Automatic signaling


P

patricka

New Member
Hello,

I will introduce myself... I am a N & On30 scale modeler, doing a lot from scratch especially scenery and electronics. I am living in Montreal, Canada.

In the past years, I developed an automatic signaling system all electronics as would be a real system. Maybe some of you already are aware of what I did on my YouTube channel https://www.youtube.com/channel/UCFCViYE_2bIZRC88drGYhIg

If you are interested, we can discuss it here in this forum or we can connect live in the Springfield show at the end of this month. I will be exhibiting my test layout with some examples...

Feel free to connect,
Cheers,
Patrick
 
Hello Mike,

Any chance you would attend the Springfield Show (Massachusetts)? We could discuss it in details with a live example...

Anyway, here are few words on the prototype. The layout is divided into independent sections, with 2 or 3 sub-zones in each. With this configuration, the whole system is able to 1) draw all signals depending on the occupancy state of each section (in the video example, it is depicting french signaling system, but it could be adapted to US or Canada) 2) execute some scripts to drive automatic trains on the layout without any risk of collision. The automatic trains will follow the signaling principles, change speed depending on the aspect and stop if signal is red (closed).

The sub-system 1) is autonomous (and can be installed alone on a layout) while the sub-system 2) needs the first one to be implanted...

The whole is running with electronic boards only (no computer) - no software to configure for the user, only CV to program as with a regular decoder. Each "section" board is made to be connected to sub-zones inside each section. So each section needs a board. If the layout is already divided (electrically isolated) into sub-zones and sections, the system if fairly easy to install.

It is also interesting to mention that it works with Digitrax-Loconet bus for now. But I can probably adapt it to other bus protocols if the documentation is available somewhere... The data bus is used only to generate the DCC signal in the tracks. The prototype is using its own bus/protocol to manage the layout and the boards.

Also, this is only a prototype. It works well but it has not been tested on a "large" layout. My test layout is really small compare to a real one. This is one of my next steps to validate the principles.

That is what I can describe in few words. If you have specific questions, let me know. It would be my pleasure to answer them...
Cheers,
Patrick
 
AS too Springfield I will not going mainly cause of distance, I live in Illinois. AS to Digitrax I use that system now. How large a layout are you looking for?
Mike
 
I guess that a mainline with around 10 consecutive sections could be enough to test the signaling...

Then we can add some branchlines to manage the turnouts and different routes...

Maybe a total of 20 sections would be ideal.
 
20230311_141617.jpg
20230311_131235.jpg
 
@patricka: I am in process with this also except I am using RS485 for comm's between control points, and I2C from control points to devices. The 485 stuff can be cat5 net cable. I don't know about ribbon cable for this - could be a ton of cross talk depending on system speed - have you tested that? In my case I have at least 22Ft between control points and no more than 36" from the control point to device. RS485 does not breath hard with 22Ft distance, and I2C is still pretty fast at 36".

Also, looks like that engine is to the left of the block boundary. Should not that signal at the previous block boundry to the right ( protecting ) be yellow?

Later
 
@ctclibby I have tested several communication channels and RS485, CanBus, etc... are really difficult to make them robust in such environment... I understand it is a paradox because they are supposed to be robust in noisy environment... Anyway I have used simple Serial between modules in this prototype and it is really efficient. I will probably end up with something else if I commercialize this product... The ribbons are used to connect the signals masts to the boards.

By the way, I2C is not supposed to run long length I guess... But if it works on your side, it is fine.

About your remark (you have good eyes ;) ), the picture has been taken right at the moment the loc is entering the left curve section. So the signal there is still yellow from the next section, and the previous one to the right is still red because the section at the top was occupied. Anyway, I can assure you it works perfectly :D:D:D

Hope to discuss it in the future either at Saint-Albans VT or Springfield MA...
Cheers
Patrick
 
You must log in or register to reply here.



Back
Top