DIY bike dynamo

[carloc]

hello all, wonder if someone can help. I am new to DIY magnet/coil generators. Have read some info. on net and made a simple prototype with 8 mags & 8 coils. i placed coils in stator in all the same direction, mags were placed N-S-N-S- etc. Coils were wired in series, ie 1st coil wire in connected to 2nd coil wire in, 2nd coil wire out connected to 3rd coil wire out, 3rd coil wire in connected to 4th coil wire in etc - when spun by hand i had an acceptable 2-3 volts - this was all proof of concept for what i really wanted to do was to build my own bicycle light dynamo - However the only place i could fit it was on the front forks on the hub of the wheel but due to the bike forks i had to leave out 2 coils - see attachment - on my prototype I mimicked this set up and still got a healthy though intermittent 2v..but on my bike set up i got .9 - 1v only, can anyone suggest another wiring set up ? I am using 0.16mm wire and neo mags diameter 10mm 5mm thick.. coils have 26ohm resistance 700 turns, core 10mm wide, coil diameter 17mm, coil leg width 5mm - is my relationship between coilwire and magnets faulty ? .. sorry if this all sounds a bit vague as i said am new to this wonderful phenomenon and truly fascinated .. thanks all ..

[soundiceuk]

Learn From The Master

If you look through Nikola Tesla's patents, the information you need is there.

Nikola Tesla Universe

[carloc]

hi thanks for prompt reply... have been trawling through the patents but as yet still haven't found anything that seems to be what i am after .. its more to do with my ignorance of the subject rather than Tesla having the same problem with his bike dynamo he seems to have been occupied with larger projects... could you perhaps point me in the right direction .. thanks again ..

[WeThePeople_USA]

If your going to wire them in series,
you will need to alternate the direction the coils are mounted/wired
so the ones inline with the N are one way, and the S the other,
or you'll waste it all in self cancellation by their arrangment.

I can see the coils in your pic,
but not the (number of) magnets.

They should be arranged so that an
equal number of alternating magnets
line up with an equal number of coils,
with every other magnet being N.

so as the N-S-N-S... magnetic flux swing goes by each coils pole,
the even numbered coils are wired one way,
and the odd numbered coils are wired opposite.

Hope that made sense to you.

Are you using a FWBR to rectify both directions
of the coil's accumulated alternating flow ?
(Full Wave Bridge Rectifier...)

[carloc]

thanks, I de soldered and rearranged the coils but still got the same results ie .9-1.2 volts ? is this possible or have i really misunderstood coil direction. Have attached photo of rotor, there are 8 magnets. I think I may have made my coils too thick and not wide enough. Am going to re wind some more today and see how that works. Making them wider and not so thick, my magnets are 5mm thick so will make coils 5mm thick also, I have also got a lower gauge wire 0.1mm (AWG 38) - Magnets and coils are lined up center to center. No I have not as yet put in a FWBR, but will do once I have got a useful amount of volts, I imagined putting 4 diodes in my system would wipe out any volts i may have ? Will also add a steel core to new coils as someone suggested that this might attract flux better, is this correct ?....

[WeThePeople_USA]

Is this possible or have i really misunderstood coil direction.

Possibly, invision the magnetic flux as a water flow to keep things simple when visualizing the flux pathes.
There is no complete loop in this settup (See end).

Have attached photo of rotor, there are 8 magnets.
Magnets and coils are lined up center to center.

Good, that they all line up at the same time on center was a concern.

I think I may have made my coils too thick and not wide enough. Am going to re wind some more today and see how that works. Making them wider and not so thick, my magnets are 5mm thick so will make coils 5mm thick also, I have also got a lower gauge wire 0.1mm (AWG 38)

Unlikely this is your projects issue.

I de soldered and rearranged the coils but still got the same results ie .9-1.2 volts
No I have not as yet put in a FWBR, but will do once I have got a useful amount of volts,
I imagined putting 4 diodes in my system would wipe out any volts i may have ?

eBay yourself a Schottky Barrier Diode Full Wave Bride,
the voltage drop across them is the smallest if that is a concern for you.

Will also add a steel core to new coils as someone suggested that this might attract flux better, is this correct ?....

Sad that you hadn't yet, yes that is essential here.



Referring to the first quote, try to invision a complete path, a loop.

V2.0:
First, cut and place strips of ferrous metal across the backs of magnets,
(the end opposite of the coils) so they are grouped into pairs of magnets.

Next, bend/form four "U" shaped coil forms out of some ferrous rod stock,
so they match center-to-center the distance of an adjacent pair of magnets.

You have now two options here,
you can wind two smaller coils on the short sides of the "U",
or you can wind one long coil on the long bottom of the "U".



Now, if you visualize the flux pathes (as water if it helps...),
you can clearly see a complete loop back to the starting point.

With all eight poles aligned in one postition of the two modes,
the path is through two magnets and one "U" including It's winding(s).

With all eight poles aligned in the next postition of the two modes,
the path is now through all magnets and all U's including all windings
in a successive path around the whole circle involving all pieces.

This will help you to see why one coil core has flux going one way,
and the next one has it going the opposite way, hope this helps.



I would stop where you are now,
un-wire all coils from each other,
and connect just two together.

Place a very small load across it,
light a flashlight bulb or something,
or around a 500-Ohm resistor or so.

Test just those two coils wired the first way,
then reverse the wires of just one of the coils,
and go with the way that reads the highest V
on the AC setting of the voltmeter your using.


Once you get that sorted out as reference,
and understand the need to alternate coils,
start on BikeGen V2.0 ...
(One with ferrous cores U's, and magnet shunts)

[WeThePeople_USA]

I'm not sure if this information is to simplified:
What is a magnetic field?



