Generator with ferrite magnets

[ChrisOlson]

Chris ...you said building big diameter rotors will fail because of the "flex" under high load conditions in the real world,..I think this only applies to hawts, a vawt , doesn't experience the same conditions??

Hi Art,

I don't know about a VAWT.  Here's what I'm saying, put a different way.  If you go to the Otherpower website you should be able to find the story about their 20 foot turbine build.  Take a look at that generator please.  For a 3.8 or 4.0 meter turbine with ferrite magnets - direct drive, three phase - we're talking about a generator that's bigger and heavier than that 20 foot build had.

So, for those who have built turbines, tell me, how big can you safely go and keep the air gap at 18 mm (or about 3/4") without contacting the stator in real world operation in high winds?  I know that even a tiny amount of "play" in the bearings translates to being amplified several times at the perimeter of those big rotors where the magnets are mounted.  And if the magnets rub the stator it isn't very long and you got junk.

So, you have some options if you want to go big.  Make the stator thicker than 12 mm so you can fit the turns in it, and increase the air gap to about 25 mm so it doesn't rub.  But with that big air gap with ferrites, my builds have shown me that it reduces the flux in the air gap to 90% of what it is at 18 mm.  Only 90% of the flux means you have to add 10% more turns.  10% more turns probably means 10% more resistance.  10% more resistance means reduced performance.

When everything is said and done, you could've put the materials in two stacked stators instead, at a smaller diameter where you can more precisely control the air gap, and get better performance.  And that's part of the reason why I did what did.  When I started looking at the alternatives to what I did, it became clear that with ferrite magnets on this size machine that it's going to be hard to match it with a single stator.

Max says he doesn't understand my thinking a lot of the times.  And Hugh says he's frustrated thinking that people are going to build stacked ferrite generators just because I had good results with them.  But, guys - good results is what it's all about.  Right?  I have demonstrated a 3.8 meter axial ferrite machine that works and is reliable.  Nobody else has.  I know beyond any doubt I could do it in a direct drive dual stator too, but the direct drive would be bigger, heavier and more expensive, and it will not perform as well as the geared machine.  But it will work, that I have no doubt about.  I have pored over ferrite designs for hours, including big three-phase ones, and I always come back to the same thing when I start designing a big three-phase direct drive - it's not practical and it's going to be hard to get it to hold up.

So that's just my opinion on it.  I don't think all this FUD over the dual stator vs single stator is even justified.  Especially when I took the challenge of building one of these things for a 3.8 meter machine, tackled it with a fresh sheet of paper and came up with something that is TOTALLY not in the book - but it works.  And it has been flying since the day I posted the first one on this thread without a single glitch.  And it gives up nothing in performance to the neo unit that I replaced with it.  And that was my goal - to build a generator with ferrite magnets that equaled or exceeded the neo generator's performance on my geared turbine.  I did it and I am extremely happy with it.

I'm glad folks are interested in building ferrite machines.  It's a challenge and it's fun.  But I'm not going to argue the point about this dual stator thing anymore until I see somebody put their money where their mouth is and build a big single stator ferrite so we can compare notes.  Otherwise, attempts to downplay my dual stator generator as the "wrong" way to do it is pure FUD.  Until then, for folks who want to build one, I think you can draw your conclusions between theory and practical application.
--
Chris

[Menelaos]

well said :-)

[ChrisOlson]

well said :-)

Max,

I must say I'm looking forward to seeing your design.  It might give me some insight on how to build a two phase in a smaller size yet 

One of the experiments I tried was stacking overlapped coils for a two phase in a single stator.  I pretty much ran into the scenario I outlined above - couldn't keep the stator thickness below 18 mm.  With the thicker stator I had to increase the air gap.  Now I had a problem with low voltage and needed more turns.  More turns meant making the stator thicker yet, or going to smaller wire to get it to fit, and then I reduce the amp capacity.

