84 turns of 3 in hand AWG 20 2sq inch N50 Magnets

[s4w2099]

Hello

I have in mind making a mill with stuff that I have around. I have 7 pounds of 20gage, 24 N50 2x1x0.5 magnets on two 11" disks, and a 9' fiberglass prop with the following specs: 1:10 TSR no load, 1:8 with some load, 1:5 heavy loaded. 12/9. To be charging a 12V system

I made a test coil with 84 turns and the resistance is about 0.38 Ohms, the output is about 2.4Vac at about 100 RPM. This thing will cut at about 90RPM for 12V.

The way I see it is that it would make a very nice 24V charger with the stator does not burn out. I want to charge 12V at very low RPM.

With this data I know for sure that it would stall badly charging 12V but if I use PWM to limit the output and make it easier for the prop to turn I think that it should be a nice charger regardless.

I am not looking to charge faster than 5-10 amps. All that I am looking for is to make a mill that will charge the 12V bank without going over 250RPM.

[willib]

i would think about using two or three strands of #20 wire , you can cut the resistance in half with 85 turns of two strands , and still have plenty of room , i think?

i would go for three myself.

[willib]

crap !

i should have read the TITLE a little closer..

Three stramds of #20 wire , and they are still 0.38 Ohms?

how are you measuring them?

[s4w2099]

Ohm meter from Rata Shack. The total area of 3 #20 is almost the same as a #15. But the advantage is that its way easier to wind. 3 strands fit nicely.

[willib]

use your pwm , set it up to draw the current through the coil ( use low current) and measure the voltage across the coil .

then use ohms law to calculate the resistance.

[willib]

or just use a AA battery, and accurately measure the current through the coil and battery with one meter , and the voltage across the coil with a voltmeter.

[s4w2099]

ok, using that method it comes up to 0.19ohms

[s4w2099]

Thanx man, I guess that the meter cant read such low values. I used a little dashboard incandecent bulb it draws 80.9mA and the drop accorss the coil is 15.1mA.

R=15.1/80.9

R=0.186650=0.19ohms.


Thats a big change from my previous reading with the meter.

:-O

[DanB]

Hi -

"I have in mind making a mill with stuff that I have around. I have 7 pounds of 20gage, 24 N50 2x1x0.5 magnets on two 11" disks, and a 9' fiberglass prop with the following specs: 1:10 TSR no load, 1:8 with some load, 1:5 heavy loaded. 12/9. To be charging a 12V system"

Yes - I've seen those blades advertised and I believe I have a set of them.  Those figures are probably not too far off - although I honestly think they pull those numbers out of the air (along with the claimed Cp of 48%)

"I made a test coil with 84 turns and the resistance is about 0.38 Ohms, the output is about 2.4Vac at about 100 RPM. This thing will cut at about 90RPM for 12V."

If your figure of 2.4VAC at 100 rpm is accurate, then I expect youll have 12VDC cutin at more like 75 rpm.

That's way too slow for a 9' machine with normal blades on it - it will be stalling in very low winds and Im afraid that if you got it past stall it would burn out in higher winds (bit with such a low cutin I doubt it'll ever get out of stall in the first place).  For 12 Volts, you should have much lower resistance and lots fewer turns if you're planning on a Star connection.  For a star connection I expect you would want something more like 25 - 30 turns per coil and you would have to wind with 9 or 10 strands of that 20 gage wire I think.  A reasonable alternative might be to use maybe 3 strands and about 75 turns or so and rectify each phase seperately (Jerry rigged - this would be like delta but you'd not be connecting the phases)

"The way I see it is that it would make a very nice 24V charger with the stator does not burn out. I want to charge 12V at very low RPM."

With this data I know for sure that it would stall badly charging 12V but if I use PWM to limit the output and make it easier for the prop to turn I think that it should be a nice charger regardless."

I don't think so  - it will be trying to produce power in winds that are much too low.  It will also have very high resistance for a 9' blade. If you do get it out of stall at startup then I think it will run away because the alternator will not hold the speed down with such high resistance.

"I am not looking to charge faster than 5-10 amps. All that I am looking for is to make a mill that will charge the 12V bank without going over 250RPM."

I guess it could be an interesting experiment but it seems to me like you have the materials and resources there to get a lot more power than that.  Im afraid that if it does actually produce 5 - 10 amps then it might run a way because the resistance is so high.  I think with those N50 grade magnets you could have a really powerful 9' machine that would hold up well and produce quite a bit of power in low and high winds.  Your blades are designed for that - I don't understand why you'd want to hold them back so slow and get so little power here.

[Flux]

I am with Dan on this.

If you really know what you are doing and by pwm you mean a true buck converter with inductor you can do as you propose, but even then I would start with nearly 1/3 of those turns to make the converter easier.

You can do it as you propose with just pwm, but not with fixed pwm and once you start messing about that way then you might as well go the whole way to a true mppt converter and make something spectacular instead of highly indifferent as you are heading at the moment.

