Author Topic: Math formula calculating number of magnets per coil ratio for a phase  (Read 61347 times)

0 Members and 1 Guest are viewing this topic.

Yachtsman

  • Newbie
  • *
  • Posts: 1
It would be really nice to see a simple formula that describes what is the optimal ratio of coils to magnets for a given phase. I would like to confirm what I believe to be correct based upon my research:

The Phase (P) really needs to be a prime number; 1, 2, 3, 5, 7, etc. Optimization occurs quickly up to 3-Phase, and increases become less with 5-phase and even fewer increases with 7.

The number of Coils (C) used for the stator must be a multiple of P; it must be evenly divisible by the number of phases. It can be an even or odd number of coils. e.g. 5-Phase could have 20 coils.

The number of Magnets (M) laid out evenly spaced onto the rotor disc MUST be an even number (because we match North and South poles); it's best (optimal) if there are more magnets than coils (so M is greater than C); and the most optimization is found when M is evenly divisible by P + 1; divisible by P - 1 can work, but the most optimized is P + 1; e.g. 5-Phase, with 20 Coils, would be optimal using 24 Magnets; 32 or 36 can be used, but optimal is 24.

Therefore a formula to calculate the optimal number of Magnets (M) placed onto a single rotor would be:  M = (P + 1) * C / P

A dual rotor design that sandwiches the stator in between two rotors would obviously require twice as many magnets.

I'm trying to learn and make something that is factually accurate and simple to use. I see examples where fewer magnets are used and the gaps are filled with extra coils--this works, but it's not optimal. I also see examples where too many magnets are used--this is expensive and it seems that it becomes counter productive to what you are trying to do in order to create the alternating sine waves in the phases.


kitestrings

  • Hero Member
  • *****
  • Posts: 1376
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #1 on: August 03, 2012, 10:36:34 AM »
Yachtsman,

Welcome.  I think the reason you may not have gotten alot of response on this post is that it simple too broad.  I don't believe there is a single, simple formula that will meet all conditions. 

Others here can speak to the theory much better than I can, but much of the alternator design hinges on trying to get a good power match between the rotor (prop), for a given size, and the alternator.  There are some good rules of thumb if you google this site or grab one of the Hugh's or the Dans' books.  And, for an alternator with a known number of poles and magnet size, for example, there are some useful formulas for determing the total flux /pole, frequency, rms voltage and number of coil turns to obtain a desired cut-in voltage - so it is not all design trial and error.

Broadly, 3-phase seems to be the prevailing PMA design, though Hugh and others have documented other workable designs.  Single-phase appears to have some limitations and problems with undesirable vibration as I undertand it.  The most common relationship of magnets to coils seems to be a 4:3 ratio, e.g 12 magets (24 in a dual-axis rotor) with 9 coils, though again other possiblies exist.  I can't say as I've ever seen a 2 or 7-phase design here.

If you can narrow up your goal, and maybe offer some specifics (power requirements, voltage, rotor size, etc.) you might find you'll get better feedback.

~kitestrings

ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #2 on: August 03, 2012, 01:19:18 PM »
I can't say as I've ever seen a 2 or 7-phase design here.

My 4.0 meter turbines have a 10 pole two-phase generator with ferrite magnets that produces 2 kW @ 62% efficiency on a 24 volt system without using MPPT.  It also utilizes dual stators, which has been hammered on as being "inefficient" by Hugh and some other folks.  But I have not seen anybody match it for performance yet in a ferrite magnet axial generator design.  I have built some 3 kW ferrite magnet three phase axial generators, but those use MPPT.







Basically, you can make any number of poles and coils work out with any number of phases.  And you can even use a unipolar design that doesn't have alternating N-S poles, and thereby use an odd number of magnets.  So no mathematical formula is going to work to describe it because if you do come up with a formula I can build a generator that will defy it.
--
Chris

definitionofis

  • Jr. Member
  • **
  • Posts: 58
  • Country: is
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #3 on: August 03, 2012, 02:16:38 PM »
I was thinking any number of phases could exist. I could not see why not.

I checked in on this thread to see if that was confirmed.

