Large f&p type opinions?

[cdog]

Hello all, as usual my mind is wandering.
Does anyone have opinions on building a large f&p type alternator, capable of say 10kw.
I read on the back shed they have converted a f&p to 7 phase and virtually eliminated cogging.
The converted stator has 48 mags and 42 poles.
One could spin the daylites out of this and not worry about the mags in the next county, I would also think that heating issues would be minimal?
I suspect this wouldn't be as good use of raw materials as an axial, but perhaps ceramic mags and an "olson" gearbox?
Just throwing this out there, don't burn me too bad!!
Thanks as always,
Cdog.

[jlt]

I am not a big fan for geared up alternators .But it will give you more output,after you get it to start turning. That is the main problem. If you decide to go that route  you could try using the jack shaft and chain drive that a lot of snowmobiles use I think It is about a three to one  ratio. Plus old snowmobiles are easy to get.and cheap .Around here . When my mind wanders it is always in direct drive.I a thinking of  putting 3 or 4 stators together and making the rotor from a large pipe. and using neo mags with it. that will get more output. with a large air gap.

[TomW]

Those "Backshedians" over on thebackshed seem to be the ones who have worked these F&P's nine ways from Sunday as the saying goes.
I suspect someone there has tried nearly anything you can imagine?

They were far more common down under so folks under the brim have a lot of experience.


I had one for awhile and it was interesting but I din't tinker much with it and sent it to Jerry who fooled with it some.

I remember 42 coils and mags on the inner rim of the drum thingy. Didn't seem very hearty with the all plastic construction other than the shaft and bearings. There was some upper limit on power out, too, based on the original motor model. Seems 2KW was the number?

Just my limited experience and not much help.

Tom

[cdog]

I was thinking like a 2' to 3' pipe with 48 large neos, and a stator to match, but I have no idea of dimensions, windings, wire gage etc.
Is there any reasonable way to calculate this, or would it all be exspensive guess work?

[Flux]

The main virtue of the f & P is that it is mass produced and cheap, particularly when surplus.

To start building something like that seems pointless, it was designed as a synchronous motor and is not a good starting point for an alternator. It has all the snags of slotted iron cores plus a lot of extra ones from its salient pole construction. You can with care reduce cog to a low level but that may not be good enough if you use much speed increase. At least using speed increase does partly solve the problem of reactance limiting but it will still be the final limit on output.

In many ways it is inferior to a motor conversion except that it is a better mechanical arrangement to spin the magnet rotor outside the stator. Unless you can adapt the standard unit then making the special core will not be easy or cost effective.

I don't know how far you can push the standard unit, it is designed for low speed and the construction may not be good enough to raise the speed a great deal. I doubt that you can get 10kW at a speed where it will hold together mechanically.

If you must go down the iron cored route with external drum magnet then I would suggest you build something similar to Hugh Piggotts brake drum machine. Slots are not such a virtue as you may first suppose and once you accept that slots are not necessary then you can use the outside of a standard induction motor core pack, which is readily available and made of high quality iron. You now have no cog at all and the core loss is quite small so it would lend itself to speed increase if you so wanted.

Flux

[cdog]

Is this the same Piggott machine that the coils appeared quite crispy, an older idea before he started down the axial route?
Do you have a ballpark idea of what size motor to look for to get large enough laminates to pull this off?

[Flux]

Yes that's the one. The snag from Hugh's point of view was the difficulty of obtaining scrap parts to get a sensible size air gap.

I would only suggest this for someone with machining facilities but it would be easier than a F & P based copy.

If you are going to gear it then the core of something like a 5hp motor would be good starting point. you would only need about 1/3 of the core length.  If you can't machine a round drum for the magnets then see if you can find a post by Oztules, he made something in the form of a polygon with flats by notching and bending a flat bar and welding it, pretty crafty idea. I think his was for a slotted core but the same magnet drum idea would work well.

Flux

[cdog]

I have access to most any machining ability.
How about milling slots on the outside of the laminates and wind the coils the same as a motor conversion, but on the outside?

[cdog]

If one did cut outer slots to bury the wire, would the laminates need to be skewed to reduce coggin, similar to a motor conversion?
I have a 50hp donor motor, it should be large enough to cut slots for the 42 poles?

[joestue]

i would look into selling that 50hp motor and look into getting new iron.

the issue is that induction motor laminations are designed to do just that...
http://en.wikipedia.org/wiki/File:Motor_laminations_by_Zureks.jpg
also, are you sure your lamination stack is a circle? most new motors are semi square, with holes for the bolts to go through.

