Signal anomaly with diagram

[laldfordo]

Abandoning the previous thread, here is a diagram for my problem. Please take on board that the route is signalled correctly and mostly it works correctly. So I'm looking for an explanation of / solution to some strange behaviour of the signals.

A red signal (in Auran) with its back to the direction of travel always goes yellow when a train passes it and then goes back to red when the train goes past the next signal it encounters. You need to know and accept that to understand the problem.





So, the train passes signal A, which goes red, and B, which goes yellow. When the train passes signal C, B should go back to being red. (And A should go yellow if another train is coming). But it doesn't. B stays yellow, therefore A stays red.

Eventually B goes red when the train goes past signal I, after the next trailing junction, and turns I yellow.

The distance B to C is about 250 metres, B to I about 1150 metres.

Hope it is clear. As I said before, my solution is to remove signal B, but then the route is not properly signalled.

Thanks to ImageShack for Free Image Hosting

 

[nismit]

Signal aspects for track sections used in both directions

Posting Deleted.

 

[Jerker]

G'day laldfordo,

I'm sorry but I can't see where you're problem lies. As far as I am concerned (given the inaccuracies and vagaries of the Auran Signaling rules), everything is working perfectly! Certainly, as you say "A red signal (in Auran) with its back to the direction of travel always goes yellow when a train passes it and then goes back to red when the train goes past the next signal it encounters" is true but you seem to be forgetting that this only applies IN THE DIRECTION OF TRAVEL. The next applicable signal for a train traveling towards the bottom of the 'map' after it passes signal 'B' is signal 'I', so signal 'B' will remain at 'caution' until the train clears signal 'I', which it does. Mind you, in the real world, this would never occur, since signal 'B' is an 'absolute' signal protecting the junction in advance of it and would require several steps for the "dispatcher" to undertake to change the direction of travel of the train (for the signaling system), so as to clear "B' to proceed, steps which could not take place whilst the train was still in the single line section (basic safe-working). Only 'permissive' signals that divide long sections of single line that DO NOT control movement over junctions act in this manner.

That notwithstanding, I can see a few problems that might occur due to unprotected junctions in your diagram but before I say anything, I would need to know some background regarding the distances and purpose of the various 'branches' shown in the diagram.

Jerker {}

 

[laldfordo]

Hi Jerker and thanks.

I'm not making myself clear again.

The diagram is for UK operation with the signals visible to the driver's left. So the signals for travel from top to bottom are A, C, D, E etc, not B and I, which protect the junctions in the opposite direction.

So B is red before the train passes it, yellow after it passes and normally goes back to red when the train passes signal C. The anomaly sometimes occurs and the signal B remains yellow when the train passes C. And remains yellow until the train passes I, which is red, and turns that yellow.

Further on it's all normal again. ie when the train passes J, the next signal facing it, signal I goes back to red, as it should in normal Auran behaviour.

I showed the three junctions before the one protected by signals A and B because they might be the cause of the problem. They are close together (all within 100 metres) and the only places on the layout I get this anomalous behaviour is at the last junction of several close together.

Quite why that should upset the standard Auran signal inter-relationships, if indeed that is the cause, I don't know.

As I said earlier, my solution is to remove signal B, but then the layout is not properly signalled for the reverse direction.

Other distances in my original post.

 

[p-dehnert]

Jerker had the right explanation for the behavior of signal B.

For signal B the whole stretch of track between signal I and signal B is a single block (for the upward direction). This signal block is independend of the signal blocks created by the signals for the other direction.

If you put a signal X (for the upward direction) between signal I and signal B, signal B will change its state when the train passes the new signal X

Peter

 

[Jerker]

G'day laldfordo,

Thank you p-dehnert. Laldfordo, I did not misunderstand you, although I had not imagined that the 'arrangement' was English based (as the signaling did not 'indicate' this). Nonetheless, I may have confused matters with the "IN THE DIRECTION OF TRAVEL" reference after mentioning travel in the downward direction. I realised that signals 'B' and 'I' apply for 'upward' movements and my comments were dependent upon that assumption.

Now that we have established that this is an English based route, might I ask which signals protect the second and third junctions from the top of the diagram (in both directions of travel) and to what do these junctions lead?

