Hi, having aired my interest in braking modules and braking and acceleration sections I would like to open up this topic to anyone who has entered a computer controlled system/ program, how they find it, how easy it is to programs it and what type of programs are available for an entry starter with other words a person who has limited knowledge.

What I hear it is the way to go and some of you have already in part explained or shown computer control programs but could we have a more detailed explanation of how it is managed, what programs you are using, how user friendly are they etc. etc.

I also believe there are 2 ways of detecting signals or pulses (sensors) from either the middle track or the outside track and we may can understand these options a bit more by explaining the good and the bad or preferred option.

another module I've prepared for my braking and acceleration sections are fade in fade out signal modules and I hope and most probably they can also be initiated via computer control instead of having a separate module for it.

it may also be important to lay out the run down procedures, similar to a review of a loco, so we can identify different aspects what is needed, brand names of programs,

hopefully some of us will take the advice, help and support which is needed to start off with and in the end saves us money on several different modules


I'm thanking you in advance for your input.

John

Hello.

Adrian is right. If you use new technology like CS2 and it's newest software versions you inevitably runs into problems, large or small. You have to accept these problems. BUT ... at the same time when you try to fix problems you learn very much about possibilities which new technology brings with.
When you use old signals and other rely-based devices with new hardware and software you have to face many problems which can be solved though. The best teacher is practice and solution of difficult problems.

Then to the automation.

I would add that Keep everything as simple as possible. Think simple, if you have ( as most of you have) complicated layout, break the layout in simple functional parts. Make a good sketch for automation and test it carefully " on the table" before making it in real layout.

I think that the most important is that when you start to plan new layout you think during the planning all the time also how you are going to automate it. I know that old and ready layout is very difficult to automate.

Waiting impatiently for a new CS3 + controller, and opportunities brought by it and potential problems.

Happy Training everyone

Regards

Jukka

Hi John,

I can only agree with previous comments regarding computer control.

If you have some familiarity with a typical windows based program, it's fairly straight forward to do some simple programming.

Many people refrain from computer control because they assume it's all about just pushing a button and then sit back and let the computer do all the work. But is far from that. Many users would have split operation with e.g. some trains running on computer schedules while others are run manually by the user. It gives a whole new fun aspect.

And you will also be able to get much better speed control that looks proto typically.

What I have learned so far with computer control:

Avoid using the central station for feedback. Bypass system by having LDT-HSI. Faster and more reliable.

Make som planning on how many and where to place the blocks with sensors. A small simulation on paper will easily show if you are missing a block section.

Read the manuals first! I highly recommend to have a look through Traincontroller manual (e.g. the bronze version). Very well written and the concept is similiar for most softwares.

Defiantly start out simple. Don't start out with complicating automation of shunting freight wagons. Start with simple automation of trains on the mainline.

The computer only knows what it knows. There is some work involved in letting the software know the speed behaviours of each locomotive (speed profiling). And if you have made some changes on the decoder you will have to do it all over again.

Have you thought about just making small cuts in the K track to isolate the outer rail. Should be fairly easy with a dremel, even though the track is ballasted? I guess the main challenge will be to solder the S88 wires on the rail.

Brgds Lasse

Hi John,

First of all, welcome to the world of computer driven layout !

I strongly suggest you to read the beginning of the thread https://www.marklin-user...ftware---Tips-and-tricks
and especially the post #2 which describes the basic concepts.

Apart that, most of important answers have already been written by other members in the posts above.
I'm going to try to bring some complements and/or explain things from my point of view...

What is the purpose of a computer driven layout ?

-To allow SAFE and COMPLEX automation of MANY trains, much more than your mind can handle and much much more than the ridiculous features provided natively by the CS2 (it's my opinion). Don't get me wrong: the word "ridiculous" refers to the set of features supported by the CS2 in terms of automation, not to the people trying to use them to achieve acceptable automation...
-To allow manual management of trains NOT already driven automatically, i.e. provide interactive play with some locos which are not under running automation: manage sidings interaction, manage a steam depot, prepare trains for running, insert them into a planned traffic, etc... The possibilities are close to the infinite !
-Be astonished while the course of some trains is chosen at random by the software when running in automatic mode, and depending on the available free routes
-Create schedules (one time or recurrent) of trains running thru various locations of your layout, and run them
-Create animations of some scenes of your layout while trains arrive is special locations
-and many more...

What is not the purpose of a computer driven layout ?

