HDD Magnet Generator - help needed

[claude]

Hello everybody,

I'm posting this after carefully reading for about 3 weeks the board (the old one too) in search of a complete answer. My findings ended up in only bits and pieces of information on this matter. I want to build a axial flux generator but it seems that the closest I can get to Neodynium magnets are HDD drives. So I got closer and after many trips to local computer shops I managed to get 81 magnets. The quest was long and tidious, in some places having to convince people that i'm not a freek building some illegal instalations (they get suspicious when someone needs so much high-tech garbage) but some of them were patient enough to find why I need them. Some even wished me luck.

But it seems to me I need more than that since I couldn't find any plans on this matter yet. Some people tried different setups with HDD magnets but there's only bits of info's out here. I may have to make them myself and that would be fine with me but when you deal with questions like how many magnets, what kind of placement, how many coils and how many windigs things get complicated.

After all this reading I would try to build a dual rotor axial flux generator. Here comes the questions ... How many magnets should I use? I'm thinking of 12. How many poles? Well, since I gathered them so hard and they are so many sizes/thicknesses/shapes I don't feel comfortable breaking them in half. Or maybe I should, i'm not sure yet. 24 poles with the magnets lined along the circumference sounds good to me. A dual rotor with 12 magnets, 24 poles, 3-stacked should be fine. The problem is that the magnets have different shapes & thicknesses. I could use them wisely to match the total magnets height but will the smaller ones cause any field loss?

As for coils I think I'll go with 9. Is this OK? There will be 24 poles ...

The shape of the coil would be round but I'd rather make them larger (longer legs). That's another matter I need expert advice at. As well as the number of windings. Of course I'll make a test coil first and post the results out on this board for further debates.

My aim is to charge a 12V battery to run some 12 V lights and a small inverter for a couple hours a day. It shouldn't be more than 150W load I presume. All this should be done at low wind speed. The average windspeed here is around 12 mph.

I try to keep my expectations low here.

Since I'll order my rotor/bearings/etc to be made in a factory out here, I will need exact plans on the whole thing. Plans I'll have to make by looking at the pictures posted on this board. There has to be a better source... Can anybody help?

My main concern is the low magnetic flux due to smaller magnets, coil shape and rotor/bearing/furling related mechanics. Unfortunately, Hugh Piggott's book is out of reach here in Romania. I have to say I always liked to build things but that doesn't mean I got to welding or anywhere near a lathe. On top of that, everything's metric here so wish me luck, I think I'll need it.

Thanks,

Claude

[kitno455]

1. break them in half. the neighboring N and S poles on the plate should be separated by a gap.
   
2. stack them to make same height in each pile, dont worry about slight differences in area.
   
3. dual rotor, even if there are no mags on one side, will make more power.
   
4. you have very little magnet material, you will have to make a very thin stator with fine wire to reach cutin.
   
5. single phase or jerry rig will reach cutin with less turns, but requires more coils (one per pole)
   
6. motorcycle brake rotors or sprockets (if not warped) make good magnet rotors for this size gen.
   
   

allan

[claude]

Hello Alan,

So I should break them, huh? Assuming I will, would it be possible to place - say - 3 half pieces of N pole side by side toward the disk center in order to increase the magnet size and the same with S while stacking them. up to 0.7" in height in order to boost the field and increase the coil size and thickness? That would require all my magnets for one rotor but I would then use a blank disc as a second rotor. Perhaps in time I could gather some more magnets to fill it up. Thanks for your suggestions but as I said, the rotors will be manufactured according to my specs - which unfortunately I don't have. Yet.

However I've seen great results with HD magnets used as they are, unbroken. The advantage is a greater numbers of poles, the main disadvantage would be the coil size. But smaller coils allow more of them on the disc. Or with an iron core ( in dual rotor with magnets) - a greater thickness, so... I really don't know what to say right now.

[ghurd]

"after many trips", Free magnets get expensive!

No plans are available because the magnets vary so greatly, and people using HDD magnets will often want to use other recycled parts. No one will have the same pieces.

Proportions are important. Scale the parts to match the magnets.

I believe more poles help too. 81 magnets, broken in half, is 81 magnets per disk.

With 16 stacks that is about 5 per stack. The nicer HDD magnets I have here would average about 12mm deep. Plenty of room for the stator.

With 16 stacks, it could be single phase with 8 coils, or 3 phase with 12 coils.

Computer monitors and TVs have a degausing coil around them in the back, in black tape.

It is made of magnet wire and easier to use than transformer wire. Looks like a nice size wire for a project like this.

G-

[electrondady1]

"after many trips", Free magnets get expensive!

thats for shure gurd .

and after the first hundred harddrives the store owner expects you buy stuff!!

layng out the mags and trying to get some kind of uniformity is  very tedious

the one i'm doing now is a dual rotor.