It has to be kept in mind (and again, I have never forgiven the Auran programmers for this incredible lack of consistency), that although Trainz signals work on the real world basis of 'signal' to 'signal' for train movements, track block sections work on a 'junction' to 'junction' basis (as opposed to the real world situation where track block sections run from signal to signal). If that makes any sense. Thus, it is almost impossible to get Trainz signals to operate as they should. It should be noted that these comments apply to colour light signals - Auran's Semaphore signaling is a totally different disaster situation, all together!

Jerker {}

 

[laldfordo]

Hi p-dehnert and, later, jerker.

Thank you for your inputs to this thread and I have re-read the earlier posts knowing that we are all talking about the same directions of movement.

The revelation about track block sections being between junctions is interesting and probably has some bearing on my problem. But can you explain, jerker, when and how the Auran logic uses a junction-to-junction criterion in contrast to a signal-to-signal criterion to govern junction and signal settings?

Whether influenced by junction-to-junction or by signal-to-signal, I can see that there is one block from I to B for a train going upwards. And, if it is the passing of the block end (I) by a train going downwards that is needed to turn signal B back to red, then adding signal X, as p-dehnert suggests, would have the same effect.

However, that is not normally necessary. Normally, the passing of signal C turns B back to red. Still, it's a great solution. All I have to do is to put an extra signal at the C position but facing the other way - and ditto for the other parts of the layout where the problem occurs.

But, as an intellectual exercise, do understand that this is an intermittent problem. Much of the time signal B does go red when the train passes C. Then the logic gets into a different mode and the trains all have to get to I before B changes back.

What are the branches for at the top of the diagram and why do they have insufficient signals? The top facing branch is to a siding. The other three branches are tracks with a similar status to the track that goes straight down. You will see that the branches are correctly signalled but there are three missing reverse-direction signals and two missing downward-direction signals on the main track. I have had these in place but, with the junctions very close together, it gets cluttered. So I took them away as it seems to make no difference to the operations.

No doubt you will tell me why I need to put them back!

 

[Jerker]

G'day laldfordo,

I am intrigued that signal 'B' should return to stop when a downward heading train clears signal 'C' (this should not occur as that signal does not apply for train movements in that direction). Before we go much further, may I clear up a small terminology confusion. In the real world, when a train is said to have passed a signal, this means that only the locomotive has move beyond the signal. When the entire train has passed beyond the signal, then it is said to have cleared the signal.

I cannot explain why the 'good' people at Auran chose to take the totally ILLOGICAL path they did with regard to the operation of their signaling rules. Regrettably, I was not privy to the discussions that took place during the development stages of the program (rest assured, were that the case, I would have quite vehemently shown them the error of their ways). It would be interesting to ascertain the source of the their research in this area.

Given the information you have provided, I am not going to advise to add any further signals to your route, although I will say that as long as the signal that you are using for the unmarked one at the top of the image is a 'scripted' one (other than the standard ones that come with the program), that CAN determine the need to display a 'slow' aspect for train movements into the dead ended siding, along with a reduced 'single yellow' aspect for the diverge to the branch and then a 'clear' aspect for train movements along the main line, then you have it all covered. Similarly, signal 'B' would need to distinguish between the three alternative routes available to it and give an indication to the driver of which route is set. If it does so, then all is well.

I will take you to task, however, over what seems to be an inordinate number of intermediate signals for train movements 'down' the page between signal 'A' and signal 'H', especially when there are none for the reverse direction between signal 'I' and 'B'. I would also question the location of signal 'A', as at present, the 'trailing junction' (for downward train movements), third from the top, has no protection for movements in that direction. If they are sufficiently close enough together (within a few hundred meters, say), I would shift signal 'A' to the opposite side of that third unction and let it 'protect' both trailing junctions. Of course, given the workings of the Auran programming, this may not be possible (getting that one signal to protect two junctions, that is). I shall leave you to your own devices and further advice from those who know more about the inner (albeit flawed) workings of the Auran signaling (I will gladly advise on signaling matters pertaining to the real world)

Jerker {}

 

[stagecoach]

Although you show a lot of downward signals, which will cause a problem on upward trains entering the the block and coming head to head, you show only the two upward signals. Where is the next upward signal above the diagram? If another downward train is passing or has passed the unseen upward signal then B will go red, the same will happen if you have a consist at any of the top branches and the junction is switched to that branch.
If you tried this with only one train then signal B would go green when you pass signal I as signals dont reconise direction of travel.

