Contactless dynamo for my road bike - help with the magnetics please

[andyboy]

Hi all,

Would appreciate some help with the magnetics in a contactless dynamo I'm trying to build for my road bike to power some LED lights.

I've got 8 rare-earth magnets mounted on the bike wheel, near the hub. They're flat, round magnets, around 2cm dia., mounted all the way around the wheel. I've then got a coil (approx 5000 turns) mounted on the fork with a ferrite core, which is perpendicular to the magnets, and as close as possible to the magnets. It kinda works, but not as well as I hoped. I'm going to try a bigger coil, but I have some questions about my set-up.

I understand that I'll get the most induced voltage in the coil when it cuts through the most lines of magnetic flux. To do this, I've got the magnets closely spaced (almost touching), and of alternating polarity (so the coil alternately passes a N pole, then an S pole, etc, as the wheel spins). Is this correct?

 

[jim hardy]

andy

the magnetic path includes the return path for flux from other end of your core to opposite side of your magnet.
Is that in iron also, or in air?

you need the air part of path to be as short as possible.
Study at the coil on a lawnmower engine- it has two ends that are bridged by the magnet. So the air path is only the clearance, not clearance plus distance between ends.
I suspect (but do not know from description) your setup includes a reurn path through air. Or is your Ferrite core C-shaped?

 

[NascentOxygen]

Need a couple of photos.

 

[maimonides]

You have two main problems: the "back sides" of the magnets (you need ferrite/iron there too) and the size of the air gap.
If you want a working dynamo, buy a dynamo hub, the cheaper ones are quite affordable (even as complete front wheel). You´ll spend less than building one.
If you´re in for the experience: you´re welcome.

You can find advanced (and expensive) hub dynamos at http://www.nabendynamo.de/english/index.html

 

[andyboy]

Hi,

Thanks for the replies, the info has been very useful and I've revised my design. Don't have photos yet, but hopefully soon I'll build a prototype and put some pics up. Until then you'll have to use your imagination. So my new design has a plastic disc mounted to the hub on the side of the wheel (imagine it mounted in place of a disc brake, it's not where it'll be, but it's close), with magnets mounted around the edge with alternating poles. The generator coil will be wound on a U-shaped core, with the prongs around the disc, and fixed to the bike frame.

As for the expense, I think I have most of the parts I need, and it's cost less than a tenner. I don't think I'll get the output of a proper dynamo system, but it'll be fun to try.

 

[jim hardy]

""I don't think I'll get the output of a proper dynamo system, but it'll be fun to try.""

the lights may flicker as magnets go by.

you've seen those flashlights that you shake to make go?
they have a supercapacitor to smooth out light between shakes.

 

[maimonides]

Sounds reasonable now.
Even commercial units have flickering lights at very low speeds (pushing the bicycle).
The voltage may rise uncomfortably high (for the LEDs) at high speeds; rectification and a supercap will help a bit; but you may need overvoltage protection. (depends on the LEDs you use resp. the driving electronics).
Don´t expect much power, most commercial units (in Europe) are 3W.

 

[NascentOxygen]

I've got the magnets closely spaced (almost touching), and of alternating polarity (so the coil alternately passes a N pole, then an S pole, etc, as the wheel spins). Is this correct?

I'm undecided whether this is the best arrangement or not. By having the magnets almost touching, it seems you might be losing some of the field lines into adjacent magnets before the lines can cut across your coil. Maybe a gap between the magnets would be better? Others might be able to comment. Otherwise, you could experiment by adjusting the magnet spacing.

 

[andyboy]

I'm undecided whether this is the best arrangement or not. By having the magnets almost touching, it seems you might be losing some of the field lines into adjacent magnets before the lines can cut across your coil. Maybe a gap between the magnets would be better? Others might be able to comment. Otherwise, you could experiment by adjusting the magnet spacing.

This is something I'm not sure about, and was hoping someone would be able to give an answer. I'll have to do some experiments.

 

[sophiecentaur]

I think the best way to get maximum flux through the core of the coil would be to to use a U shaped core which bridges between adjacent N poles and S poles with very little clearance between magnets and core as they go past. The reluctance in the N-core-S path needs to be much less than the reluctance of the n-air space-S path. I should imagine that the magnets should have their 'other' poles held on an iron ring round the outside to increase the flux at the business end.

 

[andyboy]

I think the best way to get maximum flux through the core of the coil would be to to use a U shaped core which bridges between adjacent N poles and S poles with very little clearance between magnets and core as they go past. The reluctance in the N-core-S path needs to be much less than the reluctance of the n-air space-S path. I should imagine that the magnets should have their 'other' poles held on an iron ring round the outside to increase the flux at the business end.

I follow all of that apart from the last bit about the iron ring, not sure where it would go in my design. I've made a rough sketch of it, hope it makes things a bit clearer.

Apologies for sideways-ness.

 

[sophiecentaur]

I get it now. I was assuming a design like the old dynohub with the coil inside and an outside drum of magnets.
My point about making sure that the gap between the magnets and the core must be much less than the gaps between successive magnets. There will be a compromise between maximising the number of magnets and keeping the gap ratio large enough.
Looks a fun project, though. Save yourself loadsamoney.