DIY Contactless Magnetic Induction Bike Dynamo




I'm in test phase of this project and i want to share with you the progresses!

First of all, inspiration:

...and its patent:

Then, principles involved in this project:
- Eddy current

- Magnetic induction

Ready? GO!

Step 1: TEST #1 Magnets Rotor, Rim and Eddy Currents!

Bike upside down, i assembled a magnetic rotor and fixed it to the frame.

The rotor is made up of six neodymium magnets applied on a bolt, settled in a HDD bearing.
Here it is the video, put subtitles on! ;)

Step 2: TEST #2: Adding a Little Coil

I added a very little coil disassembled from a relay for some tests in my last project, and i fixed it on the stirrup, really close to the magnetic rotor.
Here it is the video, put subtitles on! :)

Step 3: TEST #3: Selfmade Coil...

I winded up a very rough coil made up of more or less fivehundred turns of 0.1mm copper wire, and i soldered the terminals to a blu led...

Here it is the video, put subtitles on ;)

Step 4: Winding Up a New Coil...

Okay, first coil worked, but now it's time to draw the max current from that rotor...

I've collected some informations around:

"thinner is always better
go for a few layers; too many will take away from the conductance of the copper"

So the new coil have:
- more turns
- larger windings in order to have less layers
- more precise shape to get closer to the magnets rotor...

I took a little square-shaped ink container and winded on it one turn of paper tape with the sticky part outside; then started to wind up the 0.1mm insulated copper wire on it.
After more or less 700 turns i covered the exposed copper with other turns of paper tape, and slipped it out from the ink bottle.
Once removed the enamel on the two terminals, i soldered one to one led, and tested with 2x1,2 AAA batteries if led would be turned on.
In the end i soldered the second terminal and wrapped all with tape.
Note i blended and left out the led terminals in order to check voltage while working :)

Hope will be more efficient, tomorrow i'll know!

Step 5:

I decide to try something similar to that would be the final result.

I took the inside of a small roll of scotch, the classic for office, and I worked it with cutter and heat gun to get a square ring. Inside there I have stayed the magnetic rotor, and I proceeded with the winding of copper wire. About 400 windings, the thread is always 0.1mm.
After fixing it on the fork, I did the measurements that you can see in the video. The tester was connected after the diode bridge. Here are the results:

- Without the capacitor:
2.8V @ 25kph
2.6V @ 20kph
2.4V @ 15kph
2.0V @ 10kph

- With a 470uF capacitor:
3.1 @ 22kph
2.9 @ 15kph
2.2 @ 10kph

Objectives for the next tests:
Obtain higher voltages
- Probably using a magnetic rotor more often and with greater magnetic power you can get a good increase. For now I think I will keep this rotor, because I have no other magnets (these I am been given away).
- I understand what going to change in the coil to obtain an increase in voltage. if anyone has suggestions about it I would be very grateful!
- An alternative method would be to use a joule thief, I just found out and I can imagine influre positively.

increased stability at low speed:
- I'm waiting for a 1 Farad ultracapacitor to see if it can make up to voltage sags when the wheel turns slowly ...

Hints and collaboration are welcome!

Step 6:

- Ok, i winded up a new coil of quite 1300 turns and seems working better.
- Anyway i need more volts. For this i'm waiting to receive the components to test a joule thief. I'll also try a DC-DC boost to 5V... We'll see.
- I received the capacitor, the circuit i've done makes the capacitor charge while led is already fed by the primary alternate current. As soon the capacitor is full enough, it contributes to light up the led.

At this point voltage is really manageable, so i don't need resistor or voltage regulator for now.
Hints are welcome!

Step 7: Working on New Rotor...

Those are N52 Neodymium Magnets, 10*2mm... I suppose are stronger than the others i used before.

The bearing is from an HDD.

Waiting to have time to design a lego machine to wrap the copper wire..!



    • Paint Challenge

      Paint Challenge
    • Planter Challenge

      Planter Challenge
    • Games Contest

      Games Contest

    18 Discussions


    Question 5 weeks ago on Step 1

    Hi, your project is amazing. I was thinking
    that you can use this not for only lighting
    but for charging things. I tried this but my
    rotor doesn,t spin i want to know why.


    9 months ago

    Hi. Your project is amazing. Please keep working on this. I will be looking.

    (i dont speak english)


    3 years ago

    I cannot tell how old this guide is.

    It seems you are on the right track:

    1 reply

    4 years ago on Introduction

    In your first step you showed a picture of your magnets, how are the magnetic poles located? I am guessing that they are located n and s ends towards and away from the nut, is that correct? Some more clarity on how you arranged them would be nice. Thanks! Also is there a down side to having the shaft that the barring is riding on go through the copper winding? I just thought that it might be nice to have it centered.

    1 reply

    Reply 4 years ago on Introduction

    Hi Jflake!
    Magnetic field has to be alternatively N-S-N-S-N-S along the radial direction, i guess you wrote the same :) In that first model they are just magnetically sticked on the bolt faces. Now i have another model that i could show you as i'll come back home, that's stronger and i suppose better shaped. I haven't got your second question, for sure you can find good informations in this patent:


    4 years ago on Introduction

    Hi! Really nice instructable, thanks!
    I am thiking of different options for magnetic/eddy currents bike lights for a workshop with design students to build their own. Since you made a magnetic-reelight like light in another instrucable and hence, you have experience with both, do you find more interesting the eddy currents option? It needs better/more magnets, is that right? Also, how do the reelights hold the light so long? I've been trying with different caps and won't be able to hold the LED on full brightness for more than a few seconds! Have you already tried the 1F supercap?

    Thanks for the clear explanations and schemes!


    4 years ago on Introduction

    This is AWESOME! We're looking for someone to help us develop some really great power solutions like this for bikes. Would love to have you join the team. Check out what we're doing here: If you have any questions, just email us at Thanks and hope to hear back from you! :)


    4 years ago on Introduction

    I like this generator very much. I would like to build it, but first I wanted to ask: do you know if the metal which the wheel is made of will have any affect on how well the dynamo operates? Is a steel wheel best? Or an aluminum wheel? Or does it not matter? Thanks.

    2 replies

    Reply 4 years ago on Introduction

    Hi Karl!
    wold be very appreciate if we can share our goals!
    For the wheel: absolutely aluminum! As i know Eddy currents happen between a magnet and a non ferromagnetic, electrocondictive material :)


    Reply 4 years ago on Introduction

    I suspected that was the case, given the demonstration with the aluminum can. I will have to find a set of aluminum alloy wheels at some point. Right now I have steel rims and stainless steel spokes. Thank you for clarification.

    I like this idea. How much resistance (mechanical - not electronic) do the eddy currents create? I used one of those physical contact dynamos that touched the tire back in the 80s, and the mechanical resistance and friction made climbing hills torture.

    2 replies

    I don't know how to measure it, but it's quite strong. In my rim there is also a ferromagnetic material in the junction point, and this help a lot to give a strong spin to the rotor, that then keeps velocity thanks to eddy currents. Not so good at high speeds, because it creates a small "turbolence" in electric flow.

    I think the real point is how to optimize the energy from the rotor, i don't think that eddy currents will be too weak to spin it. But are just suppositions :)

    This is just spit-balling, but maybe take a 5 kg weight (or whatever is handy), tied to a pedal, just off top-dead-center, then let go, and time how long the rear wheel spins with the rotor attached. Then do the same without the rotor, and time that. The eddy currents will create a drag on the wheel, but how much is another story. Obviously, this is only if that's something that interests you. If the drag doesn't matter to you, or if you like the extra effort (get stronger faster!), then hakuna matata.