Building stacked up axial alternators

A lot of people ask me if it’s a good idea to stack the axial flux alternator with 3 magnet disks and two stators. The middle magnet disk has magnets on both sides like this one by Chris Olsen:

On the face of it this should be a good idea – it makes for a more compact design – and Chris for one is very pleased with the results.

After all two alternators can make twice as much power as one.  However if you use the same materials to make one big alternator then it can be a lot more powerful still.  Or if you use 2/3 of the magnets and 2/3 of the coils to make a slightly larger one then it is just as efficient as the stacked version.  Having all of the magnets interact with all of the coils gives a big advantage to the single-stator unit.

I have different design ideas from Chris (for example I use 3-phase ‘star’or ‘wye’ connected windings) and I cannot easily design an alternator like his, but I have designed and built many to my own ‘recipe‘ so I can produce designs for various sizes with predictable outputs.  I decided to test this using my alternator design for the 3.6 metre diameter, 48-volt recipe.  I put together a design for a stacked alternator with two stators instead of one. I used the following specifications:

dual stator unit single stator unit
no of magnets 12×4=48 16×2=32
no of coils per stator 9 12
Total number of coils 18 12
Weight of copper 7.1kg 4.75kg
diameter of magnet disks 323mm 400mm
number of disks 3 2
combined weight of disks 19.3kg 19.7kg
cut in speed for 48VDC 141 rpm 141 rpm
resistance of stator 1.1 ohms 1.0 ohms
wire size (diameter) 1.4 1.8
turns per coil 128 75

I chose the disk diameters so that the coils are exactly the same size and shape.  But each coil in the stacked alternator only has 12 magnets passing whereas the ones in the single stator get 16 magnets passing.  The result is that you get (almost) the same resistance for the same cut-in rpm, so overall the two alternators behave the same.  The only real difference is that the cost of the magnets and copper is much lower for the single stator machine.  Both can handle around 1000 watts maximum continuous output with good cooling.

I am happy to answer questions about this and all my other pages and posts.



35 Responses to Building stacked up axial alternators

  1. Pingback: Axial alternators with dual stators | Hugh Piggott's blog

  2. sanjeev says:

    As you mentioned the cost of copper is bit lower.

    Chris was using Ferrites, are you using the same? is the size of those magnets same?
    and size(diameter) of the coils are mentioned in mm right?

  3. Chris Olson says:

    I made a comment on this on The Back Shed in this thread:

    In a nutshell I have designed and built a new (bigger and more powerful) turbine for the ferrite magnet generator. Some of the concepts I used, and my reasoning, are explained in that post on The Back Shed dot com.



  4. admin says:

    Chris uses ferrite magnets, which is a good choice for low cost and durability. He overcomes the low flux problem by using a gearbox (chain drive system). I can never pin down the numbers for his designs because he uses single-phase and each description is different, so I can’t specify exactly how he could gain by using a single stator with his magnets, but it is clear to me that he would. Instead I give an example using the magnets and coil shapes that I work with, to illustrate the benefits.

    I have great admiration for what Chris is doing, but I just needed to correct his allegation that a single stator had to be 2.5 times the size. I doubt if we shall ever agree on this but I have given an example of two alternators using the same size coils where the single-stator one uses less magnets/copper and is also more efficient. It can put out higher power in relation to the power that is used to drive it.

    Chris has reasonably pointed out that even though it’s more efficient it will get hotter due to having less surface. It’s losing less energy but it has to lose it over a very small surface area. Or I think that’s what he was driving at on the backshed. And it is true. I played with that aspect in the design and checked out how much bigger it had to be to be able to use the same energy dissipation over its surface.

    By increasing the larger disk size to 430 instead of 400 mm I can fit 2 wires at 1.4 mm diameter into the coil same as the dual stator machine. It still uses less copper overall, the steel disks are ony 15% heavier, and the resistance is down to 0.4 ohms compared with 1 ohm for the dual-stator version. So the single-stator version wins in every way apart from having larger diameter. The rotor is 1/3 larger diameter than with the dual alternator. And it will also stall sooner, due to being more efficient but I’d prefer to have a series heater in the shed (or my office) rather than overheating the stator on the windmill.

