Jump to content
 

Eastwood Town - A tribute to Gordon's modelling.


gordon s
 Share

Recommended Posts

Morning all.....

 

Seems the simplest solution didn't work. I won't bore you with the details but quirks within the ECoS route system meant the led's went on and off no problem with just one line in or out of ET, but failed miserably once you introduced the complication of multiple lines in and out.

 

ECoS now have an extended control function which I'm sure will solve the problem, but their data sheet is somewhat unclear in setting conditions. I spent a couple of hours earlier this week setting up the conditions and then deliberately set one of the conditions incorrectly, but I could still select a route, so clearly I'm doing something wrong.

 

http://www.esu.eu/en/support/tips-tricks/ECoS/erweiterte-fahrwegesteuerung/

 

I did ask on RMweb but I guess it's such specialised subject no one had any real ideas how to deal with it. I've now asked on the ECoS forum, with no luck as yet, but still early days....however....

 

Woke at 3am this morning and couldn't get back to sleep thinking about how to resolve this and remembered my friends comment (Eldavo), 'diodes are your friend' so grabbed a bit of paper and started scribbling. You will know by now electronics is not one of my strengths, but I know enough to get by. Started thinking about the platform led's and if they could be switched by pointwork conditions from either end of the approaches. I guess this was exactly what Martin was talking about, but all of this could be done within the panel.

 

Each platform has 1-3 leds that should only come on if the route is selected. Because they are both inbound and outbound routes the leds will need to be switched from either end and at 3am this morning I was scribbling away with diodes each end of a parallel run of led's so they could be switched from either end without screwing up everything else.

 

Half an hour later and a dozen scribbles and I had a plan and could then get back to sleep. Of course come morning I was up with my soldering iron lashing up diodes and led's and it works.....:dance_mini:

 

I can feed from either end and the diodes stop the adjacent  led's switching on if the approach turnout is not selected properly. The key for me now is to find exclusive trigger points within the turnouts in each route which will light the platform led's. That's appears quite simple on the face of it, but the led's must go off when a route is not selected and that can only happen when the trigger turnout is changed. Unfortunately in the ECoS route setting process it will only change turnouts within a route if they are common to another route. If a turnout is unique to one particular route, it ignores it and leaves it set in that position, so even though the route is not required the platform led's will stay on as the trigger is still there.

 

The secret now is to find common turnouts that will change when routes are changed and try to establish a combination of conditions such that the platforms led's will come on and go off as routes are selected.

 

I may be wrong here, Martin, but having now spent hours on this problem, I suspect the same problem would be there with relays direct from the Tortoise motors. If the controller is not requesting a change of turnout position, then the relay controlling the led's wouldn't change either.

 

Of course this is a quirk of the ECoS system as in a perfect world, changing a route should change all the turnouts in that route back to a default position irrespective if they are involved in another route or unique to one.

 

Perhaps there may be a possibility of one button to select the route and then a second that resets the turnouts in that route to their default position. That would certainly make life a whole lot easier. Maybe I should suggest something along this lines to the ECoS (ESU) people.....

 

Onward and upward...

 

 

Edited by gordon s
  • Like 2
  • Friendly/supportive 4
Link to post
Share on other sites

Hi Gordon,

 

I am unable to offer any in sight to the workings of ECoS but it occurs to me that a route is selected by signalling and not by turnouts, if the signal is at danger then the route is not set regardless of what position the turnout is set! Considering this, if you are able to plan the signalling then the leds for the route could be incorporated to illuminate only when the signal changes to clear.

 

Just a thought

 

Ian

Link to post
Share on other sites

59 minutes ago, Ian_H said:

Hi Gordon,

 

I am unable to offer any in sight to the workings of ECoS but it occurs to me that a route is selected by signalling and not by turnouts, if the signal is at danger then the route is not set regardless of what position the turnout is set! Considering this, if you are able to plan the signalling then the leds for the route could be incorporated to illuminate only when the signal changes to clear.

 

Just a thought

 

Ian

 

Thanks as always, Ian and of course you are right. Just two problems, I have to learn about signalling from scratch as I know very little and of course I will then have to build them.

 

That in itself will be a year or two away and I need to resolve this now to get things running properly.

 

Can't fault the logic, but sadly some time away.

Link to post
Share on other sites

48 minutes ago, Barry O said:

Meant to say the "IF" is the king point at the end of the platform.. 

 

Baz

 

As you would expect, a quick look has shown it may not be that simple....I wish...

