Radial Generator II

by Robotobi Sep 12, 2016
Download All Files

Thing Apps Enabled

Please Login to Comment

What you have there is a 16 pole alternator, the best way to wind it in my opinion is with 15 stator coils. Each coil should occupy 24 degrees of the form wound in groups of five for three phases. This should look like "Clockwise""Counter Clockwise""Clockwise""Counter Clockwise" "Clockwise" . This will help you visualize it (for motors but exact same principal applies): http://www.bavaria-direct.co.za/info/

Dumb/noob question, the bigger/stronger the magnets the more power the generator will produce correct? Or is there a butter zone where you get high diminishing returns the further you get away from it? If so, does anyone know what that butter zone is? What is others running in terms of magnet count, size, strength/pull/force and what output are you seeing? I'm very interested in this project to get some free(ish) power.

Hi Robotobi,
Very cool piece of work, i'm very interested in your turbine and this generator.

However, my 3D printer (Replicator 2X) can't print files in the dimensions your provided, so I will have to scale them down (at least to 85% of the original size). Also, I can't find magnets with a force even remotely close to the ones you use in that new size (you seem to be using 15kg ones, according to supermagnete, while the others that would fit would be much lower).

Would you have any tip / recommendation ?
Thanks in advance !

Hi povpoliova

Makei and has printed a wind turbine with 80%. Maybe he can help you.

dm2p also printed a smaller one:

The generator is fun to print. But the generator is not very efficient and the magnets are expensive.

I have a new design that works better. But it is not quite finished yet.
And intended for larger wind turbines.

If you want to run LED’s, or something that needs little power. Maybe you could use something like this:


And please give feedback if it works or you need help getting this working

Vertical Wind Turbine VAWT
by Makei
Vertical Wind Turbine VAWT
by dm2p

Thanks for the reply !

I did just like them at 70%, and everything went smoothly for the turbine. I'll check with them how they did the generator.

Actually, I don't need much power out of it, I'd like to use it to power small esp8266's with a few sensors (~200mA max at 3.3V). Plus, I don't need to power it all day long, just a few seconds a few times a day is enough. Of course, if I had more, I could use it for more fun ! This seemed to fit perfectly in this generator's capabilities according to what i read in the comments;

I already tried small DC motors like the ones you linked to, with various 3d printed windmills (small ones), but they barely turned with the wind and even turning it mechanically didn't give good results.

Unlike the wind turbines I tried so far, yours is already turning even though it's not fully finished yet (2.5 layers out of 4) and with a little bit of wind only, so maybe i'll try with dc motors and see what happens.

The only issue I had so far is linked to Makerware poorly doing his job on exporting the Rotor part, which fails to print everytime in the first few layers (extruder catches already printed bits and gets clogged). But I don't think that's related to your design, I'll try with different options, since I only used default ones.

Have you tried using multiple generators on a single VAWT ?

Can't wait to see your next design, just love your work.


I do not know how well the generator works with the smaller VAWT.
But if you want to try it you can also buy magnets in France:

I have only operated the wind turbine with a single generator. You will not get much energy from such a small VAWT, I see no reason to use more then one Generator.
But I'm not an expert...

In cura there is a function: Z-hop when retracting. This might help

Have you considered a Darrieus turbine type? They are a bit more fiddly than your Savonius type - well, quite a lot more fiddly - but they also have a much faster rotation rate. Salvonius give very low speed, high torque, while Darrieus are high speed, low torque. More alternator-friendly.

Hi Vylbird

My VAWT is not a Savonius it's a modified Ugrinsky turbine.
My first goal was a turbine that looks good. It should begin to rotate with little wind and function even in strong winds without a brake.
If I also can use it to charge my phone, great!

If you are looking for an efficient wind turbine (at the right location!).
Take a HAVT, a big one.
(If we are seriously concerned about energy efficiency and gray energy, we should never switch on our 3D printer again.) :-)

Tomorrow I will print the new generator prototype. I hope it works ... I'll keep you posted ...

Any updates on this? I printed your VAWT and I'm looking into a generator now.

No updates, sorry.
At the moment I hardly find time to work on my 3D-Printer Projects. :-(

Maybe a bicycle hub dynamo could work... but I have never tried this myself.

Please do not forget to upload a photo of your VAWT! ;-)

Interesting. I may make one myself, but as an electronics enthusiast I can give you some advice:

  1. For maximum power output, you want a load resistance equal to the coil resistance. Note that this is not maximum efficiency: It's maximum power. You'll dissipate half your power in the coils and cause them to heat up, and cause excessive mechanical drag on the rotor, so for actual operation you'll not want to draw that much current.
  2. You wound one phase in the one shown here, but three phases means three times the power - or the same power at higher efficiency. It also opens up options for some different rectifier designs that let you get smoother output, or can double the voltage.
  3. Alternators are very much not regulated - the voltage will be all over the place depending on load and angular speed. Nothing a regulator can't fix. Suggest SEPIC type.
  4. Even with your very nice array layout (I like!) field falloff is severe. That's just the way physics works, but it means you really need to get your coils as close to the magnets as possible. Millimeters matter.
  5. have you considered using a belt drive to connect turbine and alternator? It would let you get a bit more speed, but more importantly I think would protect your delicate alternator from the high levels of vibration the turbine will produce. Speed is very important for alternators: Faster is better. Much better.

Thanks you for your helpful post

  1. The photo is just an example, to make the winding clearly visible. It is designed for three phases

  2. These SEPIC type regulators are interesting. Is there a ready to use regulator you can recommend? Maybe from Mouser?

  3. I have a new design with additional magnets outside the winding to close the field lines.
    Unfortunately there are still some problems to solve. If you are interested in pictures send me your e-mail address.