    If Chris were to build a ferrite alternator with 2 rotors 1/3 larger, then he would see the same benefits. If you were to take the materials that he put into the dual stators and make a single stator machine then you could discard the gearbox and use direct drive.

  5. admin says:

    You can see more debate on the issue here on the Fieldlines board:,145700.90.html
    You have to move down the page about half way before it gets going.

  6. Wes says:

    I tried to build a similar design with 3 stators but the middle rotors were 1 magnet thick with no steel. I was trying to get an endplate to endplate magnetic circuit hoping I would get similar flux with fewer magnets. However, the neo magnets I used were so strong they warped the thin fiberglass rotors. It might work with stronger material like carbon fiber but the extra cost would probably cancel out the advantage of buying fewer magnets.

    • admin says:

      hi Wes

      I can see how you might think there is a free lunch to be had here but there is not. The flux in a gap depends on how many magnets there are in series (e.g. one on each side or whatever) versus the width of the gap. If you put a magnet between gaps then it will have less power because it has two gaps to deal with.

      The whole dual stator thing is a poor choice in my opinion, and it is interesting that Chris Olsen is now building larger diameter, single stator machines with 3-phase output. His main argument for the dual stator was that the smaller diameter was easier to build and had lower inertia but neither of these have been a problem for me anyway.

      Using a single stator you will allow all of the magnets to interact with all of the coils whilst only having one air gap, so there is a huge benefit in terms of the usage of the magnets and the coils.

  7. Wes says:

    Ok maybe I came across a bit like a perpetual motion clown looking for a free lunch when I mentioned the idea I would get similar flux with fewer magnets but you might not think so If you take this into consideration. In My design with the magnetic circuit being continuous from end to end of the generator all of the magnets do interact with all the coils unlike the designs you mention that don’t because they are divided by a steel plate that acts as a magnetic shield.
    Do you think that this magnetic division might be the reason why your single stator design always wins by leaps and bounds? After all, any same size wire crossing the same magnetic flux at the same speed should generate the same power so where else in the stacked version could there be so much loss of efficiency?

    • admin says:

      I am sorry I am not making myself clear about “all of the magnets and all of the coils”. The voltage is related to the rpm and the number and size of magnets directly filling the air gap with flux. It’s the total flux in the air gap that matters. If you just have one stator and 2 rotors then the area of magnets is twice as large as it would be with two stators and 3 rotors, as in your design. So you will get twice as much voltage from the single stator design (per coil) at the same rpm. This is the basic logic from which I conclude that the dual stator concept is never going to be as good as the single one, unless you have some limit on the diameter of machine that you can build.


  8. Wes says:

    I appreciate when an expert like you has patience with a novice like me.

    After some research I must correct myself. The steel plate does not act as a magnetic shield. In fact the magnetic circuit does not appear to benefit by leaving the center plate out. Therefore, to get the same performance, the magnet count could not be reduced and it would be necessary to put half the coils and half the magnets in the top half of a dual stator machine and the other half in the bottom and gear it up 2 to 1 to create the same conditions that exist in the single stator design. I believe this agrees with what you are saying.

    If you are building a machine that will be geared anyway it should make little to no difference which way you arrange the same coils and magnets as long as you gear it accordingly but most small wind turbine builders don’t want to add the extra expense, weight and hassle of a gearbox nor put up with the friction losses.

    Of course this comparison is different from the one you offered where you designed two alternators that would give about the same output at the same rpm. If I’m understanding corectly, the difference in room you have for magnets around the circle on the different diameter plates while expecting the same output at the same rpm is why your dual stator alternator required more materials over all to do the same job.

