Signalling Saltfleet - on the bench

[Pannier Tank]

49 minutes ago, Dave-5-5-7 said:

Greetings, M0YAA (transferring to G7HJR when the paperwork goes through) IO93WN. 

 

73

 

Hello Dave,

 

Have you seen  https://www.rmweb.co.uk/community/index.php?/topic/153901-any-ham-radio-license-holders-out-there/

 

 

Stay Safe,

 

73 GD DX

 

de David G4NVB

 

[imt]

15 hours ago, Dave-5-5-7 said:

It's basically transcient TCB, with train describer. The code will flash up on a display, and the signalman will turn the accept button to the right and "Train coming from" will flash for a few seconds and illuminate. This mimics the system checking the circuit is correct for an authority to be issued. Box in rear locks release, signals clear, train runs. Once train arrives signalman turns switch back to centre and the system clears. In order to send, signalman enters code into TD, moves switch to left and presses "SEND" once box in rear acknowledges the locks release and MS938 can be cleared. 

 

Dave, I know you have better (certainly more enjoyable) things to do.  Thank you very much for the explanation.  Please nicely can I ask some more questions?

 

1) The "Send" button does not appear on your diagram yet - would it be near the direction switch or on the train describer?  Presumably the box in the rear "acknowledging" is that signaller putting his switch into "Train coming from" mode?

 

2) The "L" on the point switches I assume is "Locked" i.e. the governing signal has now changed so this point cannot be changed? Or am I completely wrong?

 

3) What is the white circle by the GPLS representations.  Is this a light indicating a TRTS plunger?  If it is it is presumably cancelled by returning the GPLS to "on".

 

4) I presume the SI/MI on 37, 941 and 39 will set automatically as a result of the route set by the signaller through the points?

 

5) why are there no SI/MI on the starter signals?

 

6) it looks as if traffic on the line to the Oil Terminal is handled differently to Skegness - presumably because there isn't a signaller there?

 

7) Why is there an extra light on the faceplates for the signals 940, 941 and 950.  I at first thought it might show that the JI/SI had operated, but then there would need to be similar lights on other signals too?  Then I thought it was the level crossing operated - but that's not true of 941?

 

You might not have noticed that the TC in advance of 940 is unnumbered.  I think it should be 0248 and the current 0248 should be 0300 - the gap in the numbers is there.

[Dave-5-5-7]

9 hours ago, imt said:

 

Dave, I know you have better (certainly more enjoyable) things to do.  Thank you very much for the explanation.  Please nicely can I ask some more questions?

 

1) The "Send" button does not appear on your diagram yet - would it be near the direction switch or on the train describer?  Presumably the box in the rear "acknowledging" is that signaller putting his switch into "Train coming from" mode?

 

2) The "L" on the point switches I assume is "Locked" i.e. the governing signal has now changed so this point cannot be changed? Or am I completely wrong?

 

3) What is the white circle by the GPLS representations.  Is this a light indicating a TRTS plunger?  If it is it is presumably cancelled by returning the GPLS to "on".

 

4) I presume the SI/MI on 37, 941 and 39 will set automatically as a result of the route set by the signaller through the points?

 

5) why are there no SI/MI on the starter signals?

 

6) it looks as if traffic on the line to the Oil Terminal is handled differently to Skegness - presumably because there isn't a signaller there?

 

7) Why is there an extra light on the faceplates for the signals 940, 941 and 950.  I at first thought it might show that the JI/SI had operated, but then there would need to be similar lights on other signals too?  Then I thought it was the level crossing operated - but that's not true of 941?

 

You might not have noticed that the TC in advance of 940 is unnumbered.  I think it should be 0248 and the current 0248 should be 0300 - the gap in the numbers is there.

 

happy to answer imt. 
1) The Train Describer is a separate  unit from the panel diagram. 
2) Quite correct, the points will remain locked until the signal reverts to danger, relevant track circuits and overlaps clear and the switch is replaced. 
3) The style the panel is based on has indicators on the diagram, red/green for main signals, red/white for the GPLS. At the time multicoloured LEDs were not available so the clear indication was placed above the GPL. The eventual plan is for it to be an NX panel, although I'm enjoying this as an IFS so I might adjust the diagram to make allowances for the signal indications being shown above the switches. 

