Tour : click picture below for previous

Back to Mike's Electric Stuff

Click picture below for next

smcd80.jpg (1775 bytes)pcb.gif (4162 bytes)The Ball-Bearing electric motor

Further info : Great balls of fire!   'The intriguing ball-bearing motor'

bbrun.jpg (18010 bytes)This interesting but unfortunately not very useful device produces motion from electricity without magnetism being involved. It operates purely by thermal means, so it works on AC or DC, and the motor can rotate in either direction, determined by the initial spin which is usually required to get it going.

It simply consists of two ball-bearing races on a common conductive shaft, with the outer ring of each race being connected to a high current, low voltage power supply. An alternative construction is to fit the ballraces inside a metal tube, and mount them on a shaft with a non-conductive section (e.g. two sleeves on an insulating rod). This method has the advantage that the tube will act as a flywheel.

This picture shows the motor running. There is a rectangular white label on the right-hand flywheel, being blurred by the motion.

How does it work ?

When current passes from the outer ring of the ballrace to the inner ring via each ball, heat is generated at the point of contact due to the increased resistance. This localised heating causes the ball to expand in the hot area, causing a slight elongation of the ball, pushing against the inner and outer rings of the race. If the ball were stationary, this would cause the bearing to stiffen and sieze up, but when it's rotating (from the initial spin), this elongation causes the ball to push itself further round in the direction of rotation, sustaining the movement. This action happens as a continuous process on all the balls which are in electrical contact with the inner and outer rings.

What use is it ?

Nil, zilch, zero, none whatsoever, - it's totally impractical for any real-world application. Unless you know different....

How can I make one ?


<sermon>The high currents involved mean that the motor and wires can get VERY HOT (glowing!), so ensure there are no flammable materials nearby, and avoid touching any part of the set-up until it has cooled. The plastic on insulated wire can generate very unpleasant fumes when it melts, and in this application, it probably WILL melt. Lead-acid batteries vent hydrogen when heavily discharged, which could present an explosion hazard (you WILL get sparks when running the motor)  - ensure there is sufficient ventilation. The motor can achieve substantial speeds, up to a few thousand RPM, so due care should be taken to protect against this mechanical hazard. </sermon>

The basic materials required are two small ballraces (1/2 to 1" dia), and a shaft that is a close fit inside them - this combination can often be salvaged from various scrap mechanical or electromechanical equipment - printers, copiers, head actuators from larger hard disk drives etc. They should turn freely, i.e. not be caked in grease etc. It's important that the shaft is a very good fit inside the ballrace to ensure a good electrical contact (if not, it may be possible to jam the shaft into the ballrace using copper or aluminium foil).

Some sort of flywheel is usually required to give the shaft enough momentum. If you're lucky your shaft may have a gearwheel on one end, or a thread to which a suitable wheel may be fixed. As a rough guide, the shaft should spin for at least three turns when given a small spin by hand. The motor pictured uses two large aluminium control knobs.

The two ballraces then need to be firmly mounted, so the shaft rotates freely. I used a vice to hold the bearings, with pieces of fibreglass copper-clad board to insulate the ballraces from the vice jaws and provide a convenient means of making the connections. Fibreglass is also quite heat resistant, an important consideration in this application! Slits filed in the copper isolate the ballraces from each other, and connections are made by soldering to the copper. If your vice jaws aren't quite parallel enough, put some thin card between the vice jaws and the copper-clad board to provide some slight springiness.  Take care not to distort the ballraces if a clamping mounting method like this is used.

A very high current AC or DC power supply is required, at least several tens of amps. A large (>100VA) low-voltage (3-12V) mains transformer is suitable, for example a powerful automotive battery charger transformer, or 12V low-voltage lighting transformer. Another possibility is to use a  higher voltage toroidal transformer, such as those used for audio amplifiers, and wind a secondary of 10-20 turns of thick (>4 cross-section) wire through the core to make a low voltage very high current winding. Ambitious constructors might like investigate the possibilities of spot-welders and automotive sized bearings.

