Square wave alternator

[Heretic]

I originally built this as an experimental motor, which worked quite well as a matter of fact. Nevertheless, it's properties as a generator made me think of using it in a wind turbine application.
It's main advantage is that it generates a square wave instead of a sine wave. As many of you know, a square wave, when rectified, provides far more power density per unit of time than a sine wave. This alternator produces almost pure DC after the rectifier.
Here's a picture of the frame and rotor of the first unit I built:



As you can see, the rotor consists of stacked neo magnets inside an acrylic frame. The axle extends through a couple of tubes and is supported by ball bearings.



This picture shows the end "fins" which will contain the windings and provide a platform for mounting.

Here it is with the windings completed. I used 24 awg wire because I wanted a higher impedance, but any size wire can be substituted. I mixed a couple of different spools which is the reason for the two colors.



This is the top of the unit, and the only place throughout to whole rotation of the armature where there are no windings. The rest of the time the magnets are producing current.



This unit is lightweight, easy to build, compact and can be customized to fit a wide range of space and power requirements. There is no cogging either.
I'm posting here because I would like to see it tested on a wind turbine. I have very little wind where I live, otherwise I would do it myself. Nevertheless, I have tested it on a load with another motor driving it and it performs quite well. This was the setup I used, with a second motor I built used as the prime mover.



Building a longer version can increase the efficiency since a greater percentage of the total winding is producing current. Voltage is determined by the number of turns and the speed of the armature. I could get over 100 volts at a relatively slow speed, but I haven't tried a lot of different configurations so some trail and error is need to determine what an optimal winding would be for wind applications. 

Cheers,

Ted

[tecker]

IT's interesting that a window motor should make it in the form . I've run these both on brushes monopole and dipole . You can't induce much with this large coil nor can the coil drive loads . I hope you carry on with a wind setup to have some data . I may fit up a few on the porch with the other motor generators . I have a tough time down low with wind and a motor generator with wind augmentation is very attractive .

[Heretic]

It's more of Newman, window hybrid, except this motor can actually drive a load. It has full torque for about 175+ degrees on each half cycle. Unlike the Newman motor, the magnets are closely coupled with the windings throughout the arc, and unlike the window motor, the windings extend to become one large coil instead of several smaller ones.
If wound evenly, it puts out an almost perfect square wave in alternator mode. It's far cheaper, lighter and easier to build than an axial flux generator, and has much better power density.
It's a great alternator for any application, but I thought it particularly suitable to wind turbines.

Ted

[tecker]

If set up bipolar brushes the motor has a substantial release of power without discharging the drive supply into the charging cycle . Brushes release more power than Transistors or Fets so much more it's noticeable  . I would have to see some numbers on and alternator charge . I was never able to induce enough with out the motor cycle opposite attraction. However if the back emf were to be switched in as to attract the approaching  pole the induction could be affected sort of like a bucking controller and the wind rotor mechanical advantage would take out all parasitic drag. So your running at 12 or 24 and charging at 24 to 36 volts

[joestue]

It's far cheaper, lighter and easier to build than an axial flux generator, and has much better power density.

uhm, no, on all of those. with the exception of lighter, it might be but there's a reason all motors have iron in them.

However if the back emf were to be switched in as to attract the approaching  pole the induction could be affected sort of like a bucking controller and the wind rotor mechanical advantage would take out all parasitic drag

math may be required...

but yes, that design has a Very high inductance, and all that extra copper is just asking to be used as a boost converter.

[tecker]

Probably not Boost since the Back emf is in the 200 volt range off the motor  . The Back emf is scalier but with it switched you may be able stretch it out a few micro seconds . A high speed set of blades could boost the motor speed and take over the torque overall . But your still wanting to buck the voltage to use the pulse off a fast Capacitor .

[Heretic]

uhm, no, on all of those. with the exception of lighter, it might be but there's a reason all motors have iron in them.

This device has no iron because it was designed that way. If you recall your basic physics, magnet "lines of force" cutting across a conductor is what produces current. One can accomplish the same with an iron core inductor, but at the expense of producing a sine wave. This design affords the same flux density within the windings, due to the induced magnetic field and the close proximity of the magnets to the windings.
With iron core alternators you get cogging, as well as hysteresis losses within the cores. I've tried both, and this one works a lot better. 

[Heretic]

If set up bipolar brushes the motor has a substantial release of power without discharging the drive supply into the charging cycle . Brushes release more power than Transistors or Fets so much more it's noticeable  . I would have to see some numbers on and alternator charge . I was never able to induce enough with out the motor cycle opposite attraction. However if the back emf were to be switched in as to attract the approaching  pole the induction could be affected sort of like a bucking controller and the wind rotor mechanical advantage would take out all parasitic drag. So your running at 12 or 24 and charging at 24 to 36 volts

Brushes are far less efficient than a good IGBT. Not only do they create unnecessary drag, but they have a much higher resistance. They're OK for big motors, where it doesn't make as much difference, but they are lossey in any event.
Fortunately, this alternator doesn't need any brushes, just a FWBR.

[tecker]

Ok I don't see the insulated gate on the picture . The window motors I have with large wire are in excess of of 300 volts out  when run at 48 volts they have enough torque you need to keep your fingers out of the way . I'll look around for an insulated gate but right now I use brushes from a circular saw. I like to see the amperage from the air coil .

[Mary B]

So when driven with a motor what voltage and current can you produce across a battery when charging?

[Heretic]

So when driven with a motor what voltage and current can you produce across a battery when charging?