It will work, no doubt about that, but it only gets you about half the power it could if the iron were thicker to allow more copper.

[cdog]

Not sure what you mean about the iron being thicker, I can look up the specs on the motor, I also have the opportunity to get others, anything ideal in your opinion?

[joestue]

If you look at the image on wikipedia, lets say the slots are one inch deep right now.

if you go cut slots on the outside that are .6 inches deep, aligned with the existing teeth such that the iron maintains a minimum .3 inches width,
you're left with hardly one half the original area for the copper windings, and about a third the original core area.
obviously you can trade core area for copper area, but everything is a compromise.
which is why i would suggest getting new iron.

also, things get really messy electrically when you run full pitch windings, but its your only choice if you reuse a motor's core like that.

[cdog]

Obviously I am in the fog on motors, is there an ideal setup?
Surely there is a motor of good size that the laminates are deep enough to have usefull material left?
My problem is I can build most anything, but I am short on the small inner workings

[cdog]

I also have to ask what do you mean by ''new iron''?

[joestue]

new iron as in hit up all the companies that mention prototyping on their webpages and ask them how much it would cost.
electrical steel is about $1/lb, and water jet cutting is pretty cheap. More like 2-5, depending on who you get it from. at the factory, non grain oriented iron is about 2-3 times the cost of mild steel.
http://www.protolam.com/
http://www.hydroblanking.com/
http://www.racinestamping.com/home.htm
http://www.laseroneinc.com/motor-lamination.htm
http://www.lasertechnologiesinc.com/Applications-Prototype-to-Production

it is a pretty steep learning curve, compounded by the fact that there aren't any books written specifically for low speed machines, where iron loss is negligible compared to copper losses and the magnets make up more than half the cost of the alternator.

you could probably build yourself a very nice cascaded 9:1 planetary gear box and drive a smaller 20 or even 10 hp motor conversion for less than a 10Kw direct drive diy machine.

another option is to drive multiple (like 6 or more) induction motors from a single bull gear, each with a 9-15:1 step up ratio, that idea had been proposed to solve the issues related to multi megawatt wind turbines, but they developed better transmissions instead.
the reason i brought that up was because there was a prototype built and a 20KW transmission was built for about $2,000 worth of gears from seattle gear works IIRC.

[Flux]

If you must have slots then choose a core for a 2 pole machine. There will be enough steel at the back to let you cut slots deep enough for a high pole number on the outside. For a 2 pole you will have lots of teeth sharing the back iron and it will be thick enough to deal with the few teeth used on a high pole count winding.

Yes some cores are far from round on the outside so you will have to choose suitable ones. most have at least small slots for clamping strips round the outside but these won't matter much.

If you mill slots you will short the core out with burrs so you will need to separate the lams and clean the burr off and restack so you can distribute a few rogue clamp slots round evenly and it will not matter.

Really I consider milling slots a last resort but anything is possible. I am still not that convinced that the slots are necessary for a wind machine, the iron loss requirements are very different from a constant voltage , constant speed machine. normal machines are optimised on full load and you reach a state where it is most economic when core loss equals copper loss. This is not a good starting point for a wind machine especially if you use speed increase, you need to keep core loss as small as possible.

Much depends on whether you are concerned with maximum power or best low wind performance.

Flux

[cdog]

Thanks guys, a lot to soak up..
What is used for an insulator between the laminates?
Will skewing the laminates help with cogging, like a motor conversion?
New iron seems far easier then searching the country for a donor that happens to be ideal.
Flux, if I do build something like this, I want low speed performance AND high output! I understand somewhat your concerns about the losses involved with the slots, but it appears to me that Hugh ened up fastening his copper with a glue or epoxy?
That seems like a source of trouble to me, with the slots the copper should stay put?
I would want the machine to be reliable, number one, if it takes more materials(magets,copper) to get what I am looking for, even though wastefull, then so be it.
Thanks for all the advice guys,
Cdog.

[ghurd]

It seems you are missing some of Flux's points.
There is no need to have slots and the issues they come with, like cogging.

If you can machine a decent one with slots, you would be better off making one without slots.
G-

[cdog]

Thanks Ghurd, no doubt I am missing a lot, how do I keep the wire in place well without slots?

[Flux]

It certainly is easier to wind coils in slots but it can be done without.

You will have to pre form the coils to shape and stick them on. I usually stick a band of nomex round the core with epoxy to give insulation but you could use a layer of tape and epoxy that. Then you can stick the coils down with epoxy or similar. When everything is in place you can wrap a layer of glass tape round it nice and tight and then epoxy the whole thing.