I just watched a Tesla video.  I think his motor had two phases and four coils, if I observed it correctly.
I think poly-phase, as he called it, is more difficult to grasp than relativity. haha

It was good, without too much mumbo-jumbo "hidden secrets" talk:
http://www.youtube.com/watch?NR=1&v=eoY_7mbm5ng
« Last Edit: August 03, 2012, 02:21:45 PM by definitionofis »

scoraigwind

  • Sr. Member
  • ****
  • Posts: 423
  • Country: gb
    • www.scoraigwind.co.uk
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #4 on: August 03, 2012, 02:40:08 PM »
I wrote this earlier and now I see a lot of replies but it still probably is worth posting:

I started out making single phase with one magnet per coil.  You have to reverse every second coil as the magnet polarity reverses.  I don't like the vibration and amd not convinced this is the most electrically efficient.  So since then I have mostly used 3 coils and 4 magnets (I invented for myself, this although I am sure I am not the only one to do so).  I also have used 10 coils for 12 magnets in a five phase (in 'how to build a wind turbine' 2005).  That works well but not so suitable for AC transmission as there are at least 5 wires.

During the years I have mostly been trying to make coils that are large enough to completely encircle the magnet face i.e. the hole in the middle is the same size and the magnet face.  My logic is that this makes the best use of the magnet although I may be wrong of course and a smaller hole does seem to give good results too.  Then I spaced magnets widely so as to get as much copper into the coils as possible with minimum of magnet material.  This works well with a 4 coils/3 magnets ratio.

Lately I have grown interested in using ferrites since they are cheap and durable.  My old ferrite magnet rotors from the 1990s go on forever.  Since then the ferrite/ceramic magnets have got a lot cheaper.  So it makes sense to me to cram them in on the rotor, unlike the neos where I leave gaps.  This alters the coil shape criteria a lot.  it starts to make sense to have fewer, smaller coils and use smaller holes to maximise the turns per coil and reduce the length per turn.    Big coils don't gain because they start to overlap other magnets and this is counterproductive.

Looking at choosing ratios of numbers I noticed that it's not easy to find ones that work perfectly, but the other ones are actually quite usable with care.  If the coil phases are random, but can be grouped in three, and you choose ones that are as close together as you can (reversing the wires in half the cases) then the coils turn out to be only 10 or 20 degrees (electrical phase angle) away from the correct phase which only loses between 2 and 6% of your voltage. 

I had fun designing a ferrite alternator with 10 poles and 12 coils for example.  http://scoraigwind.co.uk/2012/05/clearer-drawing-of-the-10-pole-12-coil-design/  The loss compared with a design where the coils are perfect 3-phase in this case was only 3.5% of voltage due to coils being slightly out of position.  The alternator worked really nicely and I would do this again although I didn't like gluing magnets together side by side to make the ten big poles.

Finally I can't resist hammering one more time on Chris and saying that yes of course you can improve efficiency using gearing and MPPT electronics but in the end you will get (enormously) better bang for your buck putting all of your magnets and coils into a single stator alternator than you would with a 2-stator alternator, all other things being equal and assuming that you are running at the same speed and all else is equal.  Bearing loads are the only valid argument that I hear against that.
Hugh Piggott scoraigwind.co.uk

ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #5 on: August 03, 2012, 03:10:40 PM »
Finally I can't resist hammering one more time on Chris and saying that yes of course you can improve efficiency using gearing and MPPT electronics but in the end you will get (enormously) better bang for your buck putting all of your magnets and coils into a single stator alternator than you would with a 2-stator alternator, all other things being equal and assuming that you are running at the same speed and all else is equal

<chuckle>  I enjoy making the die-hards scratch their heads   ;D

I have to agree on the ferrites though - after I got past the "conventional" designs using neos I've come up with all sorts of different configurations of poles, coil arrangements and stators that work with ferrites in an axial design - some of them surprisingly well.  I still like your 10 pole 12 coiler since the first time I saw it.
--
Chris

definitionofis

  • Jr. Member
  • **
  • Posts: 58
  • Country: is
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #6 on: August 03, 2012, 05:16:34 PM »
I just noticed those tightly spaced same-color scoraigwind coils are wound in an interesting series way. At first I thought the turns direction of touching coils was wrong. It is not. It is an interesting idea. I have never wound a stator.

ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #7 on: August 03, 2012, 06:23:58 PM »
I got a 16 pole 27 coiler about half built for a direct drive ferrite magnet 3.5 meter turbine that I built.  It's kind of the same as Hugh's 10 pole 12 coiler, except different and a lot more complicated    :)

I fiddled around with my CAD (Cardboard Assisted Drawing) system for about two weeks before I dreamed it up.  It uses the same 16 pole rotors as my geared 3 kW ferrite generators, but I expect its output to only be about half of the geared unit.
--
Chris

gww

  • Hero Member
  • *****
  • Posts: 762
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #8 on: August 03, 2012, 06:40:58 PM »
Chris
1500w for 12' ferrit mag turbine doesn't seem bad to me.  How many pound of copper, size rotors and mags?
Thanks
gww

ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #9 on: August 03, 2012, 09:54:56 PM »
The rotors are 355mm, 16 poles or 32 magnets total.  The mags are 2 x 2 x 1" thick ferrite blocks.  I don't know about the copper yet as I just ran a test coil the other day and decided I need 50 turns in the coils with the voltage losses because poles don't line up.  But haven't wound it yet.
--
Chris

gww

  • Hero Member
  • *****
  • Posts: 762
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #10 on: August 04, 2012, 02:13:29 AM »
Chris
Thanks
gww

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #11 on: August 04, 2012, 04:24:45 AM »
I skipped this question for the reason that kitestrings mentioned. I don't like reducing everything to formulae, sometimes there is only one solution and it works but I can't see it working here.

Hugh touched on the basics of one coil per magnet per phase and it starts there. Most commercial iron cored machines ( motors or alternators) divide the coil into several smaller ones ( overlapped or concentric) to fit into a reasonable sized slot in the punchings and for various reasons of waveform distortion and reducing copper.

There is a variant on this favoured in N America where the thing is based on one coil per magnet pair  ( they leave out half the coils, and don't reverse anything) it produces similar results and relies on making virtual poles from the missing coils.

Once we abandon iron cores we are going away from conventions that have been developed over 100 years although some early work was done on single phase in the late 19th century.

Without the iron core even a single coil has the same effect as distribution in a conventional motor so there is no point in splitting one coil into more so you will not normally find more coils ( per phase) than magnets.

Again you can start from the option of using one coil per two magnets and end up with half the coils.

For anything other than single phase these arrangements need overlapped coils as for the iron cored case.

If you choose to use an axial design ( again going back to ideas from the 19th century) you find that the overlap is difficult at the centre of the coils where space is badly restricted, I started off this way and without slots the thing is very difficult to wind and you have long turns with a lot of resistance.

I was intrigued to see what Hugh had done by leaving more coils out so that it fitted on a single layer without overlap. With conventional loading as for an engine driven generator with normal load this arrangement doesn't use the whole winding area and doesn't look very attractive but when you bog things down into a rectifier and battery then things change and the resistance becomes a major factor and has far more effect than the small loss of emf from the un wound areas.

It turns out that the 4 magnet 3 coil arrangement is hard to beat for direct battery charging applications.

When you take this further by considering other phase numbers I think it becomes obvious that reducing things to simple formulae doesn't work as each application is different and what works best for an iron cored engine driven alternator feeding a conventional ac load may be far from the best arrangement for an ironless direct coupled alternator charging batteries via a rectifier.

Flux

ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #12 on: August 04, 2012, 11:39:37 AM »
It turns out that the 4 magnet 3 coil arrangement is hard to beat for direct battery charging applications.

I'd like to add a clarification to this.  There are some situations where, for a given number of poles, the 4:3 configuration yields coils that are too fat for my liking.  Adding more coils for the same number of poles gets the fatness of the coils down so they can be wound tighter and help prevent vibration.

The other thing is that three-phase is not necessarily the cat's meow.  I've found that two-phase runs just as smooth and has some built-in features that make the generator less "stiff" to help provide better matching between generator load and blade power curve than three-phase typically does.  Many folks get too hung up on building a generator that's as efficient as possible, and without MPPT all those generators do is stall your blades and turn the turbine into a dog in higher winds.
--
Chris

scoraigwind

  • Sr. Member
  • ****
  • Posts: 423
  • Country: gb
    • www.scoraigwind.co.uk
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #13 on: August 04, 2012, 05:07:22 PM »
It would be really nice to see a simple formula

Here is a formula that could be useful.  Names of variables may or may not be unambiguous but try the maths.

Magnet pair pitch (physical angle)= 720 / number of magnets
coil pitch (physical angle)= 360 / number of coils
phase shift (phase angle)= 360*(Magnet pair pitch-coil pitch)/Magnet pair pitch
..... and then you work with that shift to see how many phases you get.

example.  8 magnets, 6 coils, magpitch = 90, coil pitch = 60, phase shift =360*(30/90) = 120 degrees
this gives a simple 3-phase output.

example 12 magnets, 10 coils, magpitch 60, coilpitch 36, phase shift = 360*(24/60) = 72 degrees
which gives a 5-phase output.

10 magnets, 9 coils, magpitch 72, coilpitch 40, phase shift = 360 * (32/72) = 160 degrees
which could be 9-phase. 