As much as you want, depending of course on the size of the windings, and the physical characteristics of the alternator. This particular unit will put out several hundred volts, which is not suitable for most battery applications. If it were wound with something like 18 or 16 awg wire, the voltage would drop and the current would increase.
Like I said, I haven't put this unit through any extensive testing. Nevertheless, it is an extremely efficient motor and works just as well as an alternator. I strongly believe that it would be an excellent match for a wind turbine.

[tecker]

I have tried aircoils inside and out good luck with that . The rotating buckeroo is a go . I'll work that out

[joestue]

uhm, no, on all of those. with the exception of lighter, it might be but there's a reason all motors have iron in them.

This device has no iron because it was designed that way. If you recall your basic physics, magnet "lines of force" cutting across a conductor is what produces current.

yes, i know that. Note however that in a well designed alternator the lines of flux don't go through the copper wire, they go through the iron and through the loop, which is what does the work. if the wire is small relative to the field then the flux going through the copper does no work, and only generates minimal eddy current. what this means is we can make an iron core that holds 5 kilograms of copper but only needs 1 cubic inch of magnet to make it work.  *though such a motor would be running close to the demagnetization limits of most magnets.

Also, please enlighten me how you make a buck regulator with the inductance of the motor, without raising I^2 R losses. (that is the whole point right?)

[Heretic]

A magnet passing an iron core causes cogging, hysteresis and heating losses. You also have to construct a massive frame to keep the magnets aligned and to prevent them from pulling into the cores. It's also a two stage process. One is the magnet transferring the current to the core, and then the core has to transfer that current into the coil.
My alternator is a one stage process. This is not an "air " core alternator. The permanent magnets are the core, and function more efficiently than a standard iron core.
The flux has to pass across the windings in order to induce current and voltage. Whether this is done through the core or directly through the windings, it still has to happen.
Magnetic current (B field) is drawn from the magnets according to the load placed across the coil. The limiting factors are the air gap between the magnet and the coil, the impedance of the coil, the size of the magnets and the power driving the armature. As the load increases and more current is drawn from the magnets, the flux density across the gap also increases. This is the only cause of drag on the armature. The magnets never have to pull away from a magnetized core.

Ted

[RP]

and function more efficiently than a standard iron core.

I'm curious if you have any data concerning efficiency

[ghurd]

This is not an "air " core alternator.

Wrong.

No offence intended.

The flux path is incomplete.
It is basically a single rotor air core PMA, even if you do not yet see it, that is what it is.

The square wave aspect is due to the inefficiencies (for lack of a better term).
G-

[NoSmoke]

Please explain how your device happens to produce a square wave.

[tecker]

Any Coil inductance would happen at the closest point to the magnet so I would guess there's a place where the poles are horizontal opposed and then fades quickly as it moves away from that point

[Heretic]

Please explain how your device happens to produce a square wave.

It produces a square wave by virtue of its geometry. It is wound in such a way that the magnetic lines of force are cutting across the windings for almost the complete rotation. Unlike most alternators, the distance (air gap) between the magnet and the windings never varies.
Polarity switches every 180 degrees, when the armature passes the poles of the coil. The "flatness" of the wave depends on the uniformity of thickness in the windings. My windings vary slightly so there are little bumps and dips in the waveform.

[Heretic]

Wrong.

No offence intended.

The flux path is incomplete.
It is basically a single rotor air core PMA, even if you do not yet see it, that is what it is.

The square wave aspect is due to the inefficiencies (for lack of a better term).
G-

Flux path does not define the core of an inductor. This alternator has one coil, which contains a permanent magnet within its core. If you put an iron shell around the coil (which I've tried by the way), it would complete the flux path but would still not define the core.
Efficiency is based on how much magnetic current is flowing across the air gap per unit of time. In a standard alternator, the distance in the gap is constantly varying due to the magnet approaching and leaving the coil. The result of this type of geometry is a sine wave.
A square wave is far more efficient in terms of power than is a sine wave.

[gotwind2]

Yep it certainly works as a wind turbine, I made something similar 3 yrs ago, but a lot smaller.
I call it a 2 pole radial flux alternator, I get a lot of requests from educational folk wanting to build it. Great for lighting LEDs.



Further info on my site http://www.gotwind.org/diy/12-inch-mini-turbine.htm

Good luck.
Ben.

[dinges]

A square wave is far more efficient in terms of power than is a sine wave.

You assume that the waves would have the exact same amplitude....

Anyway, I see lots of terms being bandied about loosely...  It would be interesting for you to measure the efficiency: the amount of mechanical power it takes to generate a certain amount of electrical power.

I am of the opinion that your generator/motor is neither efficient nor effective.

[Flux]

I suspect we are confusing efficiency with efficacy. No way will it be efficient with an open magnetic circuit and half a mile of wire, but if it happens to be a good match to a set of blades it may justify itself on a tiny wind turbine.

It seems to be similar to the Thomson-Houston arc light generator without commutator and a permanent magnet. The design fitted certain criteria for arc lighting but it made no headway in any other field, open magnetic circuits may be fine for a tiny machine such as Gotwind played with, where the iron losses of a conventional machine would have killed it, but it is limited to the low power end of things and wastes so much material.

You can get virtually any waveform by altering the magnet to coils proportions between a near square wave and a triangular wave depending on the phase of the harmonics. It is only relevent for single phase rectified and single phase has many problems so it's one considered virtue doesn't make up for its other shortcomings.

Flux