Levitating Solar Motor




About: We are a supplier of neodymium, rare earth magnets. We also love to conduct experiments with our magnets and build unique projects with them! We have several engineers on staff who are always thinking of new...

In this tutorial, we will create a magnetically levitated, solar powered electric motor...woah! That is a lot of cool things in one sentence. This type of motor is called a Mendocino Motor, named after the city in which it was invented, Mendocino, California.

The motor consists of a spinning shaft that is held up by repelling magnets, stabilized by resting a point against a wall. It is powered by solar panels mounted on the spinning shaft, which generate currents through coils of insulated wire.

First, let’s acknowledge that this motor isn’t very powerful. It isn’t useful for getting much work done. You’re not going to power a car with it. But it is a fun science project and a cool conversation piece. We like it because it is a great demonstration of the principles involved in most electric motors.

Teacher Notes

Teachers! Did you use this instructable in your classroom?
Add a Teacher Note to share how you incorporated it into your lesson.

Step 1: Parts List

Here is a brief list of what we used to build our Mendocino Motor. For the rotor (the spinning portion of the motor):

  • 1/2" diameter wood dowel rod, purchased from a local hardware store
  • Thin wood from a craft store, able to be cut with a hobby knife
  • Hot glue to hold the pieces together (do not use hot glue on neodymium magnets!)
  • 30 gauge insulated magnet wire, MW30-4 or MW30-8
  • Four solar cells from Futurlec, SZGD5433
  • Two RX088 ring magnets

For the base:

  • Wood for the base and wall
  • Thin piece of aluminum for the wall
  • Twelve RX033CS-N magnets. Alternatively, RX038DCB-N52 magnets might also have worked well. Still, we like planning on stacking multiple thinner magnets. This lets you adjust the strength by altering the number of magnets used in each stack.

Step 2: Step 1: Put Magnets on the Shaft

Though it isn’t the most perfectly straight or balanced motor shaft, we used a ½” diameter wood rod that is 10" in length. Wood rods like this are commonly available at hardware or home improvement stores.

For this light load, we placed RX088 ring magnets on the shaft in two locations, near either end. For this demo, the north poles of both magnets are facing the wall. For help identifying the north pole, see Which Pole is North. We used a single D68PC-RB magnet to help in the video.

In the video, we used smaller magnets on the rod and base, but this was just to levitate the rod. In order to effectively levitate the copper wire and solar cells, we need stronger magnets!

Step 3: Step 2: Put Base Magnets in Place

For this demonstration, we chose four sets of 3 RX033CS-N countersunk ring magnets in the base, as you can see in the previous drawing. These magnets are countersunk on the north pole side, which helps to identify which pole is north.

What spacing should be used for the two base magnets?

If the two magnets are very close together, the floating magnet is held higher but isn’t stable. If the two magnets are too far apart, they won’t hold much load. There’s definitely a “just right” distance in the middle. In the video below, we hold up the back end of the shaft by hand, while experimenting with various magnet-to-magnet distances.

After this distance is chosen, the video also shows how the base magnets are set a little further away from the wall than the floating magnet on the shaft. This provides stability, since the shaft tends to tip into the wall. With our setup, we found it to be about 3" center to center.

That’s it! The shaft now spins freely. We've made a pseudo-levitating shaft, which is a great start for making a Mendocino motor.

Step 4: Technical Info- Pseudo-Levitation

There is a common theory in the magnetic world called Earnshaw's Theorem, which basically states that repelling magnets are not stable, and adding more repelling magnets will not make it any more stable. You need some other stabilizing force to make a magnet float in a stable way.

But there is a loophole. Pseudo-levitation constrains the movement of the magnets using some form of a tether or wall. This works because the theorem shows only that there is some direction in which there will be an instability. Limiting movement in that direction allows for levitation with fewer than the full 3 dimensions available for movement.

If we set 2 axially magnetized disc or ring magnets side by side, with their axes parallel, there is a pocket of stability above them. A third magnet can sit in this pocket, but the shaft is free to move along its axis.

If we add a wall to stop this motion, where the levitating shaft starts to move away from the highest point, the wall stabilizes the levitating shaft. By setting the floating magnet slightly closer to the wall than the magnets in the base, the shaft tends to lean against the wall.

With two sets of magnets like this, the shaft is held levitated. It is stable with only one point of contact with the wall.

Step 5: Step 3: Winding Copper Wire

Once you get the rod to levitate and spin nicely, it is time to add the copper wire.

We constructed the rotor frame from light wood, held together with hot-glue-gun glue. It may not be the most accurate construction method, but it was a fast way to experiment with an easily modifiable prototype quickly. We used thin wood from a craft store that was easily cut with a hobby knife.

