Author Topic: Idea for ultra compact coils + easy to bending (Read 4241 times)

BigBreaker · « **on:** June 27, 2006, 02:12:43 PM »

Square magnetic wire wins praise for tight coils as round wire leaves spaces. In either case the thickness of the wire tends to require multi stranding (2 or 3 in hand) with lighter wire to handle the peak current requirements.

I was thinking about the benefits of square magnet wire maybe found an alternative. You could use strips of copper flashing, wrapped in a single roll at your coil / stator width. Instead of multiple turns per layer, you just have one turn per layer. Flashing is easy to bend, comes in the right thicknesses, isn't too expensive (since it's a building material) and would make perfectly solid coils.

See product example here: LONG link shortened

There were two obvious problems: (1)The flashing wouldn't have insulation and (2) How to connect to the "inside" leg of the coil.

OK here are my solutions: (1) For the insulation you just have to apply some dieletric resin. It's not a simple solution but you can coat one side of the flashing at once. The surface area to be protecting is MUCH smaller than regular wire so you can afford some unevenness or excessive thickness.

(2) Now for getting to that inside leg... You can put two coils side by side, the first wound outside-in and the second inside-out, both in same clockwise direction. Then you can bend the two inside legs towards each at a shallow angle and solder them together across the width of their faces. This leaves you with twices as many turns but both coil legs exposed and no space wasted in bringing a leg out. Alternatively you could drill radially, towards the center and solder a wire to that inside leg. That increases your resistance slightly but is probably easier.

The flat coils can be belt sanded to a perfect even, square shape. The compactness is fantastic and since the face is perpendicular to the flux it shouldn't cause any drag. The two coil method would cut the width in half as seen by and non orthogonal flux paths.

Cutting the flashing into strips might be tedious but I'm sure there is some method easier than tin sheers. I'm thinking a jig and a cutting tool would help. A really nifty way would be to roll a whole 12" wide sheet of flashing into the coil shape and then band-saw off each coil like you are slicing salami.

So what do the wisemen think? Maybe a new product for our sponsor? pre-made coils!

A6D9 · « **Reply #1 on:** June 27, 2006, 08:41:41 AM »

i think it is a great idea in theory, but i don't think i would do it untill i see someone else with sucess.

it just seams to good to be true.

BigBreaker · « **Reply #2 on:** June 27, 2006, 08:46:53 AM »

There are some entries about using ribbon wire, which is basically the same thing. One in April of this year by DanB and one from 2003. I don't think it is too outlandish. Maybe someone with experience using ribbon wire can comment?

Flux · « **Reply #3 on:** June 27, 2006, 08:51:51 AM »

I agree, the theory is perfect but I don't think you will manage to cut the stuff without serious burrs. I also don't think you will coat it satisfactorily.

If you can cut the strip without burrs then think about winding with a strip of mylar between the copper.

Simple in theory, but try winding some Ford model T flywheel coils and find out how deceptively simple it looks.

I would look for rectangular enamelled strip and leave the burrs and coating problems to those who know about such things.

Flux

RogerAS · « **Reply #4 on:** June 27, 2006, 10:43:17 AM »

Hi,

I've thought of the same thing. If you could find a sheet metal shop to help you could get them to clip you off consitent widths 1/2" - 3/4" or what ever you wanted. This would allow long straight ribbons with fairly smooth edges.

As for coating it, drill one end or punch tiny little holes, just big enough for fishing line, and hang these up where they're out of the wind. Get a can of good rattle can enamel. Give em 2 coats realy thin and sit for a day. I bet they would be plenty insulated enough. Then again the teflon tape would work too, but if you are casting these in resin after "winding" I think the enamel might stick better overall.

The inside leg could be folded at a 90 degee then just solder on a lead out like you said. I say please go for it and let us know how it turns out. Maybe wind up a wire coil with the same ammount of copper per weight and compare them. I know that's a bit like comparing Apples and Dells, but you (we) need a basis for comparison or what have you (we) learned?

I've looked at ribbon cabble inside old printers and thought about those, already mounted on a insulating substrate. If you could roll up a long section of that with one ends ribbon cross connected to a single conductor, and then the other. I don't see why it wouldn't work. Trouble is finding enough to even try. Maybe someone could find a surplus overrun or something.

