Coil shape for axial flux generator

[analog man]

  You will never match the energy production of a MPPT turbine with a direct hooked across all wind speeds. 

That I agree with, and is the key to what you say, but in my experience aiming for greater performance out side the 14 to 25 km wind range is a false economy, The winds needed for your mppt to shine at 2, 3 x are outside the average wind range that is typical?
Why doesnt this board have a edit feature?

[tanner0441]

Hi Chris.

I hope your boating went well. I think we have been here before........

I remember having Volts times Amps is Watts hammered into my skull at school. You soon learn any resistance is the enemy and at 12 V is a pain, up the voltage and you up the efficiency and reduce the amps so the voltage drop in the line is reduced. Otherwise why do the power boys not transmit the power at 110 in the US and 240 in the UK, they could save a fortune in 100ft pylons.

Even 50ft is a long way to send any sort of power at 12V.  The controlling factor is the voltage not whether it's AC DC single phase or polyphase. 

Brian

[ChrisOlson]

Tanner brings up a good point, but not made clearly.  DC power transmission is more efficient than AC.

The other thing that MPPT allows is perfect rotor and generator matching to run your rotor at the correct TSR at all wind speeds, and as much gain is made from that as from electrical efficiency.  That is virtually impossible to do with a direct hooked turbine.

I was a doubter for a long time when the Classic first came out.  I believed I could build a turbine that would match a MPPT one - and I could if the turbine was designed for one specific wind speed.  After trying it I found out that one MPPT turbine matches two regular ones in energy production over the long term.  The cost of the controller is really easy to justify for off-grid power.

[kitestrings]

I agree.  We ran (and are running) a 12V machine for many years ('85).  Very reliable, but even with wires the size of my thumb, the losses are awful.

We've only had the axial up for a short time, but I still can't get over how well the thing works (Classic).  We get lot's more usable production, much higher instantaneous, less heat in the rectifiers  - it's a bigger rotor, and that's a lot of it - but with MPPT you just seem to ride the wave through a much wider range of speeds.  And, you get a number of practical extras like measurement, data logging, local network, two Aux outputs (for diversion, braking/shutdown, battery compartment fans, etc.).

MPPT pretty much revolutionized PV, if you remember life before that - matching expensive, limited choice cells/panels to a battery bank voltage, big honkin' wires, no long wire runs.  With wind it's been a quieter revolution (there's so few of us here), but a milestone none-the-less.  MS has done their homework; excellent CS support.  And, I don't sell them, our have any vested interest.

~ks

[analog man]

Ok you have a whole new mill, completly different than your 12 volt  one,and on this new mill, at a different voltage even,  your using the classic on it, Do you think thats a fair way to compare the two systems? How about you put  a classic on the old 12 volt mill and note the difference before and after for battery charging.
I am not getting into a pissing match over voltage and the advantage of each, nor am I advocating for 12 volts, just what I am running. I have no problem using or wiring for 12 volts others that do should ofcourse go higher, and I do suggest a higher voltage to anyone looking at getting set up. My comment started regarding mppt and its advantage in the real world and the implication that direct charging is a failure to be abandoned.
I stand by it not being so, it can be very effective, can be done cheaply, again at the real world average wind speeds where you stand to get the power you need to keep your batt full is where it shines.

I do agree with Chris regarding getting more "production" out of a mill with mppt. Note the word production, there is production and there is usable production, and amps into your batts is usable production thats what we need to keep the batts up. You guys, unless I have read wrong are burning up your extra power in resistors, your batts are charged your mills spinning away, your recording this non batt charging power, as if its some great thing. Chris  correct me if I am wrong but arnt  you recording this in your production totals? So yes that big storm blows through your batts are charged and your system keeps dumping power, running up impressive numbers for your mppt system but the job the mill is needed for has long since been over.
I say amps into the batts at average typical wind speeds is what nets the best performance day in and out. I say direct charging does that job with ease, and I can prove it does at lower cost. I agree my mill becomes increasingly less efficient at higher output but I couldnt care less it will have done its job already, and a burnt up stator if you have that problem is a design issue to solve not a problem with direct batt charging, and I dont think for one minute that you could not solve such a problem if it was ocurring to you.I think the idea of perfect has become the enemy of good....
I am certain I could throw on a mppt and I would not notice one bit of difference in the power available from my batts.

