Best coil geometry for round magnets

[scoraigwind]

I am depressed that Chris still says that rotor stalling is 'horse hockey'

Well, I have to apologize Hugh, because many times I don't explain myself well.

..  There is no doubt you have to wind and build for peak efficiency at the average wind speed on your site and accept the compromise on either side of that.  But that's why this thing flux has is so interesting to me.
--
Chris

Thanks for the apology, Chris and I will take it that you do understand the importance of speed matching then.  And how much harder it gets as the alternator gains in efficiency.

Star/Delta switching seems like a simple solution on the face of it, but the snag is that it only suits very low and very high winds, whereas most of the useful energy comes from the ones between when the star/delta switherer turbines is not working anywhere near to its best.  The 2 speeds are basically too far apart.

[ChrisOlson]

Star/Delta switching seems like a simple solution on the face of it, but the snag is that it only suits very low and very high winds, whereas most of the useful energy comes from the ones between when the star/delta switherer turbines is not working anywhere near to its best.  The 2 speeds are basically too far apart.

As I have well found out with my 13 footer.  I'm going to call that configuration I have in that thing a dismal failure at this point.  It puts out exceptional power in high winds - I've seen it push well over 2 kW.  And at low winds it's very nice because it puts out power when the other ones are spinning but doing nothing.  But in between, where the wind blows every day, it just doesn't work right.
--
Chris

[scoraigwind]

I found this picture of the best coil shape for a round magnet!  I wonder if this answers the original question...

I wonder if I can upload it...  We'll see.

[dciolek]

Thanks Hugh.  That is a very helpful picture to go with my original question.  Attached is my first test coil with the trapezoidal configuration I was playing with.  Its obvious my coil jig needs some work because the hole on the inside is TOO BIG.  This is a tester only, and has 56 turns of 16 AWG.  I also have a 100 turn circular coil I just wound yesterday with a slightly smaller interior hole diameter (1 7/8).

I think I can get enough copper in the stator segment if I decrease the trapezoidal hole in the center and keep winding away until I stuff the available space in the segment like seems to be shown in your picture.  Picture of my rotor doesn't look like the magnet spacings are any closer than what is shown in Hugh's sketch.

[scoraigwind]

It's not my own work to be honest.  That sketch was from a guy pedro bernardo I correspond with in Portugal.  But for me it ticks a few boxes.  It tries to enclose the whole magnet area, but if it did so (as yours does) then there would be hardly any copper in it.  The legs of the coils are wider than yours to make room for more copper, giving more turns of wire or thicker wire as required.  Those extra turns miss some of the magnet face or even in the outer turns bring in some faces of magnets of opposite polarity, meaning that overall each of those turns will produce a bit lower voltage, but the 'cancellation' effect will mostly occur close to the 'zero crossing' where the coil is centred on the magnet anyway so in fact this sort of compromise works pretty well.  I prefer to use a larger magnet rotor, but as DanB has pointed out you can get reasonably good results by cramming magnets close like this in fact.

[Boss]

While most of this discussion is well above my level, I'm going to build my third stator very soon and hope to learn the basics this time instead of pulling numbers out of the book.
My question is about the trapezoidal coil configuration. How is this better than round? Or the shape found in the book for that matter?
I see on my stator that coil proximity is relevant, as I have heat marks between the coils. Would a flattened adjacent coil edge help keep the heat down between coils?

Sorry to newbe question here among the "heads," but I am trying to understand, I can see that it would behove me to get better shaped coils more accurately placed in my next stator.I am also wondering if the heat marks are an indication of magnet path and power generating zone? 

[Flux]

All this seems a long time ago, I have forgotten what this was all about.

I assume you are using round magnets with the stator in the picture. If so I think the big improvement you could make is to use all the space and wind with thicker wire or more starands so that the coils touch. if your turns were correct before keep them the same.

Coil shape is not so important. If the magnet spacing is ideal then round coils would be better but I suspect you are using fairly closely spaced round magnets and the elliptical coils that you used are a good compromise. making the things trapezoidal may link you a bit more flux but the turn length will be greater and you will have more resistance. It's all a trade off, the ideal coil shape only applies to a single turn. With real coils there is no room for all the turns to be ideal and it becomes a compromise.
 
