Author Topic: Rewinding a Proven stator? Hugh? (Read 27421 times)

Flux · « **Reply #33 on:** October 22, 2011, 10:11:44 AM »

Not doing very well am I .

Another typo in the same paragraph - the 32 should be 2 that may be more obvious.

Further down where it says nest it means rest. Not going to look again, sorry if there are more.

Flux

nconstanzo · « **Reply #34 on:** October 22, 2011, 10:35:36 AM »

I was looking how the stator was wound and it looks like they wrapped around the metal . This remind me of a Italian made wind mill that i saw. Where they took the silicon metal and made the end of the metal so they could lock together . that way they could make the coils on a winder then slide them on to the silicon steel rings. Then lock the ring ends together saving all the time it would take to wrap that huge coil around the metal core.I don't remember the name of the web sight .I well see if they still exist. Nick C.

bob golding · « **Reply #35 on:** October 22, 2011, 11:13:07 AM »

found another useful link that gives the flux density of toroidal cores. it will be a bit out if you use strips of some unknown lamination e should be in the ballpark.
http://www.wiltan.co.uk/client_files/default/wiltan_guide.pdf

going to go and attack my old variacs now.

bob golding · « **Reply #36 on:** October 22, 2011, 11:25:02 AM »

Quote from: nconstanzo on October 22, 2011, 10:35:36 AM

I was looking how the stator was wound and it looks like they wrapped around the metal . This remind me of a Italian made wind mill that i saw. Where they took the silicon metal and made the end of the metal so they could lock together . that way they could make the coils on a winder then slide them on to the silicon steel rings. Then lock the ring ends together saving all the time it would take to wrap that huge coil around the metal core.I don't remember the name of the web sight .I well see if they still exist. Nick C.

c cores are used like that. you can wind the bobbin and then slide the cores in afterwards. seem to remember the end of the cores have to be ground very flat to work properly. they usually have a metal band around them to pull them tight. could work with a ring if you can find a way to tighten them up would have to shape them into 2 half circles and grind the ends flat then have some a way of clamping them together. doable but difficult to get right. think winding the coils individually would be easier for homebrew . it is not one huge coil,but 36 small coils of around 25 turns each so not that difficult to do.

bob golding · « **Reply #37 on:** October 22, 2011, 01:39:45 PM »

just took an old variac apart, and found my idea of using the core wont work at least with the variac i have. the laminations are stamped discs rather than a wound strip so that's no good. i have lots of MOTs lying around so it looks like that is the next thing to try. i have seen variacs with wound strips so must vary depending on the manufacturer the one i took apart was a Ferranti.

bob golding · « **Reply #38 on:** October 22, 2011, 04:45:39 PM »

after the variac no go i have been checking out commercial core manufactures. seems there are at least 2 in the uk. i need some idea of what is needed. here is what i have come up with so far. assuming a core diameter of around 350mm OD and a thickness of 25 mm height 10mm that could be done with 8 off the shelf cores if they are cut in half down the middle. this works out around Â£41.00 total inc postage.
i want to email to see how much a custom core would cost. so do these dimensions sound in the ball park. i chose 350mm as my magnet rotors are 360 mm. at that price it is certainly doable. there is the the cost of the epoxy to add of course but that is still only the same as 4 2x1x.5 inch magnets!! i have a good supplier of epoxy anyway.

scoraigwind · « **Reply #39 on:** October 22, 2011, 06:23:36 PM »

If the diameter is 350 (14 inches) then you might want more radial 'height' than 25mm (1") on that core. I'd be tempted to go for 50 mm (2") so that internal is 250 (10") and then make it 15 thick. Three times the weight almost. Depends what you are planning to run. Provens use big magnets about 100 x 100 in some cases.