I'm not so good at drawing pictures with my words perhaps,
and I haven't found a free drafting program I like either.

I just tried VectorEngineer from here:
Free cad software thats easy to use

It is OK I suppose,
but the text is awful

But it may get the point across better than my text perhaps,
hope this helps



I tried a test drawing as seen below (Top view of V2.0 idea...),
it just shows what happens as the wheel turns through both modes,
it consists of eight magnets, four magnet shunts, four cores, four coils,
thus eight pole crossing points between rotary and stationary components..

This method allows all four coils to be wired in series or parallel.
without having to reverse every other one as discussed earlier.

[carloc]

-- hmmmm .. I see .. now that is a radical change ...

my magnets at the moment are seated on a steel ring then epoxyed, do you think that could be interfering with flux flow.
I can see that the mag shunt and U coil is a nice set up .. I only have room for 6 coils due to bike fork so that would mean 3 U coils... would that work too ?..

your jpg was very helpful as was your text .. but yes pictures speak better sometimes
thanks too for the magnetic field link

the coil on the U ferrous material is there a rule to how thick the wire should be and how many turns considering the mag shunt system, I read that coils should be made relative to the magnet size but that was not considering U coils......

seeing as i might have to do a complete re-build
I opened a wind up torch ((flashlight)) we have over here in euroland to see how it was done and it was a simple coil with what seems to be metal strips either side with rotating round magnet over it .. i have attached photo.. do you think i could make a larger version of this ? .. this small version powers 3 white LEDs more than enough for my purposes .. it seems less fiddly to build .. just one coil ...

thanks for your thorough replies ..

[WeThePeople_USA]

my magnets at the moment are seated on a steel ring then epoxyed, do you think that could be interfering with flux flow.

Sadly, many would jump straight to yes on that one,
as they are arranged in a N/S/N/S... fashion here.

But I don't see that as a problem,
they often get shipped that way.

Magnets need keepers as example.

I only have room for 6 coils due to bike fork so that would mean 3 U coils... would that work too ?..

Yes, that is what I meant above.

One coil (pole-pair) in front, one in back,
and one underneath the fork leg in a circle.

Place a coil (pole pair) in front and in back of the fork and measure
the distance center-to-center of both of the top poles (only).

Now make a circle that allows the additional six poles
to fit equally spaced and you have your starting point.

An inexpensive piece of scrap plexiglass would do fine
for mounting the coils on as a mounting plate perhaps.

A hole saw the size of the magnets lets you place them
through the holes in another piece of plexiglass too.

Let the six magnets grab the three shunts,
and apply epoxy...

I'll make another crappy pic if you ask.

I opened a wind up torch ((flashlight)) we have over here in euroland to see how it was done and it was a simple coil with what seems to be metal strips either side with rotating round magnet over it

It is what you don't see that makes this work.

Take a small magnet and walk your way around the ring magnet.

This is not a standard axial polarized magnet, like in a speaker.

You will find it has poles embedded in it around the circumference,
by the looks of the choice of components, it appears to be four.

You should observe two places on it 180 apart that attract,
and two places 180 apart that repel, placed 90 from the first two.

Not much advantage to using this idea,
this magnet would be to weak to reuse,
and custom magnets are sorta pricey, sorry.

(They hold a polarization on them as they are cooled)

Just rememeber this tidbit, it applies to the flashlight design,
you can spin a magnet, but the field does not spin with it.
(See homopolar generator stuff on youtube if needed).

So if that magnet gets spun, it has more than one pole.

[carloc]

--the more I learn the less I know--

yet again your answers lead to more questions--

yes please send pic, I think I understand but would like to confirm with a pic..

how thick and wide should I make the coil on the U ? should the ferrous material used be the same diameter as my magnets ?

am quite drawn to this design ---


as for Homopolar generators -- wow -- where does it end this wonderful science --- I was wondering how it is that the one coil set up in the flashlight works so well with what seems to be a very weak magnet and such a small coil-- still a lot to understand --
would magnets arranged in a circle N-S-N-S-N-S- etc behave similarly to the circular magnet in the flashlight ?
Do the metal strips arranged on the coil behave with the same principle like the ferrous U cores by drawing the magnetic flux ?
sorry if my questions seem obtuse but I still haven't fully grasped concept.

off now to look for some more plexi for bike gen v4.0 already...I joined the forum at version 2.3

many thanks

[WeThePeople_USA]

Later tonight I'll doodle a few pics and PM them to you.

I'll try to answer these questions too.

It has been too long now, that I haven't had a drafting(CAD)/drawing program I can stand the layout/menus of.

I aim to fix that over the nxt three or four weeks, and post pics with my replies from now on.

But for now, I'll just send a few "Hack's" to you for general reference to what I have said...



I'm not sure now many others have been following this thread,
but I will post pics in the above posts relevant to things said in them,
once I find a free program that supports a Win-ME (9x) enviroment,
and I get past the learning curve a bit so they come out somewhat fair.

Any suggestions on what to try are welcome.

And before you say it,
no I will not be updating to an NT-Kernel OS, ever,
I am running Win-ME kiosk boards with 9x embedded...

Simple, very fast, and can never get a virus, ever...
(They boot/reboot clean every time from a ROM)

[carloc]

-- great thanks -- have been reading and looking at diagrams with varying flux patterns getting familiar with their nature... am going to have take a pause on this project for a bit but will build the U core coil version when I return to it and let you know how it works out and will for sure have many more questions...

just to let know that I finished my original version with added ferrous cores, mounted it and have now between 1 and 2.5v depending on how fast of course i pedal, will add a FWBR & a 2.5v 10F cap and see how that works.

Look forward to your diagrams and further explorations into this (magnetic) field...