on the net as well as went through a slew of links I had on signalling, the first site being yours -- is there anything further on the matter on your site...? I guess the auto mode is indicated in some form for a driver.

 

[GB]

I guess the auto mode is indicated in some form for a driver.

Strictly speaking no as he does not need to know if a controlled signal is set to auto or not, all he cares about is the signal aspect and that it is the correct aspect. However on the signal plate it will say whether the signal is Auto, Semi Auto or Controlled etc etc

Would the controlled section of railway normally have track magnets and at automatic operation work like a non-manual sequence of blocks with incremental change towards green, but only the third magnet after the red (in the OXF case the starter) triggering the signal and making it change at once from red to green?

"Track magnets" have nothing to do with the operation of the signalling.

 

[At_traction]

"Track magnets" have nothing to do with the operation of the signalling.

What's so quotational about term "track magnet", aka balise as it's also otherwise known?

In this case we're talking they perhaps don't, but that's what I'm trying to find out. I simply assumed that as there would be a yellow(s) between the signals at danger (protecting the preceeding train) and one recently cleared - as Metroland stated, if I assumed correctly - they'd have to be triggered somehow, at the least the signal that eventually releases the red that would've been (for example) the OXF starter.

 

[ralphchadkirk]

Balise's or track magnets have nothing to do with the signalling of trains. In the Oxford area I believe it is TCB.

 

[Metroland]

Hmm, now this is in fact getting more complicated...

Do I have to understand from above that there in fact, in case of four-aspect signals, would be three vacant blocks before the signal that triggers the release of starter signal (or any signal previously at danger)?

preceeding train - red ------ yellow ------ dbl.yellow ------ green - following train

Would the controlled section of railway normally have track magnets and at automatic operation work like a non-manual sequence of blocks with incremental change towards green, but only the third magnet after the red (in the OXF case the starter) triggering the signal and making it change at once from red to green?

At a loss.

Going south from Oxford, you have a couple or four aspect signals, both of which are controlled, then a series of three aspect signals, leading to an auto section between Kennington Junction and Appleford. However all the signals leading south can be set to automatic working if so desired on straight routes.

Going north Oxford you have a series of three aspect signals: ones on the platform end, another one protecting Oxford North and finally one protecting Wolvercote Junction. Beyond that there is a two aspect auto section as far as Aynho Junction - which is completely automatic.

Again the controlled signals can be set to auto working in so desired but they can only work auto on straight routes. Any divergences toward Oxford, Bicester, Worcester and other routes need a manual intervention.

When in auto mode the signals merely work via tracks circuits and the aspect displayed ahead. This is quite common on entrance-exit panels, as opposed from IECCs which can have a full automatic route setting.

The aspect sequence for three aspect signals are green, yellow, red. And for four aspect signals, the sequence is green, double yellow, yellow, red.

The Didcot-Birmingham line is actually a complete mix of signal types, all thanks to its rather chequered history.

 

[GB]

What's so quotational about term "track magnet", aka balise

Edit: OK I see what you are talking about now. Magnets are not used to signal trains on conventional lines.

 

[At_traction]

Going south from Oxford, you have a couple or four aspect signals, both of which are controlled, then a series of three aspect signals, leading to an auto section between Kennington Junction and Appleford. However all the signals leading south can be set to automatic working if so desired on straight routes.

Going north Oxford you have a serious of three aspect signals: ones on the platform end, another one protecting Oxford North and finally one protecting Wolvercote Junction. Beyond that there is a two aspect auto section as far as Aynho Junction - which is completely automatic.

I knew that someone would/could dig these out...

But to further nag about the original dilemma: if the route was set at the signalling control (NX panel or IECC) to the whole extent of free blocks after the preceding train, could it be that in my case the starter switched straight from red to green as there were already enough blocks unoccupied ahead as the exit signal was entered into the system? Is that the solution?

Again the controlled signals can be set to auto working in so desired but they can only work auto on straight routes. Any divergences toward Oxford, Bicester, Worcester and other routes need a manual intervention.

When in auto mode the signals merely work via tracks circuits and the aspect displayed ahead. This is quite common on entrance-exit panels, as opposed from IECCs which can have a full automatic route setting.

Is the "principal" difference between a sequence of NX panel settings and the ARS routing the timetable information incorporated into the latter's operating "database"?