-To interact ONLY thru your PC screen and click on some turnouts and/or signals and/or click on sliders to increase/decrease the speed of some locos, and always stay in manual mode. I think you can get quickly disappointed, as the number of things you have to manage at the same time may increase rapidly, and thus creates a major disaster ! The only gain of such an approach is a total rejection of the computer driven approach. The manual mode must be a complement of the automatic mode, and/or a way of testing things while setting up the computer control.

A word of caution: Interacting thru a mouse and a screen on your PC with your layout is different than interacting with a hardware Central Station. Some people like it, some people really dislike it. You'll have to choose your camp after a reasonable amount of test time.

What do I need to start ?

-As already said, a simple test layout to learn the basic concepts and discover how things run. Please, NEVER start on your definitive layout if your don't master the basic concepts. This simple test layout must implements the main basic concept, i.e. blocks, complemented by feedback occupation information of each block. Then routes (oriented segment of travel from one block to another) are implemented. This provides you the mandatory things to allow automatic traffic. Please note that a block without any feedback information is... nothing !
-A software like TrainController, iTrain, WinDigipet, Rocrail, JMRI...
-Digitized locos. Analog locs are not manageable by a software.
-Digitized turnouts (K83/M83 or equivalent driven turnouts). Analog turnouts are not manageable by a software.

Do I need brake modules ?

-For automatic mode, as already replied, NO. The computer does the job.
-For sections of the layout where blocks are not implemented, thus not eligible for automated traffic, it's up to you...

Do I need to implement signals at the end of the blocks ?

-For the general rendering aspect, yes !

Do I need to cut off power in sections before signals ?

-For automatic mode, as already replied, NO. The computer does the job.
-For sections of the layout where blocks are not implemented, thus not eligible for automated traffic, it's up to you...


What is the best software ?

It depends on many things. There are commercial software you must pay for any feature to be available. There are free software/open source software you can use with almost the same features and the quality of a commercial software.
Check first the quality of the documentation, the help available thru various communities (forum, online help, software editor...), the speed of implementation of new features, the richness of features, the robustness of the software, the cost, etc...
As already said, learn first the concepts - they are all the same in the various software, only the implementation may differ, then try to quickly implement them on the test layout. Some commercial software allows a test period without paying, with all features (or limited features) enabled.

Comparing the various software would need a dedicated thread and a lot of time.

Back now to more specific questions you asked for...

Sensors (feedback information) should use middle track or external tracks ?
It depends on :
-Type of track used : For K & C track, use of contact tracks (K-2295 for instance or Peco SL11 insulators) is generally the preferred choice. It corresponds to "external tracks" and use of the standard family of S88 devices (6088, LDT RM-88-N, Viessmann 5217...). I do not recommend the use of other devices like reeds/magnets (difficult to adjust precisely and all loks should have a magnet) or momentary contacts triggered by a pickup shoe (not reliable) like K 2299. For M tracks, if you can't find contact tracks section, the alternative is to use "current consumption detection devices" (a loco is present is a section equipped with such feedback devices), like Viessmann 5233, LDT RM-GB-8-N, or Marklin S88-DC feedback module - 60882. It corresponds to "middle track". You may also mix on the same layout traditional S88 detection and current consumption detection.
-Pros and cons of traditional S88 detection (ground detection) versus current consumption detection:
* "Traditional" pros:
.. Very easy to implement in a Marklin world (K & C)
.. Cheap
.. Reliable
.. Allow occupancy detection of lost wagons/coaches inside blocks
.. Any axle establishing the flow of ground potential is detected between the 2 external rails
.. The size of the detected zone is not limited on plain tracks (do not use on turnouts)
* "Traditional" cons:
.. one of the 2 external rails (the one which is insulated between 2 K2295 or 2 Peco-SL11) is not permanently linked to ground, thus allowing less conductivity and possible problems. This can be correctly cured by the 'diode trick' (See Ross Web site)
.. an already installed layout is not very easy to modify. This can be solved by cutting one of the two rails with a dremel or similar tool
.. for K track, you need to solder a wire, in the insulated section, on a rail (difficult and risk of melting tracks), or use of K 7500 plug (not beautiful, but there is a workaround)
.. may need more ground plugs outside of the insulated sections
* "Current consumption detection" pros:
.. Soldering on the central conductor is easier than on external rail (K-track)
.. Keep the 2 external rails to ground potential
.. Cases when ground detection is not adapted (when a loco is on a turnout, for instance)
* "Current consumption detection" cons:
.. sections of central conductor must be insulated
.. only a device consuming current (i.e. a loco decoder or a light in a coach) can be detected.
.. no lost wagon/coach detection is possible
-Other devices: infrared detection devices, RFID detection devices (see LDT web site)...
-Already owned devices: I know you have experimented optical sensors detecting the passage of a slider ("middle track" like). If the output signal can be inserted in the S88 feedback module, then you can easily reuse them.
-Ease/difficulty of modification of existing layout to set up the sensors (see pros & cons above)