 

I have quickly marked up the diagram with what I see as possible triggers. As I explained earlier, you cannot use turnouts that are unique to one route as the ECoS will not change them. The led's will come on, but choosing another route won't cancel them as the turnout will remain in the same position if it's not required for another route. 

 

It's too hot outside, so I'll tinker about some more and see where that takes me...:D

 

1167124000_mimic_panel(2).jpg.baf039aba3afcc9b7d98a1d05bddd661.jpg

  • Friendly/supportive 1
Link to post
Share on other sites

  • RMweb Gold
2 hours ago, gordon s said:

I may be wrong here, Martin, but having now spent hours on this problem, I suspect the same problem would be there with relays direct from the Tortoise motors. If the controller is not requesting a change of turnout position, then the relay controlling the led's wouldn't change either.

 

 

Hi Gordon,

 

Yes and/or no. I'm still not entirely sure what you are trying to do.

 

Diodes are on-off. Relays have changeover switches. You can do anything with either, but the logic is different.

 

Here is a bit of your plan:

 

gordon_relays_leds.png.3ce333653cbe85d673d32d123482cac7.png

 

Using changeover relays, if turnout A is set straight ahead, and crossovers B & C are set normal, the platform LED at D would be off.

 

Turnout A being against it prevents any LED power reaching it, despite B and C being set for it.

 

Essentially, you feed in LED power from the ends of the panel only. The changeover relays mimic the turnouts, the relay contacts and are connected together exactly matching the track plan. If a train could reach the platform from the end of the panel. and you apply LED power on that line at the end of the panel, the platform LED will be on. If a train couldn't reach it, nor can the LED power, and the platform LED will be off.

 

I think that's the 3rd time I've explained it. As with Templot, I don't seem to be able to write stuff in a way that others can understand. :( 

 

cheers,

 

Martin.

Edited by martin_wynne
  • Friendly/supportive 2
Link to post
Share on other sites

Quote

Of course this is a quirk of the ECoS system as in a perfect world, changing a route should change all the turnouts in that route back to a default position irrespective if they are involved in another route or unique to one.

In this case ECoS is correct and is doing exactly what the prototype does, only throw points that need to throw for the route.

  • Informative/Useful 1
Link to post
Share on other sites

OK Martin, I understand that, but let's just try a subtle change. Assume you want line 3 to go to platform 4 as you have drawn above. Turnout A , B and C are all set correctly and the led at D is lit. Great!

 

What I think you are missing is that if I now select line 3 to go to platform 1 via the ECoS, none of the turnouts at  A, B or C will change as they are not required to take line 3 to platform 1 and as such the led in platform 4 will still be lit as A, B and C won't have changed.

 

This really is a fault with the ECoS route planning and I guess I'm trying to find a workaround.

 

I'm very close to saying sod it, and just switching off the leds manually. It's one touch on the touch screen and job done.

 

Just seen your post, Keith and if that is correct, then I'm trying to do something against prototype practice....

 

It won't be the first time......:D

 

 

Edited by gordon s
  • Friendly/supportive 1
Link to post
Share on other sites

  • RMweb Gold
3 minutes ago, gordon s said:

What I think you are missing is that if I now select line 3 to go to platform 1 via the ECoS, none of the turnouts at  A, B or C will change as they are not required to take line 3 to platform 1 and as such the led in platform 4 will still be lit as A, B and C won't have changed.

 

 

Gordon, if you do that, with changeover relays the LED at D will be OFF.

 

The relay on A won't be receiving any LED power from the upstream turnout on line 3, so can't pass power on to D via C and D.

 

It's very simple. IF A TRAIN COULD REACH D from the left end, the LED will be ON. If just one turnout in the route prevents a train reaching D, the LED will be OFF. The state of A, B and C is irrelevant if there isn't a route set from the left towards A.

 

To be honest I'm getting a bit fed up of trying to explain it. I will write it bigger.

 

If the turnouts are set such that a train could reach the platform, the platform LED will be ON.

If ANY ONE of the turnouts are set such that it couldn't, the LED will be OFF.

 

What I still can't fathom, is whether that's what you want?

 

cheers,

 

Martin.

Link to post
Share on other sites

  • RMweb Premium
20 minutes ago, Grovenor said:

In this case ECoS is correct and is doing exactly what the prototype does, only throw points that need to throw for the route.

Subject to no conflicting routes, flank protection and so on.

