Newbies first build

[BLUE]

I've done a LOT of searching on different machines in the past month or so.  I know the Axial is easer to build and the latest one Chris Olson built is REALLY impressive!!! But I've always liked the rotary type. I'd like to build a PMA like the pic. I think it's one of Flux's creations.
My plan is to build a 24v machine, partly off grid system. I'm looking to buy a 1,000 watt inverter http://www.solar-electric.com/exxp24vol11w.html Depending on how much power I can generate I will buy more inverters.  A 12' rotor would be around the size I'd like to build or buy http://royalwindandsolar.com/ It might be a little hard to get clean air hear in the Pittsburgh are, because of the hills and trees, their not very wind mill friendly, but what ya gonna do... Just over the trees is all I want to do, about 40-50' from the ground. Do they make telephone poles for these things...

With the axial type I seen a 3:4 ratio is standard. Or more commonly  9 coils to 12 magnets.
Do the radials have the same ratio?
I think Flux used a 10" core, a 12" drum with 40 voice coil mags and I can't find the wire size and turns (it's in my notes somewhere) but two in hand for his. I can't find any other info on it, and believe me , I've tried looking. So many Flux posts so little time.

I don't need to recreate the wheel, and something like Flux's looks like a good build for me.
Bigger is better in my book, (altho the magnet$$$$ might hurt me) maybe a 12" X 3"core...
If someone could give me a place to start:
1. Wire size. 24v
2. Number of turns.
3. Number of coils. I'd like to play around with "Y" and Delta wiring.
4. Number of magnets. 24 ? 15" drum? I found these; 3X1.5X.75 N42  http://www.magnet4less.com/product_info.php?cPath=8_39&products_id=76
5. Do I stay within the 3:4 ratio.
6. A gearbox might be an option but I'd rather be direct.

I might be a newbie on the board but I've been machining and welding for over 25 years (machinist by trade). I have a 3 phase set up (from single phase 220) powering my mill and lathes. A Millermatic 250 wire feed welder and flame wrench. I can build what ever is necessary.  How do you like me now?

Still, lots to learn. I have more questions than answers.

Tim Taylor has the right idea...  More Power!

Can anyone help a "newbie"?
Your help is much  appreciated!
Blue

[Flux]

I can't remember the details of that winding. I may have the information somewhere, if I find it I will let you know.

I have a feeling that the coils were wound with 2 in hand probably 1mm wire. The whole winding was split in two sections with each section star connected and having its own rectifier. That way I didn't have to deal with very thick wire or use parallel coils with all the associated problems.

It was 20 pole  and used a conventional concentric coil 3 phase winding. The 4:3 magnet coil arrangement works quite nicely with axial machines where winding space is very tight on the inside and overlapped coils is very difficult to do. With a radial you no longer have these problems. There are 15 coils with bent ends ( these are the coils that would be used in a single layer axial) but as you can see there is room to fit a whole new layer of the non bent coils in the gaps so that the whole winding space is used effectively. This gives 30 coils in total.

That machine would have just about handled a 12ft prop but I used 10. The voice coil magnets were far from ideal but came surplus and very cheap. I think the source has long gone but you could substitute the 2 x 1 x1/2 blocks and make a better machine. If I was going to do it again i would use the metric 46 x 30 x 10mm which are probably better but may not be available if you are in N America.

I used the 2 star half sections in parallel for normal running and a boost converter for low winds. If your grid tie inverter works from batteries then you will have issues matching the load at various wind speeds, star delta would be a possibility if you can deal with the switching. Star/ IRP( Jerry ) is much easier.

For grid tie alone if you don't need battery back up it would be better to make it star and work at higher volts with a mppt grid tie inverter.