 16poles on 10" disks.

 i'm using 3 magnet halves to get longer legs.

no test coil yet. i think i will over drive it with a belt or bike chain.

[claude]

ghurd, yes, they really get expensive. One of the vendors even took out the controller plates off them before giving them to me. Perhaps they figured I could use parts form all the disks to make some good ones... Anyway, some of them asked money for them. I wouldn't mind paying the man a pizza but they went up to 60% of the brand new disk value! How's that? No comments.

I ageee, with the variations. I identified 19 kinds of magnet among my 81's.

So you suggest 16 stacks per disk, 5 per stack and 8 or 12 coils... Hmm...you do realize that's 16 poles instead of 24 with the magnets unbroken... but still that's 5 per stack.... I'm thinking out loud here, don't mind me.

But hey, that half magnet-sized pole would need minute coils! Where's the advantage here?

As for the magnet wire, I intend to buy it new. I can't afford limited supply with so many variables hanging around. The magnets are limited enough. Thanks for the tip though..

Claude

[claude]

electrondady1, that's my idea too!

I wrote it up here, 3 magnet halves to get bigger coils. So it is possible.. How many will you stack? Could you post some photos of your work in progress? I ask that because I imagine you have the plans in your mind... Why overdrive? Do you expect low readings otherwise?

As for the magnets I found one small company (maybe the only one here) that can import them but the prices...

Here for instance : N35A502505 ( L=50; l=25; H=5 mm ) -price 12,49 EUR/piece. (about 15 bucks) + VAT (14%) + shipping from England (unknown value so far)+ delivery in 1 month. Is this a ripoff or what?

With 15 bucks i'll visit every computer shop in the two neighbouring counties and maybe fill my trunk with HDD's... The magnets are small, true, but what a puzzle I could make... :-) Tedious work is my middle name, I graduated fine arts... :-)

[ghurd]

With the magnets unbroken, the N and S will be too close together.  The flux will be shorted, so not be as good.

I pulled a degaussing coil from a 32" TV, it is 0.813mm wire, 4.5M circumference, estimate it at ~400M!

G-

[claude]

Right now, three scenarios cross my mind. I'll post the pictures here, it's worth a thousand words. Especially since my english gets rusty...














Will someone please tell me which way to go... I can't be sure of anything at this point.

[electrondady1]

calude,
build your geni like danb builds his.
 gurd and kitno laid it out for you.
your first diagram is the route i'm taking but you need to tighten up your mags
the space between mags should be the same as the width of you mag
some thing like this.







once the mags rotors are laid out. do a test coil.

[ghurd]

I like Willib's diagram of a layout.  2 coils are High, meaning a N and S magnet is over each leg, while one coil is low.

The magnets in this one are square, the coils are round.

http://www.otherpower.com/images/scimages/2965/g_alt_6C_8M.JPG

It is a better spacing of the magnets, like electrondady means.

The measurements do not matter so much, as long as the proportions are good.

G-

[claude]

Yes ghurd, I like it too but imagine small arched pieces of half HDD magnet trying to look like a big square one... And if I want some usable power i'd had to stack so many of them that I'm afraid I won't have any magnets left for the second rotor.

Danb's layout looks more promising to me. With three halves in line and stacked it should work, I guess. I'll try to find some more magnets, seems to me that the "Search of the magnets" is far from being over... despite my folks opinion that I started to have an obsession.. :-)

Anyway, I'll refer to the mechanics now since I'll obviously need more magnets. Here's a plan I made from my readings on this board. Can anyone see something I forgot? Faults... possible improvements...

[MickS]

Claude,

Here is how I do it.

I usually use anything up to 300 hard drives per genny.

I am lucky I have a good source.

You must cut them in half to get any decent results.

You must make your coils hole the same size as the mags and the width of the coil legs the same as the width of the mags to get optimal results

Mick S

<BR<BR>

[MickS]

Here is one of them up on its tower with 3 metre diameter blades.

Mick S

[claude]

Wow MickS, that's impressive!

Thank you for posting your pics. You used car brake discs I see... Hmm. Perhaps I should use them after all. Can you tell me the rotor diameter? And some output of that impressive 3m windmill? Also important would be what's the total height of your stacked magnets/pole. That way I could figure out how many years of searching hard disks I have ahead. :-)

Thanks a lot,

Claude

[claude]

Can anyone tell me what are the advantages of the coils with 2 wires in hand versus single wire? I saw Mick S using three in hand for his stator. I'm new to this so please explain me why is this better and how do you wire such thing. Do you just sum up the endings?

Claude

[MickS]

Claude,

They are not brake discs.

I use stub axles and hubs off vehicles, but the discs are cut from 5mm (sometimes 8 mm) steel plate - cut into a circle and drilled appropriately.

Most of my stuff is scrounged (it costs me nothing), some times this method uses a lot more labour, but I would rather be doing something useful instead of wasting time watching television or some other useless pasttime.