A consequence of using feedback modules is you need to implement the S88 bus, either thru the native connection to the CS/Ecos, or thru LDT HSI-88 devices connected to the PC. In all cases, you are limited to a total of 496 feedback contacts. This is quite an important number, but your layout is very big.
An other consequence of using a PC driven layout is you need to setup a small PC network allowing the CS2 and the PC to communicate together.

Fade in/fade out signal modules reuse.

To be honest, I don't know a lot. I personally use Viessmann 472X light signal series with Viessmann 5229 signal decoders, and the fade in/out feature is provided by the 5229. The only thing I'm thinking about is LDT LS-DEC family of hardware signal decoders. I know if you plug in such a decoder a Viessmann light signal of 401x family, you'll be able to get such a feature... but LDT signal decoders are not that cheap...
Other members may provide more useful information on this topic...

A word of conclusion.

Computer driven layout is a huge world which provides fantastic possibilities.
In my opinion, everybody can succeed in such a world and have a lot of pleasure.
But, first you have to master the concepts and the implementation of these concepts in the software.
Be warned that such a software is a complex tool. It's NOT only a point & click affair.

The things you have to keep is mind are:
-Go each step after each other; take the necessary time.
-Learn something, then test it immediately on the test layout
-Be assisted by a friend who already knows the stuff and/or ask for help in the forum. Some members already know some sophisticated software and may provide help to you. Also use the dedicated forum of the software you have chosen. You will find many information and valuable answers
-Decide if you have pleasure while using a software, before going to modify your layout.

Cheers
Fabrice

Fabrice, thank you for the very detailed and informative answer. I have a followup to this particular segment of your post.

Beyond the obvious of digitally controlled locomotives and turnouts and a layout separated into sensing blocks....

You need feedback (as you mentioned) which can be via an S88 solution (via a controller like a CS or some dedicated S88 module).

And the computer software running on a PC...

But you also need a way to get your computer software's instructions transmitted onto the layout tracks. So what are the options for that hardware part of the setup?

I gather that a Central Station provides the ability to interface to a computer and thus act as the hardware to push commands via the CS2 -> Track Box -> Track.

But is it possible without a complex bit of kit like a CS (or similar from other makers).. and can instead be done with just a simpler bit of hardware which only works on getting commands from the computer and patching them onto the Track Box ?

Hi John,

Lots of very interesting comments above.
I use Rocrail since 3 years with S88, it is free, works very well, has a supporting forum (also in English)
My experience is:
- preferably use occupation information ( with C Track or K Track) and avoid slider operated contact tracks (M-track)
- have all the length of the block managed by 2 S88 contact sections (avoiding length of track without any detection)
- Each block has 2 contacts: one to slow down if required and show occupancy, the second to brake smoothly and stop if needed;
This second detection length should be the same on most block (I use approx 25 cm).
This allows to select for each loco the speed you want and by adjusting the braking strength, to have a very precise stop location for all trains.
- patience to learn all functionalities is mandatory and a few months are required to freally get started

Here are two videos of my computer-controlled train:
- the computerised turntable: Video Steam depot
- the train itself: Video Train layout construction

In my case Rocrail drives 31 locos simultaneously all with sound and mostly MFX.
Enjoy the fun of computerized trains

Jean

Jean,

Does this not mean that you really have 2 blocks that work together? The first block detects occupancy and is used by the computer to slow down the train, and the second block right after it lets the computer sense when you are in the stopping block and has it run it stop-train commands?

good replies!

Start simple and test until you understand it.
No need for removing power and brake modules - that is analog stuff.
Signals are cosmetic.

See also: http://cabin-layout.blogspot.com/2006/11/computer-control.html especially the s88 positioning info...

Edited by user 22 April 2017 20:48:03(UTC)  | Reason: url change

I am a bit puzzled with this computer driven stuff. I have the CS2. What happens if I want to take a train, and run it around, let's say with the Mobile Station, among the computer controlled trains? Is this possible? I have a hard time seeing how the computer can know where my train is, without me running it through the computer (not using anything else but the computer to manage the trains)?