More on the issue of routing discrepancies, how are the rail swing bridges incorporated into routing conventions? Are the bridge sections controlled/overridden by a local signal box at the bridge in case of opening requests?

The Didcot-Birmingham line is actually a complete mix of signal types, all thanks to its rather chequered history.

Ah, the joys of being a pioneer with differing commercial and technical interests.

But thanks for all repliers so far.

 

[Tomnick]

But to further nag about the original dilemma: if the route was set at the signalling control (NX panel or IECC) to the whole extent of free blocks after the preceding train, could it be that in my case the starter switched straight from red to green as there were already enough blocks unoccupied ahead as the exit signal was entered into the system? Is that the solution?

Quite simply, yes! The signal will clear to a proceed aspect when the route is set, as long as the line is clear as far as the overlap beyond the next signal (and a few other conditions, such as all points in the route being detected in the correct position etc.). What proceed aspect the signal clears to will be determined by the aspect displayed by the next signal, to give the correct aspect sequence - but not by looking directly at TC occupation beyond the next signal's overlap. Generally, the route would be set from one signal to the next only (though I believe 'long routes' are sometimes provided?); if that signal is either automatic or a route is already set from it, then it will possibly already be displaying a proceed aspect, allowing the signal in rear to immediately step up to something less restrictive than a single yellow. I've seen signals, in the situation you describe, cycle very quickly through the whole sequence (R-Y-YY-G - in much less than a second) when cleared, rather than stepping straight to green - I understand that's a consequence of the relay interlocking checking the aspect of signals in advance (HR and HHR relays???).

Is the "principal" difference between a sequence of NX panel settings and the ARS routing the timetable information incorporated into the latter's operating "database"?

Not really, the two are quite different kettles of fish! An NX panel is one 'system' by which the Signalman controls the interlocking, mainly by calling routes between signals - much the same as IECC or a lever frame are methods of control. ARS is a system which can call routes itself, to reduce the Signalman's workload - and that can be used on different methods of control (including an NX panel?).

More on the issue of routing discrepancies, how are the rail swing bridges incorporated into routing conventions? Are the bridge sections controlled/overridden by a local signal box at the bridge in case of opening requests?

'Normal' detection on a swing bridge will be one of the conditions checked before a route can be set or before the signal can display a proceed aspect (the former, I suspect). The swing bridge at Keadby Canal Junction (west of Scunthorpe) is interesting, as it's located in a long section of auto signals supervised from Scunthorpe PSB. The protecting signals are plated as semi-automatic, so the Signalman at Keadby can replace them to Danger himself (obviously after checking that no trains are approaching) and operate the swing bridge or level crossing. Until both are normal again, the interlocking won't allow the signals to be cleared.

 

[Metroland]

I've seen signals, in the situation you describe, cycle very quickly through the whole sequence (R-Y-YY-G - in much less than a second) when cleared, rather than stepping straight to green - I understand that's a consequence of the relay interlocking checking the aspect of signals in advance (HR and HHR relays???).

Hmn, Westbury is like that for sure and I think Exeter, possibly Leicester. All late 1980s vintage RRI.
--- old post above --- --- new post below ---

Ah, the joys of being a pioneer with differing commercial and technical interests.

But thanks for all repliers so far.

Well the route is always swapping regions for a start. You have IECC control at the South end from the 1990s Swindon B, the early 1970s Oxford PSB, which is an NX panel. SSI panel in Banbury south, lever frames, including colour lights and GWR lower quadrant semaphores at Banbury itself, then the latest Westcad controlled SSI installation at Leamington Spa with 4 aspect signalling, which itself fringes with two workstations in the West Midlands signalling control centre at Saltley for the Birmingham/Coventry area.

You have sections of 2, 3 and 4 aspect, route relay interlockings, semaphores, upper and lower quadrant signals, and several generations of SSI. (Solid State interlocking)

More photos of the WR signalling here

http://photos.signallingnotices.org.uk/index.php?col=5

and Midlands Signalling here

Bits about the West Midlands control here, which will eventually take over as far south as Wolvercote, Oxford, fringing with the Thames Valley control centre in Didcot.

http://www.ccd-design.co.uk/p/Our-P...ay-Signalling-Centre-for-the-West-Midlands-2/


http://photos.signallingnotices.org.uk/index.php?col=4

 

[At_traction]

ARS is a system which can call routes itself, to reduce the Signalman's workload - and that can be used on different methods of control (including an NX panel?).