It ended up being the classic never ending circle.  These magnets are WEAK, man, compared to neos.  On the surface it's easy.  Anybody can build a ferrite generator that makes power.  Building one to match a comparable neo in output and performance is quite another kettle of fish.
--
Chris

[artv]

Chris........your work is fantastic,....why is everything designed around a central shaft?....perpherial (travellers) for a bigger dia.
I know .......shut-up....... [ Attachment Invalid Or Does Not Exist ]
sorry I don't know what I did just there........see what happens.....artv

[ChrisOlson]

why is everything designed around a central shaft?

Just because I like that design.  I can machine generator rotors to run true and install them, remove them and install them again and they still run true with no adjustments needed.  I gave up on turbines built with trailer spindles with a bunch of stuff stacked on redi-rod a long time ago.
--
Chris

[artv]

But you said doing a bigger dia will see run-out ,causing failure,...I think with the central bearings this is true....
Would outer (adjustable guide rollers)....eliminate the problem??
Sorry more questions than answers.......artv

[ghurd]

So, from my own experience with this; you can use a wedge shaped coil with big square block magnets and it looks like it will "cancel".  But in reality, the reduction in resistance of the coil (less copper used in it) more than makes up for the very slight difference in voltage you get by making the coil hole the size of, or bigger than, the magnet.

I realize this sort of goes against what is accepted as "the proper way to do it".  But I have done it the "wrong way", using a coil hole that is "too small" with great success for some time.

Willib proved this years ago.
Mr Flux typically referes to it as linkage instead of cancellation.

The 2-ph vs 3-ph IRP:  You are missing the point of sq root of 2.
Regardless of wye or IRP, 3-ph is more efficient than 2-ph.

I'm pretty sure nobody intended to say "it could not be done".  Guessing they said it could not be done cost effectively with neo prices as low as they were a while ago.

[joestue]

If you've got single layer windings, single, two or three phase is effectively the same provided you keep the coils the same size. and by the same size i mean electrically they all provide the same waveform, have the same inside and outside diameter. This won't be the case in real life, but its close. fundamentally you run into the same limitation, that is you've got a core that's about half full of copper.

with copper costing ~5 times as much as ferrite i would be looking to use a lot more magnets and forget trying to stuff it with copper.

[Flux]

Max raised some points but it was way back on the last page before I picked up on it. Joestue has basically raised it again here in a slightly different form so it seems a good time to look at this again.

A basic machine is single phase and in reality you can only effectively wind about half the winding area ( you can go from about a third to 2 thirds for different conditions with single phase)

Polyphase machines are developed by winding extra stators in the available space and spacing them electrically so that you have more than one machine in the air gap. For 2 phase the best will be with starting with the single phase with half wound. For 3 phase you start with single phase 1/3 wound. Two phase have two alternators spaced 90 electrical degrees. 3 phase has 3 spaced 120 deg, if you do it right you use all the available winding space. For a conventional loading scheme as you would have on an engine generator or on a big grid connected machine you have the potential to get nearly double the power.

In real life it works out roughly that a single phase machine is 50% bigger than it's polyphase equivalent.

This has been established so long that we accept it without question. It has taken me a very long time to realise that what we are doing here with these wind machines is totally different.

For conventional loading the load resistance is an order of magnitude higher than the internal Resistance of the alternator, the internal resistance affects the losses but has negligible effect on the load characteristics ( with iron cored machines leakage reactance is the thing that determines the output not the resistance and things are even more definite).

Once we remove reactance, the power you can get out and the load presented to the prop is entirely dependent on flux per pole and resistance. With neo we have a lot of flux per pole and it usually turns out that the resistance is too low for the prop to match, we end up having to reduce the alternator stiffness to get the prop to work ( high electrical efficiency is not possible with direct connection to a battery without some other trickery to change the loading with wind speed.)

When we come to ferrite, there is little flux available, resistance will now be too high and the approach needs to be different. Reducing resistance seems to be the way forward more than anything else.