However you can make a pretty workable 3-phase if you reverse some coils. Connect 1,-2 and 3 in series, 4,-5 and 6 in series and finally 7, -8 and 9 in series.

It helps to make a diagram of the phase angles and see how best to group them, never forgetting that you can reverse the wires to a coil and change the phase angle by 180 degrees.

all good fun
Hugh Piggott scoraigwind.co.uk

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #14 on: August 05, 2012, 03:57:07 AM »
That is rather neat and about as useful as you are going to get to satisfy the original question.

It will tell you what you will get, I still can't see how you can arrive at an optimum from it but as long as you stick to air gap alternators it should be useful.

It covers the case of one coil per pole pitch per phase and that is logical for air gap machines. With iron cores you usually have more coils per pole pitch per phase with distribution. This formula could be adapted to cover this. Hugh hinted at this when he said that the 9 phase case could be connected for 3 phase.

Similarly by reversing the connection of diametrically opposite coils in a 3 phase winding you get 6 phase, giving an example of using the 180deg phase shift.

Someone above said that polyphase is as confusing as relativity, unless you have some natural ability to see things in 3 dimensions I tend to agree.

Flux

scoraigwind

  • Sr. Member
  • ****
  • Posts: 423
  • Country: gb
    • www.scoraigwind.co.uk
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #15 on: August 05, 2012, 04:55:36 AM »
by reversing the connection of diametrically opposite coils in a 3 phase winding you get 6 phase, giving an example of using the 180deg phase shift.

I suppose I'd mostly be looking to reduce the number of phases to 3 so as to allow me to easily have a rectifier at the battery rather than on the turbine.  So I'd be more likely to use the reversal trick to reduce the number of phases from 6 (where there is 60 degrees between coils) to 3 (where there is 120 degrees). 
60 - 180 = -120 which amounts to the same thing as +120. 
60 + 180 = 240 = -120

In some cases the phases are almost 'random' but you can group a third of them to within a 60 degree 'band' by reversing some of them (+/-180 phase change).  If you connect that group all in series they can make a presentable job of a phase even though they are plus/minus up to 30 degrees off the mean.  In reality they probably only average about 15 degrees out of phase with the mean.  15 degrees out of phase means a loss of only 3 or 4% of voltage.

So it seems to me that you could use just about any ratio of coils to magnets.  The choice boils down more to what you reckon will be an optimal coil shape in relation to the distribution of magnetic flux.  Larger coils do a better job of producing maximum voltage per coil turn, but smaller ones can allow more coil turns in the stator, and they are also shorter and hence can be thinner and more numerous for the same total resistance.
Hugh Piggott scoraigwind.co.uk

electrondady1

  • SuperHero Member
  • ******
  • Posts: 3120
  • Country: ca
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #16 on: August 05, 2012, 08:44:31 AM »
last year i wanted to do a two phase stator with overlapping coils
in this drawing i  was checking out the two types of single phase
and comparing them with a 3phase coil.
the general consensus was the coil per pole method was superior.
it was a real struggle but i was able to build a two phase 32 coil 16pole overlapping stator.


ChrisOlson

  • SuperHero Member
  • ******
  • Posts: 3642
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #17 on: August 05, 2012, 09:55:12 AM »
the general consensus was the coil per pole method was superior.

I wouldn't say it's superior, just that using ferrite magnets with the "conventional" 4:3 three-phase design can tend to yield coils that are very fat and the outer turns on those coils add a lot of wire length to the winding.  Using the 1:1 ratio of coils to poles, per phase, puts more coils in series with a more manageable number of turns in each coil - and the interconnects between the coils are very short so basically no resistance is added there.

With two-phase you can either use four wire output, or connect the phases series for two-wire output to get 1.414x single phase voltage.
--
Chris

electrondady1

  • SuperHero Member
  • ******
  • Posts: 3120
  • Country: ca
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #18 on: August 05, 2012, 10:12:38 PM »
looks like i forgot to post the image doh!
i used four wires out and rectified separately
i'm happy with the stator
 
« Last Edit: August 05, 2012, 10:18:38 PM by electrondady1 »

WindyOne

  • Newbie
  • *
  • Posts: 44
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #19 on: January 25, 2013, 10:08:32 PM »
last year i wanted to do a two phase stator with overlapping coils
in this drawing i  was checking out the two types of single phase
and comparing them with a 3phase coil.
the general consensus was the coil per pole method was superior.
it was a real struggle but i was able to build a two phase 32 coil 16pole overlapping stator.
ED1,
 Did your overlapped coils cause the stator to be (approx) twice the thickness of a typical stator with side-by-side coils?