4) That is correct. The final diagram won't show route indicators, this is still a technical diagram so it holds info for me as I wire up. 
5) They've already been removed from the diagram. However for absolute aspects to supply a signal with RI and PL will be £110 per unit. 
6) Correct, Tetney Oil is a former line that was cut back to the oil terminal. It's worked as OTIS, as in One Train In Section. When the train comes back the signalman will get an audible alarm as the train approaches.

7)  The extra lamp indicates the Position Light has illuminated rather than a main aspect. 

The TC error has also been addressed, I moved the crossing and haven't realigned the Track Circuits. 

Edited April 27, 2020 by Dave-5-5-7

[Dave-5-5-7]

Latest corrections and additions. Including the control buttons for the CCTV crossing 

[imt]

It's the pest again! I am really enjoying learning more about this kind of panel and how it is operated.  I have been allowed (under close supervision needless to say) to pull levers and operate NX switches, but though I have seen photos before I have never come across a real IFS panel.  May I ask again?  Thank you for your patience............

 

1) you didn't say where that "Send" button was for the single line operations, and it doesn't appear on this update.  Is there one or have I misunderstood something?

 

2) I understand that none of the signals will illuminate on the diagram and lights will only show on their switch face plates.  The TCs will not illuminate to show a selected path nor are there any lights to indicate point settings - only the "N" or "R" on their switch faceplates.  Normal lie can (as usual) be seen from the way things are drawn on the diagram.  The TCs will I believe illuminate (or the red light at each end will) when occupied?

 

3) I think I can see how the various aspects will operate.  IF points are set (say) from P1: 1768 and 1773, then setting 942 to  proceed  (switch to the right) would illuminate the PLS for the move to the MPD (and presumable a MI or the SI with advice on the route?).  Nothing will be seen on the panel itself but the "R" and "L" lights on switches 1768 and 1773  and the switch 942 will show "L"  and the white light top right to indicate the PL.

 

4) if 3) is approximately right, then you are missing the shunt route to "Up Main" on 942 - its there on 38, 944 and 946.

 

5) Stupid question - why would you want a route from the DMU Depot to the Oil Terminal? 

 

6) You have inserted a new crossover which gives 2 new routes from 936 but they are not on the faceplate.  I presume only 1761 is required for  the Power Station since normal lie for 1759 is in that direction.  Would it not be better if normal was to Skegness?  You might also like to know that 945 and 955 are misnamed - aren't they the other way round?  Should 955 not need 1759 to reverse?

Edited April 28, 2020 by imt

[Dave-5-5-7]

4 hours ago, imt said:

It's the pest again! I am really enjoying learning more about this kind of panel and how it is operated.  I have been allowed (under close supervision needless to say) to pull levers and operate NX switches, but though I have seen photos before I have never come across a real IFS panel.  May I ask again?  Thank you for your patience............

 

1) you didn't say where that "Send" button was for the single line operations, and it doesn't appear on this update.  Is there one or have I misunderstood something?

 

2) I understand that none of the signals will illuminate on the diagram and lights will only show on their switch face plates.  The TCs will not illuminate to show a selected path nor are there any lights to indicate point settings - only the "N" or "R" on their switch faceplates.  Normal lie can (as usual) be seen from the way things are drawn on the diagram.  The TCs will I believe illuminate (or the red light at each end will) when occupied?

 

3) I think I can see how the various aspects will operate.  IF points are set (say) from P1: 1768 and 1773, then setting 942 to  proceed  (switch to the right) would illuminate the PLS for the move to the MPD (and presumable a MI or the SI with advice on the route?).  Nothing will be seen on the panel itself but the "R" and "L" lights on switches 1768 and 1773  and the switch 942 will show "L"  and the white light top right to indicate the PL.

 

4) if 3) is approximately right, then you are missing the shunt route to "Up Main" on 942 - its there on 38, 944 and 946.

