[youtube=http://www.youtube.com/watch?v=Te9X4VRj8dU]
This interesting little sculpture caught our eye. Called the Perpetual Ball Roller, it simply rolls a ball on a track. It has both manual and automatic modes with variations in the automatic mode to keep it amusing. This is very elegant, and would be fun to have sitting around to play with. The only problem is the noise. The servo that he is using is quite loud. What could he do to make it function silently?
[via Instructables]
If you spring mount the rail – you could use an electromagnet to compress one of the springs. Totally silent – no motors.
Nichrome?
Hardly perpetual tho, requires batteries!?
First, mount the track stationary and level. Then energize the tracks with a low voltage to pass a current from one rail to the other through the metal ball. Mount one or more inductors under the track, normal to the ball’s magnetic field, to create a perpendicular force that will push it around the track. No moving parts, except the ball.
I need to do the math though. Since friction in the system is pretty low (metal on metal), the current, hence force, needed to push the ball around shouldn’t be too large. Pretty cool to make it powered with a solar cell. It could be made to appear as a perpetual motion machine (but only with the lights on ;-).
If you use electromagnetics to manipulate the rail, you’ll have to watch out for influence on the movement of the (presumably) iron ball, or just use a ball of a magnetically inert material of course.
I should add: the drive inductors would need to be pulsed according to the position of the ball, as in a linear induction motor, so sensors (Hall effect?) would be needed.
OK, one more thing … pulse BOTH the rail current AND the drive inductors, to improve efficiency.
jdn: if the energy requirement was low enough, instead of using solar you could use a radiant energy receiver or a radio wave powered system.
I do like jdn’s solution but what I first thought this used was an electromagnet to gently push the ball around the track. A few capacitors and a coil would silently propel the ball, if not in noticeable spurts.
brainiac27: you’re right, the rails wouldn’t need to carry current at all.
On the other hand, back to my idea, instead of perpendicular inductors, simply use fixed magnets and pulse only the rail current.
at the moment this looks like a “dumb” device… set the oscillation rate and make the marble keep up. how about making it “smart” by isolating a few sections of track and connecting it to an arduino or something, so the device knows where the ball is and can manipulate the oscillation rate to keep the ball spinning at whatever speed you want
It should be possible to move the outer or inner ring horizontally in a vaguely circular manner, creating a wider track ahead of the ball and allowing it to perpetually roll “downhill”…
“The term perpetual motion, taken literally, refers to movement that goes on forever. However, the term more commonly refers to any device or system that perpetually (indefinitely) produces more energy than it consumes, resulting in a net output of energy for indefinite time”
Shape memory alloy wire. That would be cool to see.
A microcontroller seems to be kind of an overkill for this… why not just hook up a geared down motor?
thats not perpetual… read the definition of it.
but i like the project!
Wait, where is the arduino?
What a waste of a Hack-A-Day entry…how is this a hack? Hack of what?
it sounds like robots having sex
craig: not quite, the track has to be ever expanding to keep the ball rolling, at some point the track will be wider than the ball, and it will fall through. Maybe could incorporate a “reset” every few revolutions though?
Why not use muscle wires in stead of a servo.
When I first saw it, and even now, all I can think of is electromagnet. Silent and always entertaining. Now if you didn’t power the rails, you could make a ‘cage’ for the ball that goes over the top in an arch shape, which would make it more aesthetically pleasing for me.
Hall effect would be interesting to implement as well.
But of course he can use couple of Miga Nano-muscles and be done with it: http://www.sparkfun.com/commerce/product_info.php?products_id=8782
@poot: HACK = http://en.wikipedia.org/wiki/Hack_%28technology%29
Yeah, it would be cool to use coils and sensors to treat it like a railgun almost, constantly pulling the ball along. You could embed that all slightly below the surface of a piece of wood and it would be silent, and very cool looking. :)
-Taylor
It would be very cool looking to have a ball moving on a stationary surface, though I suspect half the charm of this is that it ISN’T stationary. What I might do is add a second pivot perpendicular to the first so that it can do a more circular smooth motion of the ball. This might also allow slower, quieter servo movements. As for the rail noise, perhaps replace the wires used with insulated plastic wire, or heat shrink tubing around them. The sound of a ball rolling on a surface like that is fairly quiet, but still makes an interesting noise. Could this be made to have a purpose? Faster = higher stock price or something?
Same as everyone else, use nichrome/memory wire/ etc.
