# Designing/Predicting Output from Permanent Magnet Motorcycle Stators

I'm hoping you can help me. I've long been into motorcycles, and have restored and customized many different bikes. While building custom motorcycles, I often add electrical accessories, and use high output stators (if available) to keep up with the increased electrical load. I've hand-wound many stators myself to increase output, and have a good feel for how changing wire gauge, number of turns, and delta/wye connection type (for 3 phase stators) will affect output vs. motor RPM. However, I often have to build multiple prototypes to get the correct configuration.

What I would like to be able to do is learn how to predict the output from a particular stator winding configuration, by either doing the math, or simulating using computer software. I've looked at FEMM, a free magnetics software package, though I have not learned how to use it yet. Any other free options?

For those of you unfamiliar with motorcycle charging systems, they are fairly simple. I have a feeling many of you are familiar with AC electric motors, which are similar. A quick crash course in the components below, I'll be using a 2006-2009 Suzuki GSX-R600 as an example for the parts:

Most modern motorcycle engines use a permanent magnet flywheel mounted on the end of the crankshaft, spinning at engine RPM. Here is an example:

It has embedded fixed magnets, alternating North/South poles along the inside edge of its OD, spaced evenly with the poles on the stator.

The stator is then mounted centered inside the flywheel, allowing the flywheel to rotate around it. Here is an example:

The stator cores are made up of thin stamped laminations, riveted together to form the stack. Most modern motorcycle stators use a three-phase winding configuration.

The most common variables to change or increase output are the wire gauge, number of turns per pole, and connection type (Delta/Wye for 3-phase stators).

What I would love some help with, is how can I measure the fixed factors (magnet strength of the flywheel, size/material of the stator core), and then calculate or simulate the variable factors (wire gauge, number of turns per pole, connection type, etc) to come up with an accurate new winding configuration. I would prefer to do the work up front on the computer rather than building multiple prototypes to test.

Thank you so much in advance for any help, insite, and info!

• V=4.44 NfBmax. f freq. N=turns. B=flux. [Note that 4.44 = sqrt(2) x Pi] || Imax is limited by coil resistance and stator magnetic saturation. Flux can be measured with a flux meter / hall cell / ... || Airgap decrease increases flux. Pole field with good grade rare earth magnet can reach 1T at about magnet thickness /2 airgap. eg a 6mm thick rare earth magnet can give 1T at about 3mm airgap. || Delta Wye: read it up. Delta Vout = coil voltage. Wye Vout = 2 coils at 120 degrees. Nov 2, 2016 at 9:51
• @RussellMcMahon Urms=4.44 fNABmax. Where is your cross section area? Jun 26, 2017 at 11:19
• @winny Excellent question. It's still in my head now in 2017, so where it was when I wrote that about 8 months ago I cannot imagine. I'll repost with Area added. Thanks. Jun 27, 2017 at 12:20
• V=4.44 NfABmax. f freq. A core cross sectional area. N=turns. B=flux. [Note that 4.44 = sqrt(2) x Pi] || Imax is limited by coil resistance and stator magnetic saturation. Flux can be measured with a flux meter / hall cell / ... || Airgap decrease increases flux. Pole field with good grade rare earth magnet can reach 1T at an airgap of about (magnet thickness)/2. eg a 6mm thick rare earth magnet can give 1T at about 3mm airgap. || Delta Wye: read it up. Delta Vout = coil voltage. Wye Vout = 2 coils at 120 degrees. Jun 27, 2017 at 12:23
• @RusselMcMahon Even the sun has its spots. -Swedish proverb. Jun 27, 2017 at 16:21

Useful link on similar designs : the LRK Torquemax "outrunner" motor is essentially identical to your generator,, both mechanically and electrically, but used as a motor instead of a generator.

There are pages here on theory as well as construction and winding.

Places to start:

Volts = 4.44 NfABmax.

f frequency. N=turns. A = core cross sectional area. B=flux.
[Note that 4.44 = sqrt(2) x Pi]

Imax is limited by coil resistance and stator magnetic saturation.

Flux can be measured with a flux meter / hall cell / ... .

Airgap decrease increases flux.

Pole field with good grade rare earth magnet can reach 1T at an airgap of about (magnet thickness)/2.
eg a 6mm thick rare earth magnet can give 1T at about 3mm airgap.

Area A of coil or ~= of core. More area = longer wondings per given turns = more resistance and larger overall machine size.

Delta Wye: Read it up. Lots on web.

• Delta Vout = coil voltage.
• Wye Vout = 2 coils at 120 degrees.

Trial design:

If your windings are 2 layer they appear to have about 30 turns.

Assume pole piece is 10mm x 10mm - you can measure this. A = 0.01m x 0.01m = 1E-4 m^2

Assume 1 T at pole surface - although you MAY get 2+ times this - and will probably get less. Deep-ends on magnets used, airgap, ... .

Try at f = 1800 Rpm = 30 rps.
V = 4.44 N f A Bmax = 4.44 x 30 x 30 x 1E-4 x 1 = 0.4V per coil.
There are 18 coils. They are probably 3 phase connected (see 3 yellow wires AND usual practice).
6 x 0.4 = 2.4V per phase = too low.
Double revs = 4.8V.
Double core side = 4 x area = 19V
Change ....? :-).

The above are per phase voltages.
For delta connection Vout = Vphase
For Wye connection, vout = Sqrt(3) x Vphase. So Vout in above example = 1.732 x the values given.

ie the above values are "in the order of right" and the final result depends on changing the above assumptions to actual values.

• @ Russell McMahon.Double check your calcs .Your volts seem too low Remember that the electrical machine has lots of poles . Aug 17, 2019 at 11:54