250W Axial Flux Alternator Tests

[SparWeb]

My drill press was very useful for testing this alternator.  The hub's cylindrical dust cap (they're usually domed) allows a belt to be placed around the end of the hub and the chuck of the drill press is able to turn it.  No pulleys required.  The mount was cobbled together a bit to mount in the vise, hold the hub, and hold the stator in position.  You can see the bicycle spedometer in the lower center which gives me a RPM reading.

The ½ HP motor of the drill press turned the alternator well in most tests.  In some tests, the power of the press was not enough to maintain speed.  Despite the press' inability to turn the alternator faster than this, the charging current curve up to this point was straight enough to expect it to be constant up to the stator wiring's rated current (about 20A).

With airflow on the stator, its encapsulation in an epoxy board, and the thinness of the stator, it's hard to tell in advance whether the stator will be able to deliver more or less current than that, without burning out.  Only experience will tell.

Resting with the 20A maximum current assumption, then the alternator can deliver up to 250 Watts charging a 12V battery.  

Star Cut-in is at 150 RPM and rated speed is about 550 RPM.  

Delta Cut-in is at 280 RPM and rated speed is about 680 RPM.  

I haven't computed efficiency yet, but it was immediately obvious that the Delta connection required much less torque to turn than Star, as evidenced by the drill press pooping out at 240 RPM when turning the genny in star, while in Delta I got it up to 400).

(No, belting up the ratios faster doesn't help.  When you're at 100% power, increasing RPM just lowers the torque - the belts either slip or slow down).

Oh, yeah, stats:

Coils = 12

Phases  = 3

Wire gauge = 18

# turns per coil = 56

Stator thickness = 3/8"

Magnet size = 1" round 1/4" thick

Magnets per rotor = 16

Air gap = 1/2"

Magnet circle = 7.5" diameter

Coil size = 1.75" OD x 0.75" ID

Encapsulation: waterproof Epoxy compound, with fiberglass laminations

Warning: do not mix the epoxy aggressively - I accidentally introduced too many air bubbles into my mix.  If I had known that there was enough unmixed epoxy left over to start again, I would have.

In the real world, I will need to space out the rotors because there is no way an actual windmill can tolerate 1/16" spacing between rotors and stator as it yaws this way and that.

Thanks again, everyone for the inspiration, feedback, and for sharing your own successes & failures.

[Flux]

Nice results.

Power out would suit a 6ft prop. Bit slow in star for 12v, just a bit fast for 24v.

Looks to be a perfect candidate for 12V star/delta or 24V star with boost converter.

Remember that in delta you can handle 1.7 x current in star for same winding current.

Delta will be easier to drive as the torque is reduced. If you devise a way of measuring input power you will find star is more efficient, but delta is perfectly good enough.

If you do think about switching then star/Jerry is easier to do and may be a bit more efficient. If you have decent bearings I don't see why you can't manage with 1/16" air gaps on a machine that size.

Can I reply to your other post here about cogging and starting torque on the motor conversion. You have almost devised a way of measuring iron loss and with your tacho you can do it.

If you set the tacho to measure rpm, you have the torque from your pull test. Using shaft diameter and pull you can get torque, then with speed you can measure power in.

It would be very interesting to add a pulley the same diameter as the hub on the axial so you can do a no load power comparison on the 2 machines.You may need a long bit of string to avoid confusion from the large inertia of the axial machine.

Flux

[SparWeb]

Flux,

"...almost devised a way of measuring iron loss and with your tacho you can do it."

OOOOh, another test!  I am filled with excitement!

Thanks for the delta current observation.  I remember reading that before, but of course I didn't remember to remember it until now.

Thanks.

[WXYZCIENCE]

SparWeb, the bicycle speedometer, how do you get an actual rpm reading from mph? You talk about increased spacing between the rotors this should also allow you to adjust the blade vs alternator match. Progress looks good, carry on.


Joseph.

[ghurd]

One way, set the speedometer to kph, set the wheel size to 167cm cir.

A reading of 12.6kph = 126RPMs.

G-

[SparWeb]

G beat me to it, and that's exactly how it's done.  It's a ridiculous wheel size (unless you own a penny-farthing), but the computer seems to accept it.  It may ask for milimeters, so 1667mm is what I entered into mine.  The sensor is a reed switch so you need to mount a magnet on a rotating part of the rotor.  I use a ring magnet on one of the hub bolts usually.

[willib]

sparweb those 1"dia x 1/2" thick  magnets a pretty powerfull and cheap you could double your power , i bet

$1.38 ea.

[SparWeb]

Will, that's unlikely.  The rotor plates are 1/8" thick.  If I put bigger magnets on I would lose a lot of flux out the back.  There might be a lower cut-in speed.  Given the size of prop I'm going to put on, I think I already have well matched prop speed/cut-in speed/top speed balance.  One of those things I haven't gotten around to checking yet...