magnets-rotor

Magnet theory & rotor construction

Select from the list below to jump to different topics on this page.

Making the rotor

Importance of a balanced rotor

Positioning of rotor magnets

Positioning of stator magnets

Alignment of rotor magnets

Levitation theory as applied to this type of motor:

Before we go much further you ought to understand how magnets work. Wikipedia has everything you probably wanted to know about magnetic fields. Magnetic fields are used to cause levitation and run all electric motors.

The first thing we need to do to build our levitation motor is learn how to make the motor shaft (called the "rotor") actually levitate. That's why understanding magnetic fields is helpful. You've probably seen pictures on YouTube or perhaps you have seen a levitation motor in person and you noticed some magnets at each end of the motor base with the rotor being held up in the air nice and stable.

Almost any two magnets can be made to levitate. But the trick is to have the levitation stable so the magnets don't flip around and stick to each other, north to south (N-S).

The picture below shows a magnet in unstable levitation. The upper neodymium magnet is being levitated by the three magnets at the bottom of the picture. The levitated magnet and the set of three magnets below are arranged so like poles face each other. Either N-N or S-S will work fine.

The vertical dowel is made of wood. The aluminum rectangular base has a hole drilled in it that the dowel is inserted into to hold the single levitated magnet in position. The wooden ruler is for reference.

One lower magnet will easily levitate the upper magnet but I wanted a good bit of levitation so you can see the affect more clearly. The upper magnet is tilted slightly forward because it is not in a stable mode. If the wood dowel was removed slowly in an upward direction the upper magnet would remain levitated until it wasn't touching the dowel. As soon as the magnet was free of the dowel it would flip over so it was in a N-S relation to the bottom magnets and then would violently crash into the three magnets below.

Allowing magnets to crash into each other is not good for them and neodymium magnets are quite brittle and will sometimes shatter if they crash together. So be careful, you certainly don't need a fragment of a magnet in your eye and they can painfully pinch your fingers if you get in the way of them when they are attaching to each other.

Tip:
To separate the magnets I use a 1/4" diameter aluminum rod that I insert into the center of the ring magnets far enough to slide off to one side however many I need. Otherwise it can be a real fight to pull one or two magnets apart. And the magnets I'm using aren't even the strongest ones available.

Return to top of page

To see an enlarged view of most pictures, left click on a picture or right click and select "View Image".

One of the first things we need to do is prepare the rotor shaft.

These are two pictures of my completed 5" long rotor made from a steel 5" long 1/4" x 20 threaded screw. It could be made shorter but I found that to create a very low current motor having a longer shaft helped the motor run with less current draw. I'll explain that later. I also wanted leave room to place the driver coil under the rotor. If you can find an aluminum threaded rod that would be the best material to use for lightness.

The right end of the screw had a flat blade screwdriver slot in it. I machined it off in a lathe and reduced it's thickness so it now acts as a washer on the end of the screw and also made the rotor shaft lighter. Notice that there is no magnet retaining nut on the right side of the shaft. The magnets there are held in place by the washer end of the rotor shaft. If you use a non magnetic rotor you will have to use a nut(s) on that end of the rod. Aluminum end nuts would further reduce the weight of the rotor.

When the rotor is levitated it is held in side to side position by the levitation magnetic field. But there is nothing to hold it in place axially down the length of the rotor shaft. In affect the rotor is free to move axially back and forth if it's not restrained somehow. The rotor magnet on the left side of the rotor shaft in the left picture is positioned in relation to the stator magnets so there is a small magnetic force pushing the shaft up against a vertical positioning device. A small bearing is needed at that end of the shaft to reduce friction. I will cover the positioning of the magnets further along in this article.

The bearing usually consists of a hard steel ball (some motors use the tip of a ball point pen) that lightly touches the vertical positioning surface when the rotor is levitated. For this motor I used a 1/8" steel ball which I removed from a larger ball bearing. I chucked the rotor shaft in a lathe (using the hex nuts to protect the threads) and drilled a 1/8" diameter hole about 3/32" deep in the end of the shaft.

I then placed the ball in the hole and center punched the shaft on each side of the ball. If you look closely at the enlarged picture on the right you can see the punch marks on either side of the ball. The punch marks deform the metal of the screw slightly and hold the ball in place. The ball is actually loose in the hole and that might contribute to the small off center I have of the shaft. Every once in awhile after running the motor I put a drop of Marvel Mystery oil on the ball to lubricate it. Since the ball is loose some of the oil runs into the hole.

Marvel Mystery oil is a real product that has been around forever. It is basically a highly refined friction reducing oil of about 10 wt. viscosity.