    If this is correct would it also be correct to say that at the smaller diameter and number of poles, if the magnets and coils in the dual stator were rearranged so the magnets were stacked at the ends and the stators were stacked together like a single stator, the performance would remain about the same? If so it would seem that where the gearing option is out of the question, the ratio of performance to required materials (at least in wire and magnets) is determined more by the diameter of the alternator (if the space within that diameter is optimally used) than how many stators or rotors it has.

    Now before it is assumed that I see no difference between single stator and multiple stator designs as long as the diameter is the same, I must mention that there are flux leakage problems with magnets between two air gaps where some of the flux wraps around back on the same magnet or to adjacent magnets without crossing the gap. Minimizing these losses requires optimal placement of the magnets, which minimizes space for coils. These problems and others have to be dealt with if you cast magnets in nonmagnetic rotors. If you use steel plates for the center rotors, you’re just using more plate material to do about the same job as the single stator machine.

    The only way I see the potential for a gain is if flux jumping the shorter gaps would reduce the magnetic distortion created as the alternator spins and allow a little more flux to link the coils. Do you know of any testing or simulations that would show if there is any gain to be found in that department?
    Really, with fewer parts to build and less time casting discs I’m starting to think that single stator is the way to go.

    • admin says:


      You seem to be agreeing with me now which shows I am making progress. The comparison in the original posting uses less magnet and less coils and produces the same out put as the dual stator example because of the inherently better value of the single stator. The single stator may need a slightly larger clearance than the dual (to allow for bearing slop etc) and this will impact very slightly on the flux density you are right so in the real world nothing is quite so clearly cut and dried. As for optimum magnet size there was some work done on that in Athens and they concluded that the size I use is the best somehow but I never followed the logic of it.
      There are a lot of good documents (mostly free) on the windempowerment site.


  9. Randy says:

    Similarly I am increasing the size of my alternator to hopefully better match my lenz2 style VAWT tied to 24 volt bank. It is 51 ft sq cross section turbine. Currently I have slightly over 10″ dia dual rotors with total 24 1.5″ disc mags 70# pull each. I have used two different stators three phase star. Each nine coil one with 18AWG about 1.6 ohm per phase and one with 16AWG about 1.1 ohm per phase. I will stick with the 10″ discs. I understand the cost effectiveness you have mentioned however going to larger diameter rotor versus my investment already being made I can add on in sections using the dual stator approach so it seems to have experimental advantages in cases like mine. I will share some observations with you in hopes you might have some pointer on AFA configuration for my VAWT application. 10 to 20 MPH wind my 16 AWG stator breaks quite hard but continues to turn slowly while the 18 AWG stator does not break hard and spins with medium speed. Otherwise both appear to perform fairly similarly charging other than the 18 AWG having a less steep power curve is more consistent output. I have not observed stall. I start charging around 10 MPH. My goal to hit 100 watt and run my pellet stove around 8 MPH wind. There would appear to be surplus torque and too little speed for the current motor configurations. My plan to add layer by layer at a time to my AFA starting with an additional layer stator and one row of magnets and continue to add rows of magnets and stators. If you were to guess what configuration would you anticipate being optimal?

  10. admin says:

    I find the description a bit baffling to be honest. However I will summarise what I think the dynamics are when matching a turbine (vertical or horizontal) to the battery via a stator. If the wires is thin (18 AWG?) then you will have more turns in each coil. This makes for more voltage at lower speed and so you get an earlier cut-in. You will have more internal losses so the slope of power/speed is more gradual. This can make a good match to the variable speed of the blades but there will be poor efficiency. Using thicker wire (16 AWG) means less turns but also less resistance in the stator. You will cut in at higher speed. The speed will be more constant. This will work very well at a particular windspeed but will be too fast in low winds, and may stall the rotor as the wind passes the optimum level. You can also fine-tune the speed by changing the air gap, or even by changing the voltage if you are using for example a classic controller from midnite solar.