This is where the ployphase machine runs into difficulty. Conventionally the aim has to be to keep the air gap the same as dingle phase so that meant overlapped ends to the coils so they could cross each other ( done in all common motors and generators). Now we are in a situation where resistance is the dominant factor we loose out here with the long end connections so much of what you try to gain from winding the extra space is defeated by the increased copper length.

If you leave out coils you can get a 3 phase winding on a single layer, it can now have short turns in the coils but you partly run back to the old issue of not having all the space wound and you have got your 3 phase winding by leaving coils out.

This seems to be the point that Chris was making about having more coils to get the same voltage in the single phase case. ( a conventional commercial engine driven alternator has the same number of turns in each phase of the winding as the basic single phase starting point).

The issue as I see it is this:-  can you use the wasted unwound space of a single phase winding with more copper to make a polypase machine with a lower overall internal resistance?

This is where the mathematicians need to come in, I can argue it from logic up to a point but we have too many variables interacting  and when you go off looking at one point you easily miss the obvious somewhere else.

I have kept clear of any effects of heating and any mechanical considerations,these are also important but things are complicated enough at present without clouding the issue.

This has become long and rambling, I am still thinking about all this, but it seems that if you change the way of loading a machine from conventional resistive loading where it is efficiency you are going for, to a situation where the criteria is how much you can screw out of the thing regardless of efficiency you have to deal with very different factors.

I probably took a long time to get here because I have been looking at other ways to match the prop to the alternator at high efficiency and in that case the conventional ideas still apply.

Flux

[ChrisOlson]

Regardless of wye or IRP, 3-ph is more efficient than 2-ph.

Three phase is only more efficient in the transmission of the power because you can move the same power with three wires instead of four.  From the generator's standpoint it does not make one bit of difference.  So that distinction needs to be made.  Most people are going to assume from your statement that a three-phase generator is inherently more efficient than a two-phase, which is wrong.  The one and only advantage is in power transmission efficiency and the amount of conductor mass required.
--
Chris

[XXLRay]

Menelaos, did you have time to compare your FEMM simulation with five magnet poles with a real world setup? Were you able to make a FEMM model with thicker steel backings and compare it to a measurement?

[ChrisOlson]

When we come to ferrite, there is little flux available, resistance will now be too high and the approach needs to be different. Reducing resistance seems to be the way forward more than anything else.

Flux,

The above quoted part is why my design with ferrites ended up like it did.  Using a 1:1 pole/coil ratio per phase proved to be the most efficient way to get the most voltage from the weak flux with the least possible resistance in copper.  Adding the second phase in the unwound space, while keeping the stator thin enough to keep the air gap flux the same (to prevent having to add more copper), proved to be a bigger challenge.

I think as other people try to build a ferrite generator with the amp capacity of the ones I got without ending up with something that you need an overhead crane to install on the tower, they're going to find this out.
--
Chris

Edit:
I found the sheet of paper I did my calculations on in my toolbox.  For the flat three phase with a .4375 gear ratio I figured I'd need 20 poles and 15 coils.  I could use 1.732x less turns per coil if I used the wye configuration, but when I connect two phases in series wye my resistance ended up around .5 ohm instead of .31

So to reduce the resistance of the wye winding I had to wind it two-in-hand.  This means that, effectively, I'm now using the amount of copper to wind 30 coils instead of 15, but connecting it parallel to reduce the resistance and get the amp capacity.  Now the resistance ends up at .25 ohm, which is too low to match the shaft power.  So I needed to use one size smaller wire to get the resistance up to .31-.34.  But I still got a problem - this two-in-hand winding won't fit in a 12 mm thick stator, and now it won't carry 60 amps with the smaller wire.

It was a real interesting challenge to get the resistance where I needed it, get the 60 amp continuous capacity, and still have that elusive match to the shaft power without having to "tune" with external line resistance.