I've found that if you tear a ball point pen tip up far enough you will find that the small ball and the metal tip of the pen can be reduced down to a very small part that could also be inserted into the rotor shaft. I like the steel ball better because I feel the psi on the actual part of the ball that touches the vertical part of the bass plate would be less for the 1/8" ball than the smaller diameter ball point pen tip. Sort of like would you rather be stuck with a pin point or the sharp end of a pencil?

Return to top of page

Like the tires on your car, having a balanced rotor will make for a smooth running motor. As you place things on the rotor shaft (drive magnets etc) strive for a balanced rotor when it is levitated. I would test the balance of the rotor shaft without the rotor magnets installed. The reason being that the stator magnetic fields will attract the either the N or S pole of the rotor magnets and it will appear to be out of balance when it isn't. i.e. the rotor may always stop with either a N or S pole magnet closest to the baseplate. I spent several hours trying to balance my first motor not realizing this.

Since there is so little friction with the rotor levitated the least bit of out of balance shows up. Most rotors are going to be a little bit out of balance. The slightest out of balance of the rotor will cause the rotor to bounce around at some rotation speeds similar to car tires. Some speeds are much worse than others. I've noticed that the out of balance oscillations of my motor are over a rather small range of low rpm. There seems to be another high speed out of balanced condition at very high speeds.

When running the motor on low voltages this bouncing around can cause the motor to not be able to speed up above the out of balance speed. The bouncing around of the rotor can cause the rotor to fly off the levitation field. When this happens the rotor will crash onto the stator magnets.

I have found a method of suppressing much of the out of balance rotor affect at the free end of the rotor so that at low speeds the motor can usually speed up and at higher speeds to completely stabilize the rotor. Again this is in accordance with known magnetic principles (magnetic eddy currents). I haven't seen anyone else using this technique on YouTube.

Since the rotor has two pairs of stator magnets levitating the rotor at each end, the whole rotor assembly can't flip around because each end of the rotor holds the other end in alignment.

Return to top of page

Finally we're going to put something together!

In order to avoid confusion later, let's call the right end of the rotor shaft the "washer end" for later use in the assembly Let's call the other end of the rotor shaft the "bearing end". I used two aluminum nuts to hold the bearing end of levitation magnets in place. The aluminum hex nuts helped to lighten the bearing end of the rotor.

By using a threaded rotor it is very easy to adjust the position of the various parts. When I build another motor I will probably position the magnets on the bearing end a little closer to the end of the shaft. Ideally I'd like to have the balance point of the rotor shaft located between the vertical centers of the stator magnets.

The four ring magnets in the center of the rotor are magnetically attached to two 1/4" x 20 square steel nuts. Just thread the two square nuts on the shaft and tighten them only enough so the nuts form a nice square surface for the magnets to lay on. If the two nuts came from a different batch or manufacturer try several of them till you find a pair that will be flat when slightly tightened. Don't over tighten the nuts. Don't leave the nuts loose or the magnet assembly may move on the shaft when you spin it by hand.

Positioning the rotor magnets on the rotor shaft.

This picture shows how the four rotor magnets are attached to the square nuts. The four rotor magnets are placed on the nuts N-S-N-S. To keep the magnets from jumping towards each other I usually place two like pole magnets on opposite sides of the 7/16" nuts and then place the other two magnets with the opposite poles facing out.

The 1/2" diameter ring magnets are 1/16" wider than the width of the 7/16" square nuts. As you place the magnets on the square nuts move each magnet around until it slightly overlaps the next magnet on one side by 1/16". Look closely at the small picture and you'll see what I mean. The top magnet is overlapping the vertical magnet on the right. By doing this with all the magnets they will fit on the smaller nuts and the balance of the shaft will be maintained. Also move the magnets so they are centered over the line where the two nuts meet. This will stop longitudinal resonance affects due to the rotor magnets reacting to the fixed levitation magnet fields.

If you are going to run the motor on very low voltage you will probably find that there is a slight difference in the input current required when the motor rotates clockwise or counter clockwise. This is caused by the slight magnet overlap and a small reaction of the rotor magnets to the stator magnetic field. On my motor when looking from the washer end towards the bearing end with the magnets offset to the right, clockwise rotation usually requires slightly less input current. So try starting your motor in both directions to determine which draws less power. If you are interested in high speed, higher voltage motor don't worry about the direction of rotation.

If you want to experiment you can use just two magnets on opposite sides of the nuts in a N-N facing outwards configuration. Not permanently fastening the magnets to the rotor shaft and letting them hang on for dear life allows you to experiment with different configurations of the motor.

When running the motor at high speeds (5 volts) I usually wrap a strip of scotch tape around the magnets so they can't fly away. You really don't need to get hit in the eye by a high speed magnet!!! Be careful. The tape also slightly lowers the aerodynamic drag which allows the motor to run faster.

