Simple generator

[gyulasun]

Hi Broli,

Perhaps you could ask the Magnetec people what or how they suggest the cutting should be done?  Diamond or maybe laser cutters are preferred??  be very careful and maybe look for a good machine shop once you know the how to.

Gyula

[broli]

It has been a while since the birth of this concept, sadly nothing much has happened since.

However I'm trying to get the ball rolling again, or better yet, the generator spinning. I got rid of the idea that using a toroidal core might do the trick. The Achilles heel of this concept is unwanted airgaps, and when you coalesce multiple materials together you are playing with so many variables. Hence the new idea to stamp a custom stator core to reduce bridging materials. It makes the generator even more simple I believe as you only need one airgap to worry about and it's on the outside too.

I sort of reached the limitations of what simulation software has to offer. The results don't like to converge and become almost meaningless when reaching small airgap sizes. That and the simulation time increases exponentially with element size.
However in general the torque does decrease the smaller the airgap gets. Femm being the most stable simulation software confirms this fact.

Currently I sent a few quote requests to some Chinese suppliers on alibaba for custom rotor laminates as attached per pdf, to ease the study of the above.
Any other suggestions are welcome as it seems like Europe no longer does anything in regards to stamping motor cores.

[broli]

While waiting on responses from the couple core stamping companies, one would like to charge a wopping 3500$ to create the mold alone, in china I was thinking of cheaper ways on building a generator out of the same principle. One way that came up is using cheap EI audio transformers that can be bought everywhere. These can then be arranged in such a way to form a circle. Certainly it has its own engineering challegens but this is a much cheaper alternative than stamping custom laminates.

Here's a video showing the concept: https://www.youtube.com/watch?v=K_NRIuTrjrQ&feature=youtu.be

I also added a 2D simulation result to show that indeed most of the flux going through the transformer is going through the leg that the magnet is covering.

[kEhYo77]

Hi.
I've been thinking about that type of a generator as well.

[gyulasun]

Hi Broli,

You mentioned audio transformers as off the shelf components and this remembered me to common mode line filters or chokes used at the mains input of many switch mode power supplies. There are such chokes with two ferrite C cores back to back or two ferrite E cores also back to back, usually with a little air gap between the facing surfaces. See some here and these are available in most component shops, I chose Digikey, they have several mechanical sizes with several coil inductance values.

One with C core, horizontally mountable: http://www.digikey.com/product-detail/en/B82733F2232B1/495-2787-ND/1243551

With E cores, horizontally mountable: http://www.digikey.com/product-detail/en/EH20-0.3-02-33M/817-1000-ND/1928579 and
http://www.digikey.com/product-detail/en/EH24-1.0-02-10M/817-1009-ND/1928588

With E cores, vertically mountable: http://www.digikey.com/product-detail/en/EV35-2.0-02-20M/817-1054-ND/1928633

These chokes have double windings next to each other so you could connect them in parallel or in series, or use each separately in case of the C cores.

I am a bit surprised that the simulation shows most of the flux going through the leg that the magnet is just covering, I would have expected more flux going via the side legs because those side legs actually shunt the center leg.  It is true though that the side legs each has half as much cross section area than the center leg does (but the two side legs alltogether have equal cross section to that of the center leg).

Maybe the C cores (that would have two legs only vs the 3 legs of the E cores) could also give useful induction, as per common sense, because then there would be two shunting flux pathes vs the three.

So I think these common mode ferrite chokes would serve well to build a cheap low power prototype from this setup, it would not need expensive lamination cuttings.

Gyula