 

electrondady1

  • SuperHero Member
  • ******
  • Posts: 3120
  • Country: ca
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #20 on: January 26, 2013, 08:03:25 AM »
no ! it's  about the same as one layer.




Frank S

  • Hero Member
  • *****
  • Posts: 1901
  • Country: us
  • Home with a view of Double mountain
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #21 on: January 26, 2013, 08:45:15 AM »
While we are on this subject I was wondering if anyone ever built a rotor where the magnets were embedded in the rotor exposing the N & S poles on opposite sides?
 See drawings. If so would it be better to use steel , aluminum, or other material for the rotor
 And would it be better to align the coils direct opposite of each other in the N/S or parallel series them 











 As soon as either the CNC plasma or the water jet machines have some slack time. I might as well cut out several disc from various materials. I won't be able to build any rotors until after I relocate, but at least while I have access to these machines it would be a good idea to make use of them. For the retaining clips I can either cast them from aluminum or machine them. After I leave here it might be a year before I am in position to buy a CNC Laser machine like I want.
I live so far outside of the box, when I die they will stretch my carcass over the coffin

tanner0441

  • Hero Member
  • *****
  • Posts: 1098
  • Country: wales
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #22 on: January 26, 2013, 10:26:47 AM »
Frank

If you have access to all the machinery could you not make the clamps out of stainless could you cut the blanks with your water jet or laser cutter or even a guillotine and as it is only a small run get away with the forming punch and die out of something like EN97. Even without heat treating it should last long enough to more than do the number you want, and you could put a bit of set on them to clamp the magnets. You could fabricate the die and use compressed air for the ejector.

I would think it is a lot easier than die casting moulds.

Brian.


Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #23 on: January 26, 2013, 11:07:08 AM »
This one keeps cropping up. The basic essential is to have as near as possible a closed magnetic circuit. The arrangement as you show it with coils on either side of the magnet is bad. To make it useable you need two more steel discs to close the magnetic circuit.

If you do as Chris Olson has shown above you basically have two dual rotor machines stacked together. You could use your idea for the centre rotor and just have one magnet instead of the two that Chris used on the middle rotor.

If you share the centre magnets you then get 3 magnets in series with 2 air gaps. This will work fine but give less flux than the arrangement chris used which has 4 magnets in series and 2 gaps.

I just can't see the virtue in making a complicated mechanical arrangement to have a non magnetic centre disc just to do this. If you embed the magnets in a steel disc it is virtually useless but if you stick one each side then the steel disc works, it's simple to do and you get more flux than sharing the centre magnets between 2 air gaps.

There seems to be some mistaken idea that by doing as you suggest you use both poles of the magnet. When you resolve the thing down into magnets and iron circuits you will find that both sides of the magnets are always used,

The one exception is a single rotor with no return flux disc, your idea just makes two poor single rotors back to back with flux having a longer and less useful path than the miserable single rotor.

 If you want to save magnet material and consider 3 magnets in series to be enough for your dual stator arrangement then just use half thickness magnets on the centre disc and stick them on the surface rather than making a complicated non magnetic disc with the thicker magnets.

Flux

Frank S

  • Hero Member
  • *****
  • Posts: 1901
  • Country: us
  • Home with a view of Double mountain
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #24 on: January 26, 2013, 12:33:05 PM »
Flux you were the exact one I was hoping to have jump in.
 I understand that center mounting the magnets in a steel disc pretty much kills a flux pathway.
 What I was wondering about and didn't explain it very well was having the magnets mounted on non magnetic disc and possibly even the coils with a core in the holes of the coils. after you explained it it seams that even should the coils be mounted on steel discs the pathways would not be pure as the discs would have to be some kind of tricky spiral laminate concoction.
 saving magnet material is not much concern to me I can use 12 ,24 or 188 for that matter. However mounting does concern me. machining a mounting wedge system or caps is fine for magnets and coils but I stopped using glue & paste when I was in the 1st grade. Resins and glues are either too permanent or too  temporary there is no in-between and changing after a build is not a good option with them on NEOS.
 I've seen breaking systems made out of magnets and aluminum discs by causing the eddy currents to fight each other.
  The manufacture stated that these NEO40s have a 4300 Gauss or 245 lbs pull on contact with a 3/8" plate
 this would seam to be about 4300 Gauss if 2 magnets were placed on a steel disc with a .95 gap between them . if I were to embed the copper in a single disc between 2 disc with magnets logic says this would be the better way to do things but I am more Klingon than Vulcan.
 Hence the thought of 1 magnet between 2 thinner coils with a core in the coils and a double .1 gap.
 As long as I can tear things apart and start over it doesn't cost me  anything. once glued it is like a General Motors door skin it either fits correctly or forever is wrong           
I live so far outside of the box, when I die they will stretch my carcass over the coffin