RogerAS

windstuffnow · « **Reply #5 on:** June 27, 2006, 11:13:12 AM »

It does work, I've built one a long time ago. The problems I found were eddy currents in the flat stock creating lots of heat. I did find out that there is a theoretical limit to the thickness and height of the ribbon, seems like it was something like 40% of its thickness. I'm not sure about that it was a long time ago but 40% seems to stick in my mind.

The inside wire can be bent down then folded upward to present a "tang" for soldering wire to it but you have to make sure the tang is insulated from touching the corresponding coiled section.

You can use some pretty thin copper to make what would be the equivilant to a pretty good size round wire keeping the resistance very low and still get lots of turns in a given area.

My unit produced pretty high output but would only do so for short periods of time because of the heat from eddy currents. Ribbon wire works better but it has its limits. Square wire has the same problem when used in an alternator, lots of heat. From my experience the round wire, with all its limitations works the best... so far anyway..

Another way I thought would be a good idea was to punch the flat plate into wave type windings and stack them, quite similar to the above only flat. This also produced very high output but the heat was unmanageable because of the surface area exposed to the flux. A good way to make electric while heating your house if you could find a way to insulate the plates so the alternator could be submerged in a tank.

I used a roll of .002 delrin shim stock as an insulator between the copper sheet which worked well as long as you didn't overheat it. You'd be surprised how much space that alone uses up.

It will work but there are a lot of problems to overcome.

.

BigBreaker · « **Reply #6 on:** June 27, 2006, 11:13:43 AM »

If you cut off sections of a rolled copper tube you shouldn't get serious burrs. The resin between the layers of copper should solve any chatter that would cause burrs (method and reasoning explained below). If you do get some nasty edges you can just put some fine grit sand paper in a belt or orbital sander and smooth over the faces of the coil cylinder.

I agree that ragged edges would be a problem before you rolled it, but after... seems like a small problem.

I like the mylar idea for a sure barrier. I was thinking two coats of resin. The first coat goes on and dries. Then the copper is unrolled for conductivity tests with a some salt water, a small flat piece of copper scrap and multimeter set for a conductivity test. The scrap copper is attached to the logic probe and with the meter's ground attached to the big copper roll. Then you can move the copper scrap, keeping a film of salt water in the interface to find any "holes".

Then you dry the copper roll and use a second coat of resin as an adhesive, wet and tacky, when you roll it up for the coil. Once that unit is rolled up and dry I think you'll find it VERY strong and reasonable to cut across its length. More and more it seems that rolling one big tube is the way to go.

I guess the last question is how easy it is to cut off the individual coils? Are we talking free hand with a bandsaw, only with a special jig or waterjet territory?

BigBreaker · « **Reply #7 on:** June 27, 2006, 11:21:59 AM »

I thought about the twist method of pulling out the inside leg. That causes a copper area to cut right through the key magnetic flux. You'd need to make thin cuts going the length of that area. That breaks the ribbon up into thinner strips and will squash the power robbing eddies. Definitely easier than using TWO coils in each stator slot. I bet the extra material along the coil face is not much more space lost than an isulation barrier between the two coils in the other version.

TomW · « **Reply #8 on:** June 27, 2006, 12:04:40 PM »

Uh, you may want to look for a posting where DanB used ribbon wire [enameled] to wind a stator for a fuel powered alternator. Too bad the built in search sucks on this board or I would have looked it up for you. I cannot use the google search the board for some reason either??

Bottom line was eddy currents smoked those coils in very short order. It may or may not apply.

Just something I remember from awhile back. Some days the alshiemers is not as prevalent...

Cheers.

TomW

Flux · « **Reply #9 on:** June 27, 2006, 12:36:49 PM »

Interesting stuff. First thoughts suggest that wide and thin tape should be ok, this makes the assumption that the flux lines are along the thin edge of the tape.

First reality is that fringing between the magnets makes some flux lines run at an angle. I thought that the strip Dan used should have been ok, and it may have been at lower speed and frequency, but it did change my thoughts on this and I lost a lot of interest in the idea.

I am not sure square wire will give significantly better stacking factor than round when you have to keep the strands small to avoid eddy loss. Ed's point about the space that the insulation between the tape wastes brings home the fact that the gain is not going to be spectacular anyway. Round wire can be squeezed into a stacking factor of about 60%. I doubt that tape would manage much better than 80% and this would again be reduced if it had to be wound in sections to reduce eddies.

