Alternator Design

[willib]

If anyone was wondering how to design a Three Phase alternator.
I have a way that i thought i would share with the community.


Just remember 3, 5 , 7, 9, 11, 13...
Thats odd

Code: [Select]

3    4    3
6    8    5
9   12    7
12  16    9
15  20   11
18  24   13
column 1 is the number of coils desired
column 2 is the number of poles
column 3 is the magnet DIAMETER multiplier

example 1 : Row 3 is a 9 coil 12 pole machine , with a magnet diameter multiplier of 7
example 2 : Row 4 is a 12 coil 16 pole machine with a magnet diameter multiplier of 9

Ok William what the heck is a magnet diameter multiplier ?

 it lets you select your rotor diameter and hence your whole machine.

Lets get started , i found some magnets at a good price , that are 1.5" in diameter.
and i can only afford 24 of them.
so that means i'm looking at a 9 coil 12 pole machine.(dual rotor)
In row three is a 9 coil 12 pole machine with a magnet diameter multiplier of 7.

1.5"(magnet diameter) * 7 = 10.5 inches in diameter rotor.
ok Now what??
Rule 1 Your coils shal be twice your magnet diameter.Or there abouts, But i like to make them twice the magnet diameter.
ok thats three inches in diameter. Thats doable because the way i wind coils is with resin or epoxy mixed in when i wind the coil. And they come out thnner for the same number of turns per coil as opposed to a coil with a smaller diameter.

what do we have now..
we have a 10.5 inch rotor diameter with 12 poles on it.( two of them)
we have a stator with 9 coils that are 3 inches in diameter.

How can i check if this actually all fits together ??

I use emachineshop's free online program
If your net speed is good , it works fine.
oh your magnets will come to the outer edge of the rotors.

[Bluevitz-RS]

Bump, great info, but could use a bit more.  I got an ametek motor, but decided I'd be better off building my own alternator for better power production in the limited wind that I have.  I'm only wanting to build a trickle charger to help my solar out on my tiny off grid system.

So the stuff I have as of now: I got 10--> 3/8x7/8x1-7/8 Ceramic Magnets, and a spool of 300' of aluminum magnet wire.  I know the aluminum isn't ideal but should work for my project as it was so much cheaper.

I'm thinking a 6 coil 8 magnet machine, but don't know how many turns I'll need to achieve about a 250-300rpm cut in speed at 12v nominal.

[Bluevitz-RS]

Can't anyone help me out?

Would it be better to use 18mmx3mm Neo Mags instead of the big ceramics and build a 9 coil small mill instead?  I basicly want to build functional yard art and maybe get 20W max from it.  Just a trickle charger.

[Flux]

Not really possible to help with no information.

Your 10 ferrite magnets are not going to do a lot but you could mount all 10 on  a single rotor with a blank disc to return the flux.

No details of your wire size and no idea of the rpm you hope to get or what input power you expect.

With those 10 magnets on a disc, with suitable wire and a reasonable speed you may be able to get 20w but it seems odd to buy aluminium wire to save money when in reality it may be money wasted if it is not the right size. I seriously doubt that you will make reliable connections to it anyway.

Changing to neo magnets may make life easier but again there is no point in buying magnets without having some idea whether they will do what you want

If it is yard art then we could be looking at something very slow or perhaps quite fast depending on what it is. Then factor in the fact that there is often no useful wind at low level you have a lot of uncertainties.

You could perhaps get 6 more ferrite magnets to make an 8 pole 6 coil axial alternator. By wind turbine magnet size those are really tiny but it could work.

Flux

[Flux]

Having read your previous post again you do give a bit more information.

Your 300rpm is higher than I expected and I would think that using a 8 pole 6 coil design with the ferrite magnerts could work. It would be far easier to use 16 magnets and make it 8 pole dual rotor, those magnets are really small to try to use 8 of them as a single rotor, even with a spinning blank steel disc too return the flux but it may work if you can sort out decent bearings and get things to run true.

For such a tiny thing I would tend to go to 10 pole ( using all the magnets) and use 5 or 10 coils single phase, it is too small to worry about the vibration issues of single phase.If I get time I will try to see what size disc would take those 10 magnets and try to see what winding you would need to get 12v at 250 rpm.

