Nice idea, but I want to see how he's doing it. What's he using to keep his phone in the right position without flipping over or rotating?

@newtboy ah finally see what you mean. And yes you are right a Coke can would be stable and it could rotate. It is no longer considered a spinning top I think, so that is why the contestants didn't make it that way but for sure it would work.

If the can rotates I think the torque (force due to rotation) is in the same direction as gravity. (Where in the normal spinning top case gravity pulls the cg off center and torque back on.) In the can case both would move the cg back to equilibrium, Ie on center. there would be no precession at all. Every time some small Bump would make the cg move of the center axis it would be pulled back instantly.

I think it would work, and that it would take away the challenge ;-)

I think you still misunderstand. I don't mean it would hang from above, it would balance on it's contact point. In that way, it would 'hang', but the point of contact would be pointing down in contact with the top of the 2cm cylinder and the bulk of the mass in a ring hanging below that point. To exaggerate for clarity, think of a soda can with the top cut off, turned upside down, and balanced on a pin touching the exact center of the inner can 'bottom'...then spin it.
My idea is a top EXACTLY as they made it, except the weighted ring is much lower, so the CG is below the point. Then, when set on the pedestal, it would be stable when stationary (when set on a counter, the point would not touch). I can't see why that would change when spun as long as the CG stays below the point and balanced/centered.
It would balance when stationary, no question. My only question is what might happen when you spin it, would the rotation make the CG 'want' to be above the contact point for some reason, or would nothing happen. I don't have a lathe to make one myself to try, so I thought someone well versed in rotational physics might know.

Good question. I think that the entire device is unstable no matter what, its impossible to keep it straight no matter where the point of gravity as long as it needs to balance on a single tip. So zero speed would mean tip over in all cases unless you make a more stable tip (square) which would mean it cant spin very well which means you havent made a spinning top.

You can find more about the physics of the spinning top here:
http://hyperphysics.phy-astr.gsu.edu/hbase/top.html
and here:
http://physics.stackexchange.com/questions/271/why-dont-spinning-tops-fall-over

So what is the optimal distance of the center of gravity to the tip?

There are several things working at the same time. Most importantly is the energy. You need to store as much energy as possible so that the top can spin for as long as possible. When the top slows spinning the friction at the tip becomes larger (the precession becomes bigger) so it starts to lose more energy and slows the spinning even more. You store energy by adding weight with a center of mass that is further away from the tip. When the top then "falls" the center of gravity moves down and reduces potential energy. Due to energy conservation kinetic energy goes up meaning speed of precession or of spinning goes up and creates a force pushing the top back up.

Off course, more mass means more friction at the tip, so there is for sure an optimal here, most likely depending on mass, size and shape of spinning top, etc.

Last but not least, more rotation speed I assume also means more friction, so its a trade-off.

If you move the center of mass down below the tip, well, if you move it as far off as you would above, the energy you can save is about the same, but the entire thing would be harder to build and you would need to make sure the sides fit around the ground plateau. Also, when the precessions become bigger the sides will hit the plateau, meaning game over.

In the end you are better of keeping the center of gravity above the tip point.

That is fantastic. I would start the first leap, rotate forward, and face plant on the 3rd step.

I don't want to ban guns, but I would like to see the following:

1. License gun owners like Germany licenses drivers. Six months of classes, followed by multiple exams.

2. Mandatory psychological examination for any person wanting to purchase a gun.

3. Require that a minimum of $1,000,000 death or dismemberment insurance be carried at all times.

4. Automatic 25-year sentence for ANY crime involving a gun.

Limit guns to:

- Shotguns: Single-barrel breach-load.

- Pistols: 6-shot rotating cylinder; hammer-cocked firing mechanism. No semi-automatic pistols.

- Rifles: Single-shot, bolt-action.

Ahhh, but that's part of the beauty of the thing. The grinder gears, as they're called, are just one outer ring of n gears and one inner ring of n-1 gears. The pressure of the planetary gear causes rotation of the grinder gear by meshing one notch over each full rotation. It's an incredible increasing of the gear ratio. I believe in the video he says 1 : 17,000.

Looked to me like he rotated away from home. His step was away from, not towards, home..

Might look weird because he's left handed.

It's a real phenomenon for vortices measuring hundreds of miles in diameter, yes. At that size, the coriolis effect (the difference in rotational speed of the northern and southern extremes) is a huge factor.

For a vortex measuring a few inches in diameter, not so much. It takes rather extreme measures to make the coriolis effect the largest factor.

That steak had crosshatch grill marks even though he never rotated it. Fake, obviously.

doesn't look that dangerous, really. the (what i assume is) stepper bit wouldn't have to rotate very fast to be effective. it would just have to have, and clearly does have, high torque.

Hits a rock on the bottom, spins up, takes off like a wheel from a badly maintained semi...

I doubt an explosive would impart that kind of rotation to just one rock. It could happen, but I doubt it.

That's the title from the youtube video.

I didn't actually count the rotations myself, but looking into it, it seems like their counting the vertical rotation (4 flips = 1440), which is kinda bullshit.

Still an epic trick though

It occurs to me you could probably manage it without too much in the way of Segway-esque magic. Have the outer shell as you'd expect, basically a big hamster ball, then a second ball inside it with all the clever stuff.

For the inside inside have most of the lower half be ballast of some sort - battery packs would be ideal for that - then 8 or so wheels pushing against the outer shell so you can move in a decent number of directions (might have to retract the wheels perpendicular to the direction of movement unless there's a more elegant solution I'm missing).

Then finally for the upper half make a very smooth dome, put a little cart on it whose position you can adjust with a couple of cables, stick a couple of strong magnets on the cart (need a motor on the cart too so you can rotate the head).

This is all based on my years of not having done anything remotely connected to model building!