|
Whats funny is the other day I was thinking of building a marble machine with
LEGO. After reading about this contraption system, I got super-motivated and
built myself a ball pump last night. I only spent a few hours on it, but I think
I have it pretty much perfected . . . its jam-proof if the marbles come once
every second or so. And even if it does jam, a clutch gear protects my old 9V
motor (which is geared down considerably). Its built pretty solidly; all I have
to test my stuff are glass marbles, which are pretty heavy in large numbers.
Tonight I started a few conveyor belt designs, and keeping tension in the tread
links is proving difficult. Any ideas on how to do that?
Im also working on a rollercoaster-type track with the ribbed tubes from my
Mindstorms sets. Building a framework to support it is trickier than I thought
it would be.
I definitely plan to attend Brickfest to add my contribution to the Great Ball
Contraption. Im going to need someone to send me some soccer balls, please. I
dont have any LEGO Sports stuff at all.
One suggestion for people: I dont like mystery mechanisms. I like to see how
stuff works, so please build your stuff with as many internal parts visible as
possible. For example, the stairs are mysterious until you see them from the
bottom -- a cam shaft is the secret. Get rid of the walls and Ill be happier.
:)
|
|
|
In lugnet.robotics, Jordan Bradford wrote:
|
I got super-motivated and built myself a
ball pump last night.
|
Wow, nearly untouched territory... pumping Steves... oh, never mind.
On a more pratical note, does your pump meet the height requirement for the GBC?
Ive got two ball pumps working, but both have the same problem - allowing room
under the hopper for the piston mechanism makes it hard to have a deep hopper
(and Id really hate to use a high hopper for the ball pump that has to be
filled from the standard input via another mechanism).
|
its jam-proof if the marbles come once every
second or so.
|
Mine has no problem with jamming, but occassionally problems with clotting in
the hopper starving the pump itself. Can yours handle a crate of balls dumped
all at once? I wonder what the ultimate percentage of continuous output vs.
batch output modules will end up at.
|
keeping tension in the tread links is proving
difficult. Any ideas on how to do that?
|
If you need constant tension, one way to do it is tensioner. A third gear in the
chain, that can be moved in and out changing the shape (& therefore length) of
the triangular chain path. This doesnt have to be hand-tuned either - use a
idler gear on a small free-swingin arm that is held in tension by a weight or
rubber band.
|
One suggestion for people: I dont like mystery
mechanisms.
|
Well, one of the ball pump designs Ive built can have one side of the
mechanism built out of clear panels... that is, if I could *get* any of them.
After I built it, I found S@H had run out, and Bricklink has (currently) *one*.
Sigh... Also, most of my designs are ending up rather skeletal, due to the lack
of pieces left for building.
--
Brian Davis
|
|
|
> I definitely plan to attend Brickfest to add my contribution to the Great Ball
> Contraption.
We look forward to seeing you there...
> I'm going to need someone to send me some soccer balls, please. I
> don't have any LEGO Sports stuff at all.
E-mail me your address.
> One suggestion for people: I don't like mystery mechanisms. I like to see how
> stuff works, so please build your stuff with as many internal parts visible as
> possible. For example, the stairs are mysterious until you see them from the
> bottom -- a cam shaft is the secret. Get rid of the walls and I'll be happier.
I really like to build my stuff with a very clean look. That usually means hiding
all the inner workings. That's why the stairs looked like this in my first build:
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=1055888
However, when I rebuilt it to be 32x10, I decided to open the bottom as much as
possible, to allow people to see in:
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=1068037
This gives it more of a "contraption" look to it.
Steve
|
|
|
In lugnet.robotics, Brian Davis wrote:
|
In lugnet.robotics, Jordan Bradford wrote:
|
I got super-motivated and built myself a
ball pump last night.
|
Wow, nearly untouched territory... pumping Steves... oh, never mind.
On a more pratical note, does your pump meet the height requirement for the
GBC? Ive got two ball pumps working, but both have the same problem -
allowing room under the hopper for the piston mechanism makes it hard to have
a deep hopper (and Id really hate to use a high hopper for the ball pump
that has to be filled from the standard input via another mechanism).
|
I dont know if it fits into the standard or not. I didnt plan on using it as
my first device, anyway. Im going to build a simple large hopper for the
previous persons contraption to feed into mine, and then Ill have a feed chute
or something. Maybe Ill use a conveyor/bucket mechanism to get balls out of the
initial input.
|
|
its jam-proof if the marbles come once every
second or so.
|
Mine has no problem with jamming, but occassionally problems with clotting
in the hopper starving the pump itself. Can yours handle a crate of balls
dumped all at once? I wonder what the ultimate percentage of continuous
output vs. batch output modules will end up at.
|
My tiny little hopper for the pump does clot as you say, which is a problem
everyones going to encounter sooner or later. But I can guarantee nothing will
get stuck if the balls come one at a time.
|
|
keeping tension in the tread links is proving
difficult. Any ideas on how to do that?
|
If you need constant tension, one way to do it is tensioner. A third gear in
the chain, that can be moved in and out changing the shape (& therefore
length) of the triangular chain path. This doesnt have to be hand-tuned
either - use a idler gear on a small free-swingin arm that is held in tension
by a weight or rubber band.
|
Heh, once again my over-engineering defeats me. I tried using two shock
absorbers to push a gear into the hanging part of the chain, but they just
torqued the gear so that the chain wouldnt move. I should always try the
simplest approach from now on. Thanks for the tip.
|
|
One suggestion for people: I dont like mystery
mechanisms.
|
Well, one of the ball pump designs Ive built can have one side of the
mechanism built out of clear panels... that is, if I could *get* any of them.
