Subject:
|
Re: Automatic decoupler designs?
|
Newsgroups:
|
lugnet.trains
|
Date:
|
Mon, 10 Jan 2000 14:51:36 GMT
|
Viewed:
|
1490 times
|
| |
|
|
In lugnet.trains, Ben Jackson writes:
> I built a version of the decoupler mentioned earlier in this thread. It works
> remarkably well! I don't have the exact brick mentioned on the web page, but • I
> do have an R2D2 foot, which is very similar: a short technic beam with a
> 4-plate high 30 deg ramp on the end of it. It doesn't matter that the bottom
> (actually side in the decoupler) is flat. Even with the flat side right
> against one buffer it will poke just one of the knuckles and decouple the • cars.
>
> As long as the decoupler starts somewhere between the two buffers it can go at
> almost any speed and will not derail the cars. If it's moving fast it will
> send one car rolling away (if it's unsecured) because of the ramp on the
> knuckle-poker. If it moves slowly the cars will actually still hold each • other
> gently. If the decoupler retracts at that point the cars will rejoin.
>
> For safety what it really needs is a way to tell if it hits something short of
> where there should be a knuckle (in which case it hit a car or a buffer). It
> can't just stall in that case because the geartrain needs a lot of torque to
> actually pop the knuckle in the good case. Maybe I could drive the motor at
> low power (with an RCX) until it clears a certain point and then power up when
> it appears that the cars are lined up right.
>
> --Ben
Ben,
I think you're correct in using the RCX in that way. I use one in a similar
manner with my remote powered switch, and it elimates any stalls. I actually
perform the opposite functionality - the motor is full power for a short burst,
then low power, then coast. This gets me over the built in detents in the
switch/point housing. In your case, you need the motor to feel out a good
decoupling position, then accellerate rapidly from that point to perform the
uncoupling routine. I haven't had a chance to play with the uncoupler design
talked about here yet, but I plan an using an RCX to control it. I'll
probably use either a light sensor or touch sensor (*shudder*) to tell the RCX
when a "good" position is for the uncoupling operation.
Ed
|
|
Message is in Reply To:
 | | Re: Automatic decoupler designs?
|
| I built a version of the decoupler mentioned earlier in this thread. It works remarkably well! I don't have the exact brick mentioned on the web page, but I do have an R2D2 foot, which is very similar: a short technic beam with a 4-plate high 30 deg (...) (25 years ago, 10-Jan-00, to lugnet.trains)
|
19 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