Flux

[artv]

Hi Blue excellent find are there more pics of this unit?Is the pic " Thomas1" the completed stator of pic "Big5".....thanx ....artv

[BLUE]

Hi Flux, I remember now that you mention it, 1mm wire (.118) two in hand.  Closest AWG would be #9 (.114),  #8(.128) is more common. Two rows of 15, 30 coils, thats a lot of copper, ouch!!
If I use a bigger core, (12") stretching the width of the coil so it wraps around the core evenly is ok?
How important is it to have the magnet fit the inside turns of the coil?
20 pole? 20 magnets total?
If I use a 15" drum I can get 24  1.5 wide magnets with a .5 gap around the inside. But I loose the 4:3 magnet to coil ratio. Is that a big deal??

or use parallel coils with all the associated problems.

Not sure what the associated problems are but I suppose I'll find out.

MPPT-- Maximum Power Point Tracking is an electronic circuit to maximize output power, generally allowing for solar to operate at a higher voltage than the battery, and the circuit supplies more amps to the battery than in a direct connection.

Not suer what this means either but I'm in it for the fun.

Anybody know roughly how many turns to get 24v using #8 AWG two in hand?
Thats 5 coils in series?
Two #8 sounds a little big, but, if thats what it takes... Like I said before "more power"
Thanks again
Blue

Hi artv, this is all I have on Flux's build:

Not sure who made this:

[BLUE]

Is there a modify button?
I don't think Thomas 1 and 2 are Flux's.
bigS and 13 are. I think.

[Bruce S]

Is there a modify button?
I don't think Thomas 1 and 2 are Flux's.
bigS and 13 are. I think.

Blue;
 Look over on the right side there should be on top right corner modify {maybe} otherwise on lower right corner there is an edit button, tho not marked edit it'll get you back into your post, if not let me know what you would like fixed and I'll help where I can.

Cheers
Bruce S

[ghurd]

Maybe its just me... but it sure looks like a lot of work compared to other alternatives.

The advantages of the rotating outer parts could be achieved with a motor conversion...  I think.
Instead of only turning down the armature, turn the inner part of the core to get rid of the teeth/slots.
An induction conversion donor motor comes as a complete package.
etc

It would still be able to throw a neo, but that does not happen very often in a well made unit.

IMHO, Zubbly's papers would be very related, and a Zubbly conversion would be a lot easier to make work well on the 1st try.
G-

[BLUE]

Hi Flux, I remember now that you mention it, 1mm wire (.118)
Anybody know roughly how many turns to get 24v using #8 AWG two in hand?

OOPS, Didn't check my math. Calculator error. More like .039 AWG 18. But the question remains, how many turns with two in hand 24v?

Hears another question:

Poles-- A way of picturing magnetic phenomena. All magnets are considered to be "dipoles", having both a North pole (which would point North if used in a compass) and a South pole (which would point South if used in a compass. In an alternator, generator, or motor the number of Poles is a measure of how many coils, permanent magnets or electromagnets are in the armature or stator.

  The poles are the magnets? Why does the definition say "the number of Poles is a measure of how many coils..."

IMHO, Zubbly's papers would be very related, and a Zubbly conversion would be a lot easier to make work well on the 1st try.
G-

ghurd, I found Zubbly's build in http://www.greenbits.com/User+Pages/zubbly  and I'm still trying to figure things out.

I don't see 4 poles here. But each loop around is 1 coil with x amount of turns?
"This winding having 2 coil groups per
phase, produces 2 poles per coil group, and is connected in a
consequent pole fashion."


Are there 4 poles in this?



I can see this being 4 poles



Things are sinking in. Just need a little help. Pics are a plus.
Thanks! Blue

[Flux]

Those pictures labeled thomas are not mine but the machine design is very similar. I am not sure what magnets he used.

Things work out best when the gap between magnets is  about half magnet width. That does basically end you up with the magnets lining up with the hole size in the coil. If you keep to that then the diameter will decide the number of poles of your chosen magnet size.

Certain pole numbers will end you up with a strange dog leg coil but if you choose numbers divisible by 4 you will be ok.

There is no hope of guessing numbers of turns, you will either need to estimate your total flux and calculate it to suit your cut in speed or you will need to do a test coil and scale it from that.

