There are three main components--an input magbrake, an output magbrake, and an oval racetrack.

Each magbrake is simply a set of two permanent magnet discs on a shaft with a narrow gap between them.

The oval racetrack runs along inside the narrow gap, like a belt running between two pulleys. Within this racetrack, aluminum washers are free to roll around.

As the input shaft rotates, it induces eddy currents in the washers, propelling them up to speed. These coast along the racetrack until they encounter the output shaft, where induced eddy currents cause them to transfer momentum and energy to the output shaft.

The gear ratio is adjusted by shifting the oval racetrack's position. Moving it toward the output shaft increases mechanical advantage while moving it toward the input shaft decreases mechanical advantage.

Compared to mechanical friction CVTs, this CVT is very simple. There are no twisting or narrow frictional surfaces, nor does gear shifting involve fast rotating parts. Lubrication can be extremely simple, using a totally enclosed racetrack with a small amount of oil within.

There are some energy losses in leaving/entering the magnetic fields, so it may be desirable to use electromagnet loops to adjust the radius of the magnetic fields. This increases the cost of the CVT, of course.

Also, this system works best at high RPMs. It may be desirable to combine this CVT with a traditional geared transmission to translate high output RPMs to lower speeds. The output magbrake can be used as a clutch by adding an electromagnet.
-- IJK, Mar 26 2007

Magnetic Inductance Clutch Magnetic_20Induction_20Clutch
my similar idea [discontinuuity, Mar 26 2007]

Baked (so boners eat your bones) http://www.magnomat...y/Magnetic-CVT.aspx
Magnomatics CVT magnetic gear [pashute, Mar 01 2011]

magentic transmission http://spectrum.iee...otors-goes-to-china
Motor inside wheels. [travbm, Oct 29 2015]

You're not going to get much torque at the output, but I like it.

If the racetrack goes around *outside* the input and output shaft, you could have another racetrack that goes around *inside* the input and output shafts. This would increase power transmitted.
-- xaviergisz, Mar 26 2007


Check my link to a similar idea. Also, I'm not really sure I understand this whole "eddy current" thing, but I know next to nothing about electromagnetics. That's next semester for me. How do the aluminum washers affect anything?
-- discontinuuity, Mar 26 2007


are the washers and racetrack necessary? couldn't you just have a rotating disc with discrete aluminium elements (the disc can be moved toward input or output shaft).
-- xaviergisz, Mar 26 2007


Having a fixed disc means that their path of motion is circular. This will have worse losses since the angular rotation will be more highly non-constant around a shaft.

The low torque is one reason why you really need to operate it at high RPMs. Power is equal to torque times angular velocity, so if you're limited to low torque the way to get high power is high RPMs.
-- IJK, Mar 26 2007


many Speedometers use a similar technique to decouple the speedo needle from the cable(for smoother speed indication)
-- jhomrighaus, Mar 26 2007


It took some imagination to understand how this works, but if I'm right, then the washers act sort of like a chain, and how they interact with the magnetic end of the shafts is like a sprocket with no fixed radius, so the radius of either input or output can be adjusted at will. It crossed my eyes, but I got it.

Those washers are gonna get hot. You might oughta circulate that teeny bit of oil, and cool it.

Your shafts will never lock. When slip goes away, inductance goes away, and your system stops working. Bummer.

The geared output is right: the first CVT I ever read about had about a 6:1 rearend, but over a 5:1 gear spread in the CVT with a .5 overdrive at top, it worked great. I could see how you could get something like that out of this.

I'd bone it for the necessity of constant slippage, but I'm bunning it because I like madly whirling magnets. Very unique app.
-- elhigh, Mar 30 2007


I think ferrite magnets would be better in a magnetic field.
-- travbm, Oct 29 2015



random, halfbakery