Another source of a suitably high current is a lead-acid battery, either a 'gel-cell' / 'dryfit', or a conventional wet-cell battery, although larger batteries like car batteries probably put out a bit too much current - it may be necessary to limit the current somehow - a foot or two of fairly thick (2mm) bare steel wire (wire hanger) might do the trick - a few headlamp bulbs wired in parallel might also work.

Remember that whatever power source you use, it is likely to be heavily (possibly fatally) overloaded by the virtual short-circuit presented by the motor, so you should only run for a few seconds at a time. Very heavy wiring is recommended (> 4mm cross-section), unlike the feeble wire shown in the pictures, which starts melting after a few seconds!

How do I run it?

To run the motor, first ensure the shaft rotates freely and smoothly. Arrange the electrical connections so you can connect and (especially) disconnect the supply very quickly. Holding one of the wires onto the transformer/battery terminal satisfies the quick-disconnect criterion, but be VERY careful not to burn yourself - holding the connection in pliers is a good move.

Give the motor a hand-spin, then connect the supply while the shaft is still turning. You should see (and hear) the shaft accelerate as soon as power is applied, possibly accompanied by a few sparks, and almost certainly a smell of 'hot metal'. Don't run the motor for more than a few seconds at a time, and if it doesn't start immediately, remove the power quickly to prevent siezing. The motor will reach its maximum speed fairly quickly (depending on the flywheel size), and then start to slow as the bearings heat up and start to sieze..

bb3.jpg (19488 bytes)Overall view : two ballraces clamped in a vice between two pieces of fibreglass copper-clad board, with the copper slits to isolate the ballrace outer rings from each other. Shaft with flywheels passing through bearings. As the shaft was not a tight fit, copper foil shims were used to improve the connection between the shaft and the bearing inner rings.

bbclose.jpg (13919 bytes)Close-up view. The two pieces of copper-clad board are connected together to improve the connection to the ballrace outer rings.

The copper-clad boards were stuck to the vice jaws with double-sided tape to keep them in place when clamping the ballraces.

The following article is reproduced with permission from the April 1989 issue of Electronics and Wireless World magazine, and is copyright 1989 Reed Business Publishing.

Great Balls of Fire!

The contents of Dr Stefan Marinov's travel-weary holdall did little to dispel the scepticism which greeted the man and his theories during a visit to our editorial offices. We politely listened to a rambling discourse on ball-bearing electric motors which rotated without magnetism and provided work in defiance of energy conservation theory. Dr Marinov unburdened himself as a man proselytising a deeply held yet widely ridiculed conviction.

The two ball races, one set into each end of a tube, didn't look to be the starting point from which new theories are forged. Neither did the thin PVC-covered wire connecting up to the blocks at each end of the tube supporting the ball race inner sections.
    "Stefan, how much current do you need to make the races turn... Would 5A he enough?"
    "No, you need a lot more than that"
    "No, Much more."
    "How much more?"
    "Have you got a car battery?"
    "Only the one fitted in my car."
    "Get it I show you...''
    "Stefan, if you connect up a car battery to your machine using those wires, the ball races will present an almost perfect short circuit and the wire wilI vaporise in a puff of acrid smoke..."
    "They get warm, sure, But I show you. where is your car?"
    "In the multi-storey..."
But Dr Marinov never heard the rest of my protest. He was already down the corridor and halfway out of the building. I headed him off at the revolving doors in the lobby.
     "Stefan, the multi-story is no place to advance science, let's see if we can borrow a battery from the motor transport department."
We set off across the road, Marinov clutching his holdall. I went upstairs to get permission from the garage manager. When I returned, Dr. Marinov was nowhere to be seen. I went into the garage to enquire of the duty mechanics the whereabouts of my Bulgarian friend.
    "Have you seen a foreigner with a battery fixation?"
The huddle of mechanics pointed to a figure crouched over a stack of batteries in a corner of the garage. The figure looked up without surprise.
    "I think this shouId work. Put your thumb on the hearing tube and, when I connect up to the battery, give it a flick."
I looked doubtfully at the battery, the machine and the wires in turn.
    "l'm telling you, Stefan, those wires wilI simply melt"
He didn't answer, He forced the bare ends of the wires hard against the battery terminals. There was a shower of sparks and an eruption of smoke from the bIistering cable ends. I gave the tube a flick. It took up a life of its own which all but had the skin off my thumb. The tube connecting the bearing outers spun up to what must have been at least 1500 rev/min before the connecting wire, unequal to the enormous current, disintegrated.
    "You see it turn?"
I looked at the burgeoning friction burn on my thumb.
I also looked at the acrid blue haze of PVC smoke which was rolling across the floor towards the group of curious mechanics.
    "Your bearing motor certainly works but I shall need a bit more convincing about it being a net producer of energy."

