Thanks for this, I'm going to go and try out the final part ASAP.
However, I must point out that tt was not even apparent to me that the RHS points in the pics were the isolating ones.
You're quite correct to chide me for impatience, and I stand so... chided. The forums I do frequent as a rule are probably far too quick to return answers... but then you seem to know what computer folk are like so I think it must also be apparent to you.
I do admire the truth table approach. It is so long since I saw that... I'll bet you remember CP/M too... and the great old Z80 processors. I prefer the last bit though.
I confess to forgetting entirely about some little pieces of paper which fell out of the points boxes... these would be the insulators mentioned, and I do recall... now that you and Gary have stimulated the neurons... reading somewhere that they were to be employed in a fashion as described. I discounted them I think since I was using some self-adhesive green baize for the base... and it is of course electrically inert... but then of course so is a wooden board.
I have decided to approach the business along the lines you detail... for which I thank you for your advice.
I will firstly isolate the inner loop from the middle... which is where the problem is most apparent. I actually isolated the sidings at 1,2 by removing the very short straight into that area as it was a sort of minor issue for the time being and better back burnered for now.
I think a run at 5,2 is the first step. Then if that returns some positives, perhaps the other cardinal... as you identified... in that area at 2,3
Bear with me however. I'm no longer nimble and it may take a day or so.
I do have now three locos which seem eminently capable of returning the results... if there are any. One of the others turned out to be a bit on the jaundiced side itself and tended to indicate faults by it's own intertia. At one stage it even smoked... something I've not been allowed to do for years now.
Both of the problem locos came from a bod on eBay. To be fair though, only one of them was priced at a level where you might expect to get a reasonable runner I apprehend. And I have not taken the matter up with him, preferring instead to wait and have them completely serviced by a chap who is currently on holiday.
So with two fully functional Duchesses... Athol and Montrose... and a little green 0-6-2 to work with, I may yet resolve this with your help.
One thing I seem to recall finding most disconcerting was that the isolating points registered current every which way, no matter which way I set them. On reflection of course that has to be the fact that even if you remove power from one loop, the power from the other is still naturally present.
dublover... currently ascending mount attic, to change some points.
QUOTE (dwb @ 21 Oct 2006, 12:44)
<{POST_SNAPBACK}>>I do not know electronics... believe me I wish I did...
Forget the electronics - you don't need to know anything about them in this context.
>and I can work out simple circuits
That's all you need to do here.
>I spent my time at uni doing something which is now relatively ubiquitous... they called it computing... but I can solder
OK. Here's a suggestion to see if we can translate this electrical problem to the computing domain you are familiar with.
Each point has 3 connections in and 2 sets of 3 connections out. Lets call them RailA, RailB and RailC where RailC is the centre rail.
You have two types of point which have two different types of truth table. I can't see the connections for the two outside rails, so I don't have enough information to create the truth table for RailA and RailB but if I get the idea across, you can do that for yourself. I believe in the context of 3 rail it doesn't really matter since they carry the same polarity - is this correct??
The truth table for the permanently live points (That's the two on the left of your photo) is as
follows:-
Straight ahead: RailC (StraightAhead) = RailC (In); Rail C (Turnout) = RailC(in)
Turn out: RailC (StraightAhead) = RailC (In); Rail C (Turnout) = RailC(in)
The truth table for the isolating points (That's the three on the right of your photo) is as
follows:-
Straight ahead: RailC (StraightAhead) = RailC (In); Rail C(Turnout) = OFF, Nothing, Niet, Rien, Nada
Turn out: RailC (StraightAhead) = OFF ; Rail C(Turnout) = RailC (In)
Now pick a location on your middle loop and build a tree of the connections from that location. Each time you encounter a point, introduce a branch. Depending on the type of point used, apply the appropriate truth table. Once you have built the tree, you can work out what rails are powered for each combination of settings.
Having written all this out, I think you can make your layout work if you ensure that all your crossings have at least one isolating point in them. So that's locations (1,2); (2,3); (5,3); (4,1). Co-ordinates are x,y and reference the origin of the containing square.
>I had hoped for more than that from this forum.
I don't know what forums you used to, but here it can take a few days for people who feel confident to answer coming across your post and making a suggestion. This is a fairly quiet forum but I think it has a good signal to noise ratio compared with other busier ones.
Let us know if it works.
David