Single Rotor or Dual Rotor Axial Flux PM Alternator

[arc]

I will soon have in my possession 32- 1/2"x12" cylindrical N48 neos.

They were cheap ($1.15 ea.), so I opted for getting these instead of some that would perhaps be better suited to an axial flux alternator.

My first thought was to go with a small diameter, dual rotor design, but with typically low average wind speed, have second thoughts and may opt for a larger diameter, single rotor instead.

From what I have culled from this wonderful website is that, under these, low wind, conditions a 3 phase would likely give me the best efficiency.

I'm restricted to 5'-6' diameter blades and will most likely go for 4 blades to get slightly more torque. I have a nice (works pretty well with the treadmill genny) 5' (4 blade) PVC rotor that I will try initially, but will make some (more precise) wood blades as time and money permit.

I would like to be able to maintain a charge 3- 12 volt deep cycle marine batteries that get occasional use for lighting and/or use in the RV. 60-100 watts would be great if I can achieve that, so I guess my goal is around 6 Amps @ 13VDC.

I would really like to come up with something similar to Ed Windstuff's early 3 phase, but designed with the 1/2" neos in mind. I'm thinking approx. 8" diameter 1/4" mild steel rotor which would space my neos about 1/4" apart. The coils would be circular as well (?) using either #22 AWG or maybe #20 AWG magnet wire. I figure the coils should have an inside diameter slightly less than 1/2" and an outside diameter of about 1", this should place the mean width of the coils right a the 3/4" magnet spacing.

Would 32- 1/2"x1/2" cylindrical neos be better used on a single rotor with 72- overlapped coils (3 phase) or would I be better off with 16 magnet dual rotors36 coils?

With fairly low wind conditions (8-15 mph average), I'd like to keep the cogging down, but should I consider a laminated backing for the coils, or should I just stick with an air core?

Am I in the ballpark with the spacing for these little neo cylinders?

How about a single rotor, with 16 magnets around and 16 more placed just inside the other 16 (2 per radius) and use 36 oval shaped coils? (I have very little idea of the flux this would create or if it's even worth considering).

Thanks great amounts for any suggestions, comments, ridicule or otherwise.

Dave

[Flux]

Dave

Single rotor could give more power but only consider it if you can get really low loss core material such as silicon steel or the tape core from a torroidal transformer, otherwise the losses will compromise a small alternator in low winds.

For single rotor I would be tempted to go for 16 poles 2 magnets per pole one outside the other.

For dual rotor 16 pole I would use 12 coils. To overlap coils on a stator that ought to be over 1/2" thick would be difficult and would probably end up with higher resistance.

For a thinner winding for single rotor the overlapping coils may be better.

You should get your 120 Watts at a speed suitable for a 6 ft prop.

Flux

[Gary D]

Hi arc, I have 32 1/2" cube neos sitting on a 10" brake disc. There is plenty of room for 200 turns of #24 wire Between each magnet. If you go air core, you may want to space the magnets farther apart and put a disc with no magnets on to help pull flux thru the coils? Possibly you could get more wattage even with eddy losses by going with a single stator and banding? I have yet to get anything up and running, so hopefully someone with experience will chime in here. I have't found a supplier for any better material to use. I'm thinking I need an 11" brake disc to use all 40 that I have. Don't know if any of this helps. Good luck! Gary D.

[arc]

Helpful information Flux, I do want to squeeze what I can out of these little neos, so the tape core sounds like it would be worth using should I go with the single rotor. After I posted the original message here, I stumbled across Internet Fred's site which also recommends using multiple smaller magnets if larger are not available. It would seem then, that a greater amount of induction into a lesser number of coils would be more efficient than spreading things too thin, essentially compounding the losses.

I'm glad that 100+ Watts seems doable with these neos, I appreciate the feedback and will likely be back soon with more questions as the project progresses.

Thanks,

Dave

[arc]

Hello Gary,

Is that a single phase setup using the 10" disc?

How is the #24 wire working out, I would think you'd get less resistance thus less heat using slightly thicker wire? Perhaps the difference is neglegable.

By banding, are you refering to a tape core, I'm pretty green and not familiar with the term? Sorry 'bout all the questions, I appreciate your input and wish you luck on your project.

Thanks,

Dave

(arc)

 

[Gary D]

Hi arc (Dave), the magnets are just sitting on a brake disc at the moment. I haven't found the proper materials yet to slap together a small diameter waterwheel. Too many honey do projects keep poping up. The banding material I said about is for laminants (behind the coils). It isn't the best material by far, but might serve my purposes. I have #16 gauge wire for the project, but will test to see voltages at different speeds with the thinner wire.