[Bruce S]

that's not nice fab....

this isn't a  5 year Thesis Submission from a university....which nobody will ever understand and will be stuck in pdf,s on the net

lets all roll with it for a while ...

niall2 TRY to understand a few things before lashing out too quick.
The problem with the internet is that you cannot see people smiling or giggling when they post.
Artv was and is making a joke about it and FAB is helping.
FAB grab one of those smiley faces there just above the text box next time so newbies can understand you're only joking and carrying on with artv. 
carry on
Bruce S
 

[Menelaos]

@ xxlray

I did do a simulation with thicker steel plates but there was no noticable improvement of flux density. Anyways those 3mm bars were only to get a quick setup for practical measurements. For the real rotor discs I will again use 6mm steel, just like I did for the last alternator kit. Meanwile Uwe and I have ordered another 500 neodymium magnets for an acceptable price so we will not run out of material for the next month which gives me enough time to get this ferrit stuff worked out :-)

I did also not use a 5 magnet setup to compare with the one I did with only 3 magnet poles as there will only be minor changes to the flux density of the middle magnet pair.

When I use these wedge shaped magnets, I cannot make a femm- simulation on that anyway so I will have to do a testcoil to find out about which voltage I will end up with...
I still did not get an reply from those polish guys about the costs for those costumised wedge shaped ferrits which the whole thing kind of depends on. I fear that it will be quite expensive for those small quanteties. It might make sense if we can order enough of this stuff to get 30 or 40 alternators built...but first of all the whole setup has to work out :-)

@ chris

"Now the resistance ends up at .25 ohm, which is too low to match the shaft power."

I do not see why this should be a problem. I aleays try to get my alternators as efficient as possible. If the resistance turns out to bee too low for a good match, you can still get it up by increasing the wire run or diamter of the line resistance in whatever way. That will come up with the same results except that the heat loss is then not concentrated in the stator but in the line which is a good thing...I guess yo know that, so whats the problem?

Max

[electrondady1]

from max
"« Reply #139 on: October 01, 2011, 11:37:12 AM »
   Reply with quote
I have my own theory about the coil design. ..........
 I will make the coils a lot smaller than the magnets......
 I now can easily fit 12 coils on the same space that was taen for 9 coils before.....
 I think thats pritty cool :-)"


not wanting to flog a dead horse, but that is the post that led to the misunderstanding.

Canadians like to hold their ground.
i don't want to be buffaloed 'cause someone else was rude.


i build little drag verticals(getting bigger and bigger) because i like the way they look.
not exactly main stream for this forum.
they turn slowly so i normally wind single phase stators

I'm inspired by guys like windstuff ed and Jerry,
who go there own way.
a two phase stator with overlapping coils that doesn't increase the stator thickness or require extra coil length has been my challenge.
i came up with something and will do a show and tell soon.

  
in substituting ferrite mags for neo's your getting something like 30% of  the flux density.
so you need more magnet, ether from larger poles or more poles
or gearing up and spinning the alternator faster.

i cant understand the fuss about going large dia. direct drive if you separate the bending moment and deflection of the prop from the alternator rotors.
i've seen lots of hawt machines with the alternator placed on the back side of the bearings.

[Menelaos]

I a sorry about that misunderstanding but I think it is now clear what I mean :-)
Sometimes it is not so easy for me to express myself in a foreign language, especially when it comes to technical vocabulary... :-(


Max

[TomW]

I a sorry about that misunderstanding but I think it is now clear what I mean :-)
Sometimes it is not so easy for me to express myself in a foreign language, especially when it comes to technical vocabulary... :-(


Max

Not to worry, Max. I only speak English and Bad English and I seem to be in the same position not making things understandable.

Appreciate your efforts and your sharing them here.

Tom

[electrondady1]

max you are a clever man and you build big beautiful alternators.
it's difficult not to admire someone like that.
your a real asset to this forum.

[ChrisOlson]

I do not see why this should be a problem. I aleays try to get my alternators as efficient as possible.

