View Full Version : Mendocino Motor Magnet Spacing

05-25-2012, 09:36 PM
I have seen a lot of information on Mendocino Motors but have never seen anything on how to determine the magnet spacing of the stationary magnets in the base. Anyone have any idea where that information can be found?

No matter what I try, I don't seem to be able to get the rotating shaft to levitate as is shown in many of the pictures that are on the web.

05-25-2012, 11:02 PM
I have seen a lot of information on Mendocino Motors but have never seen anything on how to determine the magnet spacing of the stationary magnets in the base. Anyone have any idea where that information can be found?

No matter what I try, I don't seem to be able to get the rotating shaft to levitate as is shown in many of the pictures that are on the web.

there are 7 magnets, 2 on shaft, 4 on support rods and 1 under rotor. I would think that you'd need to get the 4 on the support rods facing the same way as the 2 on the shaft so they would repel. I am ASSUMING that the N and S poles are the flat surfaces of the disks, I can't see any other way to get them to levitate.

05-25-2012, 11:26 PM
There are a couple of tricks to note in setting this up.

If you look at the Wikpedia photo, you see the magnets on the base are cylinders. They are polarized with the north and south poles along the axis of the cylinder. The magnets on the motor shaft are oriented the same way and in the same direction. This would cause the motor shaft to want to pull axially in one direction and attach to the opposite magnetic pole on the base. You can see a clear end plate attached to the base in the Wikpedia photo. This prevents the rotor from moving axially. As a result, the north pole on the rotor is held above the north pole on the base and prevents the rotor from dropping.

The north pole on the rotor needs to be just a little closer to the clear end plate than the north pole in the base magnet. Given this, the spacing is irrelevant.

05-26-2012, 08:55 AM
Thanks for the comments, but I do think I have all of those details covered.
Here is the web site that I have been working from.


Once you get there you can change to English language, unless you can read German.

You can download his seminar lecture pdf paper from the site. It does contain all the design engineering details, but is all in German, and I am having trouble understand some of his design properties and explanations. Here is the address of the PDF:


Here are the dimensions that I am working from.


You would think that with this much information, it would be easy to duplicate!

I am at the point shown in the second picture from the top, the one with the levitating shaft against the hard drive disk, the fixed magnets below.
No matter what I try, the outer end of the rotating shaft moves either left or right and then drops down.

Seems like a really fun project! Now if I just get it to work as shown!


Lew Hartswick
05-26-2012, 09:47 AM
<No matter what I try, the outer end of the rotating shaft moves either left or right and then drops down.>

Well if it can move either way THAT is your problem. It must be constrained
from moving the one way and be just in the direction that it wants to move that way.

michigan doug
05-26-2012, 11:42 AM
I suspect that your base magnets are too close together. If you space them too close, there isn't a big enough/deep enough magnetic "valley" to trap the axle.

If you space them too far apart, you won't be able to support the axle, but I bet you can space them further apart than they are now. The ideal distance apart will be highly dependent on how strong the magnets are. Some of the older mendocino motors used the relatively weak ceramic magnets, and consequently had to be pretty close together to hold up the axle.

A picture of your device would be interesting and helpful.

Finest regards,


05-26-2012, 12:53 PM
Sometimes the most simple things are the answer to a problem.

I read, "the weight of the armature determines just how deep it's settles into the magnetic field of the two sets of magnets. The deeper it sets, the more stable it is."

Not having built the armature yet, I was trying to levitate just the armature shaft with the two magnets. NOT HEAVY ENOUGH! Slip a brass wheel on the shaft and problem solved.

Houston, we have levitation!

Here is a picture of the project so far.
To give you an idea of scale, the armature shaft is 1/8 brass rod.


Now off to start designing the armature block. Going to cut it out of Styrofoam. Bet that will really be easy to mill!

Thanks for all the suggestions,

05-27-2012, 01:15 PM
This is the Armature Core for the motor. I machined it out of rigid pink foam. Boy, does that stuff ever machine nice!