Start winding the copper wire around the rotor. We made ten turns while keeping the wire on one side of the 1/2" diameter shaft and then ten turns on the opposite side of the shaft.

Winding the wire, we kept a tally on paper to avoid losing count. Make the same winding on the opposite position, crossing the first winding.

We chose 30 gauge (30AWG) magnet wire and used about 1,000 turns in each coil. This was more turns of wire, and heavier than most motors we’ve seen. There are some great looking motors online that use as little as 100 turns. That’s a good thing if you don’t want to use such big magnets – our floating piece weighed in at half a pound. Copper is heavy!

You can find 30 gauge wire like this as MW30-8 in our Magnet Wire section.

Step 6: Step 4: Wire Solar Cells

Once the wire is wound, label the wires so that you can keep track of the direction of the coil and which wire is which. Here we show the first solar cell with wires soldered in place. The tape is only there to prevent tugs on the solder joint during assembly.

We chose a solar panel with higher voltage and lower current ratings. We ordered them from a distributor called Futurlec. We don't have a relationship with them; they're just a place we found solar panels online for a great price.

We wired the panels as shown in the sketch. The sketch shows just one set of panels with a single coil of wire. The coil only shows a few turns, for clarity. In the motor we constructed, we added a second set of panels and coil of wire in the same fashion.

Step 7: Step 5: Technical Details!

What solar panels should be used? How many turns in the coils of wire? What wire gauge should be used? This is where things get complicated. The answers to all these questions are interrelated in all sorts of interesting ways.

We found a few Mendocino motor examples online, where builders include details about what kind of wire they used and how many turns. Many of them show some really dazzling workmanship, much better looking than our rough example! It seems like 100 turns of wire is most common in the descriptions we’ve seen online. What we couldn’t find was a justification that explained why that many turns were used.

By the way, magnet wire is the common name for the single-strand, solid copper wire that has a laquered-on insulation around it, commonly used for motors, transformers, etc.

Before ordering parts for the motor, we had to somehow decide: What solar panels should we get? What gauge of wire? We did some theoretical analysis, comparing various solar cell specifications with various wire gauges and number of wire turns. The results really depended on the interesting ways solar panels work, as well as the shape and design of the motor.

For more technical details, including some analysis and mathematical things, check out our full article here.

Step 8: Step 6: Watch It Go!

In the video below, we set the completed rotor on top of the motor base. Because the rotor assembly was so heavy, a bit over half a pound, we increased the size of the magnets used for both the rotor and the base. The rotor uses two RX088 ring magnets, 1" outside diameter x 1/2" inside diameter x 1/2" thick. On the base, we stacked three RX033CS-N magnets together to form a taller magnet, in four locations.

We show two different magnets set beneath the coil to provide the magnetic field. By hand for our initial testing, see the 1-1/4" diameter x 1" thick DX4X0 magnet held in place. A shorter, 1-1/4" diameter x 1/2" thick DX48 magnet was attached more permanently to the base.

There is a lot more research that can be done on these motors, but hopefully this Instructable can give you some insight into how it works and how to build a cheap version of one! We think its very cool.

First Time Author Contest

Participated in the
First Time Author Contest



    • Indoor Lighting Contest

      Indoor Lighting Contest
    • Metal Contest

      Metal Contest
    • Make It Fly Challenge

      Make It Fly Challenge

    23 Discussions


    2 years ago

    Hey! I really liked the idea and decided to make it. I have got the shaft levitating(with the winding, magnets and panels). I have also made all the electrical connections to the panels. But it does not rotate on its own even after giving a push on either side. What could the possible reasons be?

    Is it possible to determine the number of turns required?

    Could anyone please post the details of the electrical connections?


    2 replies

    Reply 1 year ago

    Same thing here ! I'm susbecting the number of turns .


    Reply 5 months ago

    1. Rotor not balanced well enough.
    2. The illuminated Solar cell is not energizing the right coil. (ie the one close to the bottom but not at the bottom.


    3 years ago

    nice implementation - are you aware of the Bell Labs Experiment Kit #2 from the early 1960's? The guys from the past would have loved to have access to our days' strong magnets!

    3 replies

    Reply 1 year ago

    Here is a better link that is more pertinent. I've never studied the history of the solar cell. : http://www.beatriceco.com/bti/porticus/bell/belll...
    Man, can probably make a solar cell now days. I have been encouraged to work for Bell Labs and also study solar cell technology. My focus of encouragement for the solar cells was in regards to my organic synthesis and biochemistry background. Even the foundry experience may be pertinent coupled with Process Analytical Technologies work experience too. DIY high purity silicone would be something else. I need to study electrochemistry, electromagnetism, math and physics more however.


