72 channels 3x12 coils

[zvizdic]

Search the board for info and found some.So to summarize, Stator winding is one of the key factors in windmill design. More cooper less space is the rule,thicker wire more amps less heat,heat that burns the coils .My solution 72 groves 12 coils 3 phase each coil 35 turns for 48v system wire size #16 AWG.
 http://www.fieldlines.com/images/scimages/16148/stator.JPG
 Now winding technique cw ccw cw ccw or wave windings?
http://www.fieldlines.com/images/scimages/16148/winding.JPG
Now 72 grows plate is machined. I'm ready to finish it and I'm set on cw ccw style but I'm open for suggestions. I would like to hear any comments and ideas.

My rotors are 11 1/2" with 12 magnets each (2x1x1/2).

[ghurd]

I take it you mean 12 coils per phase, 36 overlapping coils?

And a dual rotor?

I figure there will be some discussion that your coils are too large, and they should be the size of magnet.

The coils are magnetically linked, there is no dead copper, so smaller coils are better.

At least that's my take on it.

G-

[zvizdic]

Sorry !  

http://www.fieldlines.com/images/scimages/16148/stator.JPG

http://www.fieldlines.com/images/scimages/16148/winding.JPG

[zvizdic]

 Yes 12 coils per phase, 36 overlapping coils

Now winding technique cw ccw cw ccw or (  WAVE  ) windings?

What do you think about wave windings?

[Flux]

I think either way you are going to struggle with the overlap especially on the inside.

I don't think the wave winding helps much in that respect.

I agree with your theory about winding all the space but it really will be a difficult job to get more copper in there than in the usual single layer winding with 9 coils for 12 magnets.

Your full pitch coils will have more resistance than the inherent short pitch that results from doing it the other way.

For that type of winding you really need wedge shaped magnets with a gap between them of half magnet width. That way a full pitch winding works out better.

I tend to agree with Ghurd that short pitching would give a better result but whether you could wind it is another matter.

I did several windings like that, two as a 2 layer winding and one as a concentric with 3 different coil shapes. They were the most difficult things I have attempted and the end results were satisfactory but not really better than the nasty easy method.

I still prefer that type of winding for radials but never again would I try to do an axial that way.

Good luck and keep us informed on how it works out.

Flux

[zvizdic]

Magnets are surrounded with 6 1/8 spring pins and no resin only paint so I have room to play with wire on the inside of stator.Maybe thick ring on one side(inside stator)to ad some strength and fit the wire.Same thing on the outside .

Now magnets are set and no change to it. Stator plate machined so winding is a last ting to do.

I am cnc programmer and machinist so making components is no problem .

Hub is made of milling machine bearings taken from a spindle very expensive.

Oil lubricated hub spin so easily.Pics are coming.        

[GWatPE]

I have to agree with flux on overlapping coils.

My own 4phase design, with 3mm magnet plate gap, required a 12mm thickness outside the magnets to fit the wires into, that had to cross over each other.  I think the resistance of the 4phases worked out to give 30% lower resistance to a single coil.  The overlapping coils did impart more strength to the stator, but far more effort was required to get all the coils in precise alignment.

I think that you need more magnet to make the process better than a single layer without overlapping coils.  

Overlapping coil designs satisfy a need to prove it can be done.

Gordon.

[zvizdic]

Hi Gordon,

My idea was to make the coils so they're spaced and thick enough so its not going to burn. Since furling is not working or at least not in high winds and I can see that on my small windmill with DC motor I need to make stator that can take some abuse. I'm not trying to achieve thin and I'm planning 12mm thickness or 1/2" and overlapping on the outside and inside to be about 7/8 which is going to form two reinforcing rings to make stator as tough as it can be. I'm basically trying to learn form others mistakes and make something thats going to work.

[Flux]

I found similar things to Gordon. Even a thin stator has terrible problems with the end winding cross over. If you try a 12mm thick stator you will have a massive tangle of end windings.

Even in the radial case of a standard motor more copper is in the end connections than in the slots. With iron cored alternators and reactance limiting this doesn't really matter, the reactance and not the resistance determines the maximum current. The resistance mainly gives you losses and heat but the thing is very well cooled with direct heat transfer to the slots and then to the case. It has a limited current out and very good cooling. It has a good chance of survival in any storm even if furling doesn't work if you can live with the frightening helicopter noises from the blades as they self protect due to drag.

Once you remove the iron your current is determined by resistance alone and the lowest resistance winding will give the greatest output. It will also have less heat for a given output but will produce higher currents in high winds at speeds within the blade capability so you can never protect it without some form of furling or power limiting such as making the alternator big enough to stall the blades under the worst wind conditions.

At best you can only hope for about a 20% reduction in resistance by using the most elaborate winding. This is enough to give a better performance in the working region and in that respect is worth chasing. It will certainly not have any significant impact on the ability to resist burn out. There is no way the cooling can be got to match that of an iron cored construction and the only possible way to make it self protecting would be to include external reactors to limit current.

Furling can work, it does work and thousands of machines use it very successfully. There have been two cases here where there seems to be an inherent problem and I think we have come to the conclusion that it is a site problem. These have also been big machines where experience is more limited.