    If the rotor has plenty of torque, and the speed does not drop much when you connect, then yes it seems like you need a bigger alternator. But you could simply try using 20 AWG and winding more turns to obtain a lower cut-in rpm. If you only want 100 watts out (4 amps) then the higher internal resistance may not be a worry. It would be easier to follow the story if I knew the actual rpm and the actual turns per coil.

    You tell us that “10 to 20 MPH wind my 16 AWG stator breaks quite hard but continues to turn slowly while the 18 AWG stator does not break hard and spins with medium speed.” This signifies to me that the 16AWG stator is causing stall, and yet you say you have not seen stall. To be frank with you the VAWT is going to have a poor efficiency at best and fiddling with the alternator is not necessarily going to fix that.

    • Randy says:

      Perfect, that gives me some added confidence moving forward. Thank you for your trouble. I am sorry for any confusion and so respect all I have learned from you in the past by reading your literature and works. Having only made crude measurements the turbine seems to perform quite closely to the literature I have read on it. 1.8 TSR open circuit and 0.8 charging. About 60RPM at 10 mph winds where about it currently begins charging. 18 AWG coils are 150 turns each. By breaks I mean shorting out the motor, I have not observed stalling up to this point while charging that I am aware of. In considering a single finer wire stator I believe I would reduce the ability to break the machine even further which I need to increase. I am more confident now I am moving in the proper direction. I was working on my own assumptions and limited experience such as negating internal losses in the amperage operating range I pursue. So it helps much to hear from you.

  11. dan says:

    So while imight be asking just basic noob style questions, i am hoping you can help me with a few basic questions. Are the coils of the dual stators wrapped around the magnets, some of the magnets or what? I am orderkg up n50 neo magnets 2x1x1/2 and am planning on using 18awg wire with around 150-175 turns per coil. Will be wiring in 7 phase star. But was thinking of doing one star ad one delta wired. What if any issues do you see with this basic setup? Would i be better changing one of of wires to either 12. or 14 awg wire or no? Average wind speed her in wi where im is 10mph at 30 feet and 14-16mph at 90 feet. Also i was going to be using a 3 blade set up with a tilt axis of approximately 4-6 degrees. However… could i also place either a 5 blade system say at 60 or 70 feet so that i could get lower cut in speed and spend less money on metal tubing. Possible? Worth the effort? What about putting a vertical 4 blade system in same general area rather then 5 blade system? Would the pole in the middle of the verticalsystem disrupt the wind flow so that it doesnt pass through the system efficiently enough to warrent consideration? Any other overall suggestions would be very much apreciated’ thanks

  12. admin says:

    Are the coils of the dual stators wrapped around the magnets, some of the magnets or what?

    +(hugh) The magnets are on the rotors and the coils are in the stator.

    I am orderkg up n50 neo magnets 2x1x1/2 and am planning on using 18awg wire with around 150-175 turns per coil. Will be wiring in 7 phase star.

    +You need to tell us the number of coils and the number of magnets per rotor.

    But was thinking of doing one star ad one delta wired. What if any issues do you see with this basic setup?

    + I dislike delta due to the losses. Putting 2 windings in parallel with different characteristics seems a good idea at first but the star winding will have high losses at high speed and the delta will have parasitic losses at low speed and the whole thing is overcomplicated.

    Would i be better changing one of of wires to either 12. or 14 awg wire or no?

    + best to wind in star and use 2 wire sizes, but even so I am not sure that different stator characteristics is a great idea.

    Average wind speed her in wi where im is 10mph at 30 feet and 14-16mph at 90 feet. Also i was going to be using a 3 blade set up with a tilt axis of approximately 4-6 degrees. However… could i also place either a 5 blade system say at 60 or 70 feet so that i could get lower cut in speed and spend less money on metal tubing. Possible?

    + you get less wind at the lower height. Using more blades makes them run more slowly and reduces the efficiency of the alternator.

    Worth the effort?

    + 3 blades are better than 5.

    What about putting a vertical 4 blade system in same general area rather then 5 blade system?
    + lost me now.