You would have to keep the coils thin with magnets only one one side, probably something near 1/4" so it wouldn't leave a lot of winding space.

Flux

[Flux]

Just had a quick try at this, I have had to make a fairly wild guess at the flux from the ferrite magnets but it should be near.

With 16 magnets (8 pole) on an 8" diameter disc you should be able to get in 300t of 21AWG wire for the 6 coils.

This should get your cut in and should be able to manage 20W if you can get it fast enough.

I think you can also do it with 10 magnets on a 9" disc with another blank 9" disc spinning to conduct the flux.

10 coils of about 19AWG wire should do it connected in series single phase.

In either case you will have to work hard to get enough space at the centre of the discs for the coils, I think you will have to make a hub or find something a lot smaller than the normal trailer hub..

Best of luck
Flux

[Bluevitz-RS]

OK, looking back at my ebay history, the wire I have is 18AWG 600ft aluminum magnet wire.

Do you think the smaller Neo magnets would work better than the larger ceramic ones?  They felt a whole lot more powerful on the rack.

Edit: it's no big deal to bring back the magnets as I got them locally.  So I can get more to make a dual rotor if It makes it that much more efficient.  Same goes for the smaller neo mags, I'm just stuck with the wire I have.

[Frank S]

This could be interesting to see the results of coils done in aluminum comparatively speaking that is.
 I remember when Linclon welders went to copper clad aluminum wire in their SA 200 machines. the first ones were horrible but the second generation had larger wire to get their 60% rating back this was sometime in the mid to late 60s Copper at the time was going out of sight and weight reduction for portables were also a concern .

[Flux]

In general it is easier to use neo magnets but in this case they are so small that I suspect you will need a lot more of them.

With magnets only 3mm thick you are looking at coils aboul  5mm thick at the most. Unless you have good facilities for the mechanical side I doubt that you will get the mechanical side good enough to deal with such small clearances. Without using a lot and soubleing up on the thickness I think you will struggle.

I did run a few figures and couldn't see an easy way to get this to work with 18 gauge wire. If you had a choice of wire then perhaps it would have worked.

I think the best option is to get more ceramic magnets and make it 8 pole ( 16 magnets).

Make up your rotors on 8 or 9" discs and then make a test coil. I think your best option is to use 8 coils and make it single phase, so from that you can determine your coil size.

Turn your rotor at 120 rpm ( hand crank)  and measure volts on your test coil. 12v peak will be about 8v rms so that is one volt rms per coil at cut in speed. You will be looking at about 1/2v on your test coil at 120 rpm.  You may need to borrow a reasonable meter to measure such a low ac voltage, the chep ones are not very good on such low voltages.

Try with gap between magnets about 1/2" and coils about 5/16 thick. perhaps 80 to 100 turnswill go in.

For the final thing connect the 8 coils in series and you need to reverse alternate ones. S F F S S F F etc.

Even if your cut in is above 300 rpm it may still be ok, that is slow for a tiny mill with reasonable blades.

Flux

[Bluevitz-RS]

I have access to a metal lathe and wood working equipment and welder, so fabrication and precision is no problem.  I graduated from tool and die just never entered the job market to use it.

What size should the inside of the coil be?

Also, why a single phase?  I thought 3 phase is more efficient and also give me the option to adjust output if I need a better startup voltage or if that's fine better output.

[Flux]

You can make the inside of the coils a bit smaller than the magnet, again it may give more winding space if you make the coil holes trapezoidal, you loose space at the centre of the coils and that is where they touch first.

Why single phase, simply because you are stuck with one wire size and I thought it best suited single phase with 8 coils in series.

If you use the 6 coil 3 phase arrangement you can only have 2 coils in series per phase, that and the 1.7 factor in star will probably mean using thinner wire.

Is 3 phase more efficent ? good question, unless you can wind it effectively to use the space not used in a single phase winding, then probably it isn't. The main objection to single phase for wind turbines is the vibration caused by torque variation that is absent in multi phase windings.  Does efficiency or a little vibration matter in a garden ornament?

Regarding adjusting output, I can only see the option of star or delta with 3 phase, no more versatile than series or series /[parallel with single phase, where you have 2 options.

If you can wind for enough volts your best option for changing the characteristics is by widening the gap between magnet rotors.

Flux