Flux

  • Super Hero Member Plus
  • *******
  • Posts: 6275
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #25 on: January 26, 2013, 02:09:22 PM »
The wedge clamps are a good way to hold the magnets if you don't happen to have the magnets with holes in them.

Using fixed laminated discs behind the coils will work but it needs high quality magnetic strip wound into a torus. Works fine but is difficult to get and when built specially very costly. This is the real limitation to single rotors with a fixed magnetic circuit. At least your version would be partly self balancing for the axial forces which are very high with this arrangement.

You need to consider how you make the tape wound core strong enough to resist being pulled apart like a fly catcher ( if you use those things in your part of the world).

I would still go for magnets each side of a steel disc rather than common flux and a non magnetic support.

If you can get good quality core material and make it physically strong it will work fine and with the coils stuck on to fixed iron circuits it cools better than the normal air gap design.

I infinitely prefer the radial version with coils stuck on to a smooth laminated iron core and a magnet can rotating outside it. The core can be a dirt cheap scrap motor core and it will be good quality material. The magnets don't fly off when stuck on the inside of a drum and again the cooling is good. I have built several this way and it is near ideal. The iron loss is much lower than with slotted cores, there is no cog, it has most of the advantages of the air gap machine and is more robust and better cooled.

It is not really an option for those with little facilities but with a lathe it is easy to build. The only snag is that you have to get the winding right first time as you can't play with the air gap to compensate, but with a test coil this is no real problem.

For anyone contemplating using the Classic controller it seems the ideal way as you are no longer constarined by having the turns exactly right.

Flux

WindyOne

  • Newbie
  • *
  • Posts: 44
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #26 on: January 26, 2013, 04:09:37 PM »
no ! it's  about the same as one layer.
ED1,
 How did you stack one coil over the other coil and still maintain the same thickness a "normal" stator?

 

Frank S

  • Hero Member
  • *****
  • Posts: 1901
  • Country: us
  • Home with a view of Double mountain
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #27 on: January 26, 2013, 04:26:34 PM »
The wedge clamps are a good way to hold the magnets if you don't happen to have the magnets with holes in them.

Using fixed laminated discs behind the coils will work but it needs high quality magnetic strip wound into a torus. Works fine but is difficult to get and when built specially very costly. This is the real limitation to single rotors with a fixed magnetic circuit. At least your version would be partly self balancing for the axial forces which are very high with this arrangement.

You need to consider how you make the tape wound core strong enough to resist being pulled apart like a fly catcher ( if you use those things in your part of the world).

I would still go for magnets each side of a steel disc rather than common flux and a non magnetic support.

If you can get good quality core material and make it physically strong it will work fine and with the coils stuck on to fixed iron circuits it cools better than the normal air gap design.

I infinitely prefer the radial version with coils stuck on to a smooth laminated iron core and a magnet can rotating outside it. The core can be a dirt cheap scrap motor core and it will be good quality material. The magnets don't fly off when stuck on the inside of a drum and again the cooling is good. I have built several this way and it is near ideal. The iron loss is much lower than with slotted cores, there is no cog, it has most of the advantages of the air gap machine and is more robust and better cooled.

It is not really an option for those with little facilities but with a lathe it is easy to build. The only snag is that you have to get the winding right first time as you can't play with the air gap to compensate, but with a test coil this is no real problem.

For anyone contemplating using the Classic controller it seems the ideal way as you are no longer constarined by having the turns exactly right.

Flux

To keep from hijacking this thread further I will carry our conversation over to my thread
http://www.fieldlines.com/index.php/topic,146738.0.html this way the question that the OP asked won't get buried
I live so far outside of the box, when I die they will stretch my carcass over the coffin

coldspot

  • Hero Member
  • *****
  • Posts: 843
  • Country: us
Re: Math formula calculating number of magnets per coil ratio for a phase
« Reply #28 on: January 28, 2013, 01:50:12 PM »
Maybe this could help ?

"a post by willib"
"What Rotor Size do I use"

Way back machine to
Jan 2010

??
$0.02