[joestue]

I am certain I could throw on a mppt and I would not notice one bit of difference in the power available from my batts.

 i can see the points you made in a certain perspective but this statement here is false.

when you draw a straight line under the cubic power curve, you can taylor the slope of the line to suit the windspeed averages you have, but you won't ever get more than about half the total wind power available.

under the conditions most folks here operate in, i would expect the actual KWH into the batteries at a minimum to double with an MPPT converter.

regarding how much such a MPPT converter should cost, i set the bar at 10 cents a watt.

[ChrisOlson]

unless I have read wrong are burning up your extra power in resistors, your batts are charged your mills spinning away, your recording this non batt charging power, as if its some great thing. Chris  correct me if I am wrong but arnt  you recording this in your production totals?

Actually you are very wrong.  The "dump" as you call it (voltage clipper is what it really is and it shuts the turbine down automatically when the batteries are charged) is three-phase AC and the power to it is not measured.  Only the actual DC kWh that goes to the battery is measured and logged.

[analog man]

unless I have read wrong are burning up your extra power in resistors, your batts are charged your mills spinning away, your recording this non batt charging power, as if its some great thing. Chris  correct me if I am wrong but arnt  you recording this in your production totals?

Actually you are very wrong.

Ok, I stand corrected I will go back and do some reading on what your doing, must have mixed you up with someone else then.

[ChrisOlson]

I have tried to harvest the energy that could be generated when the clipper is being operated in the past by trying to heat water with it.  But I found it was too complicated with three-phase wiring and commonly available single phase water heater elements.  So I gave up on that.

We went to heating water with electric as a normal thing, always on, like any home that has electric hot water.  Use hot water and the water heater kicks in and re-heats it.  Our system has grown to the point where it can easily power it.  So I finally decided that when the batteries are charged there is no need for the power from the wind turbine.  The best is shut it down and save the wear and tear on it until it is needed again.  When the controller drops out of Float and to the battery Recharge Volts, then it starts the turbine up again and it goes back to work.

[Flux]

I think this argument will run forever. It all depends on what you want your turbine for.

If you can get enough power for your needs with direct charging then there is not so much incentive to change to mppt.

Direct battery charging works best in high winds where the power curve is steep and you can get a reasonable match to the prop. If you live in a high wind site and need comparatuively little power the cut in at 12mph and you will do fairly well.

In a low wind site with a cut in of 12mph you may get very little except in times of storm. This surprisingly is where mppt comes into its own. The power curve from 6mph up to about 12mph is very flat and you will not get a good match in this low slope region. With direct charging attempting to cut in under 8mph is not a good idea even in a really poor wind area. If you aim for 6mph and have any reasonable electrical efficiency you will be hard stalled by 12mph and stuck with a machine running at constant speed and current limited. If this supplies your needs then fine, but in low wind areas you need all you can get if you use any significant power.

For a long time I assumed that direct charging was efficient in low winds and I was rather surprised to find that the gain with mppt takes place all the way up the wind speed range and I would agree with Chris that you can double your energy capture in the lower wind speeds.

Yes you may have fully charged batteries, if that is the case there is no need for mppt but with wind you need to use the energy when it is available so instead of dumping power it is better to use it, there will be plenty of days when you need to conserve as much as possible.

For a hobby system or a small installation with energy managed carefully and helped by solar there may be a case for not considering mppt but for a serious wind user the advantage of mppt is overwhelming.

Regarding cost, yes mppt costs money in the beginning, the same applies to an inverter but I bet there are few installations that use 12v only to the loads except small cabins with led lights.

Off grid power is expensive but if you are serious then do consider mppt.

[tanner0441]

Hi

One point he seems to be missing is his 12V mill is only 12V on a 12V system. the same mill albeit running faster would charge a 24V system, if he open circuits it, it will produce enough voltage to electrocute him.

Brian

[Flux]

Yes I think most people have difficulty understanding efficiency. When dealing with mains electricity at constant voltage the source impedance of the supply is increadibly low, the load resistance is always high in relation.

This is the only way to get high electrical efficiency. A wind turbine needs to rotate at a speed proportional to wind speed so as you say it is a 12v machine at cut in, at twice cut in speed it ought to be loaded at nearer 24v, at 3 times cut in speed it ought to be at 36v and in high winds near furling it will better match 48v.

If you bog it down to 12v in high winds the internal resistance will be far to high to get any reasonable efficiency , at 48v it would have to produce only 1/4 the current so if you bog it down to 12v the losses are 16 times higher than at 48v, the electrical efficiency drops very low.