As Ghurd said it's a lot of work, don't go this route unless you are prepared for the work needed. Unless you accept a standard winding on a motor conversion I personally think a motor conversion is even more work and I believe the outer magnet slotless design has many advantages over the motor conversion.

The winding is no more difficult really than the normal axial but you need a different approach. Most of the work is in the mechanical side but as long as you can make the magnet drum fit the core with the correct clearance there is no real difficulty.

If you have to go the test coil route then you need to sort the mechanics first but if you can determine your number of turns by other means then there is a lot of sense in winding the stator first then you can make the magnet rotor fit your coils with a small and safe clearance. Otherwise you have to get the coils fairly exact. This is the only real virtue if an axial as far as I can see, you can cover a lot of mistakes by altering the air gap.

Flux

[ghurd]

Flux,

I meant possibly starting with a motor, turning the core to remove the slots, and ending up with an inside out brake drum PMA.
In my mind, that would work out the same (magnetically and electrically) as the outer magnet slotless design. Except for the magnets being on the inside.

What am I missing?
G-

[Flux]

I probably misunderstood what you were suggesting. That would work, removing the teeth is not as simple as turning them off but you can strip the corepack and remove them.

That way you have room for a bigger rotor and much more room for magnets which is a big limitation with normal motor conversions.

The good thing will be a more enclosed and protected machine, keeping the original bearings, and possibly a bit better stator cooling.

The down side is the tricky removal of teeth, the inevitable trouble of motor conversion and the retaining of magnets on the outside of a rotating object ( can be done satisfactorily as shown by many conversions). It is also a lot more difficult to wind and secure a stator on the inside of a smooth drum compared with doing it on the outside.

I have built such a machine successfully but the carrot was an existing magnet rotor from a commercial synchronous generator exciter, Without the existing rotor I would never have done it that way but certainly it is possible.

Flux

[cdog]

Flux, in a radial machine is it possible to use too much magnet, what I mean is can there be more than enough magnet used to do the job and also cause unnessesary attraction and waste into the laminates?
Or is the stronger the flux the better, to a point?
I understand much of this is by feel, trail and error, but is there and rough uneducated guidelines?
I'm sure this is almost a ridiculous question, but it was mentioned about not overgearing past 2:1, keeping that in mind can we get away with weaker mags and spin the hell out of it?

[SparWeb]

Hi Blue,
You seem to be getting great advice and suggestions but I want to welcome you to the board.
I love motor conversions myself and if you're starved for reading material I've detailed some in my user-diary and personal website stuff: www.sparweb.ca
Looks like you've got the drive and creativity to do something new or different so I don't want to say anything that would hem you in!

[Flux]

Cdog not sure where this will turn up.



View Profile Personal Message (Offline)
   
   
Re: Newbies first build
« Reply #11 on: December 29, 2010, 08:46:15 PM »
   Reply with quote
Flux, in a radial machine is it possible to use too much magnet, what I mean is can there be more than enough magnet used to do the job and also cause unnecessary attraction and waste into the laminates?
Or is the stronger the flux the better, to a point?
I understand much of this is by feel, trail and error, but is there and rough uneducated guidelines?
I'm sure this is almost a ridiculous question, but it was mentioned about not overgearing past 2:1, keeping that in mind can we get away with weaker mags and spin the hell out of it?


Some interesting questions.  For slotted cores with teeth and normal applications there is no real limit on the amount of magnet  but with wind applications there is no advantage in using more than needed to slightly saturate the teeth. Beyond this the core loss becomes excessive and it messes up the low wind end, you still gain at the full power rating.

With air gap designs including slotless radials you will never reach core saturation and you would have to go crazy to reach a point where you cause more harm than good. It's best to keep the poles in the correct place so the half magnet space between poles still holds although you still get more out up to the point where magnets touch. Just not best use of magnet material.

