Hi Torbjörn,

Your first suggestion is correct: CS2 only takes an action once the contact track is triggered (by entering of course).

Even better - I just found out this weekend. Latest issue of Märklin Magazin (okt - nov) explains how to set up blocks. They explain that in the current softwareversion of 60213/60214 a contact track is either "occupied" or "free", and you can have your CS2 to activate a route whenever this state changes. That is: your route is activated when the train enters the contact track (as usual), OR: when the last car of the train leaves the contact track (that's new, at least to me).

I had to try this on my test layout today, so I installed 2 contact tracks: one just before my "station", one just after. The first ones triggers a route when entering - this route turns a light bulb on (as I said, I'm just playing around). The second contact track triggers another route when leaving (that is, after the whole train has passed the contact track), turning the light bulb off again. It worked perfectly [^]! (and I was very proud of myself ).

To use this second option, triggering a route when leaving the contact track: when you define a route, you have to identify which S88 contact is used. On the right side of that, there is a check box that you can turn ON or OFF. Default is ON, when OFF the route is triggered when leaving the contact track.

Long answer to a short question, but I don't think this option has been discussed here before.

Peter

You're welcome.

You will have to buy MM #6 as well. They announced that will contain a following article about CS2 and shadow stations. BTW, this article is written by Frank Mayer.

Peter

Is it possible to have the "CS1" in the above posts defined, please. Are we talking CS1 or CS1 Reloaded (ESU) or maybe both.
Does anybody know?

Hi,

what kind of problems do you have? How's your setup? What do you want to do?

Hi Peter,

How does the S88 distinguish the last car from the rest of a train?

This becomes very imnportant if the route also activates a turnout on the same track when the last car leaves the track.

BR

Anders Pihl Knudsen
2

Dear Pavle,
the situation you describe is clear but then the question is : in case of constant presence on circuit track, does in some situation the CS 2.0.4 recalculate the routes? I think usually not, but in some situation yes.

Let's have on oval with 4 blocks. In principle you can run 3 trains and you need 4 routes, 3 of them are triggered once by occupancy but LOGICALLY only 2 are TRUE at a time, one is necessarily FALSE and then the CS recalculates. So the problem is not the lenght of the trains but the logics of the scheme.

If my explanation is true, the question is : how can hundreds of people use blocks? (contact tracks or circuit tracks or reeds does not matter) in other words: were am I wrong?

Thanks to all.

Hi,

It does matter whether you use contact tracks or reeds (with CS1 = 60212, that's what you are using, right?). A reed is activated by the magnet on the loc, creating a very short contact. Contact track is activated as long as axles are on that piece of track, so quite a long contact (even longer contact by longer or slower trains).

I am not sure how CS1 "recalculates" the routes, maybe one of the experts of this forum can tell us. For me it was enough to understand that as long as the contact is "true", CS1 tries to set the route accordingly. So with a "constant presence" on contact track, CS1 constantly tries to change the routes. That's why very short contacts (reeds) were preferred.

With CS2 it's a different ballgame, much better. A route is not triggered by the contact as such (on the contact track), but by the change of state: "free" or "occupied" as Jeehring described.

A very safe way to define blocks is to have a contact track on each "border" between blocks with two routes associated to it:
- route 1 sets the signal which is positioned just before this block to "red", so no other train can enter the block. This route should be triggered as by default (change from "free" to "occupied")
- route 2 sets the signal before the previous block to "green", saying the previous block is free, another train can enter. This route should be triggered by the change from "occupied" to "free".

In your oval with 4 blocks this will work perfectly for 3 trains - assuming each block is longer then the longest train.

I hope this makes any sense... Anyway, I will advice you to make a little test layout like the oval you mentioned, and just try. The best way to learn is to make many mistakes... (it worked for me)

Peter

I don't think there is any calculation. The contact track is acting as a "virtual" relay.S88 module is designed to accept impulses or long lasting contacts....
With the old Memory , a route stayed locked as long as a train is stopped on a contact track but it worked with the "locked/unlocked" logical-system ( which was quite interesting , on the S88 module :from 16 contacts, 8 contacts are for locking, and 8 contacts for unlocking.....a route.)
About a CS1 , I don't know, ...I never heard about locking/unlocking....