After I built it, I found S@H had run out, and Bricklink has (currently)
*one*. Sigh... Also, most of my designs are ending up rather skeletal, due to
the lack of pieces left for building.
|
Lack of parts. Yep. Most of my contraptions will be skeletal for the same
reason. However, I have a 4561 (4160), and that has a lot of transparent panels
for the windows. I used one of those for the part of the pump that I think
people would consider mysterious. Now they can see each ball being lowered under
the column of balls and pushed up into it.
|
|
|
> > > keeping tension in the tread links is proving
> > > difficult. Any ideas on how to do that?
> >
> > If you need constant tension, one way to do it is tensioner. A third gear in
> > the chain, that can be moved in and out changing the shape (& therefore
> > length) of the triangular chain path. This doesn't have to be hand-tuned
> > either - use a idler gear on a small free-swingin arm that is held in tension
> > by a weight or rubber band.
>
> Heh, once again my over-engineering defeats me. I tried using two shock
> absorbers to push a gear into the hanging part of the chain, but they just
> torqued the gear so that the chain wouldn't move. I should always try the
> simplest approach from now on. Thanks for the tip.
I'm not exactly sure how many "teeth" = 1 stud, but I know it's not a whole number.
So, I've found if you can change the length by 1 or 2 studs, and add a few chain
links, you don't need to add a tensioner.
Steve
|
|
|
In lugnet.robotics, Steve Hassenplug wrote:
> I'm not exactly sure how many "teeth" = 1 stud, but I know it's not a whole number.
It's exactly 2.5. 10 teeth on a gear rack, divided by 4.
Steve
|
|
|
In lugnet.robotics, Steve Lane wrote:
|
In lugnet.robotics, Steve Hassenplug wrote:
|
Im not exactly sure how many teeth = 1 stud, but I know its not a whole
number.
|
Its exactly 2.5. 10 teeth on a gear rack, divided by 4.
Steve
|
Im using two upside-down 1x16 TECHNIC beams to straddle 24-tooth gears on each
end (see my great ASCII art below, top-down view, not to scale). I might not
even need tension in the chain at all; the tread links slide on the flat bottom
part of the beams, and that seems to work fine. I havent yet built a support
structure to test the chain for real.
*
[]*[]
[]*[]
[] []
[] []
[] []
[] []
[] []
[] []
[] []
[] []
[]*[]
[]*[]
*
My buckets are simple and work pretty well, even at steep angles. Every fourth
tread link has a 1x4 plate on it, and on each plate are two of those 2x2
right-angle plates making a squarish area for the ball to rest in. Again, some
beautiful ASCII art:
Step 1:
_______
|_______| <- that's a 1x4 plate
Step 2:
_ _
| |_ _| |
|___|___| <- those are right-angle plates
What other bucket designs do people have?
|
|
|
> In lugnet.robotics, Steve Lane wrote:
> > In lugnet.robotics, Steve Hassenplug wrote:
> >
> > > I'm not exactly sure how many "teeth" = 1 stud, but I know it's not a whole
> > > number.
> >
> > It's exactly 2.5. 10 teeth on a gear rack, divided by 4.
> >
> > Steve
>
> I'm using two upside-down 1x16 TECHNIC beams to straddle 24-tooth gears on each
> end (see my great ASCII art below, top-down view, not to scale). I might not
> even need tension in the chain at all; the tread links slide on the flat bottom
> part of the beams, and that seems to work fine. I haven't yet built a support
> structure to test the chain for real.
Well, my lift for loading the train car uses both chain links, and tracks. Each
track has a 1x4 tile on it, and the balls roll up.
http://www.brickshelf.com/gallery/Hassenplug/GBC/GBC2/p1160046.jpg
That's close to what John did on his roller coaster:
http://www.brickshelf.com/gallery/Hassenplug/GBC/05johnrollercoaster.jpg
This red module: http://www.brickshelf.com/gallery/Hassenplug/GBC/10stevechainlift.jpg
uses a pair of axle joiners: http://www.peeron.com/inv/parts/6536 with some #3
axles. Brian came up with this, and it works very well.
|
|
|
In lugnet.robotics, Jordan Bradford wrote:
|
In lugnet.robotics, Steve Lane wrote:
|
In lugnet.robotics, Steve Hassenplug wrote:
|
Im not exactly sure how many teeth = 1 stud, but I know its not a whole
number.
|
Its exactly 2.5. 10 teeth on a gear rack, divided by 4.
|
Im using two upside-down 1x16 TECHNIC beams to straddle 24-tooth gears on
each end (see my great ASCII art below, top-down view, not to scale). I might
not even need tension in the chain at all; the tread links slide on the flat
bottom part of the beams, and that seems to work fine. I havent yet built a
support structure to test the chain for real.
|
I have just placed wheels regularly spaced along my chain to support it.
|
My buckets are simple and work pretty well, even at steep angles. Every
fourth tread link has a 1x4 plate on it, and on each plate are two of those
2x2 right-angle plates making a squarish area for the ball to rest in. Again,
some beautiful ASCII art:
Step 1:
_______
|_______| <- that's a 1x4 plate
Step 2:
_ _
| |_ _| |
|___|___| <- those are right-angle plates
What other bucket designs do people have?
|
Mine are similar, but use a 1x3 plate with a 1x1 plate with tooth on each end,
at a slight angle, and another 1x3 plate on top. I was finding that even with an
accurate feeder, they were occasionally bouncing over the side when I used 1x2
plates on each end.
ROSCO
|
|
|