  13. Randy says:

    I finished adding a second stator this weekend. The two are in series. The single stator tested alone inside the dual rotor produced about 18 V at about 60RPM not under load of coarse. The two together 24 V at about 60 RPM. All would seem to be working as anticipated. If you will be so kind. What say you % wise would one anticipate adding an armature rotor increase voltage. I am considering adding one as for no other reason a safety feature so as to shield the magnets on the stator having only the single rotor of magnets. Second will I experience any unforeseen effects due to their being one stator with dual rotor magnets and the second being induced by a single layer of magnets?

  14. Randy says:

    Sorry I tried to post a pic from photobucket and i seemed to fail. One stator is sandwiched between dual rotors and then the second stator has only one rotor of magnets acting on it. I guess you get the picture clear enough.

  15. admin says:


    It’s funny that I started this thread trying to stop people copying the dual stator idea but that message got lost somewhere and you guys are still doing it but anyway.

    Yes you ought to get a substantial improvement. Given that you get 18 volts off the first stator I would expect to get about 9 volts off the second (total 27) but the total only reached 24 volts so maybe it was too far away from the magnets. Anyway I see no reason why adding a second rotor to the second stator should not get you 18 again and thus a total of 36 volts. So long as you get them in phase!



  16. Autovic23 says:

    Hi guys, windpower is all pretty new to me, before I read this page I was thinking of building a dual stator alternator and wiring one star and one delta for lower and higher rpm depending on wind speed. Is this a good plan??? Or am I missing something??? I would be very greatfull for any advice

    • admin says:

      hi, it does seem like a good plan and I have done this sometimes (Bergey windpower tried it too on an 1kW early model) but I discovered that the low speed winding becomes very inefficient/wasteful as the wind gets stronger and there is not much you can easily do to stop it without making things complex. So it’s not as simple as it looks. I concluded that it’s best to use the whole stator all the time where possible. Using part of it reduces the efficiency.

  17. leontius dwi mesantono says:

    could you share me how to build the generator ? i mean the procedure, the mechanical calculation till electrical calculation of the design ? thank you

    my research last year got a problem that, i got losses a lot on the mechanical division. there were too many friction problem. etc

    • admin says:

      hi lenotius
      Have you read my Recipe Book, including the part at the end about how to design the generator? Do you have questions about that? I am happy to answer questions. I don’t want to write it all again for you if you did not read it.


  18. Leland paynter says:

    Looks like the recipe book is important to bring me up to speed. Where can I get access to it?



    • admin says:

      hi Leland,

      Click the link “Wind Turbine Plans” in the black bar above and select “all books”. The recipe book can be ordered in hard copy using paypal, by selecting the shipping option that applies to you. Or there are ebook versions on the right. Note that you need to get either a metric or an “English units” edition to suit your location and sources of magnets and winding wire.

      Have fun

  19. Bernardino says:

    I am new on this thing. Can anyone explain how to find the number of paires pole and the nominal frequency of the single phase permagnet DC brushless generator for a small wind turbine? Is the rotor wind turbine speed the same as the generator speed?if not, how can I find? How can I do the mathematical modelling of the single phase Dc generator?The only information I have, are: 160 W, Rotor speed- 1200 rpm, 12 V and stator 16 AWG 17 turns


    • admin says:

      Frequency is related to the number of poles and the rpm. Each pair of poles creates a cycle so the number of pairs equates to the frequency at one rev/second (60 rpm). 12 poles = 6 pairs = 6hz at 60 rpm, 120 hz at 1200 rpm.
      Single phase is not the best as it vibrates. 1200 rpm is a very high speed for a wind turbine. I am not sure what you are doing here but you may have the wrong equipment. Voltage varies with speed so at 120 rpm you will only get 1.2 volts. Maybe more if the nominal speed is full power rather than cut-in.