Finally What happens on load even if the eddies are ok off load. The armature reaction field will be at right angles to the main magnet field and the resultant will now most likely cut the strip at an angle and worsen the problem.

This problem is one that was always an issue when this type of alternator was in common use and needed elaborate transposed wires to reduce the effect for large machines.

Rather than spending time devising ideas that may or may not work, someone with time to spare could do us all a service by trying various sizes of wire and find out at what size wire this issue becomes significant. I have not noticed any trouble with wire up to 1.2mm thick but I am not prepared to use thicker partly for fear of this and partly because it is easier to use 2 or 3 in hand to do the same job and get a better stacking factor with less effort.

My instinct is that wire over 2mm thick may be near the limit before this issue arises but it would be nice to have confirmation.

Flux

Flux · « **Reply #10 on:** June 27, 2006, 12:44:37 PM »

I have used copper tape to bring out the centre connection for years. There is no measurable loss in a tape 1/4" wide and 0.01" thick ( 12 of them). If you take this to extreme it would no doubt be an issue, but it is an easy way to bring the lead out without a lot of increased air gap.

Bringing the end out as a single layer coil wound in the other direction is also possible but a lot more trouble.

All these windings suffer eddy loss, but until you can measure the effect it doesn't matter.

Flux

BigBreaker · « **Reply #11 on:** June 27, 2006, 01:00:42 PM »

The story on ribbon wire is here: http://www.fieldlines.com/story/2005/4/12/165759/437

Not to knock DanB but he cuts the flux with the flat face of the ribbon in that article. Of course there was too much heat at high rpm!

You can test this by leaving the coils open circuit (no connections) and spinning the rotor. If it turned longer without the stator than it does with the stator, you have eddy current problems. The real question is whether the field diverges from parallel in the air gap. The sine of that divergence angle (approximately equal to the angle in radians) is the percentage of the flux that is effectively causing eddy currents. That spreading will get worse as the rpm speeds up and the coil starts to generate a significant contra field.

Flux · « **Reply #12 on:** June 27, 2006, 01:10:01 PM »

"Not to knock DanB but he cuts the flux with the flat face of the ribbon in that article."

No, it is quite clear in the pictures that he got it the right way. Not easy to wind that section on edge with that size coil.

Had it been the wrong way then I would have expected trouble.

Flux

BigBreaker · « **Reply #13 on:** June 27, 2006, 02:06:09 PM »

I see the inside leg twisted and coming up the side, flat against the coil. That section pointing up from the twist is facing the flux head-on rather than edge-on. I think that could create eddy problems in that piece of the wire. In your other post you mentioned that .25" ribbon didn't cause noticable problems. I'd think .75" or 1 inch might cause a severe problem. Even though it represents a small portion of the coil's length, localized heat can cause thermal runaway.

One of those laser temperature probes could answer the question. Run up a gen with some serious amps, stop it and then measure the twisted leg's teemperature against the bulk of the coil.

I agree, Flux, on the flux spreading due to reverse field and flux dragging due to rotation. There is a significant angle to worry about. It's not exactly edge on.

inode buddha · « **Reply #14 on:** June 27, 2006, 02:54:52 PM »

You can actually buy such coils ready-made here [partsexpress.com] but they're not especially cheap. Figure about 10 -20 $ each. Power from 350 to 650 watts each. Some of the big ring neos might work nice with these, but the parts site has some other nifty stuff too.

RP · « **Reply #15 on:** June 27, 2006, 04:37:10 PM »

This story by DanB has links to the previous ones on the diesel generator Tom mentioned.

http://www.fieldlines.com/story/2005/9/20/16716/4310

craig110 · « **Reply #16 on:** June 27, 2006, 05:10:24 PM »

Very tiny holes in those coils, though.

Titantornado · « **Reply #17 on:** June 27, 2006, 06:05:15 PM »

Instead of enamel, how about a paper material sandwiched between each wrap. ( I think paper burns above 400`F, but I don't know at what temp it begins breaking down) Hmmmm, perhaps tar paper or Tyvek house wrap? Seems almost appropriate for tar paper or Tyvek to go with flashing!

terry5732 · « **Reply #18 on:** June 27, 2006, 09:01:45 PM »

Cheap, thin cellophane tape works well for insulation, but the flat idea gives NO output. It seems to ALL disappear in eddy heat. I think smaller round wire is best for density. Save the foils for building capacitors.