They are only ferrites, but those rotors can bite off a lot of skin if they don't have proper jacking screws (and Provens dont!)

bob golding · « **Reply #40 on:** October 22, 2011, 07:46:30 PM »

bet provens are fun to put together. i found some 75x50x20 ferrite's which look as if they fit my rotors ,and are a reasonable price. i am using my current rotor discs and working a design around what i have already that so i don't have to change too much. still thinking of chrises gearbox design,but would have to get someone else to make most of the parts for that. strip cores and epoxy i can do ok in the middle of a field with hand tools.

nconstanzo · « **Reply #41 on:** October 22, 2011, 11:19:26 PM »

I been looking for a home source replacement for silicon steel and came across this old telegraph wire hang on fence post, while fixing the cattle fence. So I took it home and tested it for magnetic memory. From what I could see it showed no memory at all. My question is would this be a good core material, or is it coated with copper which would possible cause heat. I scraped the rust off a portion of the wire and found no traces of copper. But I remember dad saying it was coated with copper maybe after a hundred and thiryt years the copper has rusted off.

scoraigwind · « **Reply #42 on:** October 23, 2011, 03:06:59 AM »

Quote from: Flux on October 22, 2011, 04:46:08 AM

Keep the flux density in the core below 1.5T and idea ll lower than that if you want low core loss for low wind start up. When you consider that you only have about 0.3T on either side you will find that with nice long narrow magnets the core can be quite thin. If you go for short wide magnets you will not get the flux density up in the core high enough to benefit from the shorter turns.

Proven design uses both long and wide magnets. That makes the coils an easier shape. But it dictates a thicker core to carry the flux. You can say that the core cross section has to carry all of the flux of one magnet (half of 2 magnets) so if a magnet is 4" x 4" then the core section needs to be about 4" x 1" to multiply the flux from 300mT to 1200mT (1.2T). If the core is only 10 mm thick then the magnets can't be more than 40-50 mm wide or the flux exceeds Flux's recommended limit. And the coils would become very narrow.

I seem to remember Gordon Proven told me that he made the magnets wide to avoid leakage. But that idea seems to have gone out of fashion on this board.

Flux · « **Reply #43 on:** October 23, 2011, 04:20:55 AM »

I agree that with 14" core it should be at least 2" deep. The longer the magnets the better but if you make the hole in the middle too small then with your coils touching on the inside, the outer edge is widely spaced.

2" seems a nice compromise.

Magnet width may be decided by available block sizes. Lots of small coils are a bit easier in that the winding bobbin needs less wire but you suffer when you come to connect all the ends up if you go for too many poles. If you have an oxy/acetylene torch you can braze the connections with silphos very quickly with no need to skin the enamel off, if you have to solder it gets tedious.

Most of the earlier variacs including Zenith used stamped discs, it seems to be the smaller ones and perhaps very late ones that use strip cores.

If you could find a 30A 3 phase one with strip cores you might have enough core material for a decent alternator. The difference in scrap price and new is really dramatic when considering cores. There wouldn't be many motor conversions if you had to pay the true price of the core.

How about trying a radial version with a large diameter motor core, cut the teeth off the inside and have magnets inside like a motor conversion and outside like a drum alternator. You may have to use smaller magnets on the inside but they would still help. Without the inside it would become a gramme ring winding and still work but with higher resistance for a given output.

Flux

phil b · « **Reply #44 on:** October 23, 2011, 04:24:36 AM »

So the idea is to permeate the slot less iron from both sides in order to get the most flux to travel to the north to the south pole. 1.5T is a combination of all the magnetically induced flux. From what I'm reading it would be similar to the slotless design illustrated in figures 3,4,and 5 in this paper. http://etd.lsu.edu/docs/available/etd-06142006-234710/unrestricted/Choppa_thesis.pdf

This person(s) is using large magnets and three windings for each magnet. Figure 6 is showing the flux paths as they enter the stator from both rotors. Now I see why each coil has to be turned opposite. Even though they are using NdFeb magnets, the principle should carry to any magnet at hand.

In a torus core that is made from one spiral piece, the winding would have to be large enough to carry the flux.

How was the winding kept from unraveling? When I constructed a torus type winding, that was one of the difficult parts of the unit. I used 1/4" bolts going all the way through and nutted on both sides. Proven's cores must have had welded ends or bolted?