 

5) Stupid question - why would you want a route from the DMU Depot to the Oil Terminal? 

 

6) You have inserted a new crossover which gives 2 new routes from 936 but they are not on the faceplate.  I presume only 1761 is required for  the Power Station since normal lie for 1759 is in that direction.  Would it not be better if normal was to Skegness?  You might also like to know that 945 and 955 are misnamed - aren't they the other way round?  Should 955 not need 1759 to reverse?

Pest? no, in fact it's helpful having a second pair of eyes over the updates, 

1) The Send button is on the train describer which is a completely different piece of equipment, as in not on the panel itself. 
2) It's become a bit of a hybrid. On an NX panel the signals on the diagram light up, on (this style) of IFS panel, the indications on the switches do and the diagram is just a diagram. The TC's will illuminate and show the position of the train. 
3) Correct
4) Ran out of space on the description. Going to see if it's possible to adjust the position of the diagram. 
5) You wouldn't. however until a few years ago history wise, the Up/Down Tetney ran all the way to Cleethorpes. As such, the route still exists. 
6) Not altered the faceplate descriptions yet, and yes, 945/955 are in error. It's been an off scene requirement as my coal trains are too long for the fiddle yard lines accessible from P3 at Skidbrooke so the trains need to run on the main. 

[imt]

One last diversion and I'll keep quiet for a while - after all it's your thread!  Happy to take a second look at any further upgrades.

 

You will have seen the photo of my Single Button Route Setting panel above.  I was very happy with my electronics but (for an exercise) I was trying to automate it using JMRI which I have just about achieved.  I am fascinated by Simon Paley's Collingwood but NX is inappropriate for such a small layout.  On the other hand your diagram etc has inpired me to see if I could IFS my layout.  Of course it wouldn't be using relays as you will be, but I need to have some idea of the relay logic which would be behind the diagram and switches so I can program that into JMRI.

 

1) One thing I don't understand (ho, ho one of many!) is how incorrect changing of panel switches is prevented (if at all).  When points have been set and the signal set to proceed, and all have their "L" lights on I presume they are not physically locked on the panel (they will be in the logic) and so panel switches could be moved.  If a panel switch was moved whilst locked, the physically point (say) won't move, but presumably it will when the covering signal is set on again?????  Is there a buzzer to warn against this action, or is it OK?

 

2) Presumably changing a signal switch cancels everything whether the train has set off or no?  If not how is that prevented, or is this just signaller discipline and process as it would have been in a manual box.

 

3) When points are being set is there any notice taken by the logic of TCs??  Would 0302 being occupied preven 1702 from being set?

 

4) As for switching a signal to proceed, presumably then the logic has to check all possible routes to see which points are set before locking those points and selecting the appropriate aspect - presumably also appropriate TCs will be checked so if (for example) 0302 is occupied the aspect on (say) 946 will not change.

 

Its a whole different way of looking at things  ...........

[St. Simon]

43 minutes ago, imt said:

One last diversion and I'll keep quiet for a while - after all it's your thread!  Happy to take a second look at any further upgrades.

 

You will have seen the photo of my Single Button Route Setting panel above.  I was very happy with my electronics but (for an exercise) I was trying to automate it using JMRI which I have just about achieved.  I am fascinated by Simon Paley's Collingwood but NX is inappropriate for such a small layout.  On the other hand your diagram etc has inpired me to see if I could IFS my layout.  Of course it wouldn't be using relays as you will be, but I need to have some idea of the relay logic which would be behind the diagram and switches so I can program that into JMRI.

 

1) One thing I don't understand (ho, ho one of many!) is how incorrect changing of panel switches is prevented (if at all).  When points have been set and the signal set to proceed, and all have their "L" lights on I presume they are not physically locked on the panel (they will be in the logic) and so panel switches could be moved.  If a panel switch was moved whilst locked, the physically point (say) won't move, but presumably it will when the covering signal is set on again?????  Is there a buzzer to warn against this action, or is it OK?