The use of the "On to Off" transition of the s88 sensor to detect when a train leaves a block is very protoypical, however in the model world the system becomes a little unreliable if you consider.
- Some coaches/wagons (usually 4 wheelers) are light and can bouce/wobble - loosing contact
- - This includes "rocking" around a curve
- Some are long and can span the contact section.

While the CS2 does have the feature to react selectively to either the On>Off or Off>On state change of a sensor, (and I have found it generally reliable in testing), I have had prior experience attempting this with older controller interfaces (Printer/605x/IB/CS1) and a PC.The PC scheduling allows for the insertion of a delay (say 2 seconds) after the sensor changes its state to "unnocupied/clear" before it "reacts" and sets signals etc. This allows for short bounce of the stock if it looses contact. If the event is indeed a ghost/momentary bounce, and the wagon lands back on the track, then the 2 second timer is reset before any action is taken.

In this setting I would prefer to use the contact track method by the front of the train to set the block occupied, and the reed switch method (with the magnet on the last wagon..) to more precisely and thus safely indicate the train has left the block.
This sort of 2 trigger method also helps should there be an inadvertent uncoupling of the train. While traditionally (as used in examples above and for over 50 years by Marklin modellers) the loco entering block is used to assume the rest of its train has exited the block behind it - and thus used to set that signal green. (Note: Personally I think there should always be a "fully empty" block between trains - none of this assumption that the train in the processes of leaving a block will be out of the way before the train, just released from the block behind it, catches up.

Neither of these is foolproof when it comes to bi-directional train movement. Even something like the LISSY or RailCom system which can detect and differentiate between which (loco/caboose) has past a point AND in which direction of travel, still needs some external (Train operation) logic to know if this correct. (e.g. Terminus running when the locomotive is at the rear of the train pushing - i.e. which Tail (Caboose) ID is meant to be joined with which Head (Loco) ID to make up parts of the same train)

So ask yourself. (for both now and the future)
- Do you want to allow for dual direction running
- Are you wanting an infrastructure for trains running push/pull
- How happy are you with the occasional rear end collision (caused by an uncoupling)
- Will an intermix of sequence controlled running be mixed with Operator controls.
- When can we expect LISSY type capabilities from Marklin

This afternoon the owner of a CS2 just has made test :with a 40 cm long isolated track, he did it between 30/40 times with diffferent wagons, not a single failure, it works well.

About spanning the contact section : for safety the contact section must be longer than the longest vehicule in your possession.
We have same problem with the old Memory : in the middle of a train if you use 2 or 3 TRIX coaches together with isolated wheels ( not making contact between the 2 rails), the former MEMORY could react as if there were 2 trains, one following the other one.
I have been using isolated track section with MEMORY, it works pretty well.
Necessary conditions are:
- only use cars & coaches with 3 rails axles
- only use isolated sections that are long enough. ( as indicated by PAVLE : Marklin says 36 cm . There is a reason. The longer it is, the more reliable it is.
In case of failure : check your axles !( I had several TRIX coaches & wagons with isolated wheels me too, you can run them easily on a 3 rails layout excepted if you use isolated track sections with Memory - or CS2 as shown above...)
Warning : when buying second hand or on E-Bay you may also find some Marklin cars & coaches with isolated wheels for 2rails...

BTW : in DCC world they use to build very long isolated sections. It works with current consumption detection.(more complicated and less flexible when the train must be stopped - like in stations...you need extra electronic device...)

Hi,

When I read the first example of how to use blocs with the CS in the oct/nov MM, I thougth there was a translation error in the French version of MM. In fact, that example uses the principle that a loco entering a block allows a "green signal" for the just-left block! As if there were no wagons/coaches pulled by the loco!
But, as Clapcott says, it seems it's a traditionnal method for Marklin!! I understand now why F.Meyer writes that the block lenght must be "at least" 1.5 times the longest train! And also why he wrotes that the method of detecting the "OFF" contact gives "the maximum security". In fact it's just "a better" security than the previous principle, which is quite unsecure !
I personnally use the principle that a lok detected at the end of a block means that the train has left the previous block. That implies that all blocks are just longer than the longest train (more exactly the lenght between the beginning of the block and the lok detector at the end of the block are just longer than the longest train). It's a 100% secure principle. Why Marklin doesn't use that? May be linked to the use of the Memory? (I don't know the Memory for I use a PC to control my layout).
Do you think of any drawback to that principle?