Flux · « **Reply #45 on:** October 23, 2011, 04:26:23 AM »

Since posting that, I had the thought that the inner rotor could be a claw rotor ( like the car alternator) energised by a stack of ring ferrite loudspeaker magnets, you could probably get the same amount of ferrite material in there as the outside magnets then. The claw rotor would also be a flux concentrator.

Flux

Flux · « **Reply #46 on:** October 23, 2011, 04:34:16 AM »

phil b

I am glad you found that. I had a copy of something very similar by a different author and I couldn't find it. This covers all the options that are going to crop up in this discussion before long. None of this is new but it is either being rediscovered or people are searching through old books ( you have to have the books, this early stuff is not on the net).

If i can find the papers I have and they are still on the net I will post the links. if they are not on the net any more I can't scan them and post them with this new board.

Thanks for that.
Flux

Flux · « **Reply #47 on:** October 23, 2011, 05:07:49 AM »

Here is a nice easy one to get started with.

http://www.elkraft.ntnu.no/eno/projects/PublStev/toroide.pdf

Flux

Flux · « **Reply #48 on:** October 23, 2011, 05:23:48 AM »

Here's another to send you off down a dead alley.

http://www.nrel.gov/docs/legosti/fy98/24996.pdf

Sturmey cycle dynamo in disguise? If you want something to reactance limit this is the ultimate.

Flux

Flux · « **Reply #49 on:** October 23, 2011, 05:48:44 AM »

Here's another. google search finds things easily without me having found the original titles and authors.

The list at the end of this will probably turn up more hits, especially Spooner.

Maybe that some of this won't mean much, but you can usually ignore the maths and get a general idea especially about construction. Most of this is of course based on neo but it still is relevant to ferrite.

http://www-eng.lbl.gov/~rasson/windsail/alternator/ICEM2000.pdf

Flux

bob golding · « **Reply #50 on:** October 23, 2011, 08:19:15 AM »

mmm can see a few problems coming up here. not insurmountable but need addressing. the biggest one seems to be weight. haven't worked out how heavy a turbine with a 350mm OD core 50mm wide and 15mm thick will be yet. will do that later. with the extra weight of the magnets as well this is going to be a challenge for the bearing doing the yawing. dont think a welded cap on the end of a bit of scaffolding tube will be up to the job. probably going to need a longer tail as well. not sure the volvo hub would be up to it either. what would be nice would be a shaft drive so the alternator is on the ground. cant think of an easy way to do that though and incorporate the furling as well. any one got any ideas. be back to vertical axis designs soon.

Flux · « **Reply #51 on:** October 23, 2011, 09:02:01 AM »

Yes weight bothers some people ( including me) others seem able and happy to get several hundredweight up in the air.

That was the big virtue of neo. Even a conventional wound field slotted core alternator for any power at very low speed becomes a fair monster to the point that I used a gearbox for my early attempts. the gearbox was a failure and best forgotten about but the chain drive does look to be much more promising with proper weather protection and oil bath. If you don't want a monster then it has a lot going for it especially in your high wind area.

The torus is well Worth trying and it may possibly turn out lighter than some other forms of ferrite generator. I don't know how it will compare with a conventional ferrite magnet slotted core alternator. Cores are a problem all round for home construction.

Flux

bob golding · « **Reply #52 on:** October 23, 2011, 02:38:12 PM »

just found another usa core manufacturer. hopefully be helpful to someone.

http://www.toroidcore.com/reqinfo.html

B529 · « **Reply #53 on:** October 23, 2011, 04:08:14 PM »

Removed the rear magnet rotor and stator. Took Hugh's statement regarding the magnet rotor biting skin very seriously, all fingers still intact!

bob golding · « **Reply #54 on:** October 23, 2011, 04:16:45 PM »

very helpfull, but now i am confused. if the stator has 36 coils 3 sets of 12 coils per phase? why does it only have 8 magnets? i know the flux path is differant but having trouble visualising how it works.

bob golding · « **Reply #55 on:** October 23, 2011, 04:20:57 PM »

found a site with a core weight calculator after much searching.