You could cast the coil assembly in a solid block of epoxy if you are in a wet climate and are using high voltage but I haven't found it necessary as the magnet drum gives a fair bit of weather protection.

If you do go down the slot route then you can skew the lams to reduce cog. Various materials are used between the lams but most modern machines trend to use an oxide coating formed during the annealing process. if you strip the stack and just de burr the edges and reassemble you will probably only have a few point contacts and the voltage is very tiny so it will be ok. If it gives peace of mind you could varnish the lams but without the facilities of stoving used by the manufacturers it will most likely be soft and the points may still prick through. I am sure it will help a bit. Even letting them outside to rust on the de burred edges will be good enough.

Flux

[picmacmillan]

not sure if this is relevant..but i think zubbly scewed the magnets,and used new material for the drum(wrong terminology) on his conversions not the wire.

  like i said i apologize if this is redundant, i have seen the very big stator hugh piggott built, it had a name..oh ya nervona..    hi hugh     maybe what youre building is a whole different animal....take care...pickster

[cdog]

Flux, I follow what you are saying, but if I were to epoxy the coils in place, one may as well go with an axial from the heat point of view?
There must be a non~cogging way to secure the copper and leave it unburied??

[joestue]

cut two disks of plastic just larger in diameter than your core stack.
cut the slots in the plastic.
insert coils into the plastic.
insert ~2mm thick plastic strips between the coils, or don't.

wrap the circumference in clear plastic.
dam up the bottom with a ring of silicon right next to the coil, to prevent the epoxy from draining out.
heat the entire assembly (electrically) to about 100C
pour epoxy in it, no vacuum required.

remove the plastic strips after the epoxy is cured (will approximately double surface area for cooling.)

[Flux]

If heat is your main concern then the best construction to get rid of heat is a motor conversion or something built similarly with the stator on the outside. If you put the magnet drum round the outside then cooling will never be so good.

A slotted core design with outer rotating magnet drum will be slightly better cooled than one with windings stuck on to a smooth core but the difference is not that great, the coils in this type of design are thin and there is quite good heat transfer to the massive core, certainly an order of magnitude better than the normal axial construction.

This heat problem is only a concern if you don't match things properly, normal axials work ok down to about 60% efficiency and you can get away with lower. With a motor construction you may be able to get even lower for wind duty but I fail to see any virtue in having a machine dissipating 3/4 the energy you produce. If you go down that route and you want 10kW you will need 40kw input power from the prop and it will be much bigger and expensive than a decently matched design.

If it's the top end power you want then match that region at decent efficiency and heat will not be a problem.

If you must retain the low wind performance and still have the 10kW rating then you will need some form of load matching.

I have said many times here that I consider cogging to be a red herring as far as wind machines is concerned, yes it can be a serious problem if you don't tackle it, but there are ways of keeping it low enough not to be an issue.

In the pre-neo days I built several geared pmgs and found that cog was not the limiting issue. If you made a sensible attempt they would all start to rotate in practical winds but the iron loss in the core caused so much drag that it took much too high a wind speed to break the prop out of stall and get it up to running speed. Reducing cog to absolute zero wouldn't have helped in the slightest.

For direct drive on motor conversions with the comparatively small amount of magnet you can squeeze in there , if you moderate cog they will probably start ok. If you go to an outer magnet drum and keep the slots and saturate the teeth the iron loss will be higher and you might start to hit trouble, if you introduce gearing I know you will.

Your choice in the end, choose an efficient design and match the load or dissipate huge amounts of heat in the alternator and choose the design that best suits a heater rather than a generator and use a larger prop to power it.

You might have a look at what Chris Olson is doing, he has tried many things and a lot of his findings confirm much of what I did in the early days but at least on his geared machine he has kept away from slotted iron cores and won't fall into the iron loss during start up trap.

I think you may well be advised to stick with larger machines and direct drive if you want to go the slotted route, it works absolutely fine for AWP and others but it tends to be a big alternator for the higher powers and changing from ceramic magnets to neo will reduce size but you will increase the iron loss so it may be a compromise on the maximum flux in the end.

Flux

[joestue]

once you machine a proper block for the rotor and bury the magnets underneath pole pieces the flux density is entirely up to you.

the root problem here is that you're going to try and get 1/4th the design power out of a motor... at one tenth design rpm.
when you consider that at 90% efficiency, you've got 37Kw out, and  41 in (for 50 hp), that's 4Kw waste heat.
to get 10Kw out at say 17 in..
personally i can't see this as practical.
7Kw of waste heat, most of it in the copper, only bonded on one side to the core with epoxy...

combine this with the constriction i described yesterday if you go and cut slots into the outside of the stator and its not looking doable.