  20. William Deitsch says:

    Hey all you knowledgeable experts, I have a newbie question. I’m buying my first house and I want to build a VAWT. I’m not “overly” concerned about the lack of efficiency of VAWT’s compared to HAWT’s, as it’s mostly a hobby DIY project that I can have fun building. I have chosen a Savonius design for my first build using 25″ plywood rounds and 6″ stove pipe, mostly for the simplicity of the build and they seem to product high amounts of torque. My question(s) is(are): Most of the youtube videos I’ve seen involved stators/rotors that are in the 12″ to 18″ size range. Is there any reason for not building a stator/rotor at the edge of my 25″ diameter? I’m thinking more copper/more magnets=more power. Also, I’ve seen professional generators (like the Freedom II PMG, and others, for example) where the stator coils seem to overlap one another. Is there any reason for not doing this in a DIY project? Again, it seems to me more copper/more magnets=more power. Also, I’ve seen, “professionals” again, slanting either the coils or magnets or both in an effort to reduce cogging. Is this something that I should do in a DIY project? OK, enough questions for now. Again, I want my first build to be fairly simple, yet effective. I will play around with more advanced designs as my experience grows.

    • admin says:

      hi william,
      Yes the savonius is simple and reliable and the efficiency is terrible and you will need a very large alternator to work at the very low speed. So yes a larger diameter is a good plan for the magnet rotor but bear in mind that the power you will produce depends on the wind speed (cubed) and on the size and efficiency of the rotor so if you have a small and inefficient rotor in low winds then a big alternator still can’t help you as there is no power to be converted to electricity.
      Overlapping of coils can be done in an axial flux design but it does not usually bring much benefit as the end windings need to be rather long and this cancels much of the benefit. Cogging is not an issue if there is no iron in the stator.
      Have fun

  21. Ali says:

    750 Watt Axial Flux Alternator from Helibug:

  22. Tony O says:

    Hello Hugh,

    I am working on a project that requires a high amp alternator and since these are expensive, I decided to build one. It was while researching that I stumbled upon your excellent turbine plans. First of all thank you so much for giving the world a legendary gift of your talent that will immortalize you in the households of alternative energy enthusiast.

    Now, my indended use of your alternator design will entail being driven indoors by a machine that will run anywhere within 500 – 700 rpm at most. I intend using 24 pieces of N52 neodymium magnets (@12 per rotor) and with a 9 coil stator. I intend using several wires in hand (5 @ 1.8mm) so that the stator will not get overheated since it won’t be cooled in the wind as in a proper wind turbine.

    Now, to get high amp from an alternator, it is necessary to choose delta connection but I have already read in your blog why we should avoid delta connections for hand made alternator. In another of your blog, you recommended connection in parallel but that each coil should be rectified separately. I am going to stick to your recommendation but please let me know if using 5@1.8mm is a good choice that will handle up to 100amps continuously at 18v indoors.

    I understand that several wires in hand will lead to fewer turns thereby little voltage at low rpm. But I can compensate for the voltage by the available high rpm as well as by adjusting the gap between rotors. So my main concerns are designing the alternator with THICK wires so as to COMFORTABLY handle the large current and the effective means to cool the alternator indoors. I am thinking installing a dc fan will be an option but I will be pleased to hear from you the expert.


    Tony O

    • admin says:

      hi tony,

      Delta will give a lower voltage and a higher current carrying capacity, but you can achieve the same result using fewer turns per coil and avoid the parasitic currents you get in delta with a rectifier. The main thing is to get high rpm I would say and avoid the losses like that. A fan might help but a bit more rpm will help more. 40 % more rpm means half the resistance and half the loss and half the heat in the windings.
      Higher rpm will allow you to handle higher output. If you can’t do higher rpm then use more magnets and more coils. Cooling will have limited benefits and a delta winding will not help at all.

  23. Ali says:

    something to share, this one is a 1.4 kilowatt 48v alternator…
    1000+ Watt at 900 rpm !

Leave a Reply

Your email address will not be published. Required fields are marked *