Encyclopaedic Compendia    

Hullquist Central

    Ecclectic Interchange




My Replication of the Adams Motor

mpeg video (7 mb file recommend broadband connection)

I wanted to quickly test the claims of this device and chose the Bedini model, probably the simplest version around. I needed a rotating nonmagnetic plater using a low-friction ball-bearing. I found a garage door pulley wheel at Home Depot that looked suitable and then picked up the following hardware:

3" long 1/4" diameter hex-head bolt with a 2" unthreaded shank,
3 matching nuts
1 1/2" long 3/8" diameter hex-head bolt,
2"x3" aluminum right-angle bracket

some washers, and a
electric-drill sanding disk kit that had a 6" diameter rigid plastic disk--just the thing on which to mount my magnets.

Magnets? I bought a bag of 10 from Radio Shack, 3/4" x 1 7/8" x 3/8" domino-sized ceramics that are quite strong. I enlarged the central hole in the plastic sanding disk to clear the 3/8" bolt and pulley hub. I then glued the disk to the pulley with industrial strength contact cement. Once dry, I cemented the magnets with all north faces up at 90 degree intervals on the plastic disk as shown here:

As you can see, the pulley-disk-magnet assembly was bolted to a 3/4 x 4 1/5 x 9 1/2" pine board.
I simply drilled a 3/8" hole just short of all the way through 3" in from one end and self-tapped the bolt into the hole until tight.

I mounted a 4 1/2" long piece of 2x4 as shown so that it just cleared the rotor using 2 1/2" long wood screws from the bottom.

The aluminum angle was screwed to the upright 2/4. Now it's time to wind the coil.

The coil was wound on the 1/4" bolt by first puting a single thickness of black electrical tape around the unthreaded shank.
I cut out two stiff cardboard washers and put them on both ends of the coil area as shown here:

Electronic Parts
I bought 2 bags of "Assorted Enamel-coated Magnet Wire" from Radio Shack
Each bag contains a spool of 26 ga (75 ft) and 22 ga (40 ft) wire. (a 3rd spool of 30ga wire is not used in this project).
A strand of each was taped at the head end and wound together to the opposite end, then repeated until I ran out of 22 ga after only 200 turns. I spliced the scraped off end by soldering it to the beginning of the 2nd roll of 22 ga that came in the second bag. I ran out of the 26 ga just after 420 turns. Here's what it looked like after the first row:

I tried to bunch most of the windings near the bolt head where most of the coil would be in the strongest part of the magnetic field.
The coil was finished by covering the outside with a final layer of black electrical tape.

While at the Shack, pick up:
1 TIP3055 switching transistor (get the tab case style, not the can)
1 1N914 switching diode
1 1/2 watt 10 ohm resistor
1 1K ohm variable resistor
1 9v alkaline battery
1 9v battery connector (comes in a pack of 4)

The TIP3055 switching transister and heat sink (which I later realized was totally unnecessary) were mounted to the end of the base board with a small wood screw. The rest of the parts were then soldered together with 22ga vinyl insulated wire following this schematic:

As you can see, Bedini's design is a vertical ferris-wheel rotor arrangment mounted over a fixed vertical pole stator.
I've modified the specs to provide for a variable resistance which I found by trial and error to be about 50 ohms for my coil-magnet combination. If the resistance is too high, the transistor won't fire when the approaching magnets induce a current in the 26 ga winding. On the other hand if the resistance is too low, the transistor could be damaged by excessive currents.

I mounted the coil with two nuts added to the end of the bolt before inserting it through the hole drilled in the aluminum angle piece. This was adjusted to provide about 1-2 mm clearance (air gap) between the bolt head and the magnet faces as they rotate past it.

Next, scrape the ends of the four copper stator leads and tin with solder. Next take the two 22 ga (thicker, heavier wire) leads and determine what polarity is required to produce a repulsive field. Touch the leads to the 9V battery terminals until you see repulsion, then mark the lead connected to the positive battery terminal when the repulsive effect was produced. Solder the red positive battery terminal lead to this 22 ga wire. Solder the other 22 ga lead to the emitter lead-diode-black negative battery lead-ground connection.

The transistors outside terminals are B (base) and E (emitter) as shown here. The diode is connected between these outside terminals with its black band side connected to the B terminal.

The mass of fixed resistors here should be replaced with the 1K variable resister in series with a single 10 ohm fixed resistor.


The smaller 26 ga wire on the same side as the + 9V lead, goes to the Base terminal resistor(s).
The smaller 26 ga wire on the opposite side goes to the negative-ground-emitter-diode connection.
The final thicker 22 ga wire, also on the opposite side, goes to the transistor's middle collector terminal.

Starting the Motor
The first trial of the motor was using the original 680 ohm fixed resistor. A brisk spin of the rotor just spun a dozen or so times and came to a stop. I suspected that perhaps the resistance was too high, and not having a variable resistor handy, just added another 680 ohm fixed resistor in parallel to cut the effective resistance in half. There seemed to be just a slight improvement in how many spins the rotor made before finally coming to a stop. This encouraged me to keep adding parallel resistors as the spinning become more and more prolonged with each addition. After the resistance had dropped to about 60 ohms it was definitely spinning much longer than it had to begin with. After I got it down to about 50 ohms, and a squirt of WD-40 on the bearings, the next spin just kept going and going!

I quickly measured the voltage drop across the battery under load. It was just at 7v at 2:43 pm 9-28-2004. Within a few minutes the battery case was noticably warm to the touch. But the coil, drive wires, and transistor remained at room temperature of even slightly cool. The battery voltage steadily dropped over the following 70 minutes that the motor ran before finally shutting down with a final voltage of 5.5v.

Possible Improvements
Although other accounts of 13 year old science fair projects built on this design were said to run uninterrupted for days, I was pleased at least to get the motor running. My opinion is that there is a prohibative amount of drag in the heavy-duty garage door bearings that I used. A more precision axel-bearing-plater would certainly produce superior performance.

I have doubts, however, with the current circuit that there would be any over-unity achieved in that back-EMF doesn't appear to be collected, at least by the battery. Tim Harwood and John Jankowski's excellent treatment of the Adams Motor imply that the back-EMF is used by the stator itself to enhance its repulsive field. In that case, my coil design might need to be optimized: better core (Bedini suggests using multiple lengths of welding rod), more turnss (some designs specify 450-800 turns). Bifilar windings are mentioned often, but simply winding both wires together doesn't really qualify as bifilar since they constitute two separate windings.