You may need to move the magnet assembly to level the rotor shaft. Make sure the baseplate is level first. To make fine adjustments in the left-right position of the rotor magnets just remove the magnets, loosen the nuts and rotate them to the new position, slightly tighten the nuts and put the magnets back on.

Return to top of page

The following information needs to be presented somewhere and this seems to be as good a place as any.

There are many considerations and compromises that need to be decided upon at the same time as you initially build a motor from scratch. For instance if you want to permanently mount the drive coil you have to decide do I want a low current motor or a high speed motor. Also, do you want to mount the coil under the rotor or off to one side?

The spacing of the drive coil to the rotor magnets has a direct influence on how much current the motor draws. The closer the drive coil is to the rotor magnets the motor will run on lower voltage and less current. If you run the motor on solar cells this will allow you to buy smaller sized cells. For a high speed motor with a wider spaced coil it also means the motor will not want to run under low light conditions that a close spaced coil motor will run.

For a low voltage/current motor the drive coil needs to be very close to the rotor magnets because there is less drive energy available at low voltage. For a high speed motor running on higher voltages the drive coil should be further away so the motor can develop higher speeds at lower current drain compared to a close spaced coil.

I've also found that a close space coil running at high voltage will tend to demagnetize the rotor magnets over time. I check the strength of magnets by using a ruler to measure the distance several magnets of the same shape and power will deflect a compas neadle 90°. You will also notice that the motor will no longer run on as low a voltage as it did with new rotor magnets in place. My motor originally would run fine on 0.6 V input. When the magnets got weaker it wouldn't run below 0.8 Vdc. I change the rotor matnets and it runs on 0.6Vdc again.

By having the coil loose you can change a spacer under it to vary the clearance to the rotor magnets. This way you can have a low power requirement motor and a high speed motor just by changing the coil spacing to the rotor magnets.

The motor was running when this picture was taken so the hex nuts on the rotor appear blurry

Determining the proper placement of the all the magnets on your motor is very important, so understand what needs to be done before you start. I can't give you exact dimensions because each person's home design will have a different rotor weight, different size and/or strength magnets etc.

I am going to call the magnets mounted on the base plate of the motor the "stator" magnets because they represent the part of a more conventional motor that has that name. On my motor the stator magnets are mounted on the beige nylon posts in the above picture. Each builder comes up with his own method of holding the magnets in place. Some methods are very artistic using clear plastic etc. Look at some of the videos on YouTube for ideas.

You can use steel bolts and nuts instead of the plastic ones I used. I didn't test my system with steel bolts but they should work OK.

Each pair of stator magnets creates a magnetic field similar to bar magnets. At the same time because there are two stator magnet pairs located near each other, there is another, more complicated field created between the pair of stator magnets. This field, if looked at in a vertical cut away view running from the center of one stator magnet pair to another stator magnet pair would have a vague figure 8 pattern laying on it's side if the pairs are separated the correct amount.

The trick is to separate the two stator magnets just far enough from each other so this figure 8 field has a slight dip in the middle (1/8" dip in my motor).

As usual, a picture is [sometimes] worth a thousand words.

This picture illustrates the shape of the combined vertical magnetic force fields that exists between the two pairs of stator magnets at one end of the rotor. The pencil line on the vertical aluminum wall is the vague figure 8 field mentioned above. In affect the pencil line represents an equal strength magnetic field around the sator pairs.

I am resting the end of the rotor shaft on my hand so the rotor shaft is horizontal and is free to move up and down where it touches the vertical aluminum. The rotor is centered between the stator magnets even though it appears that it isn't in the picture. The end of the shaft is lightly pressing against the vertical aluminum panel. The steel ball is in the end of the shaft.

Let me explain what we are looking at here.
I used this test set up to determine where to place the various magnets for levitation and actually running my first motor. I made a temporary base plate from a 5" wide x 6-1/2" long piece of sheet aluminum that has a 90° degree vertical bend at the far end behind the ruler. The vertical part is 5" tall. The measurements are not critical at all. It was the first piece of aluminum I found in my cut off pieces of aluminum that had a nice 90° bend in it.

I applied two strips of scotch tape to the vertical part of the aluminum so I could record where the rotor shaft touched the plate as I guided the it from left to right over the stator magnets with my other hand. I put a little oil on the ball in the end of the rotor shaft so it would leave a trace on the scotch tape. I then darkened the line that the rotor drew on the tape with a pencil so the dip in the magnetic field is clearly visible. The dip is 1/8" lower than the peaks on either side. That's the sweet spot where the rotor will ride when it is levitated (as it is doing in this picture).

You want to have all the parts mounted on the rotor shaft so it will have the final weight the motor will run with. Holding the washer end of the rotor loosely with your hand represents the stator magnets on that end holding the rotor up. Since you are holding the other end of the rotor in place (the same as that pair of stator magnets will later do) the rotor is in a stable levitating mode. Using only one set of stator magnets at first will allow you to concentrate on getting only one end of the rotor shaft levitated rather than fighting both ends at once. Once you have found the dimensions for this end of the rotor shaft you can duplicate the magnet spacings for the other end of the rotor.