In addition to this we need to load the prop to constant tsr to keep the efficiency there and this requires the applied load to increase as wind speed cubed, you can't get these conflicting requirements anything like right with direct connection if you choose a low cut in speed. In fact the prop matching is far more important of the two and it works out better to compromise the electrical efficiency to keep the prop out of hard stall, but this brings back the old stator heating problem.

The true efficiency is watts into the battery/ watts available in the wind. We start at a big disadvantage with the Betz limit of 63%. Add in the best prop properly loaded and you are down to about 45%. For a small alternator ideally loaded you are down to near 40%, then add in line and rectifier losses and a mppt converter and you may see 35%. Compare this with direct charging with prop off its power curve and 50% electrical loss you are struggling to get 10%.

Some power sources such as hydro can give you 80%. others such as solar are likely to be under 15%

With wind and mppt it is comparable with fairly large mains supplies from turbine or diesel engine. without mppt you need  much larger machine to get the same energy production.

I think there is a lot of confusion between efficiency and value for money, that is a personal thing and can't easily be analysed as everyone has different variables.
Flux

[kitestrings]

Ok you have a whole new mill, completly different than your 12 volt  one,and on this new mill, at a different voltage even,  your using the classic on it, Do you think thats a fair way to compare the two systems?

AM,

No, and that is not what I was suggesting at all.  What I was saying is that if you look at the losses on a 12V system, especially with any distance involved they quickly start to add up.  And, if you have to add line resistance to improve the load match, well that's not usable power either.

Oddly, I've long thought 12V to be a good choice for small systems.  Until recently the choices of inverters, end-use equipment, lighting and refrigeration was much better for 12V.  Our Trace inverter is still plugging away, in the shop now, after over 25 years.  We also have quite a bit of 12V lighting & a fridge.  I would have never been solely reliant on an inverter for the essentials (light, water, refrigeration) in years past.  I'm not questioning your choice there.

I do believe you'd see more power out of your 12V turbine with MPPT (and we would ours, if we choose); likely there's more opportunity with 12V than say 48V.  There was a line earlier about building a MPPT controller.  To me, unless you have a passion for this sort of thing it's been done and done well already; at an affordable price.

When we have a 'clipper' role through like the one we had yesterday - short lived but hellish winds and snow - and the controller springs to life, and sends some of it streaming to the water heater without any user intervention, it's hard not to appreciate that someone got things right.

Best, ~ks

[ChrisOlson]

I've been testing a high-end dynamic tracking solar controller on a wind turbine.  This type of tracking does not unload the turbine to do a "sweep".  Amazingly it works really well.  It does not track as fast in wildly varying wind conditions as a Classic does but it adjusts very well to the average wind speed for the day and adjusts the turbine's operating speed to match it.

It works very well with a high voltage generator and high-torque, high lift, low rpm GOE222 blades.  The GOE222's are very forgiving on tip speed ratio and refuse to stall.  So if the turbine is operating at 300-400V in 15 mph and a gust comes where the rotor should immediately speed up and the controller should adjust the voltage up to 550V, the torque from the GOE222's takes over and "grabs" the power in the wind gust without the controller having to adjust the input voltage very much.  It adjusts the other way also and starts making power to the battery as soon as the blades are spinning, even in 3 mph wind it is putting out 45-50 watts.

If the wind has a higher sustained speed, then the controller gradually tracks the input voltage up and lets the turbine speed up to match it.  I really like it because it is smooth steady output from the turbine without the ammeter needle jumping all over the place.  And with the GOE222's the turbine runs nice and slow and quiet.

[lifer]

Sorry to interrupt this interesting discussion, but I have a question related to the main subject: what's the optimal airgap between magnets for an axial flux PMG?

I want to build a 12 poles / 9 coils alternator using rectangular neo magnets (2" x 1" x 1/2"). Even on this thread there was various suggestions regarding magnets spacing (one magnet width, half magnet width or even 3/16 of the magnet width).

I've build a cardboard "prototype" to play with but still have doubts. If the airgap is a magnet width (like in the thread starter's posted image), we came to this situation:



You can see from the image above that when phase 1 is at full voltage, the other two are almost zero (though in theory they should be at half the maximum (opposite) voltage).
Seems like putting the magnets closer it's somehow better but there might be some (important) flux leakage between magnets.