But it would utilize in routing the timetable data in terms of intermediate junctions and stations and thus (in case of no conflicts or unfortunate diversions outside the routed/programmed path) can be seen as a series of successive manual routing entries?

More photos of the WR signalling here
http://photos.signallingnotices.org.uk/index.php?col=5

Just to get this out of system (then to be forgotten forever): I assume this is the OXF station (diagram looking eastward), with the section "up" from the station. At least the locations of points are similar to views in aerial photos (I'll get my life) and the platform indications as well as east siding also correspond.

In the first photo there is a train in a similar situation as the ones in my case, occupying the up platform, with no route set (the starter signal indicator seems a bit worn, but I assume it's at danger -- next to that is the feathered "bypass" rail signal, at danger too).

In the second the route has been set, leading from the platform rail to the middle "bypass" up rail, with also the first signal after the starter cleared to green. I assume that the bypass rail signal to the left (where the route trail makes a divergence), with indicator extinguished, is connected to the straight route along bypass rail, preventing entrance to the section after the points?

There also seems to be a train on the extension of the up platform rail, which is a h*ll of a long section, actually. Could that be awaiting to be passed by the (more) express service?

 

[Metroland]

Yep, first photo is the north end of the station. There’s a train in the up platform, with no route/signal set.

Second picture is showing the up and down passenger loop. You can bring trains from the down direction, via that loop, into the up platform. The train in that pic could be travelling either direction, but most probably waiting to enter the up platform. You can also get onto the up line from there, toward Didcot. Trains are not usually held to pass, only very occasionally all stops or freight. Sometimes freight is run up to the first signal out of the station on the up line, and passenger trains run around it.

One thing to note about running trains through the up platform loop, apart from being 25mph all the way, you only get a yellow signal on the platform signal, with a theatre indicator. It doesn’t clear to green, even with the next signal set. This is because of the 25mph restriction back onto the main line, whereas if they go straight to the up main, it’s 25mph across the points then 90mph.

On the up line, if the train has passed the first signal out of the station on the gantry, it will show a yellow if the route is set, double yellow if it’s passed Hinksey, the signal protecting Kennington loop, and green if it’s clear of the signal at the end of the loops. Beyond that is an auto section, with the fringe the other side of Radley with Swindon B IECC.

Signalling panels are schematic and not shown to scale, in fact some of the sections shown here http://photos.signallingnotices.org.uk/photo.php?pc=66&p=S7300608.JPG are several miles long, and are shown very short. Note the fully automatic signals do not have their aspect shown. And controlled signals only show green, for off (that’s anything but the least restrictive aspect, usually red) and red.

IECCs tend to have full ARS worked from a pre-programmed timetable processor. There’s a certain amount of AI (artificial intelligence), it usually looks for trains in the area, compares the timetable and goes from there, giving certain trains, certain priorities. It doesn’t tend to be that ‘bright’, it’s just an aid really to cut down workload, so one man can operate a bigger area.

Finally, there is no such thing as ‘starter’ signals with track circuit block areas (that’s most panels/IECC), any signal capable for showing a red is called a ‘stop signal’

Home, distant, and starter/section signal is absolute block parlance, a system still in use at Banbury with block bells and at Worcester, and still quite a few secondary lines.

The starter signal, or more correctly the section signal is the last stop signal at station limits (which is the area immediately outside the signalbox). This controls access to the box in advance’s block section or an intermediate block home section, which is an additional stop signal and distant (often colour light) to break up long sections.

Oh on a very anoraky note, Oxford is a Swiss Integra panel, made up of all metal domino tiles with small bulbs, usually made by Henry Williams ltd of Darlington who are still in business. Anyway, the all metal panels tend to get very hot! Later panels are plastic and LEDs, which has now all been superseded by LCD VDUs (although Basingstoke new is actually a panel). Some of the Intega panels have worn badly, Oxford being a good example. The other WR boxes (which tend to use the same system) have faired better, even though the vast majority date from the 60s, such as Swindon http://photos.signallingnotices.org.uk/photo.php?pc=81&p=IMG_7997.JPG

They are museum pieces now, in fact if you go into the Swiss transport museum at Lucern, you will see something shockingly similar!

 

[Railsigns]

In fact, while pondering this dilemma I made a search on the net as well as went through a slew of links I had on signalling, the first site being yours -- is there anything further on the matter on your site...?