 

2) Presumably changing a signal switch cancels everything whether the train has set off or no?  If not how is that prevented, or is this just signaller discipline and process as it would have been in a manual box.

 

3) When points are being set is there any notice taken by the logic of TCs??  Would 0302 being occupied preven 1702 from being set?

 

4) As for switching a signal to proceed, presumably then the logic has to check all possible routes to see which points are set before locking those points and selecting the appropriate aspect - presumably also appropriate TCs will be checked so if (for example) 0302 is occupied the aspect on (say) 946 will not change.

 

Its a whole different way of looking at things  ...........

 

Hi,

 

Sorry to hijack Dave's thread a little, but here is my thoughts on your questions:

 

1) You're right with the fact that we don't want the signaller to inadvertently (or intentionally)  pre-select the position of a switch, so on I.F.S. or O.C.S. panels, there is 'Anti Pre-Selection' circuits added in to prevent this. If a signaller turns a switch that is locked electrically, these circuits force the signaller to return the switch to its 'normal' position and then turn the switch again. There's no warning as such, it is just that nothing will happen until you return the switch to its normal position and turn it again.

 

2) I'm not sure quite what you mean by 'cancels everything', but it does depend on what has happened to the signal between the switch being turned and then reset. If the signal has remained 'ON', then the route can cancel and everything can unlock almost immediately. If the signal has cleared to 'OFF' but a train has not gone past it into the route , then resetting the switch will cause the signal to go back to danger immediately and then the approach locking would have to time out before anything is unlocked. If the signal had cleared to 'OFF' and a train gone past it into the route, the route will be cancelled immediately (subject to the route locking). I'm going to put up a post in my thread about route release this evening.

 

3) Yes, track circuits are used in the point calling logic. If the track circuits directly over the points are occupied, then the points will deadlocked (unable to be moved) until the track circuit becomes unoccupied. Although not directly related to track circuits, points will remain locked if there is a train approaching them. See this post in my thread for more details.

 

4) On an I.F.S. panel such as Dave's, switches are provided for each route rather than each signal, i.e. a signal with 2 routes would have two switches, one for each route from that signal. So, when the switch for that route, the 'route level' of the interlocking will check that points are set in the correct position or are free to move to the correct position, whilst the 'aspect level' will check the track circuit occupation.

 

I hope that helps.

 

Simon

[bécasse]

Conversation overheard:

Driver to bobby - "Hallo signalman, driver of 2B13 standing at XX154 at xxx station. I have a green aspect without a feather and the road is set straight on, am I being required to divert?"

Bobby to driver - "Sorry driver, my mistake, set the wrong road, go forward one coach length and then set back into the platform. That will clear the locking and I can reset the route immediately. Proceed as normal once you get the aspect and feather."

[St. Simon]

22 minutes ago, bécasse said:

Conversation overheard:

Driver to bobby - "Hallo signalman, driver of 2B13 standing at XX154 at xxx station. I have a green aspect without a feather and the road is set straight on, am I being required to divert?"

Bobby to driver - "Sorry driver, my mistake, set the wrong road, go forward one coach length and then set back into the platform. That will clear the locking and I can reset the route immediately. Proceed as normal once you get the aspect and feather."

Hi,

 

Whilst it is a good story, it wouldn’t clear the locking immediately!

 

You need the berth track clear to release the locking.

 

Simon

 

[Dave-5-5-7]

2 hours ago, imt said:

One last diversion and I'll keep quiet for a while - after all it's your thread!  Happy to take a second look at any further upgrades.

 

You will have seen the photo of my Single Button Route Setting panel above.  I was very happy with my electronics but (for an exercise) I was trying to automate it using JMRI which I have just about achieved.  I am fascinated by Simon Paley's Collingwood but NX is inappropriate for such a small layout.  On the other hand your diagram etc has inpired me to see if I could IFS my layout.  Of course it wouldn't be using relays as you will be, but I need to have some idea of the relay logic which would be behind the diagram and switches so I can program that into JMRI.

