Perpetually Rotating Bar Magnet Arrangement

Lasted edited by Andrew Munsey, updated on June 14, 2016 at 9:25 pm.

  • One error has been found on this page. Administrator will correct this soon.
  • This page has been imported from the old peswiki website. This message will be removed once updated.

The following is a theory for an all-magnet motor, presented by Chris Wilshaw.

He proposes using magnets in a pattern where symmetry is slightly disjointed. The theory is that magnets naturally find a place to rest somewhere where the system is in balance. He proposes the following system/arrangement as on that deliberately has no balance point.

News Just In!!!

You need to use 3's 9's and so on and form a Torus shape. The magnets around the moving ring must not be in line with one another but on a left / right wave as you watch laterally. Imagine your orientation to bicycle wheels as they rotate under you. Now arrange magnets on that wheel to give a left/right wobble. The static magnets pushing will be no different to diagrams show previously but the effect will be incredible.

Again, I can not build this, I have no cash or facility to do it, but have a strong model in my mind and will draw it as soon as I find a way.

Main Ideal

If a system has to be started, it will surely stop. Only a thing with no beginning can have no end. This is to seek an arrangement where, on assembly, rotation begins no pushing by hand (unless tiny).

Using magnets in a pattern where symmetry is slightly disjointed. (1)

Magnets naturally find a place to rest somewhere where the system is in balance.

This system/arrangement deliberately has no balance point..

(1) New animations

Now show different setups - The animations say much more than I can say in words.

Much of this article is now talking of previous setups as the idea has evolved beyond them yet they may still offer something. I will bring the article into line soon, the page of animations is the best explanation. My current favorite being 32 angled magnets on the outer frame and 9 angled magnets on the moving ring.

The rest of this page needs updating

Thought Experimentation

I have attempted to guess the rotation force for the first 45degrees of rotation (assuming that each subsequent 45degrees would be similar). I imagined each magnet on the ring at a chosen point in time and which way it would choose to travel from rest. The system should start without assistance as at every degree I checked there was a positive clockwise intention to rotate.

There is error in the guesswork obviously as some positives results show. In most cases of high positives it is usual that I saw to the positive side of a roughly balanced situation. Yet this also accounts for low positives where I placed a negative or balance guess to a close call. - I will keep looking at the diagram to better guess but would prefer to do this in a prototype. Numerous layouts of this system can be tried using the basic principles mentioned within this article.

We all know from school that 2 similar poles repel therefore it seemed to me that any arrangement of magnets should being magnets into a position in which they could push each other from a position where they were not interfering so much. The same with attraction.

Breaking symmetry as much as possible. Using numbers that don't divide well. Four outer magnets and 4 inner would come to rest in 90degrees guaranteed. A 2 and 2 system would be similar to an magnetic compass and would be worse still. I have seen that even 360/360 system would suffer the same failure of balance. Yet a 100 inner and 99 outer certainly wakes the thinking matter as to where it would find to balance - but only for a second. I have chosen values that divide easily into 360degrees to aid drawing but still do not divide well into one another so meet my desire.

Positive placement of magnets. Arrangements that avoid the need for north to come near south unless the result is clockwise rotation. The same for south/south and north/north. All layouts attempt to give a clockwise direction of rotation with any negative (counterclockwise) rotation not equaling half the total force at work. Many points in the layouts would very much not want to turn anticlockwise even with assistance.

I had numerous ideas (and still do) where parts shuffle about to give positives and move out of the way to avoid negatives. Ellipses, soft rings which sag or swell via gravity to assist placement of desired magnets and so on. - I have so far discounted them all due to friction and manufacture.

Positive Rotation

In frame 1 there is a lesser inclination to move backwards as there is an inclination to move forwards in frame 3. That is the positive (clockwise) intention in 3 is greater than the negative (counterclockwise) intention of 1. This is one of the systems which will aid positive rotation that a side on approach is easier to overcome than a end on. So in these 3 frames, although there is inclination to rotate backwards it is less than the overall inclination to rotate forwards. The exact angles required would be down to experimentation and unique to the magnets used. The offsetting (so called disjointed symmetry) would be to arrange magnets around the ring so that this positivity is as dominate as possible and where one magnet is in a serious case of wanting to reverse, there are several others inclined to rotate forwards. Other magnets in the current theory and diagram show a pulling version of this pushing idea.

There was an error working with the wiki: Code[1]

How it moves

Using magnets fixed to a non-moving frame use both their repulsion and attraction to rotate a ring.

The ring is pushed/pulled due to it also having magnets attached to it. There is always a positive push/pull in the system which is best viewed on the animation which is linked to after this image.

Animated Version of this Diagram

The Diagram is just a sketch. The working prototype may require an additional magnet or two be added into the stationary arrangement or the moving ring of magnets to upset symmetry even more or smooth the rotation.

The aim being that say 9 doesn't divide into 8 or 16 so the ring always being out of equilibrium to avoid the ring finding a comfort point where it can rest. Thus constantly rotating. (8 inner, 9 on the ring and 16 around the outside)


It is clear to see in the 8/9/16 version shown in the animation that the theory works with those numbers. On a scaled up real world version I imagined arrangements of 73 inner pairs, 90 outer and 47 on the ring. The numbers came to me as I first envisioned the idea in my head before I got chance to draw it and only as I was picturing non dividing numbers to prevent balance these numbers could be well off or close but the point is that there bust be a deliberate attempt to avoid any balance point and the animation already shows the possibility with 8/9/16.

2 or more systems could be run along side one another, out of sync, so that a lag in one system would be aided by a corresponding surge in another.

Getting power out of the system

To obtain power in a frictionless manner the rotating section would pass through wire loops thus pulling charge along the coils in the loops. Loops would be added up until the point the actual resistance in the wires was close to preventing the rotation. The only friction thereby being electrical resistance. In fact the rotating system if it rotated fast enough would not even need an axle to rotate around, an axle may only need to be in place for the initial startup procedure.


The diagram, although it has inaccuracies, as it is a sketch, allows us to see that even when one magnet is in a place where it would have to overcome a South-South push or a North-South pull there would always be a number of other magnets benefiting from a South-South or North-North push or North-South attraction.

Official Site

Animated Version

Inventor's Website


Chris Wilshaw


email: []


See Talk:Perpetually Rotating Bar Magnet Arrangement

See also


Directory:Magnet Motors

News:Magnet Motors

Directory:All-Magnet Motors:Theory

PowerPedia:Magnetic motor - encyclopedic entry

Directory:Linear Magnet Accelerators




LEADING (including legendary status)

Directory:Muammer Yildiz Magnet Motor

Directory:Michael MacInnes' Magnet Motor

Directory:Torian Magnet Motor

Directory:Magnetic Motors:Howard Johnson

OS:Magnetic Motors:HJ

PowerPedia:Howard Johnson

Directory:Steorn Free Energy

Directory:Shinyeon Energy Research Center


Directory:Disturbed Self-Compensating Symmetry (Terawatt Research)

Directory: Mechanical Opener for V-Gate Magnet Motor

There was an error working with the wiki: Code[1]

- Directory