Fred

Good to know.

Did you test if the trigger occurred "more than once".

Usually multiple pulses go unnoticed , however in certain scenarios involving interlocking , this can be undesirable.

Allow me to add some thoughts on this as I am currently assembling in my layout an automated shadow station (24 sidings) with CS2.

A possible solution (already mentioned in this topic but at least my tests have been positive so far) is the current consumption alternative (using Viessman 5233 S88 device - oh!, when will our dear M* add such device to their portfolio? ). Not only I do not have to cut lines (I am using C lines ) but it provides a very reliable means of alerting the CS2 when the train enters & leaves each section. The red ("B") feed is used here instead of the brown ("0") electric connection for current detection.

For this to work you have to feed each section (a block should have various sections, typically start-transit-brakeage-stop) individually through the S88 detector.
It gives some additional work but in my humble opinion it is very reliable and paves the way for further computer-assisted automation.

Hi Jan,
If I may comment on your reply in sections.

As always I reference terminology
"Contact Track" as 24995 (or equivalent) - Uses metal of wheels/axles to short the 2 rails
"Circuit Track" as 24994 (or equivalent) - uses pickup shoe to activate a switch arm in the middle of the track


This should not be taken as a generic statement for all situations. The detection mechanism is just the same, it is what is done (triggered) that may have limitations


This is indeed a powerful feature available with the 6043 <memory>. It is poorly understood and , in my experience (biased by the PC <interface>), rarely used.
However your point in relation to the CS2 memory (or CS1 routes) is extremely valid. There is no blocking/interlocking


Taken as read, and is no different form the CS1 or PC driven logic


From my <English Version> of 2009MM6 , Under "Principles..." does use the term triggers a momentary contact and Under "Practical Tops".. It is best to use reed switches

However this does not exclude the use of Contact Tracks as an option.

What must be avoided at all costs is any multiple or re triggering of memory/route sequences, most notable "the next time round the cycle" when the train is set to leave but, because part of the train it is still bouncing over the contact track, it will/may trigger the start sequence for another yard thus causing multiple trains moving out onto the main line.

This situation is very similar to that caused by "Circuit Tracks" and trains with multiple pickups

The articles section "Tricks for the Circuit Track" explains one option to mitigate this situation by disabling the sensor input as the train moves off (i.e. when the signal is green and "stop section" is live) and waits until Reed3 when the train must be fully out of the siding to re-enable/re-arm the sensor at the same time it sets the signal to red. There is a (very low) risk when substituting a contact track option in this situation, that vibration of the train waiting in the yard will cause it to rock and bounce the contact causing an unwanted re-trigger of a memory sequence. Personally I prefer specific triggering, however if migrating an existing layout then my take is that this is an acceptable risk.

This solution can work just as well with contact racks (with one proviso). In the article the sensors (Reeds 6,7,8) are located at the start of yards with the aim of being triggered by a magnet on the last wagon. This is an ideal solution for capturing any unexpected uncouplings.
In most implementations however the trigger is located just before the stop section and assumes the train is intact and clear of all turnouts into the yard.


The answer is probably "no" to the rebuild, however the solution will depend on the details of your current wiring.
If you are using contact tracks for a short space (say one track piece length) directly in front of the stop section, you can employ the "Circuit Track trick"
If your contact track is the length of the yard then your main issue (as it would have been previously) is the issue of NOT changing the entry points until after the whole train has cleared them.

