| |
In lugnet.technic, Duane Hess wrote:
> In lugnet.technic, Duane Hess wrote:
> > In lugnet.technic, Chris Phillips wrote:
> > >
> > > You also might check out a construct called a Directional Transmission. This is
> > > a simple setup using a sliding worm gear that does pretty much what you are
> > > describing. Michael Powell has a web page describing his design that I used
> > > successfully to build a robot arm that used a single motor to perform two
> > > actions: (1) grip+release and (2) lift+drop.
> > >
> > > http://www.sonic.net/~rci/transmission.htm
>
>
> Well, I think this is going to work. Do you have any pictures of the
> logic/gearing used downline from the transmission? I am concerned with how the
> worm gear makes the transition from front to back and the resulting interaction
> with the drive gears. I only have MLcad until I can get home from work to play
> with the real brick and experiment.
The key to making this work is to ensure that there is tension at either end of
the run that keeps the worm gear at one end of the axle or the other until it
hits a hard stop. For example, my robot arm uses rubber bands to hold the
gripper closed, and gravity to hold the arm down in the absence of direct force
applied by the worm gear. So the worm gear at one end is pushing against the
rubber bands to open the claw, and at the other end is pushing against the
weight of the arm itself to raise the arm.
This is important for two reasons. First, when the worm gear is not in contact
with one set of drive gears, that mechanism must be held in a predictable
position, otherwise it could flop around while the reverse function is being
controlled by the motor. Second, when the motor is undoing the action at one
end, the tension in the gear train caused by the rubber band or gravity holds
the worm gear up against the stop until that action is complete before allowing
the worm gear to slide over to the other position.
This is fine for my robot arm application because I never need to raise the arm
with the gripper open, and I never need to open the gripper with the arm in the
raised position. If I did need to do either of these things, I probably
couldn't have used the Directional Transmission approach.[1]
I have one picture on BrickShelf that shows this creation, but it is a massive
image (before I learned to scale them down to 640x480 for uploading) and it is
taken from far enough away that it doesn't show much detail. Still, for the
bandwidth-wealthy: http://www.brickshelf.com/cgi-bin/gallery.cgi?i=1109 I also
have a smaller, closer picture on my web site, but it does not show much more:
http://www.drvegetable.com/pic_franken01.html
I'll try to post some close-up pictures today of the robot arm if I get the
chance. Unfortunately, that particular creation is not very elegant (being one
of my first MOCs out of my Dark Age) so it is not really easy to see the
functional mechanisms.
- Chris.
[1] This may not be entirely true. After thinking about this more, I realize
that if there was a second worm gear downstream in each gear train, the motor
could control the arm or gripper, but the gripper or arm could not spin the
motor or axle in reverse through the gear train. The trade-off being that you
can only drive each function in one direction, so you'd have to make it so that
the function at each end was cyclical. ie: Keep spinning the motor forward and
it would repeatedly open/close the grip. Spin it in reverse and it would
repeatedly raise/lower the arm. This is because the tensioning effect could not
travel backwards through the gear train to prevent the worm from sliding over to
the reverse function. Unless you can build the tensioning mechanism into the
gear train between the directional transmission and the secondary worm gear...
Just thinking out loud here, so I don't know if I am making any sense. I'll try
to post a picture worth 1,000 words.
|
|
Message is in Reply To:
 | | Re: One Input - Two Outputs?
|
| (...) Well, I think this is going to work. Do you have any pictures of the logic/gearing used downline from the transmission? I am concerned with how the worm gear makes the transition from front to back and the resulting interaction with the drive (...) (21 years ago, 12-Nov-03, to lugnet.technic, lugnet.robotics)
|
7 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