Max, it's maybe because of a different design philosophy where I like to build the match into the generator designed with a 25 m wire drop on the tower, rectify at the tower base and use two big conductors to transmit the DC power to the load (2/0 aluminum underground on my machines).

But even with the .25 ohm two-in-hand winding I still had a problem.  That would not fit (even with one size smaller wire) in my 12mm thick stator.  I arrived at the conclusion that with 2 mm wire, if I wanted to keep the coil shape and size to what I had determined worked the best, that I could have 48 strands of wire in each coil.

With the 20 pole 15 coil three-phase configuration, even in wye I had 66 strands of wire per coil wound two-in-hand.  That was not going to work no matter what i did unless I dropped the wire size drastically, and then it would not handle the 60 amps I was looking for from it.

The end result is what I posted earlier in this thread - the dual stator.  It handles the 60 amps with ease.  Uses the same 40 magnets.  The internal resistance ended up at .31 ohm wound with 2 mm wire.  The match to a 4 meter rotor is perfect - it cuts in around 3.5 m/s and runs at 400 rpm @ 12 m/s.  My latest version of it blows the previous neo right out of the water for power and the "perfect" TSR match for the rotor.

I just don't believe in throwing the most efficient generator possible at it, then trying to match it with line resistance.  All you're going to achieve is tuning it for one specific wind speed range by doing that.  You're not going to keep that rotor at its ideal TSR thruout the full range of wind speeds from cut-in to full rated power like I have achieved with my latest one.  That's just my opinion on it, and I realize everybody has different goals.  I believed for a very long time that the neo generators were too powerful and only hurt overall performance of the turbine.  They are so "stiff" that you can only really tune for one wind speed range with them and take what you get on either side of that.

I approached the ferrite project with the same design philosophy - build it for 6 m/s wind and see what you get on the low end and high end.  The results are very pleasing.  I still take what I get for cut-in, and 3.5 m/s is reasonable.  The surprise was how the ferrite generator will let the turbine spin at higher wind speeds and the power results are a dramatic improvement over the neo generator at those wind speeds.

I'm still mapping a real world power curve for the 4.0 meter machine.  I had a problem with it the other day when it busted the tower mast in high winds and I had to shut it down.  I got the tower fixed and the machine running again and am continuing to log data on it to determine what it actually develops in real world conditions.  That generator is too fast for the 3.8 meter three blade rotor.  It's almost a perfect match to the 4.0 meter two-blade running at 7-8 TSR.  So, in a nutshell, that's the "problem" as you put it, the way I see it.  And like you say, much of what I do does not make sense   
--
Chris

[niall2]

apologies to Fab and Artv......got the wrong end of the stick there ...would,nt be the first time doing that ...

i,ve noticed a little small blistering on one or two of my mags now, so the ferrite debate is very interesting ,not so much for efficiencies
but more to do with longevity of the alt ...

i,d like to know the reason why they sometimes break down ...i think in my case letting some of them get acquainted with big spanners did,nt help ...  

i touched up the first blister with  rust inhibitor and epoxy ....i put some inhibitor on a spare mag to see if any side effects showed up later

now ...just where exactly did i leave that dead fish ? ...    

[bob g]

OT, i know, but i just have to ask...

with most folks using neo's, and now all the interest in ferrite, and with machine
hovering around 17-20ft in dia. ...

i have to wonder if at these diameters, and with gearing, are we not getting close to
where a wound field rotor might start to be competitive in output?

if so, the electronics to control the load on the alternator and the blade set by extension
becomes very simple.

wouldn't it be weird to see windpower turn full circle back to wound field design?

makes one wonder, at least it makes me wonder

bob g

[fabricator]

Bob g I was thinking the same thing the other day, perfect example, Jacobs, they have been around for close to 75 years and still going strong as Chris well knows.