The notches in the corners are where the windings will go around. There are two milled out areas on each of the four sides. The upper "shelf" is where the 3" X 1" solar cell will be mounted. The deeper recess is where the hookup wires will terminate. Also will serve as a place to screw in a few small screws to balance the final assembly, if necessary.

Next step, wind about 100 or more turns in the slots on the corners of the core.


05-28-2012, 06:38 PM
Well today I got a lot more done on the project. As winding the Armature is the next project, I started by making a stand to hold the roll of wire. As I did not want the wire to freewheel, ending up with a big mess, I ran it through a piece of tubing, squashed just enough to put tension on the wire, which also helped me to wind the wire tightly on the core.

Here is a picture of the setup. Very cheap and dirty, but it worked really well!


Next it was time to wind the core. I started by drilling a series of holes from the shaft opening into the wiring area of all four sides. I also drilled a pair of thru holes between the opposite sides of the wiring area.

Here is a picture of the first set of 120 turns, alternated 10 turns at a time on each side of the shaft.


Then on to the second side and 120 more turns of #30 wire. I did slip small pieces of 3/32 tubing over the center shaft to insulate the wires from the center brass shaft.


I finished up the armature by installing and spacing the magnets. Again, using small pieces of 3/32 tubing to hold the magnets in final position. Next I had to balance the assembly. I let it levitate and spun it between my fingers. When it stopped it was obvious that it was out of balance. If you look closely you will see several screws placed to balance the armature. I can not stop it an any position and it just sits there.

The next three pictures show the project as it is of of this afternoon.


05-28-2012, 06:39 PM
Here are the last two pictures.



Next I will be sorting out the wires from the coils and establishing some solder points in the wiring areas. As soon as the solar cells arrive, I will be ready to solder them in and see what happens.

06-01-2012, 09:13 PM
Well gang, the learning curve continues and is very steep!

The armature shown in the pictures above was a complete failure, for many reasons. I had it perfectly balanced, or so I thought. You learn by doing dumb things. First dumb thing, DON'T USE STEEL SCREWS to balance the armature! Once I got the solar cells in and wired, it just would not run, no matter what I tried. Even though it was balanced, it kind of limped along as you would spin it. Finally figured out where the limp was coming from, the *&#.&% steel screws! Every time they came across the bottom of rotation, they would abruptly slow down the armature. They were passing above the three magnets in the base!

OK, now replace them with brass screws, and do the balance all over again. Put the solar cells back on the wires, only broke 2 of the 4, and try everything again.

Second dumb thing! No more rubber bands! Might as well glue the first two solar cells in place. So, take some super glue and carefully put it along the edge of the cells. Coming back in about 30 minutes....WOW....The super glue melts foam! The cells are held in a few places, but everywhere I put the super glue there are gaping holes in the foam!

Third dumb thing! Solder in the other two cells, this time use some Gorilla Glue to hold them in place. Forgot, Gorilla Glue expands when it dries. One solar cell is broken, the other three look like a Kindergartner was working on the project, really ugly! Icky brown gobs of glue everywhere!

Onward and upward, lets levitate it and add some light and see it run! But first, now with the cells and glue in place, lets check the balance. First turn off all the direct light so balance is not influenced by magnetic field of the coils with solar cell in light. Balance is like I had done nothing before. OK, now what, can't hide the brass screws under the cells. Forge ahead, poke screws in the areas between the cells, toward the windings. OK, it is kind of balanced! Will it run? Yes, Yes, Yes! So the proof of concept seems to be working, but I am so ashamed of how it looks, absolutely no pictures! Seeing I did machine several armature cores, tomorrow I will do a complete new one, hopefully without all the screw ups!

A new day!
Checked it out, still runs, but ugly! Try to salvage what I can. Only thing that can be saved is the shaft and magnets. And yes, break the core all apart and retrieve the brass screws for possible use later! Throw away the rats nest of wire! Start from scratch.