http://www.nicore.com.cn/en/03_product/product.asp?ptype=0&pid=

seems my worries about weight are unfounded. core will only weigh 5 kilos. ok a bit more with the wire and resin,but not as bad as i thought.

bob golding · « **Reply #56 on:** October 23, 2011, 07:16:19 PM »

think i have worked it out. each magnet covers 4 coils? making up for lost time here no posts for 2 years then 15 in 2 days. will shut up soon promise.

phil b · « **Reply #57 on:** October 23, 2011, 09:04:13 PM »

I'm glad to hear you do have all your fingers after seeing how the Proven is put together. That looks tricky, especially with no jacking screws. Must be a trick to it.

36 coils and 4 poles? only 6 bolts holding the stator? I would not have guessed either. It looks like the air gap is preset by the width of the magnet rotors where they slide onto the shaft.

Thanks for posting the pics B529.

jlt · « **Reply #58 on:** October 23, 2011, 10:37:04 PM »

Now I am totally confused. I would think with 8 poles you would use 28 coils.

phil b · « **Reply #59 on:** October 23, 2011, 11:11:05 PM »

I am asking a question in my last post because I'm unsure. There are 4 north and 4 south magnets on each stator, albeit 2 N together and 2 S together also. Unless this is different than I'm thinking, it would be 4 complete N/S poles.

I could be the one confused.

scoraigwind · « **Reply #60 on:** October 24, 2011, 03:37:53 AM »

Proven have built several different sizes of turbine. I think this might be a rotor from the (now obsolete) 600 watt machine? I am not sure but I think that the larger 2.5kW (or Proven 7 or now Kinspan 3kW turbine) has 12 magnets. It's hard to remember the numbers since I never work with Provens these days.

Flux · « **Reply #61 on:** October 24, 2011, 04:06:39 AM »

Yes that is obviously an 8 pole machine, in which case it will have a stator with 24 coils. The other stator picture posted with 36 coils will have the 12 pole rotor.

In the same way as an axial, the number of poles is determined by one magnet rotor. The other rotor is just feeding flux in parallel in this torus case. For the axial it is feeding flux in series.

As Hugh said, the smaller machines would likely have less poles aith less circumference to fit them in.

Flux

scoraigwind · « **Reply #62 on:** October 24, 2011, 04:22:45 AM »

If you study the stator very carefully you can see there are 24 coils in that one.

B529 · « **Reply #63 on:** October 24, 2011, 07:53:08 AM »

It's a WT2500.

ChrisOlson · « **Reply #64 on:** October 24, 2011, 01:45:03 PM »

Quote from: scoraigwind on October 22, 2011, 06:23:36 PM

They are only ferrites, but those rotors can bite off a lot of skin if they don't have proper jacking screws (and Provens dont!)

Hugh makes a good point for folks fiddling around with ferrite magnet rotors. They are weak compared to neo but even my little 11" ones on my geared turbine will pinch your fingers pretty good if you get careless with them. And that's what happens - you get careless and the next thing you know you got your fingers pinched between a rotor and stator and can't get them free because you thought it was so week you could put it together without jacking screw or puller. Two of my fingernails on my left hand are getting recovered from that now.

Anyway - carry on. Watching this thread with mild interest. I helped take down a Proven 2500 watt that had a burned out stator this spring. It is an impressive machine, and very interesting. I do not think you're going to beat the all around performance of an air core axial with an iron core machine, but it is interesting nonetheless.

And those generators on the Proven are VERY big and VERY heavy too BTW.
--
Chris

SparWeb · « **Reply #65 on:** October 24, 2011, 01:58:11 PM »

Bob,
Another link you may want to pursue: http://www.cartech.com/techarticles.aspx?id=1624

Not sure how much is just marketing spin, but they clearly are a supplier of the soft iron you'd require to "roll your own" laminations.

Their datasheet for Silicon-Iron-"B": http://cartech.ides.com/datasheet.aspx?i=103&e=191&c=TechArt
Available in sheet strips...

News:

Author Topic: Rewinding a Proven stator? Hugh? (Read 27421 times)