 

1) One thing I don't understand (ho, ho one of many!) is how incorrect changing of panel switches is prevented (if at all).  When points have been set and the signal set to proceed, and all have their "L" lights on I presume they are not physically locked on the panel (they will be in the logic) and so panel switches could be moved.  If a panel switch was moved whilst locked, the physically point (say) won't move, but presumably it will when the covering signal is set on again?????  Is there a buzzer to warn against this action, or is it OK?

 

2) Presumably changing a signal switch cancels everything whether the train has set off or no?  If not how is that prevented, or is this just signaller discipline and process as it would have been in a manual box.

 

3) When points are being set is there any notice taken by the logic of TCs??  Would 0302 being occupied preven 1702 from being set?

 

4) As for switching a signal to proceed, presumably then the logic has to check all possible routes to see which points are set before locking those points and selecting the appropriate aspect - presumably also appropriate TCs will be checked so if (for example) 0302 is occupied the aspect on (say) 946 will not change.

 

Its a whole different way of looking at things  ...........

1) with the lock light on your can flick point switches to your hearts content, the points won't move and the signal will not revert to danger. The simplest way I found to do this was when the signal relay is energised the lock light works through the third set of contacts and the forth contact removed the power feed for the switch. It's not prototypical but it works. 

2) Reverting a signal will activate a "backlock" and the route remains locked for 2 minutes. 

3) eventually. At the minute the circuitry has TC inputs for the signalling system but they're grounded as the TC modules will take a while to construct. 

 

4) yes, the principle of Track Circuit Block is a signal will only show a proceed aspect when the line is clear up to and including the overlap of the next signal, and the route is set and locked. 

 

[Dave-5-5-7]

2 hours ago, imt said:

One last diversion and I'll keep quiet for a while - after all it's your thread!  Happy to take a second look at any further upgrades.

 

You will have seen the photo of my Single Button Route Setting panel above.  I was very happy with my electronics but (for an exercise) I was trying to automate it using JMRI which I have just about achieved.  I am fascinated by Simon Paley's Collingwood but NX is inappropriate for such a small layout.  On the other hand your diagram etc has inpired me to see if I could IFS my layout.  Of course it wouldn't be using relays as you will be, but I need to have some idea of the relay logic which would be behind the diagram and switches so I can program that into JMRI.

 

 

 Incidentally I'm looking into JMRI as an idea to provide an Over-ride on screen IECC type panel to work the fiddle yard instead of another operator when I'm on my own. 

 

Having trouble getting my head round the panel construction. 

[imt]

13 hours ago, Dave-5-5-7 said:

Having trouble getting my head round the panel construction. 

 

Oh yeah, tell me about it!  I had horrendous trouble trying to understand it - and even now I'm not sure.  The photo below is my current attempt to get Ullapool into JMRI.  It works but I am not happy with it.  Please if you like PM me about it, or maybe you would like a separate thread?  I know Simon Paley has had similar problems starting up and he too may be able to comment.

 

JMRI using Logix programming is infinitely powerful BUT it does need careful understanding first.  It is sequenced stateful operations control not really programming like using a programming language - BASIC or Python.  If you understnd relays and so on you should have no trouble at all.  Honestly.

[imt]

14 hours ago, St. Simon said:

4) On an I.F.S. panel such as Dave's, switches are provided for each route rather than each signal, i.e. a signal with 2 routes would have two switches, one for each route from that signal. So, when the switch for that route, the 'route level' of the interlocking will check that points are set in the correct position or are free to move to the correct position, whilst the 'aspect level' will check the track circuit occupation.

 

Sorry Simon I don't understand that bit at all.  What you are describing is a Single Button Route Setting which I have used and that isn't how Dave's panel looks - which seems to have a set of individual switches for each object (signal or point).  I may have horrendously misunderstood this whole subject and now I have really trampled on Dave's patch.  Maybe you should PM me and put me right, please very nicely!!!!!!!!

[bécasse]

13 hours ago, St. Simon said:

Hi,

 

Whilst it is a good story, it wouldn’t clear the locking immediately!

 

You need the berth track clear to release the locking.