The magnet size in the one I have shown here is 40 mm long by 20 mm wide and 20 mm high.

A hint on where to get the hard drives. Where I live there is a Recycling Centre / Rubbish tip. All recycling things (including metal) are free to dispose of at this place.

The deal is that I get all the old computers they do not want /sell , I take them home regulary, dismantle the hardrives and remove the magnets.

The same goes for microwave ovens - I remove the copper secondary windings from the transformers (I use these to wind the coils for my wind gennys).

Then I take all the leftovers back to the recycling centre and place all the non ferrous metals into their special bins for which they get lots of money from the scrap metal dealers.

So I am actually doing them a favour by dismantling all these things.

Hope this helps.

Mick S

[MickS]

Hmmm, nearly forgot,

The rotors are 300 mm diameter.

Depending upon the diameter of your wire  and how much room you have for the coils and also how many turns you need per coil decides on whether you can use 1, 2 or 3 in hand. The more strands you have mean the lower the overall resistance of your coils.

Do a test coil, to determine how many turns you need to cutin at your desired rpm cutin.

1. mm diameter wire is very easy to calculate how many turns will fit into a given area.
   
1. mm diameter wire is very common in microwave transformers. Just be aware that many of them use aluminium wire.
   
   

After awhile you can tell by looking at them. None of the aluminium cored transformers will have soldered ends. They will be all of the insulation displacement type.

Hope thia helps.

Mick S

[claude]

thanks MickS. So 3 wire coils are actually three coils in one.

The begining and ends get soldered together?

For the first time I wish I lived near a recycling center :-)

Great source you have out there. Here, the computer shops have only one or two broken disks. The second-hand retailers have more but if you take what they have now you'll have to come again in a year or so to get a significant amount.

One company was kind enough to handle me 25 disks. I also got 10 more from different sources. I know you used up to 300 disks but do you think it would be possible to use all I got on one rotor and leave the other one blank until I can gather some more? Would 81 magnets be enough for one rotor? It's a dumb question, I know...

I'm eager to start building one HDD magnet generator, so that explains the hurry... The idea of postponing everything until I get enough magnets is so depressing.

Claude.

[MickS]

I can understand your frustration Claude.

Below is an earlier genny I made, using a lot less hard drives.

These figs are from memory - I cant find where I wrote the details down.

Mag rotors are 220 mm diameter - 5mm steel plate

24. mags per rotor, using 3 hard drives per mag - just make sure the piles are all the same height.
    
15. coils (I forget the number of windings) One in hand of 1.0 mm diameter wire.
    
    

Just wind the coils big enough so that the 15 coils will just fit in the space available. I made the coils 10 mm thick.

This one has 5 coils per phase (3 phase star config).

Originally the output was a little low, so I dismantled the genny and drilled a hole through the resin in the centre of each coil, gathered up some iron filings and mixed them with fibreglass resin and filled the core holes with this mixture.

The resultant cogging does not affect startup and it has upped the output to where I easily get 10 to 15 amps into a 12 volt battery in a good breeze.

Lets see  

24 * 3 = 72 hard drives for each rotor.

Two rotors mean you are still gonna need a minimum of 144 hard drives.

(Worth the hunt I think)

Good Luck

Mick S

Pics below.

[claude]

Wow, a small one!

It has my name all over it! THANK YOU! It's much more "normal" to aim for 144 hard drives than 300, in my situation. Today I squeezed the Yellow pages /computer stores and called everyone in my area. It feels a bit embarrasing trying to explain them you actually need their garbage but my pride is on vacation right now.

I may be able to take a shortcut with the magnets since I found 3 huge and very old Micropolis SCSI drives with 4 very big magnets each. So that's 12 magnets of 4 mm thickness.

I measured all the magnets I have. The total thickness of all the mags put together is 250 mm. Divided to 48 that's 5.2 mm. So I could make now two rotors with 5.2 mm thick poles. Mission impossible - just a few magnets are the same in thickness. I clearly have to gather more drives but I can see the light now :-) It sure is worth the hunt.

One question popped into my mind looking at your fine pictures: Is the distance between the magnets the same as the width of the magnets? I can see that clearly on the template's outer side but towards the center they seem to get much closer. Is that OK?

Thanks,

Claude

[RP]

It's simply because it's a pain to wind coils with really heavy wire.  You treat the multiple strands as parallel conductors and tie them together at each end.

Also, with a stock of say 17 gauge wire you can wind single, two in hand, three in hand, etc. to create various load capacities and winding thicknesses.

[adone]

Sorry for answer in Romanian. Is a link for a site in romanian about this topic and others ...

Am incercat si eu asaceva si am obtinut cam 25 W la 12V cu 12 'magneti' (in ghilimele caci sunt cate 2-3, in functie de grosime) si 6 bobine , monofazat.

Incearca www.incalzireglobala.tk, dupa care astept o scrisoare pe adresa de e-mail de acolo.

Succes!