Does anyone have any documented suggestion? By the way, the PMG will charge a 24V battery block using a MPPT charger.

[Flux]

I think you are confusing air gap with magnet spacing. The air gap is the distance between the magnets on the front and back rotors. It is the region where the stator sits and includes stator thickness and the necessary air gaps for it to rotate. Typically for 1/2" thick magnets the air gap will be about 3/4"

These designs are very tolerant of spacing between magnets and within reason the more magnets you crowd in the more output you get.

A good compromise for that shape of magnets is to make the gap at the outside the same as the magnet width.

Don't get bogged down about magnet and coil leg positions, if you make the hole in the coil the same size as the magnet it will link all the flux. With the magnets crowded at the inner radius you can come under magnet width , say 2/3 magnet width and keep it full on the outside.

Don't go bothering yourself about the waveform of individual coils. Your coil is not a single turn but a series of distributed turns and there will be differential action as the magnet progressively links the coil.

Unless you choose something very far from the normal the phase volts will be something approaching a sine wave and when star connected the line voltage of these machines is in reality a very good sine wave.

If you have ever met that equation E = BLV forget it and also forget all you have read about magnets crossing coils at right angles. They are strictly correct if you understand how to interpret them but a single conductor doesn't exist and a coil is just a series of conducting loops with flux linking it in one direction and then the other. If you look at transformer equations they will give you a far better idea of what is happening.

Flux

[lifer]

Thank you very much, Flux! Actually, there was a misspelling ("airgap" instead of "gap") - at least I know the difference!

My design will be a classical 12 poles/9 coils (nothing fancy) but I want to run it at higher RPM's (over 600) to get a higher voltage (to decrease rectifier and line losses) as I'm going to use a MPPT controller anyway - so higher (voltage) is better.

My thought was that to get maximum voltage from a coil, both legs should be overlapped by two adjacent magnets at a time. That's where the "V" inner hole idea came from.

So, to make it clear: the gap between magnets (at the outside) should be the same as the magnet width, right? Then the coil shape will be rectangular or wedge, depending on the available space and personal option?


PS: I have another little question for you: what steel alloy to choose for the rotor back plate? I've called a CNC plasma cutting company today and they asked me about the "flavour" of steel I do prefer. I don't know what's best in this particular case. Maybe something with magnetic properties? Do you know any steel alloy code or something? Many thanks, again.

[boB]

I would love to get people's takes on serpentine winding and overlapping phase coils  Vs.  the typical single coils next to each other methods as discussed here.

boB

[Mary B]

Chris, odd idea for 3 phase water heating. Be more expensive than a single water heater but what about splitting the water into the water heater into 3 streams, run them thru a smaller 120 volt single heater tank, recombine then into the main heater.  Or dump to 3 huge peltier junctions to cool the beer fridge 

[ChrisOlson]

Chris, odd idea for 3 phase water heating. Be more expensive than a single water heater but what about splitting the water into the water heater into 3 streams, run them thru a smaller 120 volt single heater tank, recombine then into the main heater.  Or dump to 3 huge peltier junctions to cool the beer fridge 

Well, that's what I was doing at one point.  We got two water heaters and four elements.  The top element in the primary heater has always been "always on", controlled by its thermostat.  The other three I had wired in delta and was using them for a clipper and trying to heat water with it.  It didn't work all that good because the turbine basically has to be running flat out at max voltage for the clipper to come on.  So to get any hot water out of it the turbine was screaming at 10 TSR in a 15 mph breeze.  And if you don't have it screaming at max voltage, then you don't have enough power to heat any water anyway.  Being a resistive load, the elements dissipate power based on the input voltage and they aren't very efficient if they're not drawing at least 1,000 watts, and 2,000 watts is better.

One kWh has 3,400 BTU in it to the water after the minor losses in wiring and the element.  One kWh will raise the temp of the water in a standard 50 gallon residential heater only 8° F.  It takes a LOT of power to heat water with electricity.  If you're only dumping 500 watts to a 50 gallon heater it's about like spitting in the ocean.

I finally gave up on that experiment and went back to powering the elements with the inverter.  The inverter can power 'em at their full rated 2,000 watts and it heats that 50 gallons of water up from well temp to 135F in only 4 1/2 hours.