 

Simon

 

I assure you that it did clear the locking and that the train was able to depart on its correct route well before any two minute delay device would have struck in. I thought at the time that it wasn't the first time that the bobby concerned had done it. It probably saved him from a Form 1 too, although it is possible that the train was wrongly described on the describer.

[Dave-5-5-7]

1 hour ago, imt said:

 

Sorry Simon I don't understand that bit at all.  What you are describing is a Single Button Route Setting which I have used and that isn't how Dave's panel looks - which seems to have a set of individual switches for each object (signal or point).  I may have horrendously misunderstood this whole subject and now I have really trampled on Dave's patch.  Maybe you should PM me and put me right, please very nicely!!!!!!!!

The panel has become an amalgamation of a mid 80s NX panel (seen here at Pasture Street Grimsby) 

 

and a more modern variant of an IFS which is basically a lever frame with rotary switches) seen here at Immingham West, which coincidentally is the last signal I passed out to work most recently. This has pushed me towards an IFS panel. 

[imt]

23 minutes ago, Dave-5-5-7 said:

a more modern variant of an IFS which is basically a lever frame with rotary switches) seen here at Immingham West

 

OK I thought that was what we were talking about.  Those photos match what you have on your diagram and I thought I understood that.  It was Simon's "switches are provided for each route rather than each signal," that threw me.  You have switches per point/signal not per route which is what I thought IFS was all about.

[The Stationmaster]

Being of an age where OCS means much more to me than IFS could somebody explain the difference because the Immingham example has only one switch for each signal irrespective of the number of routes that lead from that signal.    Therefore there would presumably be no route locking (by which I mean locking solely between routes rather than locking points and signals by other points & signals).  So IFS effectively takes control panel design back to the early part of the 1930s before route setting  (rather than individual point and signal control) became the usual way of doing things.

 

In contrast the 'Ullapool' panel appears to be, in effect, an OCS arrangement with one switch per route from each signal.

Edited April 29, 2020 by The Stationmaster

[imt]

I agree - certainly my "toy" Ullapool is one switch (in my case illuminating button) per route.  I have only ever seen a real one like that once in my life and have no photograph so its all my own invention - though based on something real I saw once.

 

What is OCS? You see I am in amongst the professionals confusing things, misunderstanding and making a fool of myself.  Can somebody enlighten me or perhaps just tell me to p**s off.

[The Stationmaster]

34 minutes ago, imt said:

I agree - certainly my "toy" Ullapool is one switch (in my case illuminating button) per route.  I have only ever seen a real one like that once in my life and have no photograph so its all my own invention - though based on something real I saw once.

 

What is OCS? You see I am in amongst the professionals confusing things, misunderstanding and making a fool of myself.  Can somebody enlighten me or perhaps just tell me to p**s off.

OCS means One Control Switch.  It was system which as far as Britain is concerned came into use in the 1930s and was still being installed in some schemes into the early 1960s long after it had been technically superseded by the NX (eNtrance eXit) system which had first been used in Britain in a small installation in 1938,  OCS was a development, effectively a miniaturisation onto a control panel, of what was known as a route lever system where an individual lever in a miniature lever frame was used to create the route and set all points and signals in that route..  All these systems originated in the USA and then spread across the pond to Britain although route lever operation never gained much currency here - the GWR had several installations two of which (at Newport, Mon) lasted until the very early 1960s when they were replaced by an NX panel operating colour light signals

 

The essential difference between the two is that in OCS there is a separate switch for each route from a signal (watch the video linked below to see what that can mean in a complex layout) so there is still an element of learning for the Signalman as very few OCS British schemes placed the switches geographically on a mimic diagram, they were normally placed separately below the diagram  (as in the video).  In contrast the NX system placed switches/buttons geographically on the mimic diagram and is ergonomically much simpler because the Signalman operates one switch or button at the commencing signal for a route and presses another button at the end (eXit) of that route.  Nowadays mouse operation and positioning a cursor on a VDU has replaced switches and buttons.