DanB · « **Reply #19 on:** June 27, 2006, 09:37:14 PM »

Yes, the inner winding coming out was surely an issue, but the heat in the coils was throughout them - I dont believe it was the bit of flat wire coming out that caused the problem. That was almost completely outside the moving magnetic field. The eddy currents were induced in the coils throughout - I could feel that as they heated up, most of the heat was within the coil - not the lead that passed over the top.

Its a Big pain to wind with flat wire and keep things perfect. So many folks here obsess with 'perfecting' things when just building a slightly bigger machine (maybe 25% more magnet) will make things seem so much simpler. You get into flat or square magnet and the difficulty of using it - it doesnt make sense to me. Perhaps if you want to build some fairly expensive/complex manufacturing equipment and build 1000 units it might pay off. Otherwise - use round wire and adapt a more forgiving design. The extra cost in magnets and wire (if there is any) will be fairly insignificant I expect.

willib · « **Reply #20 on:** June 28, 2006, 02:08:48 AM »

why bother ?

what would be the point? of someone taking the time to prove one way or another , people are still making the inside of their coils the size of their magnets , when i proved that making the inside of the coil considerably less than the magnet size did not detract from the output in the least , it still produced a perfect sinewave .

these are the specifics

magnet dia. 0.875"

coil OD 1.5"

coil ID 0.5"

poles
" dia rotor

dist. between magnets approx 0.480" dont remember exactly

so why bother?

dinges · « **Reply #21 on:** June 28, 2006, 03:55:16 AM »

I'm not aware of any experiments of this kind that you have done, Willib, so if you've published it somewhere on this board, please provide the link.

As far as a 'perfect sinewave' goes: did you perform a Fourier analysis on it (or something else) to determine the amount of harmonics and shape of the sinewave? Because in my experience, something that looks like a beautiful sinewave sometimes isn't so perfect.

Modern digital scopes (like Tektronix TDS210) can perform a Fourier analysis/spectrum analysis) and can show you the amount of harmonics present. Just judging a sinewave 'by looks' on a scopescreen is dangerous in this case, I think.

TomW · « **Reply #22 on:** June 28, 2006, 07:24:47 AM »

dinges;

Just judging a sinewave 'by looks' on a scopescreen is dangerous in this case, I think.

I have to agree there. I try not to get involved in this stuff too much but a lot of people make claims that just do not add up. Constructing an entire theory from one data point based on one visual observation is unlikely to be accurate.

Just an opinion based on a couple decades of troubleshooting.

Carry on.

Cheers.

TomW

redeyecow · « **Reply #23 on:** June 28, 2006, 11:30:33 AM »

The Diary topic "Converters" by WXYZCIENCE has pictures and data on the aluminum version of ribbon coils as used in a McCulloch gen set. Also my cat. keith

Don Cackleberrycreations · « **Reply #24 on:** June 28, 2006, 07:51:16 PM »

Guys your ignoring a simple alternative . You are looking for copper foil tape used in making stainedglass among other things . It comes in different widths and guages .

use a good varnish and nylon roller guides to make your coils .

willib · « **Reply #25 on:** June 30, 2006, 05:16:12 PM »

yes you were not around , or i'm sure you would have commented..

this is the conclusion that i came to , you can back up if you want to ,at the time i was under the assumption by visual analysis , that there would be a cancellation effect due to the inner windings being smaller than the magnet size, but the opposite was true .

there was no cancellation because as a magnet approached the center of the coil the wave was approaching zero .

in an effort to be dialup friendly ive included a smaller size photo of the waveform , below.

30KB

this is the conclusion

http://www.fieldlines.com/comments/2006/3/16/195739/099/33#33

this is the complete discussion.

http://www.fieldlines.com/story/2006/3/16/195739/099

i dont really believe there is a need for Fourier analysis , considering the waveforms produced by skewing a conversion probably dont look half as good , and no one ever asked Zubbly to do a Fourier analysis on the wave comming from his machines , and they charge a battery perfectly well.

News:

Author Topic: Idea for ultra compact coils + easy to bending (Read 4241 times)