[fabricator]

that's not nice fab....

this isn't a  5 year Thesis Submission from a university....which nobody will ever understand and will be stuck in pdf,s on the net

lets all roll with it for a while ...

niall2 TRY to understand a few things before lashing out too quick.
The problem with the internet is that you cannot see people smiling or giggling when they post.
Artv was and is making a joke about it and FAB is helping.
FAB grab one of those smiley faces there just above the text box next time so newbies can understand you're only joking and carrying on with artv. 
carry on
Bruce S
 

I like to give people the benefit of the doubt Bruce..............

[ChrisOlson]

i have to wonder if at these diameters, and with gearing, are we not getting close to where a wound field rotor might start to be competitive in output?

Bob - I don't think so.  The Jacobs 23-10 is horrible for low wind power.  If the wind isn't blowing at least 5 m/s you may as well leave the Jake shut down because one of my 3.8 meter machines can beat it.
--
Chris

[bob g]

i was not particularly referencing what jacobs did 75 years ago, however
one has to give reverence to what was accomplished for sure.

what i am thinking is we have some very sophisticated electronics available now
that jacobs had no clue about back in the day.

you state that below 5m/s the old jacobs couldn't produce, well what if we step up
to a modern design field wound rotor, 3phase and rectified with silicon instead of
brushes and commutator, and with a modern solid state regulator instead of the old
mechanical relay regulator?

if i can make a hd truck alternator make 2.88 kwatts of output using less than 40watts of excitation power for the wound field, it just seems to me that those 40 watts lost might be a pretty good trade off for less cost and the ability to control the output and match to the blade set over the complete range of windspeeds available.

now i know that sub 10 footers likely would not compare well with neo units, but when we near 20 ft in blade diameter, my bet is there might be a very close race between the two methods.  bare in mind i can also reduce the excitation current at lower windspeeds and in doing so not loose 40watts, but perhaps 20 watts?

it would also be possible to build a hybrid wherein we use both wound field and neo's
so that the neo's could be much smaller and the wound field only powered up in winds where the losses could be accepted?

i also realize that this might be seen as sacrilegious, but thought it might bare discussion.

perhaps a different thread so as not to pollute this one?

bob g

[ChrisOlson]

you state that below 5m/s the old jacobs couldn't produce, well what if we step up to a modern design field wound rotor, 3phase and rectified with silicon instead of brushes and commutator, and with a modern solid state regulator instead of the old mechanical relay regulator?

Bob, that's what my 23-10 has got.  It's all electronic.  Brushless, three phase, 12 lead, 10 kW Fidelity generator in it.  It's one of the new grid-tie ones that I converted to battery charging.  It's not an old DC Jake.



It holds enough residual magnetism in the core to start up the field by itself.  But at low wind speed it takes 2 amps to the field to get any power out of it so it basically doesn't cut in until the wind is blowing 4.5 m/s (about 10 mph).  And at 5 m/s it's putting a whopping 85 watts while the field is pulling 60 watts.  And this is a 7.0 meter (23 foot) machine.  At 6 m/s it comes right to life.  But by the very nature of the thing, wound field generators are not low wind producers.
--
Chris

[scoraigwind]

I'm glad folks are interested in building ferrite machines.  It's a challenge and it's fun.  But I'm not going to argue the point about this dual stator thing anymore until I see somebody put their money where their mouth is and build a big single stator ferrite so we can compare notes.  Otherwise, attempts to downplay my dual stator generator as the "wrong" way to do it is pure FUD.  Until then, for folks who want to build one, I think you can draw your conclusions between theory and practical application.
--
Chris

I am also reaching the conclusion that it's pointless trying to explain the advantages of a single stator (twice as much voltage from the same materials etc). The people who get it get it, and the rest never will. Until it's built and tested.

I am slow at development and testing compared to Chris, but I recognise that the cost and fragility of neos will soon make them obsolete (as we all have for some time).  And I am realising that the very low price of ferrite magnets requires a shift in my approach to design.  I am learning that cramming magnets in and using smaller coils helps to achieve the goal now.  Thanks for that, Chris and Max!  I still see 3-phase as the best option and I am personally devoted to direct drive although I see no objection to the chain drive approach if you can do the engineering. (Doubling the speed effectively turns ferrites into neos.)   I will produce a direct drive alternative that is reasonably attractive compared to Chris's dual stator up-ratio version.  Until then I will stop being annoying.