 

Both OCS and NX employ route locking - in other words routes are interlocked with each other in addition to locking between points and signals and all are interlocked with track circuits.   Proper OCS panel design is not bad and it was cheaper than NX but it required good ergonomic design to be really efficient in use and not all OCS panels achieved that.  The interesting bit starts at c.4 minutes in.  and a detail view at c.7mins 30 - 8.15 and at 12.00 - 13.00

 

And at 18.07 is the former NER Board Room - one of the more impressive rooms in which I've attended a meeting!

 

https://player.bfi.org.uk/free/film/watch-this-is-york-1953-online 

Edited April 29, 2020 by The Stationmaster
Add more times to watch

[imt]

1 hour ago, The Stationmaster said:

OCS means One Control Switch.  It was system which as far as Britain is concerned came into use in the 1930s and was still being installed in some schemes into the early 1960s long after it had been technically superseded by the NX (eNtrance eXit) system

 

Thanks Mike.  I continue to learn.  So OCS is the right term for what I have done EXCEPT that (for ease of use by an amateur like me) I put the switches (acrually buttons) onto the face of the diagram.  Mine is indeed fully internocked and prevents conflicting routes.

 

So where does IFS fit in.  In my ignorance, because it was so close to a lever signal box, I presumed it was the next in line after lever - so I assumed lever to IFS to route setting (OCS) to full NX (which is quite beyond me).  I had seen the York panel in live operation, but didn't really understand it (too young at the time).

[imt]

2 hours ago, Dave-5-5-7 said:

Immingham West

 

Dave, after what you have taught me so far I can understand this panel. 

 

Currently there is a train standing behind signal 255 on the ore line.  Points 103 and 104 have been reversed to allow that train onto the Killingham UP line towards signal 267.  Points 102 and 105 are also locked against conflicting routes I suspect.  Signal 255 is set to proceed but not yet locked? Points 117 and 118 have been locked normal ahead of signal 255 presumably a route up the main line and signal 257 has also been set to proceed, green light is on but NOT the locked light.  Presumably the sytem was ticking through its functions at the time and it is not yet all set.  I presumed the red "hat" is to remind the signaller where all this started from? The TCs will follow the progress as the train moves Z, T, S, O, 448 and 447.  Presumably the signaller will reset 255 and then 103/104 as it passes onto (say) tc "S"  or later.

 

Did I read that right?

[The Stationmaster]

22 hours ago, imt said:

 

Thanks Mike.  I continue to learn.  So OCS is the right term for what I have done EXCEPT that (for ease of use by an amateur like me) I put the switches (acrually buttons) onto the face of the diagram.  Mine is indeed fully internocked and prevents conflicting routes.

 

So where does IFS fit in.  In my ignorance, because it was so close to a lever signal box, I presumed it was the next in line after lever - so I assumed lever to IFS to route setting (OCS) to full NX (which is quite beyond me).  I had seen the York panel in live operation, but didn't really understand it (too young at the time).

Yes, OCS is exactly what you have created albeit (and very sensibly in my view) with the route 'switches' geographically located on the 'panel',   It really comes down to what you are happy with and, if need be, what any visiting operators can quickly learn to the necessary level of competence and confidence for your normal pace of operating

 

As far as I can see IFS (Individual Function Switch) is basically a massive leap backwards as it does away with using a system based on route setting as in OCS or NX and reverts to a sort of miniaturised lever frame by having a separate switch to operate each point or  signal.  This will undoubtedly making it much simpler and cheaper to construct although it hardly strikes me as reaching the sort of ergonomic standards we ought to expect in the 21st century.   In point and signal operating terms it does no more than the Westinghouse Style L lever frame although it does it at lower cost and in a much smaller space.  I presume the system has really emerged as a means to save money in constructing a new installation rather than a means of making the job easier for the Signalman(ler).  