About the 6043 Memory : the maintenance engineer of Marklin France , one day told me that there were different kind of Memory with different serial number due to different hardwares/firmwares. He even told me that the Memory has been continuously evolving and was a Marklin item on which they have made more "changes" than on any other electronic device from the Command chain.(Externally, at least for the most important changes it can be seen only on the serial numbers)

Broadly speaking , for us customers, we mainly have to know that there were 2 generations of Memory.
The first generation doesn't work with "locked/interlocked" method, they are not able to make the difference between S88 inputs "1,2,3,4, etc" and S88 inputs" 9,10,11,12, etc...". Also : they don't have the few elements of logic we find on the second generation...
The second generation understand "the locked/interlocked" method... It is the one I'm talking about in this thread (Well... I'm trying with my very rudimentary English). Last series are the most sophisticated .
He told me how to distinguish the different serial numbers, but...I don't remember ( A matter of letters...)
The second generation came after 1998 (or after 1996, sorry, I'm not sure, don't remember the exact date. It has been evolving independantly from the rest of the command chain...

Edited by user 21 February 2010 14:45:50(UTC)  | Reason: Not specified

Thanks for your reactions sofar. But I will explain my problem a bit better. I had a 6043 Memory (more than 1) and a semi automatic shadow station, let's say 8 tracks. Each track was connected on the S88 no. 9, 10 ..till 16 and the T bus. The nos 1 till 8 were connected with each other. The program said for track 1 when entering the shadow station: railway point 'round' but if track 1 is occupied, railway point 'straight', check track 2, if track 2 is occupied check track 3 and so on. You can make it fully automatic and you only have to add: signal 1 green, all the 7 others red, or signal 2 green and all 7 others red and so on. If all tracks were occupied the signal at the beginning of the shadow station turned red and no train could enter the shadow station until one left. I changed all the old stuff for a CS2, but left the Memory program intact, connected the S88 (5 of them) to the CS and the CS kept on giving commands, changing railway points, signals and nothing could stop it, only when I removed all the trains from the contact sections. Then the CS stopped giving commands. When I read MM6 (German version) I understood a bit better what the difference was between 'old' and 'new' In the new situation, when you have a train on a contact/stop section and your program says: set the train on track 2 in motion, the train on track 2 indeed leaves the station but the command is constantly given, as the train on track 1 occupies the contact section. It doesn't matter if the contact section is connected to no 1 or no 9 on the S88! Thus the old connects are of no value anymore and I have to look for another solution.
Question: I know the Viessmann S88, but do they operate different from the Märklin S88 in connection with the CS2?
The great difference between circuit or switch tracks and contact tracks is that the first are 'moment' contacts and the latter is a 'permanent' contact, this for a better understanding.

Hallo Jan,

I will answer in English, if it's not clear please send me an e-mail.

1) You can use any of the 16 ports of an S88.
2) On the Memory screen you have to change "manual" to "automatic/manual", on the right hand side of the row of memory symbols. This is because you want your trains to do the work (automatic), and not trigger the routes by hand only (default setting is manual - I usually forget to change this, until at the first test run nothing works...)
3) Bottom part of the screen it says: "configuratie memory". Below that are three fields you can change: the name of this route, the number of the S88 contact and the "S88 on/off" - checkbox.
4) Number of S88: if this is the first S88 attached to your CS2, use numbers 1 - 16. If it is the second, use numbers 17 - 32, and so on. You can use the plus and minus to change the number.
5) The S88 on/off thing: correct, that's the check box I meant. By default this is ON (check visible). It's not very intuitive, but this means the route is triggered when a train enters the contact track. If you turn it OFF (no check mark visible), it means the route is triggered when a train leaves the contact track.

I have not yet used this in a shadow station (mine is already done by reed contacts), but I think it's very well possible. Set S88 to ON if your contact track is at the end of your shadow station, set it to OFF is your contact track is at the beginning of the shadow station track. Märklin Magazine (the latest) said one should use reeds, but they did not say why.

Hope this clarifies things...

Groetjes,
Peter

Peter

I have the same experience! It has much to do with the software updates such as 1.3 which made it possible to switch the S88 on and off. In the manual I have of the CS2 this possibility is not given. However I have to change some of my layout, because switching the signals to red and the former to green is quite different compared to my 1985 digital setup.

Jan