Dr. Marinov simply gave me a look which suggested that all his efforts had been in vain.

Frank Ogden

The following article is reproduced with permission from the April 1989 issue of Electronics and Wireless World magazine, and is copyright 1989 Reed Business Publishing.

The Intriguing ball-bearing motor

Whether or not one can accept the author's contention that it delivers energy produced from nothing, the novel electric motor he describes certainly deserves to be better known.


It is almost unknown that if direct or alternating current passes through the ball-bearings of an axle, it is set in rotation. In the few papers where this effect is discussed, the torque is explained as an electromagnetic effect. Yet the torque is due to thermal extension of the balls in their bearings at the points of contact with the bearing races.

bbmot1.gif (3810 bytes)bbphoto.jpg (5887 bytes)The arrangement of the simplest ballbearing motor is given in Fig.1, where the inner races rotate. With the same ball- bearings, a bigger torque can be obtained by rotating the outer races. In such a case the axle must be made of two electrically insulated parts, and the current goes through a metal cylinder connecting the outer races of both ball-bearings. Such are the small and big ball-bearings motors presented in Fig.2.

I have established that the ball-bearing motor is not an electromagnetic motor but a thermal engine. Here the expanding substance leading to mechanical motion is steel, while the expanding substance in all thermal engines used by humanity is gaseous. There is, however, another much more important difference; the motion in the conventional thermal engine is along the direction of expansion f the heated substance, while in the ball-bearing thermal engine it is at right angles to the direction of expansion of the heated substance. Consequently, in gaseous thermal engines, the gas cools during the expansion and the kinetic energy acquired by the "piston" is equal to the heat lost by the expanding gas. This is not the case in the ball-bearing motor. Here not the whole ball becomes hot but only that small part of it which touches the race, at a "point contact" where the ohmic resistance is much higher than the resistance across the ball. Only this small "contact part" of the ball dilates; and the dilatation is very small, only a few microns. (Of course, I have not measured the dilatation, I only presume that it is a couple of microns.) Since the balls and the races are made of very hard steel, a slightly ellipsoidal ball produces a huge torque when one of the races rotates with respect to the other.

Usually a push is needed to start the ball-bearing motor. However, on occasions it does start spontaneously (with a greater probability at greater bores) because the surface of the races is not absolutely smooth. With absolute smoothness and geometrical perfection, spontaneous starting is impossible.

During rotation the ball's "bulge" moves from the one race to the other, the local overheating is absorbed by the ball and the radius of the "bulge" becomes equal to the radius of the whole ball. At the new point of contact, when current passes and ohmic heat is produced, the radius of the contact point becomes again bigger than the radius of the whole ball and again a driving torque appears. Thus, as a result of the mechanical motion, the ball is not cooled; and consequently, in the ball-bearing thermal engine, heat is not transformed into kinetic energy. The whole heat which the current delivers remains in the metal substance of the machine and increases its temperature, If the ohmic resistance between balls and races is the same both at rest and in rotation, the heat produced and stored in the metal of the machine will be the same at rest and rotation. This resistance, however, increases in rotation; but with further increase of the velocity the increase of resistance is very slight.

I established that the hall-hearing motor produces the same amount of heat at rest and rotation in the following manner. I measured for a definite time the temperature increase in a calorimeter in which the motor was maintained at rest, applying a tension U and registering the current I. Thus the resistance of the whole motor was R = U/I. Then I started the motor and applied a tension U' such that at the new resistance R' the current I' = U'/R' was such that UI = U'I'; i.e., in both cases I applied exactly the same electric power. According to the energy conservation law, in both cases the temperature increase of the calorimeter had to be the same, as in both cases the same amount of electric energy was put in the machine.