Link to post
Share on other sites

Just read your post again Martin and I understand what you are saying, but using the info on ECoS switching I would probably need a relay for every turnout on the panel that is required to change during a route selection process. I'm just working on an Excel sheet that shows the set direction required for every route and every turnout.

 

Now I see your system would work but it would require 30-40 relays and probably 4-6 relays connected together for each route. It's not finished or checked yet, but this is the table for the left hand end only.

 

77757408_Screenshot2020-08-07at12_13_53.png.796cddd95a8c4912e860b392354f89a2.png

 

18 turnouts and 23 routes, so that's 18 relays, one for each turnout. Relay number 5 is required for 20 routes, so the wiring would be considerable.

 

Of course you know far more than I and perhaps I'm imagining the difficulties, but think it may be outside my capability and understanding to produce such a matrix.

Link to post
Share on other sites

  • RMweb Gold

Gordon,

 

Sorry I shouted.

 

You need 1 relay per turnout, no more, no less, regardless of how many possible routes.

 

The wiring is very simple. You just connect the relay contacts to one another following the track plan.

 

Martin.

  • Thanks 1
Link to post
Share on other sites

  • RMweb Premium
49 minutes ago, gordon s said:

OK Martin, I understand that, but let's just try a subtle change. Assume you want line 3 to go to platform 4 as you have drawn above. Turnout A , B and C are all set correctly and the led at D is lit. Great!

 

What I think you are missing is that if I now select line 3 to go to platform 1 via the ECoS, none of the turnouts at  A, B or C will change as they are not required to take line 3 to platform 1 and as such the led in platform 4 will still be lit as A, B and C won't have changed.

 

 

 

 


Hi Gordon,

 

I guess in this instance A is in the wrong place and for the sake of setting that route it might be 2 or three points up the line (-B if you will), so when you set the point that would send the train into platform 1 the route wouldn’t reach A to set that as a route so the LEDs in platform 4 wouldn’t light (but if you changed B the route into platform 4 would be set from wherever that line comes in from).

 

For clarity when it’s on the panel you might need to show a red on the fast/slow at the station entrance if (for example) you’ve set a train from line 1 into platform 2 it doesn’t show the usual route of line 2 as set.

 

Hope that’s of any use!  
Simon

Link to post
Share on other sites

  • RMweb Gold
2 hours ago, gordon s said:

 

Thanks as always, Ian and of course you are right. Just two problems, I have to learn about signalling from scratch as I know very little and of course I will then have to build them.

 

That in itself will be a year or two away and I need to resolve this now to get things running properly.

 

Can't fault the logic, but sadly some time away.

 

I do think that you are right to address this at an early stage. Too many modellers get into this far too late in the layout build and find it difficult.

 

I understand Martin's frustration. But if he thinks it is difficult to explain in writing here, he should try answering such questions over the phone.

 

One relay per point is certainly a reasonable solution. It will produce quite a lot of wiring but it will be quite simple wiring (but do remember to label up each wire for future troubleshooting). Others would advocate an electronics based solution rather than electrical.

 

My own preference is to cut down on the relays and try to do as much as possible with 2-way 6-pole rotary switches, each switch representing a signal box lever. You may find this a bit laborious to operate on a large layout (two signalboxes) such as this but it reflects much better the operating practices of the steam era when route-setting was in its infancy.. The wiring within the panel provides most of the interlocking and can even provide all the track feeds, so avoiding extra switches for that (apart from any insulation sections at the end of sidings, etc.

Link to post
Share on other sites

  • RMweb Premium

..  no need for relays.. you are looking at something which checks the points for the whole route (which is in the logic required to set a route, whether ECoS changes all points they must be set for that route .

 

For example using Martins diagram above if A is set , B and C are set straight then the route into that platform should be lit as should the platform.the platform LEDs should be illuminated..

 

Forget the ECoS switching, look at your route and which point needs to be where on the route for access to the platforms. Only if they are set should the platform be lit. If a point is set to do something else the train may not get to the place you want it to go.. but the "route logic" can be used to see which points you need for each route. Use the Point motor switches to set the LEDS with a bit of logic to make it work. Or can teh ECoS set teh lights from its route selection outputs?

 

The trick is.. this is DCC.

 

Baz

  • Agree 1
Link to post
Share on other sites

  • RMweb Gold

Hi Gordon,

 

I have tried to make a diagram to explain how simple this idea is.

 

1. draw the track plan on a suitable piece of board.