 

[laldfordo]

Wow, Stagecoach, that's a can of worms.

There clearly is interaction between signals facing in opposite directions (eg A won't come off while B is yellow) and between trains and signals that have their backs to the trains (eg B goes yellow when a downward train passes it).

But I disagree if you are suggesting that a train has the same relationship with "offside" signals as it does with its "onside" signals. Signal B will not go green when a downward train passes or clears I. If B is still yellow, B will go red as I goes yellow.

But normally B will go back to red when the train passes C.

The bit of track in my diagram, and in fact my whole layout, is signalled for uni-directional running. Signals such as B and I are only there because the Auran signalling manual says to protect each side of all turnouts. So the next signal for upward-travelling trains is about a mile away at the next junction.

Jerker asks why so many signals. Well, my blocks end at E and H. I follow Lewisner's principle that you put three signals close together at the end of each block. If you don't do this the trains will spend a lot of time running at half speed.

As to whether the Auran programmers could have given us different rules for the signals I would say:

Every time I turn my computer on I feel a bit like those obsequious creatures in the TRS intro. I've got this unbelievably powerful box of tricks on my desk and it will do all sorts of calculating and communicating thanks to all the people who write software. That doesn't stop one being critical but, taken as a whole, TRS is a pretty good experience for just £12.95!

I take on board that there is a stronger interaction than I realised between trains and the signals they pass the backs of, and shall explore that further. It is also useful to use cleared and passed to mean different things as, indeed, sometimes signals change when the front of the train goes by, sometimes when the rear. The same is true of speed boards.

But no-one has yet suggested an explanation for why this is an intermittent problem. On most of my layout the 'B' signals go back to red when each train clears the 'C' signal.

 

[stagecoach]

Ive not found it works well if you signal all turnouts. Ive got a complicated group of turnouts in one section of my route which would need a lot of signals and look wrong. What ive got and works well is to place signals on all tracks leading into that area but none in it. the signals will control all trains running through it only allowing one train through in each direction.
As your diagram shows a single track and no passing loops only one train can be in it at anyone time i would remove all signals on the main line and only keep the signals for lines entering it. As youve got it a train heading upwards could pass the green signal I then a train heading downwards would pass green signal A to the point where the two trains meet, unless you are using a block rule.

 

[laldfordo]

Thanks stagecoach I'll try that. I too thought that completely signalling all those junctions close together looked wrong.

 

[Jerker]

G'day laldfordo,

You might want to consider converting those intermediate downward facing signals into "invisible" ones. They would still serve the same function and to all intents and purposes, 'not be there'! Although I cannot surmise if that will have any effect on your intermittent problem. Speaking of which, I hate "intermittent problems", they're so damn difficult to deal with!

Jerker {}

 

[laldfordo]

G'day indeed jerker,

You're up early on a monday morning! We're still watching Sunday evening television here!

I might try invisible signals but I worry about invisible items as things happen and you don't understand why because you have forgotten about the invisible items.


I reckon the extra X signal in an upward direction will squash the intermittent problem - unless that signal in turn fails to go red when the train clears (I've learnt that one) C.

 

[Jerker]

G'day laldfordo,

I can only suggest you try it to see what effect it will have (placing the extra signal in between 'I' & 'B', that is). At the time of writing my last post, I had just started my shift for the Monday morning and am still here getting a few words in between 12":1' scale trains...

Jerker {}

 

[laldfordo]

Don't get on his train

In the same way as one shouldn't use a mobile phone while driving a car I don't think you should be posting to a thread while driving a train! Much as I appreciate your input. And it's Monday here now

 

[Jerker]

G'day laldfordo,

You'll probably get this upon awakening later but I did say "between 12":1' scale trains." I agree, were I on the footplate, I would not be quite so insubordinate as to attempt such folly. Alas, I am but a lowly signalman (these days) who has plenty of time to attend to such matters as these, especially at that time of the morning...

Jerker {}