[arc]

Yes, it does take some wheelin' n dealin' & time (not to mention a few duckets) to keep these projects rollin', doesn't it. Still need to paint the house, get wood for the Winter, have a yard sale, change the oil in the rigs ... etc. it's a challenge, but once we get the fever, we always seem to weave a project in there somewhere. Good luck!

arc

[Ungrounded Lightning Rod]

If you're limited by the number of magnets, here's a thought...

Try doing a dual rotor but putting all the magnets on one side, as if you had twice as many.

On the other side put steel pole pieces of about the same size and shape, but at least as thick as - or even thicker than - the magnets on the other side.  (Be sure to leave a gap between them, as you would on the magnet side.)

The pole pieces are to keep the magnetic field concentrated as it approaches the all-steel rotor, rather than spreading out after it leaves the magnet faces (and thus reducing the amount it cuts the coils).

I THINK this geometry gives more field lines between the poles than putting half the magnets on each side and reducing the pole area.  But I'm not sure.  (If it doesn't, you'd be ahead to put magnets on both sides and reduce the size of the coils, reducing their resistance.)

Because the all-steel rotor is spinning with the magnets, the field is not moving within it.  So it doesn't need to be laminated to prevent eddy current losses.  (In fact, any eddy currents help keep the field in the pole pieces rather than sliding out into the gap in response to the generated current.)

And because the steel rotor has steel pole pieces you don't have to glue them on.  Instead you can hold them on with screws (i.e. (NON-stainless) steel flatheads in countersunk holes to keep the pole piece essentially solid and flat) and leave the gap between them.

The steel rotor with the gap between the poles should also act as a more effective centrifugal fan than the flat rotor with the magnets imbedded in it.  So (if you leave some room at the hub) air will be drawn inward on the magnet side, through the hole, and outward on the steel side, cooling the stator.

Whatcha' think, folks?

[Ungrounded Lightning Rod]

By the way:  If somebody is constructing a dual-rotor alternator, I'd appreciate it if he'd try an experiment:

 1) put the second rotor on once - before adding the magnets to it - and measure the force needed to remove it

 2) Then do it again with the magnets on it (and at the same gap, magnet on disk 1 to magnet on disk 2) as he had between the magnet faces and the plate in the first tes.

A doubled force to disassemble in 2) says to go with magnets on both sides and smaller coils.

[Flux]

Interesting thoughts

You are right that the flux does fan out to a flat plate. I am sure your idea will help prevent this, it may be even better to alternate the magnets and cores on each disc. If I ever tried a single rotor with a rotating keeper plate I would try it that way.

As I see it the problem with doubling the number of poles with the same magnets is that you halve the air gap to work the magnet at its best point. The mechanical air gaps between rotor and stator then use a much larger % of the available winding space.

For a simple side by side coil single layer winding I think the dual rotor has to be the way.

If the un wound spaces of the simple stator bothers you and you try overlapping coils to wind the whole circumfrence, this becomes virtually impossible with thick stators so the single rotor with twice the poles may be a better bet.

I would be interested in the results if anyone tried it this way.

Regarding your next post, this is not a fair test as you have twice the number of magnets with the second case. The pull has to be greater.

Keep up the interesting thoughts, everything is a compromise ,there is no best

way, we all learn by others experience.

Flux

[Flux]

Sorry I didn't read this properly.

I see what you are getting at, I don't trust anything to do with pull regarding magnet strength, there are too many hidden factors it might be a valid test but would be nearly impossible to do. Even with jacking screws its all to easy for the plate to tip and hit the magnets.  If anyone is to try it I suggest a soft non magnetic disc about 1/2" thick over the first magnets to prevent a crash.

Flux

[arc]

ULR,

I really like your idea, seems like a great way to get more potential out of these neos. It makes a lot of sense to channel the flux this way.

I also like the idea of incorporating this with magnets on alternate sides, as mentioned on the next post.

If I do try this, I'll drill and tap both rotors, so I can try both methods and report back to you which seems to work best. I think the key here, as mentioned below is to keep the field coils fairly thin and allow the air in to keep things cool.

Thanks for the great posts, this input really helps out.

arc

[Ungrounded Lightning Rod]

Remember that if you alternate magnet poles you need to put the same sense (N or S) up on ALL of 'em.  (Makes assembly a bit simpler.  B-) )