Not specifically relating to what we've been discussing, no. A page on the principles of Track Circuit Block is in the pipeline, but until I get round to writing it, some of the pages linked here might help you, especially the first few:

http://www.railsigns.co.uk/info/info.html

I guess the auto mode is indicated in some form for a driver.

No, this information is of no interest to the driver.

 

[Metroland]

Yep, controlled signals set to auto are of no interest to drivers. The dead giveaway on the lineside is the signals always tend to return to their least restrictive aspect. It's always at the signallers discretion whether they are set to auto. Remember it's not ARS, so he/she must remember to pull the route and reset to a diverging line if required, a very common cause of wrong routings.

In IECC areas, ARS aims to keep 2 signals at green ahead of drivers. On the lineside, a lot of controlled signals will always be held at red, and will start stepping up to green as trains approach. This can be observed at Ealing broadway for instance, or on lines running out of Liverpool street.

Confusingly lines out of Liverpool street also have additional yellows, because the sections are so short. Unique on NR.

 

[At_traction]

Not specifically relating to what we've been discussing, no. A page on the principles of Track Circuit Block is in the pipeline, but until I get round to writing it, some of the pages linked here might help you, especially the first few:

http://www.railsigns.co.uk/info/info.html

Thanks, I will peruse through the pages. Very good info on the site in general.

Yep, controlled signals set to auto are of no interest to drivers.

As for "proper" automatic signals then, is there generally a way of informing the controlling centers/signal boxes of defective signals that are unilluminated or ones that have reverted to danger? Are their aspects or status generally shown to the controllers or do the drivers have to inform by phone/radio about abnormally waits at a red...

The dead giveaway on the lineside is the signals always tend to return to their least restrictive aspect. It's always at the signallers discretion whether they are set to auto. Remember it's not ARS, so he/she must remember to pull the route and reset to a diverging line if required, a very common cause of wrong routings.

Does this scenario refer only to situations where a "straight" route has already been set and then cancelled in favour of a diversion or also the other way around? Is it correct to assume that there has to be a route set for a train before the selected section can be set to automatic working and do the auto-set controlled signals revert to non-auto after the train has passed the set route? Or does the automatic status remain, with the signals returning to green through yellow(s) after the train? Which in turn would require controller's input in case of routings differing from the ones the preceeding train has taken?

In IECC areas, ARS aims to keep 2 signals at green ahead of drivers. On the lineside, a lot of controlled signals will always be held at red, and will start stepping up to green as trains approach.

That's interesting.

This can be observed at Ealing broadway for instance, or on lines running out of Liverpool street.

Confusingly lines out of Liverpool street also have additional yellows, because the sections are so short. Unique on NR.

One would assume that during commuter rush hours there would be little more than yellow and red aspects on the platforms due to the Bishopsgate "gap's" effect extending back... Is it two up and two down lines there?

There's rather a buzz of departures:
http://www.railforums.co.uk/attachment.php?attachmentid=2964&stc=1&d=1281368990

 

[GB]

Does this scenario refer only to situations where a "straight" route has already been set and then cancelled in favour of a diversion or also the other way around? Is it correct to assume that there has to be a route set for a train before the selected section can be set to automatic working and do the auto-set controlled signals revert to non-auto after the train has passed the set route? Or does the automatic status remain, with the signals returning to green through yellow(s) after the train? Which in turn would require controller's input in case of routings differing from the ones the preceeding train has taken

Yes, a route has to be set prior to the signaller selecting the automatic button and no, after a train passes it does not cancel the automatic function.

It will need an input from the signaller to disable the automatic working allowing him to set other routes as you rightly suggest.

 

[Metroland]

With automatic signals, if trains are held at red signals they must contact the signaller, by whatever means necessary (usually signal post telephone, or cab secure radio, or even mobile phone) to find out why they are detained at the signal. If they cannot contact the supervising box by any means they are allowed to pass it at extreme caution and phone in at the next signal.

In some IECCs the automatic signals are indicated, but the general practice with panels is they are not. The indication has to be worked out from track circuit occupation.

In older installations not every automatic signal can be replaced to danger, and then when it is, it cannot be guaranteed. Originally, in many PSBs in the 60s, none of the automatic signals could be replaced, and emergency replacement was fitted later on selected signals after a few mishaps.

All new ones, every signal can be replaced and the indication is guaranteed.

 

[At_traction]

OK, thanks for all on the in-depth info.