[Flux]

 Lifer
The various steels are chosen for their mechanical properties, The better magnetic steels used for motor cores are not available in plate form. Just go for a simple mild steel ( probably you will get that if you go for the lowest price). Many use brake discs which are cast iron, this is considerably worse than mild steel but in this application it makes little difference. The thickness for mechanical strength is even then greater than you need to carry the quite low flux linking the magnets of an axial machine.

boB
That really is an interesting question. The answer depends on the type of alternator you are building. For radial with an iron core all machines use the convention that has been established over the last 100 years. A single phase winding will have one coil group per pole, each coil may be one of several in a pole group. If you treat it as a virtual pole design you halve the number of pole groups and wind them wider.

When you go to 3 phase you just add two more windings at 120 electrical degrees and to do this the coil ends have to overlap. The coil per pole winding is usually done as a 2 layer winding, as a mush winding in small machines or with specially formed evolute windings in big ones. The virtual pole type is wound with coils of straight and variously bent ends called a concentric winding.

For single phase you can use the serpentine winding, it only needs one big single coil to wind then you thread it in and out of the pole slots. It is also common to do a similar thing but wind the poles as coil loops ( skein winding).

You can do this for 3 phase in small machines but the ends now conflict and you end up with overlaps as in a mush winding, it becomes less attractive. with iron cored slotted machines winding resistance is not a major issue, the output is limited by reactance and a bit of resistance just drops the efficiecy a bit.

Now to the axial design that everyone builds:- The first commercial alternators used many forms of construction as the pioneers found their way in the early industry, Axials were common and for the time worked well but they never evolved into the polyphase era so all we can learn from the pioneers such as Ferranti, Mordey and Siemens is only useful for single phase.
They did sort out many issues that the modern world seems to have had to re-invent such as eddy loss in thick section copper.

When you try to get two extra phases into these machines you have to use overlapped coils if you follow convention. this is no trouble on the outside where there is plenty of room but it becones a bit of a nightmare at the center where space is virtually non exixsant. Serpentine winding fares no better here as the crossings come in bunches and not all of the limited space can be used.

I did my early axial experiments with overlapped coil windings and to wind something with no slots to hold the coils in place is a nightmare, the concentric virtual pole with straight and bent coils is easier but with thick coils it is still a challenge.

I don't know why or how Hugh Piggott came up with the 4 magnet 3 coil winding but it transformed the ease of winding and unless you are a genius at overlapped coil windings in axial machines you will not beat its performance, part of the winding is not wound, but the rest is wound very effectively and with shorter turns of low resistance, the overlapped coil uses all space but adds long end connections with added resistance.

For mppt where resistance is not such a big issue you may do better with overlapped coils but the complexity of winding makes it a debatable gain.

Why do we use axails anyway? I am sure the answer is  that they are a lot more forgiving in that you can change the flux density by simply changing the air gap, with a radial you rewind it if you get it wrong. The other reason I suspect is nothing more than the absolute simplicity of the 3 phase single layer design. Only skilled people can cope with overlapped colis in radial designs and get the connections correct even if they get the coils in. Take an overlapped coil axial machine with no iron core to a skilled motor winder and I suspect he would sweat blood to wind it.

Hope that helps.

Flux

[kitestrings]

Wow.  I'm again reminded how fortunate we are to have you here Flux.  I should be paying many here for the educational benefits.

Somewhat related and something I've wondered, is there any benefit to having an axial wound where each phase is wound with a continuous conductor (or multiple conductors), such that the only terminations are at say xyz & abc?  You'd have potentially no embedded connections which might be a plus, but it sounds like the any added resistance of multiple connections points is minimal (for axial's) if I followed you.

~ks

[joestue]

A 12 magnet, 18 coil structure would certainly result in a stiffer alternator.
Probably to the tune of +50%

thanks to 3-D printers and the axial topology, it should be rather easy to make a jig to hold all 18 coils in place, while you fold all the coils over at the same time and press them into a slot to hold them where they need to go --how they insert "mush" windings into radial machines i have no idea.
then compress the coil structure as thin as you can, while you fill it with epoxy.*

given that the magnets are .5 inches thick, you have space available on the inside diameter (and outside) for the overlapping coils to stick out.

for cored alternators, the optimum inside to outside diameter is a ratio of .6 to .7, mostly due to lack of space, most of the time you can't have the coils extend past the core.

*given the amount of force it will to take, i would suggest laying the coils out on a plywood disk, with finish nails as pins to hold the coils in place.
you'll need a few tons to squash the coils into position.