It is worth a shot using this motor as a synchronous conversion however, because it will only cost you your time and 50-100$ worth of magnets.
machine a new rotor, square, toss 4 magnets on it (start with 5mm thick magnets) and machine 4 pole pieces for it, with an air gap of 1 mm, such that they form an arc 80 degrees wide (the gap between the pole pieces will be 10 degrees) bolt them on with stainless steel bolts.
then spin it at 1745 rpm and see what kind of voltage you get out of it.
from there you can increase the air gap, add more magnet or reduce the air gap, cut the arc down to 75 degrees, etc, etc.

from the data collected you can then go and do the math for what a 6 or 8 pole machine is going to look like, and how much effort its going to take to rewind it.

[cdog]

Wow guys, a lot of great info.
If I were to consider a large typical motor conversion, I have found a very old 100hp 550 volt 3 phase 885 rpm motor I could do some testing with.
The frame is huge, old iron, that I would likely build my own casing for to shave 500lbs or so!

[cdog]

The tag says type k, 6325s frame if that means anything, general elecrtric.

[cdog]

Thanks for all the advice again guys.
I think I will try taking this 100hp apart and try a test with a coil on the outside of the lams and build a rotor to go around it.
I have been reading all I can find on motor conversions to lessen my rookie questions, however I do have one more...
If I were to compare the amount of output from the mags and copper in an axial to the same amount used in a laminated machine, how much less output could I expect for an identical rpm?
This will give me some idea of the amount of raw materials to get.
Thanks again,
Cdog.

[joestue]

i've got a 200 page document that explains that...

let me try to paraphrase...

the only reason you use large axial machines is to get higher torque density.. and by higher i mean like 3 times as much torque for the same weight...
radial machines are more efficient.

did you read my post on converting that 100 hp to a synchronous motor and keeping the original winding?
this could be done in a weekend.

[cdog]

Yes, I read the post, thank you.
I had the misconception that magnets facing each other would have stronger flux than a single magnet and laminations?
Obviously to do as you required needs testing, but I am still in the fog as to how one calculates wire size to get the voltage and amp capacity one is looking for in a given coil size....
Would it be just luck if the motor was wound with the proper wire, or can you guess by its specs that it would be close?
This is a crash course, a lot to take in, but I appreciate it!
Cdog.

[joestue]

Yes, I read the post, thank you.
I had the misconception that magnets facing each other would have stronger flux than a single magnet and laminations?
Obviously to do as you required needs testing, but I am still in the fog as to how one calculates wire size to get the voltage and amp capacity one is looking for in a given coil size....
Would it be just luck if the motor was wound with the proper wire, or can you guess by its specs that it would be close?
This is a crash course, a lot to take in, but I appreciate it!
Cdog.

calculating the wire size is the easy part, the hard part is figuring out how much power you can get out of it after you convert it to a synchronous motor.
the usual 700 circular mills per amp has no bearing on this.
volts and amps are 100% interchangeable.

on the other hand, i have no idea how much magnet you need to get x amount of torque at what air gap if you just wind coils on the outside of the smooth circular laminations.  and i suppose the question isn't what amount of torque do you need, but how much copper do you need to get it at what efficiency?

if you have an idea of what a microwave oven primary looks like, picture 16-24 of those on the outside of the core ( can you give us some detailed dimensions?), with a 3 inch by 1 inch by half inch magnet over the top of each one, with the drum a solid 30 millimetres away from the core.

now try an figure out how to put it together after you pin the magets on.. each one is going to have a good 100 pounds of pull. .. and thats just for a 3 inch long magnet.. if your core is a more modest 8 inches deep and 1 foot diameter.. you'll need a 20 ton press to assist you taking it apart.

[ghurd]

Thanks for all the advice again guys.
I think I will try taking this 100hp apart and try a test with a coil on the outside of the lams and build a rotor to go around it....

http://www.scoraigwind.com/brakeupdate/index.htm

Look around the area where it says-
"The 20 magnets Les used are also neodymium ones, with a rather unsuitable curve on them, but they are nicely fitted to the rotor by painstaking machining.
This 20 pole machine can produce over 1kW at 320 rpm with approximately 70% efficiency."

[cdog]

Ghurd, thats a great example, according to Dan's 20' build, if we double the blade diameter we need 8X the alternator, comparing a 10' machine to the 20'.
It actually worked out to about 5X the copper and mags...
Can I apply this line of thinking in this application as well? If so, it would give me a much better idea of what direction I  need to go.
Cdog.