Wound the core with 150 turns in each coil. Next levitate the armature and check balance. Not to bad, a few BRASS screws here and there and it will set just about anywhere I stop it. Now to solder in the solar cells. This time hold all four in place with a couple of rubber bands. Glue can come much later, if ever! Again check for balance, not to bad. Three or four 2-56 short oval head brass screws in the right place, almost perfect balance again!

Time for the test, add some light, give it a light spin.....WOW.....it actually runs, and quite smoothly! Just a bit of a wobble at some speeds, and a bit of back and forth motion on the free end. Add a brass nut to the free end, sets it a bit deeper in the "magnetic V" and most of the back and forth motion stops.

I am pleased, it really runs great. Did not check the RPM, but will sit and run smoothly with just the light from the overhead florescent ceiling lights in the shop.

Now that I know how to build one that runs, Using the completed armature I am going to try different lower magnet spacings to see how I can improve the stability. I think that if I were to space the lower magnets farther apart, the armature magnets would set deeper into the "Magnetic V" and reduce the wobble and side play considerably. So tomorrow, I will probably design and cut a couple of new base uprights that hold the magnets with new spacing and see what happens.
I also want to try the base with only one magnet in the center, rather than the three that I now have. Maybe even a stack of several to see how it influences the operation.

As far as holding the solar cells in place, need some more ideas there. As they do sit nicely in the milled out pockets in the form, rubber bands, at least for now will have to do.

Below are two pictures of the new finished armature. The first shows one of the solar cells almost in it's niche. The other shows the armature spinning with just the room lights providing the illumination. As I make any improvements or new findings I will report back here with the details. This has been, and will be for a while, a very neat project.


George Bulliss
06-01-2012, 11:28 PM
Thanks for for the progress report George. Looks like a fun project. I've done some digging on these engines but need to do some more.


06-02-2012, 12:01 AM
Interesting project. Couple of ideas- you could probably use water based rubber cement or contact cement to hold the cells. If you used two pairs of support magnets on each side, you could probably dispense with the vertical tab. Also you might play with some light shielding to possibly improve how it runs. Some flat black paper over the bottom might improve things also.

The thing runs by having current pass through the magnetic field at right angles to it, creating a force at right angles to both those directions. The more lines of force that intersect the windings, the stronger the torque will be. Hard to say what will happen to the rpm- theory says that the back emf would be larger with stronger magnetic field, and therefore it might run slower. With less magnets it might actually run faster, but it may not have enough torque to start.

The solar cells are probably working in current mode, as the wiring resistance would be so low that the full cell voltage might not develop. As the armature turns faster, the current draw will lessen and the voltage will rise. I could be wrong, but I'm going to suggest that in this case the armature rpm will probably increase with a stronger magnetic field- up to a point of course.

Keep playing :)

06-02-2012, 06:36 AM
Hi Darryl, never thought of rubber cement, sounds like that may be a good solution. I want to coat the ledge behind the cell so nothing shows, rubber cement holds really well. I believe it is a form of contact cement, so pressing it together almost dry will probably hold them firmly in place. Great suggestion!

I did measure the voltage and current with full illumination on one. I think the coil of 150 turns is around 5 ohms. I measured .5 volts at .1 amps from the little cell, connected to the coil, illuminated by a small desk lamp. It is amazing, they are rated at .5 volts @ 250 ma with sunlight. A lot of efficiency from something so small, only about 4 mills thick!

Can you please explain your comment about using two pairs of support magnets negating the need for the vertical post?

Incidentally, commutation takes place due to the fact that opposite cells are wired in reverse of each other. So thinking about the schematic diagram, the coil ends up in parallel with the solar cells. The cells wired plus to minus of each other. Kind of neat.