[Flux]

KS

If you were producing a batch of alternators it would be worth winding each phase as a continuous winding, mainly to save the time of doing the interconnections.

For a one off it is probably not worth the time spent on making the winding jig. There will be little difference as long as the interconnections are soldered properly. For those not good at soldering it may be a worthwhile step even for one off.

Quicker easier and better than soldering is to use"silphos" if you have oxy/acetelyne available. Silphos may be a trade name but it is copper fluxed with phosphorus and should be available in N America under some name.You don't need any cleaning or stripping or extra flux and it will never melt even if you murder the stator.

Flux

[tanner0441]

Hi

We used to use a fluxed copper rod when I worked in AC. It went under the trade name of "Cupro" it would actually melt with a "Map Gas" torch.

Brian

[Flux]

I expect it is the same stuff. When I checked on Google I found Silphos, Sil-phos and Sil-Fos.

It seemas to be quite common in the AC industry for joining copper. Yes I suspect Mapp gas will melt it,

THe beauty of oxy/acetylene for joining wires is that you can do the joint quickly without burning back the enamel over a large area. Brilliant for joining larger section copper.

Flux

[xxzxxz]

SinWT must  be = 1.

понимаешь?

[DamonHD]

I don't think that many of us are going to understand Russian.

For maximum utility to all members, and for you, please keep to English as far as possible.

Rgds

Damon

[boB]

SinWT must  be = 1.

понимаешь?

понимаешь = understand   I think...

Thank you Flux for your explanations !  Very helpful !

Really appreciate your sharing of knowledge here.
boB

[TDC]

what most people do  is build their rotors and then do some test coils .

Just had an idea for testing coil and magnet shapes.  Rubber magnets,  those flexible things on your fridge!  Cut your you test magnet with scissors and slap it on a thin rotor plate.  Then wind a few turns and slap that on a thin plastic stator disc.  Iron shavings could go on the plastic stator disc to observe flux patterns.  Am I nuts? Could it  show general trends or is it to far from the real deal?   

[Flux]

This would be a good idea for anyone studying the various waveforms with different shape magnets and coils. For building we are normally forced to use standard ( cheap) magnet blocks and most of these have been evaluated reasonably well.

Would certainly be useful for a school science project. The leakage flux from such thin magnets would bot really be typical of thick magnet blocks but as a first starting point that wouldn't matter much.

Flux

[TDC]

This would be a good idea for anyone studying the various waveforms with different shape magnets and coils. For building we are normally forced to use standard ( cheap) magnet blocks and most of these have been evaluated reasonably well.

I should have noted, I was thinking primarily of the 2" square ferrite that Chris and Hugh have used recently. I think the situation is different when the mags are packed on the rotor with the inside corners almost touching.  Do you think a trapezoid shape would gain much with the triangle shape coils used? Has anyone looked into pricing of custom ferrite? I believe a friend can charge them for me.     Has anyone tried cutting ferrite with a diamond blade tile saw? 

[ChrisOlson]

I should have noted, I was thinking primarily of the 2" square ferrite that Chris and Hugh have used recently. I think the situation is different when the mags are packed on the rotor with the inside corners almost touching.  Do you think a trapezoid shape would gain much with the triangle shape coils used? Has anyone looked into pricing of custom ferrite? I believe a friend can charge them for me.     Has anyone tried cutting ferrite with a diamond blade tile saw?

First of all, a diamond tile saw blade might cut ferrite.  But it would get pretty hot.  Ferrite magnets are made by cutting them to shape first, then they are magnetized.  Nobody I know of has had success cutting one after it's magnetized.

Secondly, you can look at the flux patter with iron shavings, or similar.  But it doesn't tell you much.  It is not the flux pattern, but the change in the flux as the rotor rotates that excites a current in your coil wires.  At one point I was of the school of thought that wedge magnets were "better" than blocks based on experiments that Ed Lenz did.  But after much experimenting I found they are not.  The only real advantage of wedge magnets is that the shape allows you to put more poles on a certain sized disc to get higher AC frequency.  The higher frequency will do more for you in developing a more powerful generator in a compact size than adding more tesla of flux will.  When I came to that realization some time back, taking a 12 pole nine coil generator that was good for ~1,000 watts at 300 rpm, putting a gearbox on it and running it at 1,000 rpm and much higher voltage and AC frequency instantly turned the 1,000 watt generator into a 3,000 watt unit.