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

[TinselKoala]

With a lot of power you can levitate non-magnetic metals like copper or aluminum or brass with an AC electromagnet running at 50/60 Hz. The changing polarity of the field induces eddy currents in the metal piece and those eddys produce a magnetic field that opposes the field from the magnet (like polarities, switching very rapidly) so the metal plate rises up above the electromagnet pole pieces. John Hutchison did some of his fake videos like this.
It does take a bit of power though, and generates heat in both the electromagnet and the levitated metal. But no sensors or switching are needed, just the AC in the electromagnet. I used to demonstrate it with a "magnetic floor sweeper" we had at the lab, using a big variac for control and a chunk of aluminum plate for the levitated object. I don't have anything with the right pole pieces now, though.

[allcanadian]

I built a hall effect based levitator a few years ago and they are fairly simple however there are a few tricks. The first issue you will encounter is hysteresis or feedback causing oscillations, next high current draw due to coil saturation if the magnet is too close or high current draw if the magnet is too far away from the coils. The most efficient layout I found was using a coil above in attraction and a weaker permanent magnet below in repulsion. The most efficient layout does not actually levitate the magnet with a coil but uses a permanent magnet to produce the levitating force and a coil to stabilize the levitated magnet. I also used a neat trick I discovered to stabilize a rotating motor/magnet being levitated, a 1/2" aluminum plate below the magnet will dampen oscillations however it will not hinder rotation. Remember the magnetic field is stationary in space and bound to the source as such rotation on axis does not produce any noticeable eddy currents or dampening effects however any other movement on any axis will.

AC

[allcanadian]

A few pics.
Maglev photo is a simple hall effect magnet levitation device and I was using a laser bounce system to map field inconsistencies.


The second photo is a magnetic bearing system I developed based on my magnet levitation experiments. Been sitting under the bench for a few years now but I was thinking of digging it out for the pulse motor build off. It works very well and is rock solid on all axis and by my estimate is a 99% passive system. Note on the far right outside column there is a silver aluminum disk in the column attached to the shaft, this is used to stabilize the shaft on all axis but does not dampen the axis of rotation as explained earlier.


The record for this bearing with only the 1" steel shaft spinning is continuous rotation for about 6 hours from 1000 RPM. By comparison a good set of ball bearings oiled with synthetic oil and the seals removed will spin approximately a minute or so. Interesting stuff and it becomes very apparent that a field is the ultimate bearing surface because it has literally no friction.


AC

[Lakes]

I`d love to see any video you have of these devices.

[TinselKoala]

@AC: in the second photo, one end of the shaft is bearing against an end plate somehow, is that correct? That is, there is one point of actual contact between the levitated shaft and the support structure?

@Lakes: at the risk of boring you silly....

Optically-sensed, Arduino controlled:
http://www.youtube.com/watch?v=xkiGTWODERo

Electrostatically rotated, passive permanent magnet levitated:
http://www.youtube.com/watch?v=tVfw-TeJ9r4

This could also be accomplished with concentric ring magnets, as I presume AC is using in his system, and which Steorn used some years ago in their "Plinth" Orbo demonstrations at the Docklands. Still will need one point of contact with the frame, though, but will work vertically (as Steorn) or horizontally as AC's demo.
I have a set of such ring magnets but don't have the ability at the moment to construct a frame with sufficient accuracy and strength to support them.