    Reply 1 year ago

    I found, there is a kit on ebay for #1 and #3. I also found this awesome link that isn't really related though on the 3D printed version of this on another instructable I noted placing in a vacuum with a magnetic clutch. Thanks a bunch for noting!


    1 year ago

    I did it using some youtube videos , I used 0.2 wire with 100 turns . Didnt work ! So Im not sure if its the wire thickness or the magnets umderneath the coil ,but I didnt find anyone talking about it maybe I should add more turns or get bigger magnet ,what do you think?


    3 years ago

    If you set your axis rod slightly aside from the basis magnets, so that it leans on the wall, wouldn't it be possible to replace the needle with a couple of weak, opposite-faced magnets to reduce friction even more ?

    It would require fine tuning of the 'wall' magnet position, but mounting it on a screw and playing with fine equilibrium should do the trick.


    3 years ago

    I guess this could be the starting point for building high debit, high diameter, low speed, low noise and extremely long life PC fans (but not very power efficient, though). Was the first thing I thought of.

    3 replies

    Reply 3 years ago

    Replace the solar panels with thermovoltaics and the heat source (think CPU) becomes your power supply. This would become a self-regulating cooling system. The hotter the CPU the faster the fan spins.


    Reply 3 years ago

    I was trying to think of a practical application as well, considering the design is obviously terribly inefficient. Cooling fans would make sense, though maybe not so much for PCs as larger applications. The only problem would be the lateral force of the air flow.
    But that presents another idea. If you could significantly decrease or prevent friction that would hinder the axle's rotation (perhaps with magnets powerful enough to overcome the friction between the surface beneath the rig and the rig itself), you could turn it into a vehicle. It would be costly to increase the size, but it could be relatively efficient for parcel transport if a max load is considered in calculations. Anything larger would require taking into account materials used, stress, environment, backup power, etc.
    Of course, I'm also thinking "Why not just mount top-facing solar panels off the axle and wire it to the on-axle motor using a circular track?" I guess that'd require an efficient conductive material that could resist high temperatures or a floating motor block. I dunno. At that rate, why even use a motor? You could just mount magnets on the floating axle and rig. Then it would run constantly without fluctuation or need for sunlight, and you could surround it with a housing to protect it from the elements. You could even mount a wireless control board with a battery to adjust the magnets on the rig, allowing you to control the rpm of the axle.
    Eh, whatever the case, it's a fun novelty machine.
    Maybe give some of my ideas a shot, tho, eh?
    I certainly don't have the resources to.


    Reply 3 years ago

    Nah, anything requiring the transmission of significant force to some other mechanical parts, as in the case of a vehicle, would increase friction and not make this any better than normal electrical motors, with friction or ball bearings.

    The big issue with PC fans is noise. In a room of just 2-3 people working together, if each of them uses a powerful desktop with 2-3 fans in it (one for the power source plus one or two for the case, maybe one for the hard disks), if the fans were not specially constructed for low noise, the noise level would constitute a significant industrial pollutant.

    Therefore, PC fan producers use all sorts of noise reduction technologies, from magnetic liquid bearings to special blade designs to reduce noise. Problem is, as dust starts to accumulate both inside the bearing and on the fan blades, these techniques become less efficient, therefore most fans start being overly noisy after just a few months of operation.

    The sources of noise are mainly two: vibrations transmitted to the case by the fan's electric motor and noise caused by the air flow.

    I think that mounting a fan on the axle of such a levitating motor, and having it rotate at slow speeds, using brushes instead of photo-voltaic cells, or maybe putting magnets on the axle and have the coils on the fixed support, would greatly help with the reduction of transmission of vibrations. The pressure in the only contact point (well, two contact points, if you want it to be able to also operate upright) would be small, and a needle-like contact piece, set in a high hardness scoop-shaped blind hole would not make a good vibration transmitter. Add to this a large diameter fan propeller rotating at small speeds, and you could get a pretty much noiseless fan.


    3 years ago

    si vous mettez un moteur brushless à l'extrémité de l'axe vous produirez de l'électricité.

    si vous mettez un moteur sans balais à l'extrémité de l'arbre vous produire de l'électricité


    3 years ago

    Very good and interesting. Thank you KJM.


    3 years ago

    Thank you so much! What great and interesting experiments to enjoy!!


    3 years ago

    Around 1990, I worked on a project team where we floated 2 ton turbine motors on magnetic bearing. It took a rack full of computers to control the custom designed circuit boards. Amazing how far we've come.