 

2. glue a relay on it over each turnout.

 

3. drill 2 holes alongside each one and take the relay coil connections through to the back where they won't get confused with the logic connections.

 

4. connect the relay contacts to follow the tracks like this. The changeover contacts mimic the changing of the points in each case:

 

gordon_relays_leds1.png.32115acf0858b73ee6dd233323d9b06f.png

 

The pale green lines are the track plan. I have set a route from X to the platform (bold line).

 

If you follow the LED power (purple lines) through the contacts, you can see that it can reach the platform LED along exactly the same route that the train can reach the platform, and the LED will be on.

 

If you were to change the points at Y, there would no longer be a route for trains into the platform. Likewise there will no longer be a route for LED power to reach the platform LED, and it will be off.

 

Laying it out on a board as I suggested makes it very easy to wire up -- the wires between the relay contacts follow the track plan.

 

(p.s. the circuit diagram doesn't show the LED resistor, which depends on the LED rating and supply voltage)

 

cheers,

 

Martin..

 

 

 

  • Agree 1
Link to post
Share on other sites

Thanks Martin, despite appearances, I did understand what you were suggesting, but kept thinking you had a much simpler version, rather than one relay per turnout for reasons I’ll explain.

 

I accept I sometimes misinterpret diagrams and proposals, but having now finished the complete table of turnout positions for every route, I can see the total number of connections required to link every relay in series in the manner you are suggesting. I really hope I’m wrong as I  would like a simple solution.

 

Taking the chart I posted earlier there are 17 turnouts (#16 is missing) and 36 routes. The total number of connections between the 17 relays is 228 or 456 solder joints. Taking turnout 3 as an example, there are 25 connections in the straight ahead position and 5 curved away from the straight. Am I right in that assumption that 30 wires have to go from relay 3 to other relays in the series?

 

Surely that can’t be right?
 

Am I missing something obvious as that seems a massive amount of work to switch on and off a few led’s?  Of course that is only the left approach so a similar amount would be needed the other side.
 

I’ve said so may times that a lot of the progress on ET has been down to you so I really appreciate your advice and certainly don’t want to lose that help and support, but am genuinely confused by the possible amount of wiring between these relays.
 

Can you see where I’m coming from? Are my number of connections totally wrong? I really hope so......:D

 

Edit: Just noticed my earlier screenshot had missed some routes as I must have has freeze frames on in Excel, so there are more routes to cover.

 

Revised route table - 36 routes...

 

2041573290_Screenshot2020-08-07at20_48_52.png.8dfafbe44e02f1fe5ee95d7502e14f98.png

 

 

Edited by gordon s
Link to post
Share on other sites

  • RMweb Gold

Hi Gordon,

 

You have lost me I'm afraid.

 

If there are 17 turnouts there are 17 relays.

 

Each relay has 5 terminals. 2 for the coil and 3 for the contacts. That's a grand total of 5 x 17 = 85 soldered connections.

 

No matter how many possible routes, there are still only 85 soldered joints.

 

There are 3 x 17 = 51 contact terminals. Each piece of connecting wire linking them has 2 ends, so the number of pieces of wire is 26 approx.

 

Does it help if you imagine the inter-connecting wires running along the centre-line of the track? At each turnout tie-bar you have a changeover switch, switching the electricity to the left or right wire, replicating the turnout setting. When you get to a platform, you connect the wire to a LED. At the ends of the panel, all the wires are connected to the incoming LED power supply.

 

If a train could start from the end, and get to a platform, so can the electricity, and the LED will be on.

 

If a train can't get to the platform, nor can the electricity, and the LED will be off.

 

You don't need any charts of route combinations.

 

cheers,

 

Martin.

Edited by martin_wynne
Link to post
Share on other sites

  • RMweb Premium
3 hours ago, gordon s said:

Taking the chart I posted earlier there are 17 turnouts (#16 is missing) and 36 routes.

Each individual turnout only has two routes, Gordon. What happens after that is determined by the next turnout, which has two routes, and the next, which has two routes, and so on. Martin's diagram shows this more clearly than any words or tables.

 

The changeovers could just be switches, not relays, depending on what point motors you are using (I'm sure you've told us but I can't remember). If, as I do, you use Tortoises or Cobalts, there are two switches per motor. One can be used for frog switching, the other for the position indication. Even if you don't, a simple microswitch would do rather than installing relays.

 

Hope this helps a little.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...