HO Switching Display
DataJ
Posts: 23
Update·on the HO Switching Display.
Attached is the Track Schematic and the source code for this display.
At this point, the train will travel left to right, reverse and travel back and forth
between two of the three sidings.·
Additional·work is needed to get the train to travel down the third siding, sensed by Detector #4.
Some of the difficulties encountered while developing the code was to determine how to
hold a count once a detector was activated.· This is where the "Flag" comes into play - it
holds the count until another detector is activated, at which point another count is added and
the first flag is cleared.
There are a few questions related to this:
· 1.· Is there a·more efficient·method that would clear each flag?
···2.··During the code development, Detectors 1, 3 and 5, were used for the initial programming.
······ Now, Detector #4 needs to be added, however, its directions·will follow detector 5.·
······
······ So far, not much success has occurred in the attempt to add Detector 4 to the activity.
······ Is it possible to do this successfully, or must each event follow in numerical sequence?
Any ideas or study references would be appreciated.
·
The reference material that was helpful in·code development is as follows:
··· 1.· Parallax Manual entitled, "Industrial Control - Version 1.1", pages 40 and 41, covering "Edge Triggering".
··· 2.· Parallax Manual entitled, "Process Control", Chapter 4, pages 104 through 109, "Sequential Process
········ Control".
Attached is the Track Schematic and the source code for this display.
At this point, the train will travel left to right, reverse and travel back and forth
between two of the three sidings.·
Additional·work is needed to get the train to travel down the third siding, sensed by Detector #4.
Some of the difficulties encountered while developing the code was to determine how to
hold a count once a detector was activated.· This is where the "Flag" comes into play - it
holds the count until another detector is activated, at which point another count is added and
the first flag is cleared.
There are a few questions related to this:
· 1.· Is there a·more efficient·method that would clear each flag?
···2.··During the code development, Detectors 1, 3 and 5, were used for the initial programming.
······ Now, Detector #4 needs to be added, however, its directions·will follow detector 5.·
······
······ So far, not much success has occurred in the attempt to add Detector 4 to the activity.
······ Is it possible to do this successfully, or must each event follow in numerical sequence?
Any ideas or study references would be appreciated.
·
The reference material that was helpful in·code development is as follows:
··· 1.· Parallax Manual entitled, "Industrial Control - Version 1.1", pages 40 and 41, covering "Edge Triggering".
··· 2.· Parallax Manual entitled, "Process Control", Chapter 4, pages 104 through 109, "Sequential Process
········ Control".
Comments
HO Switching Display - - Update #2
Attached is the update to the source code for the HO Switching Display and the associated schematics created in Cadsoft Eagle.
A method to uncouple the freight cars on each of the three sidings is still in the experimental stages.· Current off-the-shelf designs·will not·operate as needed.
Please let me know if there are any questions.
DataJ
Post Edited (DataJ) : 6/15/2009 3:45:00 PM GMT
photos.
I suppose you've tried the Kadee uncoupling solution (here)? I've used some of their stuff on a G scale layout and I like it.
What exactly is your goal in this project? It seems like you want to make an autonomous or automated display that shuffles the cars about on it's own. Also, do you want to tell where the locomotive is, or where the cars are, or where both are? I think you could get by with just the locomotive, for which I would try a different approach. I'd cut up the rail into lots of little isolated blocks, then have a sensor that tells you if the locomotive wheels are in one of those blocks. The sensing part could probably be as simple as a high pullup (100k?) that gives a high signal to the controller when no locomotive is present. When it is present, the motor pulls down on the line and gives a low to the pin. I don't think it would affect the locomotion at all. This way, if your blocks are small enough you could fit multiple locomotives (or maybe a speeder...). Since you'd have lots of blocks, you'd have to use some external logic to get more inputs.
It then starts out and picks up all the cars and returns to its beginning position on the mainline.
I haven't tried the Kadee electromagnetic unit yet, however, I have wound my own electromagnet using a 3/8" diameter bolt and 28 gage magnet wire.· The unit is a British design, however, it does not have enough energy to even move the couplers, even at 3 amps and 21 volts.
Currently I·am also·experimenting with 3/4" rare earth magnets that appear may work.· However, the NMRA HO·couplers are·designed so that a locomotive backs over a magnet, which spreads the knuckles, then the locomotive pulls forward disengaging the knuckles, which are still over the magnet.· The locomotive is then reversed, and since the couplers are pulled to the outer sides of the track they will not reconnect and the car can then be spotted and the locomotive can be pulled away.· I have used this method on my main layout in the attic, however, after re-coupling, the locomotive needs to quickly pull away, or they will de-couple.
I am hoping that will some modifications to the knuckles that a servo & rare earth magnets can be used to slide beneath the layout to decouple and then pull away.
·