You look to have wasted a lot of space that could have used more copper, that really is the only way to get more out . The heat is due to more current than the wire can handle, it has nothing to do with coil shape, all losses are resistive, nothing to do with coil shape, cancellation or anything else as long as your wire is not thick enough for eddy loss in the wire itself.

Flux

[shawn valpy]

I read this thread from the beginning to end (very informative )
And have also read on anouther fourm about useing cap doublers to harness low wind power and would be interested on your take on them the idear was to wind your coils for late cut in then use capasiters to get the low wind power which would avoid star delter switching. sizeing them corectly so as not to stall your blades then when the wind cranks along the alt is matched better !!
 

[Flux]

The voltage doubler circuit works well with high frequency alternators such as the F & P motor.

I am sure the basic concept works for the axial machines but the size and cost of the capacitors may make it less attractive. My personal preference was to wind for a cut in about 12mph and use a boost converter in low winds. This is far better than star/delta as it is proportional and with star/delta the match is usually poor in the low wind condition. The big problem with star delta is the time it spends in the wrong state on the average day, we don't just have high wind days and low wind days ( not here anyway)

I haven't tried the capacitor scheme so can't comment, but I am not a fan of electrolytic capacitors on ac, I think you will need a lot of capacitors to get reliability for many years.

I originally used a single mosfet boost converter in the dc line and that worked fine but I did later try some experiments with a mixed rectifier bridge of 3 fast diodes and 3 mosfets. This looked to be an even better scheme. I went on to a buck converter scheme that was even more promising so I didn't get any detailled results from the mixed bridge boost but I think Bergey's XL1 does that or something very similar.

Flux

[kitestrings]

Healthy discussion and mix of things here.  It took me awhile to find time to get through it all.  I had one thought regarding the orignal "rules".

In dciolek's case he's working with a set rotor diameter, and I understand his desire to use it if possible -but otherwise would the simple math suggest:

[pi * d - (mag d * #poles) ] / #poles] = space between magnets, in his example assuming the magnets are at the outer-most edge of the rotor:

[3.14 * 11 - (2 * 12)] /12 = .88", or about 7/8".

Others here could could, or have, suggested a minimum dimension between magnets before leakage is unacceptible.  To me, staring with the rotor limitations and woking toward the best stator design in this case makes sense.  I think this is what Wilbr was suggesting.  If you have a rule regarding minimum clearance between magnetics (e.g. half the magnet diameter), it would seem to define at least the minimum rotor diamter.

Am I missing something?

~kitestrings

[kite_flyer]

hi.

i think that it would be handy if the legs of the coil were radially in line so that the magnets spinning cut the coils
as close to 90 deg. as possible. and maybe the coils could be no wider than half the magnet diameter. make the coils
slightly egg shaped (to fit the radial legs of the coil) also reducing coil resistance. and the hole in the coil can be slightly
smaller than the magnet. cant remember where i read this, you could use CAD ( cardboard aided design) to lay it all out
to see how it goes. 



hans

[tecker]

Straight legs tangent to the outermost arc of the magnet will get the best charge . A good mechanical action of these magnets is they reduce buck a bit . The arc of the magnet
reverses the charge some what more on a 20 degree angle .

[fabricator]

( cardboard aided design) to lay it all out. 
hans

That is an Olson ism. 

[Boss]

Thanks Flux
I am going to change out the 1/4 inch thick plywood disk which makes up the outside plate of our coil winder, as it was warping under the stain of the 24 gauge wire. That caused the coils to be thicker than would fit in the 5/8th inch stator mold, apparently the coils were also less spread out because of this defect in coil winder design. I'm thinking a steel disk would create a more uniform coil.
I've been reading about creating a test coil. It makes sense, but I am unsure how to hold it in between the moving rotors

[Flux]

One way is to make a hole in a piece of plywood and stick the coil in it with some contact adhesive or superglue or something. you can then clamp the plywood to hold the coil central between the magnets. Just to measure volts it doesn't take a lot of holding. It will jump about if you short the coil but measuring short circuit amps is near useless anyway.

Flux