Today I did some tests with a 250 mm long 5 mm diameter threaded brass axis and a magnet arrangement as in the drawing of my last post above.

(250 mm ~ 10 inch)

Initial trials (moving the base magnets around by hand) gave good results, the set up seems to be pretty stable and the heavier axis dampens earratic movements.

Once I have finished the build I will show it.

It seems to be pretty easy and not very critical to build some levitating axis with two base magnets on both ends. The base magnets could be any rather strong magnets (cuboid, disk, ring). The magnets on the axis on both ends have to be axialy magnetised ring magnets, as far as have seen so far.

All ring magnets looks neat, but I have this strong disk magnets (diameter 20 mm, thickness 5 mm, axialy magnetised) which I wanted to put to some use.

Greetings, Conrad

Nice work on those mag bearings.

What might be interesting is to have 2 small mags repelling to replace the point of physical contact.
Not that what you have is bad in any way.     Just another level of amazing.

Mags

By the looks of that last pic, just bring a mag close to the back side of the wall that the rod is touching that repels those mags on the axle. Just enough to make the axle not touch the wall any longer.

Could end up with a total floater. 

Mags

A "total floater" is not possible with permanent magnets (as TinselKoala wrote in a post some days ago).

Citation from Wikipedia: "Earnshaw's theorem proves that using only ferromagnetic or paramagnetic materials it is impossible to stably levitate against gravity; however, servomechanisms, the use of diamagnetic materials, superconduction, or systems involving eddy currents allow to achieve that."

http://en.wikipedia.org/wiki/Magnetic_levitation

http://en.wikipedia.org/wiki/Earnshaw%27s_theorem

I looked at many "magnet bearings" on YouTube (for instance skycollection shows a nice one http://www.youtube.com/watch?v=B56zvadt2vA), and all need that pointed tip of the axis against some glass. Skycollection used two magnets (behind the glass an both ends of the axis) to stabilise the axis against one side.

When using base magnets (two on each end of the axis) there is only a very narrow area where the ring magnets (on both ends of the axis) have to be. And the axis only presses very slightly against the glass, which is bad, because it might easily move away when it is disturbed.

Although I saw many "magnet bearings" on YouTube in the last years I had some personal misconceptions. I thought that I understood magnet bearings, but that was not true. Only by building two such contraptions myself in the last weeks I slowly get a real understanding.

But once the axis floats, it looks incredibly cool and people always ask: "Does it really float?" Well, the axis floats, but needs that glass wall to lean against.

The stability of the axis is a bit precarious and it is good practice to arrange some non touching "guides" around it, to keep it from slipping away when disturbed. Attached is a photo of my "little model" (3 mm axis, 140 mm long) with a "non touching guide" on the free end. This guide keeps the axis from falling off when it is disturbed. A disturbance always happens when the axis stops turning. I think this is because of the interference of the central magnet with the base magnets (but I might be wrong). The axis bangs against the guide plate when disturbed but does not fall off.

I still have to solve this interference of the central magnet (diametrically magnetised) with the base magnets. One could ignore it, but it has a breaking effect, which I do not like, because I want to spin the axis with minimal power.

The idea behind all this is to have a diametrically magnetised ring magnet (on a "floating" axis) spin with very little power. I hope to be able to do it with about 10 Milliwatt.

Once the "diametrically magnetised ring magnet" spins fast with very little power one could try to harvest some power from it (e.g. with a pancake coil) and feed it back in order to have even less power draw.

This probably leads to nothing, but it is a nice play thing.

Greetings, Conrad

@Conrad,
You could try some mu-metal sheet to shield the centre magnet from the support magnets.
Two sheets either side with a hole in them for the shaft to go through should provide the shielding you need.


Meggerman.

@Conradelectro,


                        Jonny Davro extended his axel to perhaps twice the length of yours to eliminate that magnetic drag.

Here's the video of Jonnydavro's unit: The vertical axis design looks great!


https://www.youtube.com/watch?v=eggfplU7Abw


Jonnydavro's vertical axis:


https://www.youtube.com/watch?v=hZbscMyotJg

@synchro1 & Pirate88179:

I really like the designs from jonnydavro, specially the vertical spinner is a great idea. Thank you for pointing me to jonnydavro's videos. Just what I needed.

Fortunatelly I have got some ball magnets.

Greetings, Conrad

At the first attempt I could achieve about 13 mW and then 10 mW power draw to spin the little horizontal model at about 3000 rpm.

Power draw:  4 Volt at 3.2 mA  and then 3.6 V at 3 mA

The LM311 circuit is not very good yet, just a quick try. (Do not replicate it.)  Also the position of the coils is not yet optimal. And the LM311 is a power hog.

At the moment I am using a drive coil with iron core and a DC resistance of about 280 Ohm. The next attempt will be to use the same type of coil for triggering and for driving, namely a coil without core and a DC resistance of about 1 K. Finally I will replace the LM311 with the MAX931 which needs a fraction of the power. I also want to try AC coupling of the trigger coil, which should dampen oscillations of the comparator.

I do all the testing with the LM311 because it is cheap (do not mind damaging a few). The more expensive MAX931 comes in when everything is a bit under control.

As it looks now, I can bring down the power demand well below 10 mW. But it will have to wait a few days.

Greetings, Conrad