I recorded, however, that in the second case the temperature increase of the calorimeter was higher. Thus I concluded that in both cases the ohmic produced heat was the same; however in the second case there was also heat coming from the friction of the rotating ball-bearings. The temperature increase in the second case was about 8% while the mechanical energy produced was about 10% of the input electrical energy.

One can see immediately that the baI1- bearing motor has no back tension because there are no magnets, and the magnetic field of the current in the "stator" cannot induce electric tension in the metal of the "rotor".

Thus the firm conclusion is to be drawn that the mechanical energy delivered by the ball-bearing motor is produced from nothing, in a drastic contradiction to the energy conservation law.

With a direct current supply, the ball- bearing motor can rotate either left or right. Thus it cannot be an electromagnetic motor, since a DC electromagnetic motor rotates only in one direction, with a given direction of the current. The ball-bearing motor rotates with DC as well as with AC. With a greater current it rotates faster. It is in teresting to note that the resistance of the ball-bearing motor depends on the current, and for higher current it is lower. If the current doubles, say, the applied tension increases only, say, 1.3 times. Here I wish to avoid any confusion between the increase of resistance because of the increase of the rate of rotation, and the decrease of resistance because of the increase of current; although, obviously, a higher current leads to a higher rate of rotation. The torque disappears if the ball-bearings are replaced by box-hearings. At equal applied electrical powers and equal number and size of the balls (i.e., at equal resistance), the torque is bigger for a ball-bearing with bigger bore. A ball-bearing with two times bigger bore has two times bigger torque. Fig,2 shows two ball-bearing motors with a small and a large bore which have almost equal ohmic resistances (of course, the mechanical friction of the bigger motor is greater). By touching both motors, one can immediately feel the difference in their torques. The bigger ball-bearing has greater number of balls and consequently a bigger torque; however, its current (and power) consumption are higher.


Methods of improving efficiency in the ballbearing motor include the following:

  • The use of balls which are harder and where a smaller amount of heat leads to larger thermal extension. We know that normally a harder solid body has a lower coefficient of thermal dilatation, so that one has to find the optimal solution which nature offers.
  • Tighter ball-bearings have a better pushing force. However, at the same time they will have more friction. A compromise is needed. But even if friction is very low, there is always a maximum velocity which the motor cannot surpass. At this maximum velocity, heat from the "bulge" cannot be absorbed by the ball, and the ball retains more or less a spherical shape. It is obvious that the maximum velocity is higher for larger balls. 
  • The driving force is higher for bigger bores, as the curvature of the races is less.
  • The driving force is greater for bigger balls, as their curvature is less.


1. Milroy, R.A. Discussion, J. Appl. Mechanics, vol. 34,1967,p.525. 2.
2. Gruenberg, H. The ball bearing as a motor. Am. J. Phys., vol. 46, 1978, p.1213.
3. Weenink, M.P.H. The electromagnetic torque on axially symmetric rotating metal cylinders and spheres.Appl. Sc. Research, vol. 37, 1981, p.171.
4. van Doom, M.J.M. The electrostatic torque on a rotating conducting sphere. Appl. Sc. Research, vol. 40, 1983, p.327. 5. Mills, A.A. The ball-bearing electric motor. Phys. Educ., vol. 15, 1980, p.102.
6. Marinov, S. The perpetuum mobile is discovered. Nature, vol, 317, 26 Sept. 1985, p.xii.
7. Marinov, S. The Thorny Way of Truth, Part II. East-West, Graz, ist edition 1984, 3rd ed. 1986.

At the time of writing, Dr Marinov was at the Institue for Fundamenral Physical Problems, Mouellenfeld- gasse 16, A-8010 Graz, Austria.

More info on Stefan Marinov : Institute for New Energy    Harold Asdpen's Energy science site

smcd80.jpg (1775 bytes)pcb.gif (4162 bytes)


Click picture above for Previous

Back to Mike's Electric Stuff

Click picture above for next