I have a lot to understand about how all of this works, and I still do not have a formula or any basis for low the lower set of levitation magnets should be spaced. Am going to test extremes, narrower and wider spacing and see what I can discover.

Thanks for your comments,

06-02-2012, 04:02 PM
My comments in regards to the black paper and shielding are to do with maximizing the difference in output from the cells as the armature rotates. You would want the exposed cell to be producing current, but as it rotates 'under', you would probably want it to drop in output so the other cell coming into exposure to light would not have to fight it at all. Thus the idea of creating a 'dark' area under the bottom side of the armature. I don't know what effect it would have, but the thought is that it might run more forcefully, run with less light, etc. Something to experiment with.

As far as the magnetic suspension goes, you have one 'stand' with two magnets in it attached to the base at each end of the armature. I'm thinking that if you had two stands at each end, spaced apart by 3/8 inch or so, the suspended magnet on the armature shaft would center itself between the stands as well as remain suspended. Doing this at both ends of the armature would add a bit more centering force and would keep the armature level as well. I'm thinking that some of the wobble you're getting is probably due to the end of the armature shaft 'walking around' on that vertical tang.

On to the rubber cement- I mentioned water based as that would not melt the styrofoam. You could try a patch of solvent based rubber cement on some styro just to see if it melts it- after all the re-steps you've taken, you don't want to have to re-build the armature yet again. Good on you for sticking with the project through all the hassle, though.

06-02-2012, 04:25 PM
Do you actually need the foam once you have the wire wound? You could fashion small clips attached to the wire to hold the cells.

06-02-2012, 08:16 PM
Darryl, it is quite dark below the armature, even in sunlight, so I don't think that is an issue, but sure something to try. A piece of black paper on the bottom, with a steady rotation would prove the thought. I will try it and let you know.

On your magnetic suspension idea, sorry but you cannot levitate both vertical and horizontally at the same time, that is according to Earnshaw's theorem, Maxwell's equations prohibit this type of levitation. If you Google Earnshaw's theorem you will get all kinds of good reading material about why it can't be done. It sure would be neat, and I did cut a second set of matching end plates and tried it. It just will not work!

I got some rubber cement this afternoon, will try it on a piece of foam tomorrow and see if the foam survives the experience.

I did a lot of experimenting today, and have an even better and simpler working system. Have a bunch of pictures of the now finished project, will probably get them written up and on here sometime tomorrow. I think you will be pleased with the results that I got today, and it really looks nice to boot.

Elf, that is a really neat idea, but I do see a lot of challenges you would need to overcome. The first would be a system to balance the finished armature. I would also think it would be difficult to hold the coils in place on the armature shaft, and how would you keep the bundles of wire in a permanent shape? It would really be different, give it some thought and see what kind of solutions you can come up with. Keep in mind it is number 30 wire so very small and flexible, 150 turns in each coil.

More later,

06-02-2012, 10:15 PM
Today I really did a lot. I started by widening the end pair of magnets. Went from a center to center of 1.5 inches to 1.875. My idea was to bring the armature down closer to the end magnets as well as the base magnet. What a difference, almost immediately it started running, all by itself, in just the light in the shop. I got rid of the brass nuts (weights) on the both ends of the shaft and did a really fine balance of the armature. Made a new base with only one magnet, as the three were influencing the armature magnets to some degree. It immediately started running much smoother, and of course it sets deeper in the "Magnetic V" due to the wider spacing. Incidentally, after all the changes today, my armature weights a whole 2.5 ounces. After all the changes the armature sets just over .5 inches lower as well. The only thing left to finish is the gluing on of the photo cells. That will complete the build. The last thing I will add will be a list of materials that can be purchased to build a similar motor in your own shop. This is really a fun project! Anyone who sees it does not believe what they are seeing.

Follows, a series of pictures of how the unit looks after doing a complete rebuild of the base frame work today.