For the 2:1 speed increase I was thinking of slotted machines with about optimum flux in the core. Obviously if you reduce the flux the core loss drops and start up may be possible with higher ratios but there are other factors. With low flux you need more turns but you reactance limit at lower currents. You can reduce the turns by increasing speed and to some extent get back where you started but apart from trying to use ceramic magnets it seems a complicated way to go. I could see some advantage to use the ceramic magnets to save alternator size and weight. Beyond about 4:1 speed increase I suspect bearing friction, oil drag and windage will again be against you except for larger props where you may have enough torque and the lower input speed will keep the alternator speed within reasonable limits.

I had no trouble with 3:1 speed increase on wound field machines starting without flux but even using magnet steel magnets on pmgs the starting was poor at 3:1. Normal neo designs would have higher flux than this.

Hope that helps to explain some of the factors that arise with speed increase.

Flux

[jlt]

Just trying to post some pictures. that i took of a 15 inch radial I was messing with.years ago . Made from a pu wheel and used ferrite mags.First it was made with a plywood stator core with coils glued to it .it showed some promise. But then i tried To cast a stator using iron filings and resin. and cutting slots for the coils. I only tried three slots for three coils. to get an idea for how many turns to use.What i got was an eddy brake . it was very hard to turn.And then i started playing with air core axial. machines.I always wanted to go back and try to make another one using stronger neo  Mags . iF I only had the time.    I will try to find more pictures. they are somewhere on this %#@* dam computer .

[cdog]

Flux, thank you, I think this flux stuff may finally be penetrating my solid iron core!
Spar, I will take a look, thanks,
cdog.

[TomW]

I think this flux stuff may finally be penetrating my solid iron core!
cdog.

Unlike "normal" cores the permeability of your brain (loosely the ability to resist flux) will improve as it gets repeatedly saturated!  Allowing deeper penetration.



Tom

[SparWeb]

Unlike "normal" cores the permeability of your brain (loosely the ability to resist flux) will improve as it gets repeatedly saturated!  Allowing deeper penetration.
Tom

Despite my variable reluctance to point it out, Tom...

It help that you have a soft iron core, doesn't it?

[BLUE]

Thanks everyone for your input. I have lots of info to start with. Just need to find a donor motor and get winding!

Hi Blue,
You seem to be getting great advice and suggestions but I want to welcome you to the board.
I love motor conversions myself and if you're starved for reading material I've detailed some in my user-diary and personal website stuff: www.sparweb.ca
Looks like you've got the drive and creativity to do something new or different so I don't want to say anything that would hem you in!

SparWeb, thanks! It's going to take me some time to go threw all your projects but from what I've seen so far theirs lots of good information. You answered some questions I had. The You Tube video helps me big time. Most of my learning is hands on and watching others. If I can't be there, video is the next best thing. Have any more?

Flux, in a radial machine is it possible to use too much magnet, what I mean is can there be more than enough magnet used to do the job and also cause unnessesary attraction and waste into the laminates?
Or is the stronger the flux the better, to a point?
I understand much of this is by feel, trail and error, but is there and rough uneducated guidelines?
I'm sure this is almost a ridiculous question, but it was mentioned about not overgearing past 2:1, keeping that in mind can we get away with weaker mags and spin the hell out of it?

cdog, I found this hope it helps. Read the reply from oztules #4 and flux #6
http://fieldlines.com/board/index.php/topic,139988.0.html


Again, thanks to everyone.   

God bless!
Blue

[SparWeb]

Blue,
Maybe from your own experience, you've seen old compressors left out the back of buildings when the shop buys a new "quieter" one.  I really don't have much difficulty picking up old motors that are suitable for conversion (and some that aren't!). 

Sorry I don't have any more funny videos like those ones.  I did test the Toshiba conversion in just the same way, so you may find it interesting to compare.  The Toshiba, by the way, hasn't been used since I finished the conversion because it's 2x heavier than the Baldor.  It would require at least 12 feet of rotor, too, making it too large for the tower I have now.