I know there are reported burn outs here regularly but most of these are avoidable and the machines concerned are furling at too high a power or not furling in the burn out wind speed.

I hope you can get your coils in, if you can do it you will end up with a good alternator but make sure you do include furling that works and you will reap the benefits. If you try it without furling and you have wind it will burn out.

Flux

[GWatPE]

As you seem to want to persue this, I see a problem with the small number of magnets.  The alternator could do with probably 50% more magnet volume, with the appropriate change to the number, or shape of the windings.  This will reduce the amount of copper that is not in the flux path as well [going for the ride].  This will up the total flux through the coils per revolution, with a lower overall resistance.

I would still wind for a higher cutin rpm, probably around 160-180rpm, and use a voltage boost, to recover the low rpm cutin.  This will give even lower winding resistance.  

As Flux reitterates, top end control is about power limiting.  Furling, or dynamic electrical braking, or blade feathering have to still work, to ultimately protect the alternator from burnout.  It is possible to go completely over the top with the alternator, so it can never burn out, using 2-3-4x the magnet volume, and coils to match, but we still have to lift this unit into the air, and maybe even service it without the need to hire a crane, or have a mobile gantry on site.

Gordon.

.

[zvizdic]

Flux respecting your comments as one of biggest contributors !

Last night gust of wind 70km/h prove to me that furling is not working .

In lover winds working well no argument there .

I am positive in high winds from what I witness prop is strait in to wind and tail is like welded not moving at all .  

So eider furling or propeller redesigning or brake(I do not like).  

[Flux]

Yes perhaps your furling is not working, from your description I suspect you have something seriously wrong with it. In the other examples that failed in high wind the tail was well over towards the full furl position but the blades were still into the wind.

If your machine is straight into the wind and the tail is still straight back then something is not right.

Making a more powerful alternator will increase the wind speed at which you need to do something to protect things but you will still have to brake it or reduce power in some way. You may be able to hold it stalled longer but I seriously doubt that you have even half enough magnet to hold a 10ft machine stalled in very high winds. Those magnets are only just adequate for a 10ft machine with a normal winding and furling working. Your change of winding may get you another 5mph wind speed before you need to worry.

I am not even sure you have enough magnet to brake it from run away in very high winds but if the brake switch is very close it may be ok.

If my conversion from metric is right, 70kph is not very high. The problems that others have had have been at near twice this wind speed. I don't have details of your present set up but I strongly suspect that it has never furled at all in any wind speed.

Yesterday I was looking at a video from Hugh Piggott on the windmills on Scoraig,(you can see it on Hugh's web site http://www.scoraigwind.com/) there are dozens of machines that have been running there for years, all rely on furling and it must be one of the windiest places on earth. It certainly can work.

Flux

[Ungrounded Lightning Rod]

Non-overlapping windings are about trading extra magnets for simplified construction.

Completely paving the space between the magnets means that you're getting the most out of the magnets - but you waste a bit of it by having the current go through extra copper as the portion of the windings that is NOT between the poles has to bend and go extra distance to avoid another winding it's crossing over.

A simple compromise is to use easy-to-wind flat coils, still completely filling the space between poles (excluding mechanical clearance gap) where you DO have coil, but "wasting" the portion of the time the magnets are over empty cores and making up for it by buying a few extra magnets.  B-)

Ideal is for the area between the poles to be full of copper and the spacing between the poles to be about the sum of the thickness of the magnetic material of the opposing poles.

Widening the gap between the poles to stuff in more turns means generating more voltage due to extra turns but losing voltage per turn due to weaker field.  There's a maximum where the gap is about the thickness of the magnetic material.  Wider and you lose more voltage from losing field than you gain with extra turns.  Narrower and you lose more voltage from lost turns than you gain from stronger field.  Curve is a hump with low slope near the peak (and necessarily zero slope at it, as with any max or min on a continuous function) so the parameters are not critical.  Just ballpark it.

Amount of power generated in a given field is proportional to that copper cross-section.  Just as with transformer windings, you can cut the cross-section up very fine with lots of turns of thin wire and get a high voltage and low current, or cut it up thick with few turns of heavy wire (or many-in-hand, which is better) and get lots of current at a low voltage.  Power is product of current and voltage so it stays the same.  (Caveat:  If you get TOO thick, like by winding it with copper bar or pipe, you lose some power to eddy currents.  Winding N-in-hand with something thinner that adds up to your desired cross-section avoids that - and also is easier to bend.)

So by tweaking the wire size and still filling the same cross-section you get to pick your voltage-current tradeoff to tune the voltage/RPM to your charging circuitry, blade radius, TSR, and wind regime, without affecting the total power the device can generate.

[zvizdic]

OK my main goal is to spread coils equally around stator and hoping to distribute heat to all area of stator.Yes about 10-15% more copper but I am not looking to mace it simple but rather more efficient.    

[GWatPE]

Here is a pic of my own stator, and one of the phase coils prior to moulding.

I hope this works, as have not had much luck with picture before.


Gordon.

]