 

A good few years back I acted as a 'guinea pig' Signalman operating at the factory the new GET MCS vdu/mouse based workstation (destined for Stoke-On-Trent) for the benefit of an ergonomist plus assessing its practicality for everyday work from the Signalman's point of view.  That job took the better part of a day including testing all the various procedures the manufacturer had designed into the controls  (one of which i criticised very strongly).  It would be interesting to hear what that ergonomist would have to say about an IFS arrangement.  But what it did do was give the ergonomist a lot of information about the way in which signalling systems could be easily operated via an NX system which was readily learnt, including out-of-normal-working procedures by someone who had never worked with an NX panel but had plenty of experience working lever frames which she thought would considerably reduce training times and simplify conversion to a very different system..

[imt]

41 minutes ago, The Stationmaster said:

As far as I can see IFS (Individual Function Switch) is basically a massive leap backwards

 

I agree that is what it appears to be, but then there appears to be quite complex relay locking behind it which, on moving a signal switch to proceed, has to investigate the route set by the point switches to see if it is valid and there can be no conflicts by points or TCs before locking the point switches (logically not physically as i understand it) and finally setting and locking an appropriate aspect on the signal. Now I am not sure how such logic is set up in relays, but it would be quite complex I think if "programming" something like JMRI or some other "solid state" system.  I am hoping Dave or some other passer by can make that clear to me. Dave is busy doing this, except he is making it more difficult for himself by having limited contacts per relay (again as I understand other comments).

 

Like you, I think OCS is easier to understand.  But maybe not for professional signallers - though of course you have that experience too.  IFS is not much different from a lever frame with relay locking, or is it???

[St. Simon]

Hi,

 

 

 

3 hours ago, imt said:

 

I agree that is what it appears to be, but then there appears to be quite complex relay locking behind it which, on moving a signal switch to proceed, has to investigate the route set by the point switches to see if it is valid and there can be no conflicts by points or TCs before locking the point switches (logically not physically as i understand it) and finally setting and locking an appropriate aspect on the signal. Now I am not sure how such logic is set up in relays, but it would be quite complex I think if "programming" something like JMRI or some other "solid state" system.  I am hoping Dave or some other passer by can make that clear to me. Dave is busy doing this, except he is making it more difficult for himself by having limited contacts per relay (again as I understand other comments).

 

Like you, I think OCS is easier to understand.  But maybe not for professional signallers - though of course you have that experience too.  IFS is not much different from a lever frame with relay locking, or is it???

 

I think you've become confused between the 'Control System' and 'Interlocking'. Technically in a Mechanical Lever Frame, they are the same thing, but for all relay and computer based signalling systems, these are two very separate components of those systems, however most people thing they are all one 'package'.

 

The 'Control System' is the interface between the Signaller and the Interlocking, this can be a switch panel (O.C.S. or I.F.S. etc), a route setting panel (NX) or an IECC. The control system does NOT contain any safety critical functions or interlocking functions. The control system only allows the signaller to make requests for action from the interlocking and it provides the indications back from the interlocking. The control system, in Dave's case a I.F.S. panel, will only carry out the bit highlighted in Green in the above posts.  However, I would point out that where it says 'valid' in green, where in the control system 'valid' means that the route is actually provided rather than not in use, wouldn't apply to an I.F.S. panel as there is a switch per route (and therefore this check is done automatically). It would apply on an O.C.S. panel and an NX Panel as the control system will have to validate that the switches being turned / push constitute a real route.

 

The 'Interlocking' provides the safety critical logic that allows routes to be set, locked & released, points to be set locked & released and decides whether a signal can show a proceed aspect. The interlocking carries out the bit in red above regardless of what control system is in use. All Interlocking's, be they relay, solid state or computer based, carry out the same functions, albeit in slightly different fashions, it is not dependent on what control system it is attached to. The checks I'm describing in my thread from this post onwards are exactly the same checks that a relay interlocking would have to make, it just so happens that I'm using an IECC control system, it would react in the same way if I used an NX Panel or an OCS panel. There is no reason why a physical panel can not be connected to a solid state or computer based interlocking nor is there a reason why a VDU can't be connected to a relay interlocking (at least technically, maybe not practically). 

 

I hope that helps to clear things up a little bit.

 

Simon

Edited April 30, 2020 by St. Simon