12 VS 9 Coils

[solarengineer]

Good day all, Ok, i think i've read the whole site by now and studied just about every design. I just want to ask a few questions based on my observations and see if my final direction is good to go..

It seems that 1.5KW is the upper end pushing with approx 10-11ft blades and that heat is a major factor above this mark due to internal resistance.

Any 3kw+ machines seem to be in the 24 and 48 volt range, the higher in KW rating you go the higher the voltage.

My goal is to build a 5kw machine. even the small machines seem like they would be better using 12 coils and 16 magnets/dual rotors since you can wind fewer heavier turns  in the extra 3 coils.

What are your general thoughts on this? 80% of the machines on this list seem like 9 coil systems. Just wondering if it's because 24 magnets are cheeper than 32 and system cost is lower, I realize some using brake rotors are limited by space for magnets.

But if designing from scratch with waterjet cut rotors, to shoot for a 5kw machine would 24V with 12 coils, 16 magnets per rotor, dual rotors be the better way to go heat wise with less turns/coil and thicker wire?


Thanks

Jamie

[Flux]

For higher power you need larger discs and more magnet material.

What magnets you can find at a sensible price will finally settle the number of magnets and coils.

There is nothing magic about 12 poles and for the power you are looking at I suspect that 20 or 24 poles may be a better way to go. Keep it a multiple of 4.

24v is probably about at its limit at this power, I would be looking at 48v if starting from scratch.

Flux

[finnsawyer]

Why stop at 12 coils?  With my design you can use 24.  This means the same amount of copper is spread over twice as many coils giving slightly more turns.  It also nearly doubles the rotor size, which doesn't appear would be a problem for you.  A larger rotor means the magnets pass over the coils more quickly giving a greater voltage spike in each coil.  Consequently fewer turns and heavier wire can be used to get the same voltage at a given rpm.  The fact that the voltage pulses add linearly rather than like vectors may also help give a lower cut-in rpm.  So, taking an existing 12 coil 16 magnet design and modifying as indicated here can be made to have a reasonable cut-in rpm with higher current carrying capability.  That is, it will be a higher power alternator still operating at say 12 volts.  If you keep the number of magnets divisible by 4, you can bring out a center tap of one of the phases and use that to get a 24 volt output as indicated in my diary.

If you kept the wire size the same you should get 24 volts at nearly the same rpm.  This would basically give you a machine capable of four times the power since the resistance is unchanged.  The smaller size and wider spacing of the coils also would allow for better cooling.  Using the center tap effectively cuts the resistance in half and reduces the peak voltage by twenty five percent.  I would recommend bringing out the center tap in any case just to have it available.  That requires the number of magnets be divisible by four.

A note about Faraday's Law:  Faraday's law states that the voltage induced around a loop is equal to the time rate of change of magnetic flux crossing the plane of the loop.  If you make the loop of a conducting material such as copper, a current will flow.  This means power is dissipated in the copper as heat.  As the resistance is lowered, perhaps by going to a silver loop more power shows up in the loop.  Since power is the time rate of change of energy this has to come from somewhere.  In the case of a coil passing by a magnet it comes from whatever is making the coil pass over the magnet.  By making the resistance lower or making things happen faster you can in principle get any amount of power from the coil passing over the magnet as long as you can put it in mechanically.  This is the reason for my belief that my design can provide reasonable performance with fewer magnets.  You make the resistance lower and/or make the magnets pass by the coils faster.  It all comes down to manipulating the time rate of change of the magnetic flux through the coils as well as the resistance of the coils.

[richhagen]

Now all I need is some super conducting magnet wire and I will be all set.  Unfortunately I have none, and silver magnet wire is still a bit on the expensive side for me, so I guess I'm back to the limitations of the resistance of my copper magnet wire. Hmm, I know a lady with a lot of silver jewlery, I wonder If she'd be willing to part with a bit of it for science.  Rich

[finnsawyer]

I thought of mentioning a super conductor, but thought better of it.  All I was trying to say was that by lowering the resistance of the loop, perhaps by using a thicker copper wire you can get more power out of the loop passing over the magnet.  I feel that a lot of people think there is some specific limit that imposes a certain design when this is not true.  The real limit is imposed by the wind mill itself.  It can provide only so much power at a given wind speed.  You are free to capture that power in any way you see fit.  One size (or cookie cutter solution) does not fit all.

[richhagen]

Yes, but there is a limit if using copper.  You can only fit so much into a given volume, and it has a fixed resistance.  If you widen the gap, you need larger magnets to achieve the same change in flux.  You could effect the rate of flux change by making the diameter of the magnet rotors on an axial machine larger, but there is a practical limit to that as well, and while that changes the waveform, I don't recall that it generates much more power. (there was a post on that a while back, I'll have to review it)  For most of us there probably is some room for improvement by wrapping tighter coils, or the coil or wave winding geometry, but there is a limit based upon the materials available.  Rich

[solarengineer]

Thanks everyone, im going to hit the math a bit longer!!

lots of great information!


Jamie

[finnsawyer]

Doing some "off the head" calculations I found that using eight round magnets and twelve coils with the coil diameter twice that of the magnets gave a six inch diameter rotor for half inch diameter magnets, nine inch for three quarter inch magnets, twelve inch diameter rotor for one inch diameter magnets, and twenty four inch for two inch diameter magnets.  In doubling the magnet size you double the speed or rate of change and increase the magnetic strength  by a factor of four.  This gives the potential for an eight fold increase in voltage for a given rpm.  You also increase the wire size to get the same number of turns.  If you increase the wire diameter by a factor of two you then would be looking at a sixteen times increase in power output by doubling the magnet size and rotor size, since you need twice the length of wire which has half the resistance per foot.  You could also keep the air gap constant and not increase the wire size as much to get for a smaller increase in power output.  The wild card here is what is going to be the effect on the magnetic flux by doubling the flux path when you double the rotor diameter.  If the air gap is the limiting factor then there will not be much change.  As I mentioned, one also has the option of keeping the air gap the same.  It appears some experimentation is in order.  I do believe this design will prove viable over some range of output power.  I also feel a two foot rotor is doable, which might be of particular interest to the VAWT builders, as it may allow for a lower speed operation.  Every design has its peculiarities, which wait to be exploited.