[Flux]

Just a quick reply to Bob g.

I don't think many home constructors will have much luck with wound field machines, they are a real challenge to build and there will be little hope without the right starting point.

You are right that quite a lot can be done, it is not possible to eliminate the field loss but you can certainly reduce its impact but it makes life difficult. The field required to produce small power is not great but the field requirement goes up rapidly with the effect of armature reaction and you need to bring the field on very quickly  to get any real power. Everything with wind power is a compromise and most factors work backwards against you.

The hybrid with ferrite magnets and a wound field machine for high winds does have a good chance of working, but the wound field machine inherently suffers many of the snags of ferrite, no problem at high speed but the thing becomes a monster for direct drive.  I ended up using a gearbox rather than a hundredweight of alternator for a 10ft machine. The thing that failed was the gearbox, if chain drive solves this then speed increase may be the way to go. ( gearbox, never, it costs the earth to make it reliable and many commercial machines run into trouble there).

You certainly have much better control with wound field. For a few devoted people who really understand what they are doing this could indeed be a good way forward but it would be a serious constructor with enough facilities to tackle it.


Surprisingly , looking back at some old  Electro data they had ferrite and wound field versions available, the low wind end was not much different but the wound field came out better in high wind. The data is poor for low wind and it is probably true that the low wind end on both versions was way below anything we hope for now.

I won't hijack this thread any more, but all may not be lost.

Flux

[ChrisOlson]

........but I recognise that the cost and fragility of neos will soon make them obsolete (as we all have for some time)

Anything "rare" or "exotic" is usually expensive, whether it be magnets or women.  Ferrites are made of one of the most common materials in the earth's crust - iron.  The environmental impact of mining rare earth materials is going to keep neos in the "rare" or "exotic" category, I think.  So folks hoping the price will come down can have hope.  But I just don't see it in today's economy.

I will produce a direct drive alternative that is reasonably attractive compared to Chris's dual stator up-ratio version.  Until then I will stop being annoying.

Not annoying - just two different design approaches.  I think people appreciate both approaches, Hugh, and what can get annoying is when one design approach is hammered as being "wrong" when it has a different goal.  My design goes for heavy duty high continuous amp capacity.  If I designed for a 40 amp continuous capacity with spikes over that, depending on duty cycle, my design approach would be different than what I used.

But there is interest in ferrite generators.  I have had many emails that I have not been able to respond to all of them yet, asking about how I think some design "x" would work with ferrites.  Unfortunately, I only know about what I have built that works.  And it took some experimenting to arrive at that.  But the main thing is, folks are very interested in it.
--
Chris

[ChrisOlson]

This is going to be a challenge for me because I am not good at figuring out how to make a tabular data presentation on this board.  But I have been working on this for many weeks and I think I have good output data on my 3.8 meter machine with ferrite magnets.

First - how is the data logged?  Crude, really.  I have a recording anemometer that records the wind speed for a logging period.  It stores how much time in hours and minutes the wind spends at each increment of wind speed and averages all this data in computer software to supply an average for the logging period.  During that time I record kWh output of the turbine.  I divide the output in kWh by the hours and minutes in the logging period and that is the average output in watts for that average wind speed.

Some day maybe I'll invest in a APRS or a logger like Hugh has.  But for the present, that is how I do it.  It is important to realize that doing it my way is crude and the data may not be 100% accurate.  But it's "close enough" to give a general idea of what the turbine is capable of.

I have three machines with these generators on.  The prototype one shown in this thread had a conglomeration of stuff.  It was tested as a single phase.  Then I stacked two phases on it.  Then I replaced the three blade rotor with a four.  I played with furling.  I'm constantly fiddling with that turbine because it's on a tower that's easy to raise and lower (I never take the gin off it) and it's my test machine.  That prototype generator has one stator with 50 turns, the other with 53.  I found out that's pointless.  It works the best if the stators are evenly loaded.  The "improved" version of it has 51 turns in both stators.  And that is the one I took this data from.