It would be cool to do this with no electricity, just heated wire, possibly with a candle or something for power (large magnifying glass and the sun?)
I like the idea, keep hacking!
@haters,
hack = something built together out of available skills and parts.
What it accomplishes has nothing to do with whether it is a hack.
Hi
I am the person who made it.
Someone on the Picaxe forum told me I was on Hack-A-Day and sure enough! :D I do subscribe to the RSS feed but it’s buried at the bottom of the page so I nearly missed it.
While designing it noise of the servo was not on my mind at all, I knew it would be there but I wasn’t fussed about it. I have actually coded a mode now that uses loops to make very small changes. It is practically silent. I will upload a video of it since it seems to be an area that people first think about.
To people bothered about “perpetual” in the title should note two definitions:
2. lasting an indefinitely long time: perpetual snow.
3.
continuing or continued without intermission or interruption; ceaseless: a perpetual stream of visitors all day.
Which I think suits this piece quite well as it captures the fact it does go round and round and round.
To the comment “sounds like robot sex” that made me laugh.
A stepper motor may also work, not sure having never really used one before. Electromagnetics would take this piece away from being a very simple project which means I personally wouldn’t use them. They would give nice silent operation though.
As for making it “smart”, that was the original plan. The track wire is insulated so it would be pretty easy and there are plenty more inputs. The thing is it’s not needed at all. The ball just gets into the motion on it’s own. It may take a few seconds but it falls into place. I could use the potentiometer to vary speed but the Picaxe I am using doesn’t have the memory for all the code and I like the “random” motion. I could use a bigger Picaxe but its costs less than £2 which is cool.
As for the Arduino, I am aware of them but have never used them. I believe, living in the UK, Picaxe are quite a bit cheaper. I may be wrong so if anyone wishes to correct me, please feel free to do so.
We need an arduino in this thing
This is most certainly not perpetual. It does not last an indefinite amount of time as the length of time the batteries would last could be calculated probably down to the minute given the electrical draw and whatnot. And as such it would not continue without intermission since the power source would eventually wear out. Still pretty neat though.
Might be a gadget to keep people busy in waitingrooms.
I guess the perpetual is just a fancy way to describe the circle in this case, a perpetual track as it were since it has no start or finish, bit deceiving though but it’s not like it isn’t obvious.
ridefst: craigs idea would work.
Start with an inner and outer ring, with the ball on it. Make the inner ring move eccentrically in a circular motion (keeping inner ring in same horizontal plane as the outer ring, and not actually rotating the inner ring).
It is easy to ensure it is *impossible* for the ball to “fall through” if the maximum gap is smaller than the ball diameter.
The “gap” moves around/forward at rate x, and so long as the ball moves around/forward at rate x then all is good. Easiest to imaging ball placed half way between minimum gap and maximum gap.
Alternatives could be:
a) tilt the inner ring in a rotational manner i.e. connect center of inner ring to a ball joint, make front of inner ring go up/down in sine wave, and the left of the inner right go up and down in cosine wave.
b) rotate either the inner or outer ring (keeping the other ring static), and the ball will move at approx half the rate of the tangential velocity of the ring.
ridefst: I suggest you should always assume someone has a clue, think about why you could be wrong in your first assumptions, and try not to jump to conclusions so quickly.
one way to pulse the current would be to have sections of non-conductive or insulated wire and sections of conductive wire. the ball would conduct between wires when it is in the non insulated sections.
This is probably the best string of comments posted about such a simple project in a long time, I hope we see someone take on some of these ideas in a follow up post.
I’d like to see one using a stationary track and a single loop of wire that the track (and ball) pass through. The loop could be used as a sensor to detect when the ball passes by, and then as an electromagnet to propel the ball.
Nice project mate. The suggestions of the electromagnet 1 sound like a very nice idea, I would use the magnets to pull though (turn on 2 magnets to the left/ right of current ball postition). When I get some free time to get back into these projects I might make one myself :)
I’ve uploaded a few more videos. One of the new routines and the other two are the machine running quieter.
Servo needs a better mount
I think I’m in love…
maybe using the same kind of electromagnet system used in hard drives?
Pneumatic actuator
I agree with Paul. Some of the noise is coming from the entire track rocking back and forth on the table from the servo.
If you split the rail and apply a current you can use Lorrentz Force for an accelleration. You’d have problems with constant acceleration. So you’d need acceleration and static cycles or apply a PWM setup for a constant speed.
So happy to have you on board. Can’t wait to go to Bigstuf with you. Welcome abord