I also have a partial answer to anyone who wonders about having "too much" magnet.  My experience with 4 motor-conversions now is that the more the merrier.  Provided that you have wire connection options that permit you to select a 3-phase option in the right speed range, using a lot of magnet will do several useful things:

a) Low cut-in
b) Stiffer power range
c) Delayed onset of reactance limiting

The reason for a) is known by many but the trouble with a "bad" motor conversion would be that it needs to spin too fast to be useful.  More magnet fixes this.
For b) I refer to the rise in current produced with increasing speed.  If two generators cut-in at 200 RPM, but at 300 RPM one is making 5 Amps and the other is kicking out 20, then clearly the second one is much "stiffer" than the first.  The practical upshot of this is that the turbine doesn't run very fast throughout its power range, and at the high power range, it is more likely to stall the blades - a very effective and safe means of overspeed control.  Lastly c) is an effect I have been able to see in test data and numbers I can read in my datalogger - the generator just reaches a "plateau" and no more wind will make more power output.  I don't pretend to understand reactance limiting, but so far I've seen different generators behave differently and the key factor may be having lots of magnet packed on.

PS I have no idea how my previous message was date-stamped, but I certainly wasn't awake at 1AM last night! 

[BLUE]

Hi SparWeb, More questions answered. But, the blades stalling.

For b) I refer to the rise in current produced with increasing speed.  If two generators cut-in at 200 RPM, but at 300 RPM one is making 5 Amps and the other is kicking out 20, then clearly the second one is much "stiffer" than the first.  The practical upshot of this is that the turbine doesn't run very fast throughout its power range, and at the high power range, it is more likely to stall the blades - a very effective and safe means of overspeed control. 

Even if you had a stiff PMA with large diameter blades to turn it, it will reach a point and stall even in faster winds? Sounds like a good way to go. Like you said "overspeed control".

Blue 

[Flux]

Yes stall can be useful for overspeed control but it can also be a problem with reduced output if you take it to extreme.

The ideal situation is to cut in at a nice low wind speed and have the alternator load come on at such a rate that the prop speed follows the wind speed. This would imply that with a cut in of say 100 rpm at 7mph you would want to run at 200rpm at 14 mph and 300rpm at 21 mph.

If your generator is not very stiff you could get somewhere near this ( not ideal as the wind follows a cube law). This not very stiff generator is not very efficient and at 20mph may only be about 40% efficient so you will have less than the available prop output to charge your battery.

The next thought is to make the generator very efficient and that implies that it will be stiff. Now for cut in at 100 rpm you may get the full load power of your prop in a 20 mph wind at only 150rpm. You have now a very efficient alternator putting out most of the mechanical power you put into it but by so doing you have stalled the prop . The end result is that the prop power at 20mph will now be not much greater than at cut in. You have improved electrical efficiency but traded it for a much more serious loss in the prop power curve.

What would be nice is a less stiff alternator that gives you reasonable electrical and aerodynamic efficiency to work well up to maximum power but then it would be nice if the load then came on drastically with any further increase in speed to stall the prop and prevent overspeeding.

No normal single alternator has these characteristics. You can fiddle things to some extent by having a blade profile that goes abruptly into stall but this is specialised stuff and even then it normally only works with very stiff alternators with their characteristics set to loose much of the low wind speeds ( induction grid tie is a good example).

The larger axial machines built here do to a considerable extent rely on stall for speed limiting, with the help of furling in high winds but if they are chosen for best results at cut in they will be way off the peak power in the 25 mph wind range. For big machines the safety and quietness of operation may outweigh the loss in high end power but for anything under 10ft you will not get very good results if you fit an alternator that is very stiff just for speed control.

This is quite difficult for newcomers to follow but I hope this partly explains things.
Flux

[BLUE]

Flux
I see why axial machines are so popular. Able to adjust the air gap to find the sweet spot makes things much easer. For a radial machine (with a fixed air gap) I suppose adjusting pitch and size of the blades could accomplish the same thing.

