Subject:
|
Re: Automated shunter/switcher
|
Newsgroups:
|
lugnet.trains
|
Date:
|
Wed, 21 Mar 2007 15:05:40 GMT
|
Viewed:
|
4216 times
|
| |
|
|
In lugnet.trains, Chris Phillips wrote:
> In lugnet.trains, Matija Puzar wrote:
> > So, I got this idea of making an automated shunter for our group's
> > exhibition layout, it would be a nice eye catcher and a great oportunity
> > for the big trains' motors to rest and cool off.
> > In the simplest version, it would just come to a bunch of parked cars,
> > take them to another location, park them there and then go away somewhere
> > and give place to the big trains.
> >
> > This is most probably not a new idea, but I haven't seen any solutions
> > so far (if I missed them, I apologise).
> >
> > So, a few problems arise:
> >
> > * Detaching the cars from the locomotive - anyone with some clever and
> > proven ideas? :) A mechanical arm coming in between and separating the
> > magnets might be a solution, but it would be hard to make it unnoticable.
> > It would also be a problem with regards to the number of motors needed if
> > I'd want to be able to shuffle the cars automatically, i.e. sometimes take
> > 1, sometimes 3, etc. An electromagnet would be also an idea, but hard to
> > implement for more than just the locomotive (also, it might prove to be a
> > heavy burden for the DCC decoder and funny things might happen if the
> > engine loses power in the middle of the loop :)).
>
> I built a powered decoupler from somebody else's instructions quite awhile
> ago. A quick search brings up this LUGNET thread that might help you...
>
> <http://news.lugnet.com/trains/?n=19117>
>
> > * A lot of reed switches will be needed and multiplexing becomes a
> > necessity. Has anyone been able to make a stable multiplexer for more than
> > 3-4 sensors? If 3 is recommended, would you recommend the parallel or the
> > serial version, or does it boil down to the same? I tried making one for
> > 4 sensors, but soon found out how problematic it gets if the resistors are
> > not extremely precise.
>
> I have multiplexed up to 4 reed switches using four different resistance
> values in series with each switch, but it is tricky. (I used something like
> ~1K, ~2K, ~4K, and ~8K with 5% and stock resistance values, IIRC.) You
> usually have to calibrate the code to match the actual sensor value ranges
> that are returned from your particular setup. But you should be able to make
> this work using a single reed switch located right at the decoupler. Just
> use low speeds to count passing couplers, and be prepared to make some slight
> forward-backward micro-adjustments at a crawl. I think some of the GBC train
> modules use this technique.
why not just keep it simple? build the decoupler into the switcher, put a touch
sensor on the side or bottom and use it to trigger a program that is something
like this:
forward untill touch sensor depressed "ie attaches to train cars"
stop
wait 10seconds
reverse untill touch sensor depressed
stop
uncouple
reverse 1 second
stop
wait 40seconds
slow forward ignnoring first touch sensor depression
stop at second touch sensor depression
uncouple
forward untill touch sensor depressed..
stop
a way to keep this interesting and not just a train bouncing back and forth in a
straight line is to use switches in your favor. if you leave the switch switched
you can go through it one way in one direction and go someplace completely
different going the other direction. if your confused i'll draw you a picture.
oh, if you use a rc train motor you don't have to worrie about polarity issues
with the track either.
i hope this helps.
ondrew
|
|
Message is in Reply To:
 | | Re: Automated shunter/switcher
|
| (...) I built a powered decoupler from somebody else's instructions quite awhile ago. A quick search brings up this LUGNET thread that might help you... (URL) (...) I have multiplexed up to 4 reed switches using four different resistance values in (...) (18 years ago, 21-Mar-07, to lugnet.trains, FTX)
|
9 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
