Electronic Levitron...How the heck does this thing work?

[Nink]

They have to be using hall sensors or reed switches since they put it in a pillow etc   so I agree Pirate88179 your probably looking at a circle of magnets with the ring magnet and hall hall sensor in in the corners with coil magnets making small adjustments to the voltage to ensure ring magnet is always locked.  So maybe we need a broken microwave to salvage a large ring Magnet from and some computer cooling fans for hall switches.

EDIT:  I THINK I WAS WRONG ABOUT THIS. I posted later that there looks to be a hole in the pillow and the table also had a hole in it for an optic sensor.

[TinselKoala]

Yep. A single ratiometric Hall sensor feeding an op-amp comparator with adjustable setpoint, the comparator then drives the pushpull coil driver stage. The ratiometric Hall sensor will put out a voltage that is proportional to the sensed magnetic field strength. So you adjust the setpoint voltage so that the comparator flips at the field strength corresponding to the right levitation distance. Closer (stronger) means the push-pull output stage pushes, farther (weaker field) means that the push-pull stage pulls. (this is assuming the drive coil is beneath the levitated magnet disc). So in this manner the levitated object will be "locked" into a narrow zone, the height of which is adjustable by the setpoint of the comparator flipping state. It still must be stabilized sideways though.
For levitation coil above the object, you only need to "pull". In fact if you aren't concerned about sideways levitation, even the coil-beneath only needs to push, but for that arrangement there must be some other means, like side magnets or coils, to keep the object centered above the coil instead of slipping off like you see in the "magnet hopper" video. Of course the levitated magnet must also be prevented from flipping over in both cases; this is easily enough done by the geometry (distribution of mass wrt the distribution of lift) of the levitated object.
The only Hall sensor I have available at the moment is the one I pulled from the computer fan and it's a latching switch type, not ratiometric. Allegro Microsystems makes the ideal sensors for this purpose, they are cheap and robust and I've used them in several projects before, but I can't remember where they are at the moment. Oh... I just remembered, at least one is in the Orbette pulse motor, but I can't get to that one right now.
My Arduino-based levitator (coil on top) uses optical sensing, a really neat system that actually shuts off the illuminating LED and samples the ambient light every 100 pulses and adjusts for it automagically.
I'm not sure how the Hall sensors will work in the presence of the strong PMs and also the variable field from the coil, though. Perhaps the Hall sensor reads when the coil is momentarily off. Things can happen very fast, at multi-kHz frequencies, so you can turn coils off for a few cycles to take a Hall reading and the object will hardly move in that time. This kind of thing would require microprocessor control though, I doubt if you could do that "coil off then sample field then coil on at right strength and polarity" with op-amps alone, but maybe.

[Pirate88179]

They have to be using hall sensors or reed switches since they put it in a pillow etc   so I agree Pirate88179 your probably looking at a circle of magnets with the ring magnet and hall hall sensor in in the corners with coil magnets making small adjustments to the voltage to ensure ring magnet is always locked.  So maybe we need a broken microwave to salvage a large ring Magnet from and some computer cooling fans for hall switches.

Yes, I agree with you.  Did you see it that video above that it can (according to the specs.) hover 9 pounds?  Holy crap that is a lot of weight, and their hover height is very decent.  I would be thrilled to hover 1 pound in a stable manner.

Bill

[TinselKoala]

That has to mean that most of the levitation is done by the PMs and the coils modulate the field of the PMs. I think.  That's the only way a small current (low heating) can possibly lift that much mass. I think.

[Pirate88179]

TK:

Did you see that video?  I am still dumbfounded!

For the price these folks are probably charging (and good for them as this is the best I have yet seen) it would not surprise me if there were an Arduino or Rasperry PI style micro processor in there somewhere.  In quantities, I am sure they can be had very economically.  Looking at the version they had hidden in a table, only the center part of the base (looked to be about 1.5" dia.) was showing through the wood.  If that is true then the feedback sensors would all have to be in that small area right?

In the next day or so, I will see if I can find the patent (s) for this type of device.  They probably have a patent in Europe as it looks like that is where they are from.

Bill