The machine spec: 3.8 meter (12.5 foot) rotor.  Gottingen 222 airfoils with 10 degree pitch, 158 mm chord, no twist, no taper.  .4375 gear ratio, meaning for every revolution of the rotor, the generator turns 2.28 times.  51 turns of 1.8 mm (13 AWG) wire.  24 volt system.

Now to see if I can make a table:

Wind Speed	Watts
3.0	64
4.0	171
5.0	272
6.0	411
7.0	644
8.0	917
9.0	1209
10.0	1522
11.0	1799
12.0	1926
13.0	2285

I have the furling cranked up on this machine just to see if I can burn a stator out.  So far I haven't, even in 20 m/s wind.  But I consider this generator good for 1,500 watts sustained.  And by sustained I mean for days at a time at that power level.  I take the maximum amp capacity of the magnet wire, divide it by 1.15 and that is my maximum continuous.  That designs a 1.15 service factor into the winding to insure that it will hold up when the wind really blows.  My newest 4.0 meter machine has 2 mm (12 AWG) wire in the stators.  But that machine is not fully tested yet.

The ferrite dual stator generator is more power efficient than the neo that I originally had on this machine by about 8-9% at higher outputs.  So it creates less heat, puts more power down the tower, and is more robust than the neo delta one I had on it before.  Another consideration in my data is that our wind rarely averages in the 10-13 m/s range.  For some reason it's either below that, or when we get our Norwesters it's way above like 15-20 m/s and any trees that are not strong are laying all over and have to be cut up for firewood.  I have to extrapolate there by taking readings off the anemometer and compare against the watt meter several times, then average them to arrive at a reasonable output number.  Like I said, maybe sometime I'll invest in one of those high-dollar loggers.  But for now that's the way I do it.

In the above table, the furling is really starting to "kick in" around 11.0 m/s.  At 13 m/s this machine is pretty much folded up completely.  Another consideration is that with the dual stator, each stator is only carrying half the load, or supplying half the total power.  So you have to divide by two to see what each stator is delivering at each wind speed.

These ferrite generators have proven to be the most robust and power efficient generators I have ever built, even though they're quite heavy.
--
Chris

[ghurd]

Three phase is only more efficient in the transmission of the power because you can move the same power with three wires instead of four.  From the generator's standpoint it does not make one bit of difference.  So that distinction needs to be made.  Most people are going to assume from your statement that a three-phase generator is inherently more efficient than a two-phase, which is wrong.  The one and only advantage is in power transmission efficiency and the amount of conductor mass required.

Incorrect

[ChrisOlson]

Incorrect

I feel that I missed something profound in your post.  So I need to ask the question:

What determines the elusive "efficiency" of a generator?

• The number of phases?
• The weight vs the diameter?
• Or is it measured power in vs measured power out?

If it's the latter then the number of phases does not make one bit of difference.  If it's weight vs diameter, why not just make it out of titanium?  If it's the number of phases, then why stop at  three?

Example: My house standby gen is s simple two pole, single phase, 3,600 rpm, 240 volt unit.  It generates 6 kW/hr on 1.05 gallons of LP gas (96,180 BTU input), or 16,030 BTU/kWh.  My 250 kW genset is three phase, 480 volt, 4 pole, 1,800 rpm.  It generates 250 kW/hr on 30.7 gallons of #2 diesel (4,298,000 BTU input), or 17,192 BTU/kWh.

Which genset is more efficient?
--
Chris

[joestue]

A simple test between a three phase stator and a single phase stator would be to short both of them, wind a rope around the shaft with a weight on it, and measure the rpm as it runs out, which ever is slower wins. Compare with a plate of aluminum if you feel like it.