Here is how it looks running.
Here is a full side view, and yes I did polish all the aluminum plate!
Another side view, favoring the back end.
The small holes you see in the foam are where I have pushed in #80 brass screws while balancing the armature.
Here is a picture of the rear end rest. I machined it out of a piece of smoke lexan. It is bolted to the base with 2-56 screws in threaded holes.
In the picture above, notice the way I am now holding the rotor magnets in place. As the shaft is .125 brass, the .125 milling tool stops are an ideal way to hold the magnets vertical and in place. The stops really press on hard, and will not move once in place. Just getting both end magnets to turn without a wobble is a giant step in getting the armature to run smoothly.

06-02-2012, 10:17 PM
The next picture shows the "floating" end of the armature shaft. Note the tool stops holding the magnet in place.
Here is a pictures of the front of the frame with pressed in magnets.
Here is a picture of the rear of the frame, showing the end stop for the shaft.
Here is a close up of the 1" X 3" .5VDC 250MA solar cell. Hopefully tomorrow I will have them all in place with some rubber cement, and the rubber bands will be gone for good.

06-02-2012, 10:19 PM
And the last picture is of the bottom magnet that attracts the armature coil that drives the rotation. You are looking at the reflection of the armature in the base place. It really polished up very nicely.

I did use my optical tachometer to check the speed today. Sitting under the shop ceiling lights it ticks over at about 215-220 RPM. I exposed it to blue sky this afternoon and was amazed when it consistently showed 2180-2190 RPM, and running very smoothly at that speed. For this unit, I believe that is the maximum speed it will achieve.

For this to be a declared finished project, I will be adding a list of materials and links to the DXF files of the base unit and foam cutouts. I will also be writing up a list of dimensions for those that would like to have a starting point and do their own thing. I do think that my final dimensions are ideal, and will be very easy to duplicate. This is really a great running little motor!

06-02-2012, 10:37 PM
Nice work, George! I'm curious about the solar cells- are those amorphous? If so they must be some advanced version, since none that I have can put out that much current. I'm also intrigued by your good results with just the single base magnet. Maybe I was right about the solar cell working in current mode- with a single magnet there is probably less back emf, so the low voltage from the solar cells doesn't matter as much as to how fast the armature rotates.

Jaakko Fagerlund
06-03-2012, 04:10 AM
About the need for end plate:

What if the ends of the shaft would have tiny magnets, with for example S poles pointing outwards and the end supports would have the same but pointing inwards, this would create an ever so small opposing forces that keep the shaft from touching anything.

06-04-2012, 03:12 AM
I made one of these repulsion-suspension things today, just to see. Mine is only magnets and supports- no solar cells, etc. I tried the dual sets of magnets- sure enough I couldn't make it work either. I also tried the end magnets to keep the shaft centered- couldn't get that to work either. Earnshaw comes out 2 to my 0. :(

06-04-2012, 08:13 AM
The cells are actually rated at .55 VDC and .5 amps each. That does seem like a lot of current, but maybe that is why my concoction starts so easily and runs in very low light. I tried one of those HF LED flashlights, you know the ones they give away with a coupon, and it runs right along in a perfectly dark room.

Here is the link to information about the solar cells. Very good description of the type of cell down the page, toward the bottom.

And yes you are right, Earnshaw, got ti right! I tried several different configurations, and nothing seems to work. The one thing that I have not tried is putting 4 magnets in a square, facing each other in opposition, and then using 4 very small magnets on four sides of the spinning shaft, also in opposition to see if it will float. That is what I think that Jaakko is suggesting. But again, Earnshaw will probably win another one. I may mock that up just to prove it will not work.

Anyway, thanks for your comments, this has been a real fun thing to build and play with.

michigan doug
06-04-2012, 05:49 PM
Then there's the levitron:


Some theorists had decided this was impossible not so long ago. There's a pretty juicy story of the guy who invented it, Roy Harrigen, and the guy who stole it and made money at it:


In heaven, I will be able to levitate at will.

Finest regards,