  4. The VAWT runs very quiet, no vibrations. :-)

SEPIC has the advantage of being able to handle inputs above or below target voltage, which is exactly what you want, and are more efficient than buck-boost. Downside is that they have inherently low current capacity - but then, if you are generating that much power, it is a good problem to have.

I got this one off of eBay, maybe you can find one like it: http://birds-are-nice.me/ipfs/QmcDUuHTRK5FJXA8ex76ywdwytDZHcacTnGCbarTVvK6ER/sepic.jpg

It's only good for about ten watts, but it'll operate off of as little as 2V input.

I've some ideas of my own to try in future in this field, but right now I'm experimenting in a different direction. It involves an easy modification to improve the efficiency of car alternators at low RPM without having to rewind the stator.

Following with interest:)

500mA/5V is good enough to slowly charge your mobile phone or power a raspberry pi :)

Any suggestion for an AC-to-DC inverter?

It's not 5V: It's 5V volt at a certain speed and load. It's not regulated.

If you are using a single phase winding, then I suggest a simple bridge rectifier (Schottky diodes, remember!) and a simple SEPIC-type DC-DC converter, which is available in module form from eBay. You could improve performance by using three phases, which is important not so much for getting more power (It's micro-wind, you'll never get more than a few watts) as for allowing it to operate in lower wind speeds.

If you're looking for a practical power source, you're probably better off with solar. But if you like making generators as a hobby, wind is more fun.

you could star with bridge rectifier and big capacitor.

this looks like a really well thought design, thank you so much for publishing! Just started to print it but I can't wait to see the axial version as well :)
Keep this great work going!

Thank you for your kind words.
Currently I am working on an update of the Radial Version… I'm pretty busy right now… But I will probably have time on Sunday.
Do not forget to upload some photos :-)

Magnets ordered! :D
Do you know any good supplier for PETG in EU/UK? I have never bought this type of filament.

that's great!
Be careful with the magnets. I already destroyed three. :-(

I use the PETG filament (mostly transparent) from http://extrudr.eu/
I usually buy the big 2.3kg spools.
So far I have found no better PETG filament. The layer adhesion is excellent.
And they have a great customer service!
I Print with 220°C to 235°C, 70°C bed, on a carbon plate (from the same store)

I have a request:
I'm looking for a solution to take the rotor out without having to remove the coils.
The "combs" work well. But they make it impossible to removing the lid.
Let me know if you have an idea.

Update Parts: „Stator_1x.stl“ and „Deckel_1x.stl“
easier to print but fully compatible with old version

just a quick FYI:
I bought PETG filament from prusa shop, it prints well and the price was very good, even with the shipping costs, got 5Kg for 100£
the only "problem" is it only comes in black and orange :/

thank you, unfortunately they not have filament with 2.85mm diameter

I will! how did you destroy them? i dont want this happen to me since they are a bit pricey!!
Sure, If I think an idea, will let you know ;)

My magnets are marked with nail polish.
I laid them out to dry, with enough distance from each other… Accidentally, I bumped to one of the magnets… chain reaction… :-)

-mark one magnet (north or south pole does not matter)
-use tape to attach the magnet with the marking upwards under a wooden board.This magnet stays there until you are done.
-put a magnet on the board (on the opposite side).
-mark the same pole NS► ▒ NS ►
-put the Magnet in the rotor
-one after the other

Hey, I was on holiday, tomorrow will restart the work on this. how is it going? any updates>??? best regards...

I have moved last week, still many boxes ... But I am working on the
Next version as soon as I have time.
With kind regards Tobias

Well, Tobias, you did it! Now I need to go in search of magnets...

Hi Lofton, nice to hear from you

The axial-generator is “almost” finished. I guess it will work better on the small turbine.
I use the same magnets (30 pieces).
hope it works ... :-)

nice work - can you upload a video?

What kind of video?
At the moment the generator is disassembled. I need the magnets for the axial-generator :-)

I Would like to see it moving,even if you move the turbine with your hand.
if its easy to disassemble- please do this too,I would like a look through. :)

thanks man .

Unfortunately I can not make a video at the moment.
My stator is broken (Never play with strong magnets close to the generator ...).
… i run out of PETG filament…I'll be moving next week…

Do you work with Inventor? If that helps, i can send you the 3D files.

haha. I know what you say about strong magnets - as soon as you stop concentrating on them something happens and they break.... I do use Inventor, i would love to see the files

im going to try do something similar
I am working on a "Shake Torch" and only get around 2v when shaking it. so want to try another approach
Your design looks great and I think it will do a better job, i will just adjust it for the magnets I have.

just out of curiosity and i know you can really give me a exact answer but would like your opinion,
what % increase would you say the Halbach array gives you compared to a normal NSNSNS arrangement ?

cool man thanks
my mail address kylejardine@me.com

I'll send you the files (it's a little messy)…

I would say the Halbach array makes a big difference to NSNS
But I know too little about ... therefore I have no provable value.

If you want to know more about the arrangement of magnets I can recommend the program FEMM 4.2. http://www.femm.info/wiki/HomePage
it is very easy to use and you can import DXF files from the CAD program.
There are many tutorials about the program.

Tutorial auf Deutsch: http://www.daswindrad.de/forum/viewtopic.php?f=8&t=673

did you measure the current or connected something bigger than a 20mA LED?


load was 500mA

That sounds promising.
I might build a slightly bigger version for 5Watts ?
Or two, located at different places here.

5 watts is not a problem.

The performance depends on its application. If you run the generator with 24 windings ⌀1mm Wire at 500rpm.
You have a resistance of 0.33 Ohm at about 12v on one phase. In this case a 4.15A load (50Watt), no problem.
But I'm not an engineer… everything without guarantee :-)

for which application you want to use the generator?