Thanks, Blue

[SparWeb]

Flux
I see why axial machines are so popular. Able to adjust the air gap to find the sweet spot makes things much easer. For a radial machine (with a fixed air gap) I suppose adjusting pitch and size of the blades could accomplish the same thing.
Thanks, Blue

(If I can jump in once more) That's why I go on about the wire connections (Star/Delta/Jerry).  It allows me to do the much the same thing since I can't change the gap.  In fact, Axial-flux alternators have that capability too, provided you separate the phase end wires and bring them out of the stator. 

This all fits under the general topic of "matching the loads".  The prop has its own power curve, with some ability to run a bit faster and slower than its ideal speed with a penalty in efficiency.  The alternator has its own power curve - the input power it requires to be driven at various speeds (regardless of the output).  When these two power curves match then the system works at its peak mechanical efficiency.  It's actually very difficult to get this to happen, because these two curves have different shapes and are ruled by different forces.  But you can get close and making compromises allows you to optimize things for at least some range of good operation.  Well, Flux explains it better - in fact a few years ago he dedicated a thread to the subject.  Stuff you'll inevitably pick up as you go.

[wooferhound]

Here is plenty of information about doing Motor Conversions . . .
http://fieldlines.com/board/index.php/topic,143568.0.html

[BLUE]

SparWeb, jump in any time!!

Wooferhound, thanks for that link. Tons of information there.

Thanks guys!
Blue

[cdog]

In the very first pic of this thread there seems to be mounting tabs welded to the inside of the stator core.
How does this affect the shorting of the laminates?
I was wondering how large an effect this has on how the flux travels, it seems to defeat the purpose of scrounging for laminates instead of just using a piece of pipe for the stator?
Is it worth the extra effort to ensure the laminates aren't shorted to each other in any way, or is it not an issue?
Cdog.

[Flux]

You have to look carefully to see that those lugs are welded on the inside of a steel tube. The laminated core is the bit with the coils placed on it. It is in fact a motor core with the teeth removed from the inside to save space and weight.

You can weld the non active side of a core anyway. Many motor cores are welded on the outside, most of those not welded have wedge strips driven into slots on the outside which must short the core.

What you must not do is provide closed circuits in the eddy path, as long as the circuit is broken all is ok.

In that particular machine the toroidal core stack is slid over the steel tube and clamped between a pair of steel end cheeks which are welded to the steel tube. There is virtually no conduction through this path but in fact it wouldn't matter very much if I had cut some slots in the tube and welded the core through it. The thing not to do would be to cause any closed circuit on the outside face of the laminations where the flux is linking them.

I think people fail to understand the tiny voltages causing these eddy currents. If you punch a stack of rings and make sure the edges are free of burrs then you can stack them together with very little in the way of insulation between them. If completely bare you will only have contact at a few points on each, this is all tat is done with non critical laminated parts. For more critical things such as motor laminations the oxide formed during the annealing process is enough to do the job and steam is often used in the annealing furnace to form a better oxide layer.

Highly critical cores have the laminates coated with either a tough stoving varnish or some specialised coating. Just varnishing them with air drying varnish is probably no better than using nothing as the varnish will prick through at similar points with the clamping pressure.

In the old days paper was sometimes used between the laminates but things run too hot for that now. Although considered bad practice nowadays an old trick was just to let the things rust before assembly. It was perfectly effective electrically but is questionable for long term mechanical stability. I have to admit that the laminates used in that alternator were rescued from scrap and very rusty and that is the insulating material between them.

Far more critical than the insulation between them is the thickness of the actual laminates, things more than about 1/2mm thick are not suitable for critical ac cores but up to about 3mm is used in pole pieces of wound field machines where some form of laminating is desirable but the issue is not critical.

Hope this answers your question.

Flux

[cdog]

Thanks, I have not taken the 100hp motor apart yet, but it is likely 40 plus years old, and been outside for more than half that, likely there will be adequate rust! So there is no problem welding mounting tabs on the inside of the laminates then?
Is there a guideline to use as far as thickness of the lams left if I wanted to turn them to save weight and still have enough metal left to conduct the flux?
Cdog.