The larger diameter debate

[Devo]

There was some debate a while back about the increase in voltage out of a coil as the diameter of the rotor gets bigger with the same rpm caused by the velocity of the magnets. As I am building a "Jerry Rigged" Test windmill to try some tests of my own I wanted to give the diameter theory a straight up test.

Here where my results

I had a 17" steel disc with 8 magnets at 9 inches as in Hughs 4 foot design & I layed 8 more at 16 inches.

At somewhere between 540 & 600 rpm (hard to tell as the peice of steel hitting my finger to keep track of rpm was a little numb...) I would get 4.27 volts with 1 coil of 15 guage wire at 85 turns. this same coil on the outer ring produced 3.9 volts.

I was quite surprised that the inner diameter had a higher output , probably from the 2 magnets crossing the coil legs at the same time?

I realize the debate will be you can add twice the wire but this comes at a direct cost of copper & effecientcy.

So from this I take that it is best to keep the magnets at a spot to cross both coils at the same time than going for velocity. If I want more available output power or better effecientcy I will go with more magnets & increase the diameter as needed to suit

Just my ramblings.

By the way Ghurd if your reading this & still interested in some of those motors give me a shout atDevinH@xcelcodoton.ca I may even have a 1/2 horse permenant magnet motor with the shaft a little worn.

Devin

[Flux]

I am not at all surprised by your findings there is a big misconception due to the confusion of linear velocity and angular velocity.

The angular velocity is not changed. Average emf depends on the flux per pole (unchanged) turns ( unchanged) and frequency ( unchanged) . What has changed is the waveform of the output. With the sensible proportions the waveform is reasonably sinusoidal, when you mess the proportions it is far from it.

If you use a scope you may actually find the larger diameter gives a higher peak voltage ( but in many cases it won't),  this affects the cut in speed a bit and you may benefit to some extent with Jerry connection. If you star connect the messy waveform violates all the conventional rules that we assume ( forgetting that they only apply to sine waves) and the result with star connection is miserable.

The assumptions of star giving 1.73 times phase volts line to line is based on sine waves. When you add in the differences between rms currents into the rectifier with different waveforms and the differing conduction pattern of 3 phase and Jerry connected rectifiers it all becomes confusing.

As you say you can still gain ( particularly with Jerry connection) if you use the available space to throw more copper at it but if you throw more magnets at it and increase the angular velocity as well then the gain is far greater.

For a given set of magnets and a given speed there is little to gain by straying from what has been found to work well. If you must go this strange route then don't use star. Delta is always problematic, so for strange things jerry's method seems to be the way to go.

Flux

[wooferhound]

A lengthly discussion about the Diameter/Velocity thing

http://www.fieldlines.com/story/2006/12/31/235521/68

johnlm did a test almost exactly like your test

http://www.fieldlines.com/story/2006/1/5/23847/16961

[Devo]

Wooferhound , thanks for the links , That experiment was the same as mine & I seem to recall reading it a while back-that's how good my memory is...lol. Still like to try the things I wonder about just to fill my time , I should have used that time carving blades :-).

Flux you always leave a good read & vast knowledge-today I will look up this word in my spare time- sinusoidal so I can get a better grasp on that part.

Thanks fellas, this little 4 footer is coming together quite nicely it's amazing how much easier it gets with each build & the quality slowly starts to improve-I think a couple dozen more & I should be making some real nice ones...

Devin

[finnsawyer]

You need to use a scope.  With the greater diameter you will get voltage pulses of greater amplitude that last for a shorter time.  With the same number and size of magnets and coils you are also introducing dead times when there is no voltage out.  Your meter measures the average of all this.  You will note that in my design I introduce more coils to eliminate the dead times.  Try the eight magnets with 12 of the coils spaced properly according to my design with the coils connected in series.  That would be one proper comparison.  Since you already have nine coils, you could drop to six magnets with the proper number of coils and rotor diameter for each case.  If you do the tests this way using your voltmeter should give valid results.

[finnsawyer]

Sorry, I misread your post.  I thought you had more then one coil.  You definitely need a scope to see what the true voltage gain is.  Do not rely on RMS or any kind of averaging meter for these kinds of experiments.