If want the above, I cannot see how I can get past using braking modules in front if signals.

Also, if I want to run the entire layout manually one day, just because I feel like it - again, using a mobile controller - how will all the computer controlled braking sections react? Will they work at all? I mean, how will they know which train to slow down?

As I see it, if I want the flexibility of running automatic some days, and manually on other days, I cannot rely on the computer alone?

I might have gotten all this wrong.

Thanks for replying.


If I am away from my CS2, I some times operate a turnout using the MS2. Will I still be able to do this, and will the computer be able to see that I do this?


Oh. So if I just take my MS and operate the train, there is no guarantee that the braking sections will work as they do when I use braking modules?


I currently operate my turnouts and signals from the CS2.

Hi Minok,

No, each block needs one or multiple sensors:

In Rocrail you have multiple possibilities for block detection:
- one sensor (ENTER2IN): your train has fully entered into next block and can free the previous block. Simplified functionning with some drawbacks (less secure and less smooth speed control)
- two sensors (ENTER and IN): ENTER: your train enters in next block and start decelerating if needed. IN: your train is fully in next block, frees previous block and stops if needed.
- three sensors (Enter, PRE2IN and IN): PRE2IN allows to stop shorter trains earlier (e.g.: a railcar in the middle of the platform)
- other possibilities

Each block:
-is bi-directional but could be set to one direction
- allows direction reversing (if selected)
- allows shuttle train operation with loco pushing the train without complications
- allows selective loco type (non-electrified which cannot be used by electric locos)
- allows ramdom waiting time, specified waiting time or loco-specific waiting time or no waiting
- allows trains longer than the block
and more.
It is really fun to program all this.
Cheers

Jean

What you are describing as "track segments" really seems to describe what the computer control programs call a "block". I think where people are getting confused is in the word "Block" which in the old analog world referred to an isolated section of track that could have power to it turned on or off. In computer control of a digital layout power to the track stays constant for the layout allowing both manual and automatic train control to happen at the same time. The computer controls the locomotive(s) selected for automatic operation, the switches for the route and the signals for the route. The signals become cosmetic and do not control anything.

Peter

Hi Minok,

You are perfectly right. No overkill

No track zone, except the switches (and this can also be discussed but it's more complicated), should be left without a block being implemented.
If you left a segment without feedback, then the computer won't be able to know there is something on the track in such a zone, and as a result unable to take the good decision (prevent a train to go in this zone, trigger emergency stop, start a train in such a zone, etc...)

Reminder: A block is a segment of tracks (without any switches) when only one train/loco can be at a time. In a block, a train can be stopped if needed.
See: http://wiki.rocrail.net/doku.php?id=block-en

Cheers
Fabrice

Yes, indeed we are "on the same page" (i.e. thinking alike) here.
The LDT DiCoStation does seem to fit the bill.

An Arduino or RaspberryPi sort of solution would also be possible. However, thats more for folks that want the joy of doing the hardware and firmware writing fun.
I am, at present, completing a course in VHDL (I took a university course in Verilog about 12 years ago) - so using that technology I'd even be able to program an FPGA to create the track signalling in a very fast and reliable way. All a wonderful project. All not where I want to necessarily spend my hobby time at this point. I think I'd get more "joy by the minute" in laying out track, building substructure, mountainsides, bridges, laying out train stations, and the somewhat less 'electronics involved' tasks. I think I can do both, and maybe some day, I'll even look at building an FPGA based solution that could do mFX (assuming its not cryptographically protected).. but for now I'll stick to someone else building/testing the hardware.

With the CS3/3+ imminent, the price of a descent used CS2 may drop in the next year or two, and be more reasonable to use, as well as the LDT DiCoStation and any similar products.

I like the HSI-88 solution, as it decouples that entire electrical feedback task from the signalling/control task.

Thank you for the insightful feedback.

Thanks.

Its certainly true that you can have trains longer than the segment of track where you can detect occupancy in (which we won't call a block), a train can thus trigger occupancy signals in several sequential detection zones. Those zones, as is now clearer to me, are a subset of what it takes to make a block, but even still several blocks can be covered by a long train.

I can envision gaining some efficiencies if you make the longer detection zones the length of the shortest train, thus providing for the possibility of more traffic density when the trains vary substantially in length.

Of course such solutions require more complex planning/management but thats what a lot of the digital side of things is about.