The gray 5" long 1" x 1/2" gray 90° aluminum angle makes a perfect temporary mount for the stator magnets. It is parallel to the back wall of the base plate.and about 3/4" away from the wall. Since this picture was taken I have added 10 mm and 5 mm tic marks across the angle where the magnets are located. This allows me to reposition the magnets quickly. I have one longer mark at the center of the angle to allow me to center the rotor for testing.

All you have to do is put a magnet on either side of the vertical part of the gray angle facing N-S to each other and they will stay in position for testing. Clothespins hold the various parts in place for testing. In my case the stator magnets are spaced 1-5/16" apart center to center. The other end of the rotor should have the same dimensions for the placement of the stator magnets as the end shown in the picture (once you find the dip etc).

Since there is a 1/8" gap between the two magnets of each stator pair the magnetic field will be reduced somewhat. So it is best to use a very thin angle material or temporarily have all the magnets mounted the way they will be before you drill holes to permanently mount them. For my final placement of the stator pairs I placed two magnets on one side of the gray angle and held the magnets in place with more clothespins. Don't worry to much about the height of the stator magnets at this point but if you want to mount the coil beneath the rotor the stator magnets do need to be high enough so the rotor magnets clear the drive coil as they rotate.

Return to top of page

Alignment of bearing side levitation magnet(s):

The reason for the offset is to allow the stator magnetic field to exert a slight "push" on the rotor levitation magnets towards the left in the above picture. This forces the rotor shaft ball bearing to press lightly against the vertical positioning part of the 1" x 2" 90° aluminum angle. This is what stops the rotor shaft from moving left-right and is the final part of a stable levitated rotor.

Once you get the motor running you can adjust these rotor magnets very slightly towards the right (by loosening the hex nuts and rotating them in one direction or the other) to reduce the friction on the bearing. This will allow the motor to run faster and/or draw less current while running.

When running the motor on 5 volts (at very high speed) the rotor usually gets a seesawing and/or side to side motion as it reaches top speed. I think this is because the bearing has a slight tendency to move away from the stable sweet spot in the magnetic field and the rotor magnets enter a mechanical-magnetic resonance to the rotor and levitation magnets. This causes the rotor to oscillate fore-aft and vertically away from the bearing end vertical positioning wall.

By looking at the bearing where it contacts the vertical metal wall I get the feeling that it is not always in full contact with the wall if you have set the offset to a very small offset. As you lessen the bearing pressure against the vertical wall the motor is harder to start because your fingers tend to pull the rotor away from the wall as you spin it. As you get to the point where the rotor isn't being pushed against the vertical wall, one sixteenth of a turn of the bearing end hex nuts can make a lot of difference in the offset push. So go easy with the adjustment.

All the stator and levitating magnets on my motor have the "North seeking poles" facing -away- from the center of the motor. I've found that this lowers the minimum voltage and current (power) the motor can run at. In affect the stator field at the rotor magnets is hopefully somewhat nulled out because the stator magnet poles are S-S to each end of the motor. I haven't tried it but I see no reason why the having the south poles facing -outwards- would change the power requirement. But I have another reason why I oriented the magnets N facing outwards.

When my motor isn't running, one of the north pole rotor magnets is facing downward because it is reacting to the south magnetic field from the stator magnets below it. I want this to happen because I am trying to design a simple self starting feature for the motor. This will cause the self starter to have a known position of the rotor when it's not running. i.e. a N pole of the rotor magnet will be towards the drive coil.

As soon as you get the rotor to levitate properly with all the stator magnets you can try spinning it with your finger tips. Mine will spin for over three minutes once in awhile. Most likely you will also see the out of balance bouncing condition as the rotor slows down.

The bouncing condition brought out one of the main problems with this temporary test set up. That is, just about every time the rotor would loose levitation it would fall to one side and invariably stick to and move the stator magnets. Then I'd have to reposition everything all over again. This went on for quite a few days until I realized that a dip in the magnetic field was necessary. After that things progressed much faster.

The small red fiber washer you see next to the lower stator magnets is to bring that magnet set into alignment with the other stator set. The drill must have slipped a little when I was drilling the baseplate for the mounting hardware for the nylon post. Luckily the mounting posts have side to side clearance for the 1/4" nylon screws which allows me to make a final adjustment of the center to center distance between the stator magnets. I don't remember where those posts came from but I think they might have been wiring hold downs in a commercial CRT monitor. If you make the magnet mountings out of a plastic sheet you could slot the holes for the 1/4" bolts to allow adjustment.

Return to top of page