If you think you have the skill to do it properly, carry on and let us know how it works ( get it right and it will be spectacular). Otherwise stop now and save yourself much trouble.

With that winding you will be drastically stalled and you may not even get your 6A.

You will need to make it furl early otherwise if it does get through stall in a gale it will match the prop perfectly at about 30 mph, produce about 1kW and burn out in minutes.

It's not too difficult to wind 4 in hand. I would do that with about 60 turns, bring down the 6 leads and Jerry connect it. If you want to keep star, split the winding into 2 sections ( again 6 leads) and use 2 rectifiers in parallel.

I don't really agree that a fixed blade prop can have more than one value of tsr, but I know what you mean.

Flux

[tecker]

With around 100 ft of 20 per phase you'll be maybe 70 ft longer than machine amperage 10 to 11 amps) and maybe good for 6 amps and will get hot  at  continuous three phase.

 You'll need to wind 3 or 4 strands  to be assured of over speed protection . Voltage wise 100 rpm won't have the inertia to maintain charging I think with 9' prop will want to go faster than 100 rpm as a rule and you'll want to hold it down rather than let it go.

[s4w2099]

So basically, it is not very possible to get a nice charging current in the 100-250RPM range with a 9' prop?

What if I add another strand of 20, keep the same number of turns and jerry it?

To be sure, what would be a nice cutin for a 9' prop?

But do we agree on something? will this machine be good for 24V charging? I do no have much preference since I just started my system. I could just wire my batteries for 24volts.

[Flux]

100 rpm is too slow, in fact the figures you gave would give you a cut in of nearer 75 rpm as Dan said. That is so drastically slow for that prop that it just wouldn't produce anything.

If you want to keep it slow for some reason I don't understand then no slower than 140 rpm cut in. Lower cut in will reduce your power out even in low winds.

To keep below your 250 rpm top limit you either need to make a very powerful and oversize alternator or somehow get it to furl desperately early.

If you keep to star you will need about 6 in hand at about 40 turns. If you use about 65 turns and Jerry connect with 3 coils in series you will be similar. You may get 4 strands in.  If you are prepared to put up with 9 rectifier bridges then you have some other options. You could wind each coil with about 120 turns and connect each to a rectifier.

The other option is to use 10 coils, 5 bridges and use Hugh's 5 phase. With your constraint on wire that would make a lot of sense, buy the book and you have all the details. You will still need at least 2 strands in hand and probably you can manage 3.

Lots of choice to make a decent machine, don't go down the silly slow route and spoil the whole thing.

Flux

[s4w2099]

One more thing that I forgot. Given the same number of turns what advantages would jerry phased give me over delta? (same number of turns)

[Flux]

Cut in speed will be the same. Below cut in there will be circulating harmonic losses with delta, which may make little difference but if you are looking for low wind performance these losses will be against you. Beyond cut in delta may again be more lossy. At the high wind end Jerry connection may run cooler for the same output.

Delta works but with rectifier loads it leaves something to be desired.

Flux

[s4w2099]

Ok, I will try them all. I can test the 9 coils without casting them. Then I can remove turns to get the maximium output.

Thanx for the info.

:-D

[Flux]

Sorry I don't follow this. Do you mean that you are going to try them under wind conditions without casting them?

You realise that this is a propeller problem, you will not come to any useful conclusions by testing volts out when turning by other means. There is no problem getting the output from the alternator at low speed, the problem is that the stalled prop will not drive it.

flux

[s4w2099]

Yeah I do understand that. I will make them 60 turns and jerry them with an extra strand of wire. The amount of amps is far greater that way.

I dont quite understand how the current will add up at the end because we have 3 phases in parallel but they will not be in phase with each other.

[s4w2099]

sorry, what I really meant is that I can test the 9 coils in a test tower with the prop on and keep removing turns until the max output is generated with my wind conditions.

This could be tedious I know. But I do understand your point and really appreciate your advice. I am going as you said to start of, 60 turns per coil, jerry. Lets see how it goes.

What cutin speed should be optimal for this blade?

Thanx

[Flux]

With it Jerry connected any phase that has a peak voltage above the battery will contribute. Most of the time that means that two phases will be active, when highly loaded there is the possibility that when one is at peak, the other 2 may still be high enough to add something. That is why Jerry connection can run cooler than delta on high load.

With delta connection ( or even star), a 3 phase rectifier without inductance can't run with overlap, at any instant the current is carried by one pair of ac wires.

Flux

[Flux]

You needn't go to that much trouble, if you are near with the turns you can increase the cut in speed by increasing the air gap.

If your prop does as it is claimed then for 9ft, cut in about 140 is as slow as I would think you could go.

With N50 magnets you should hit that with about 35 turns star. For Jerry you would need to multiply that by 1.7 so the 60 turns should be a safe starting point, you can widen the gap if it still stalls.

If you can get 4 strands in at 60 turns then you will have a very decent machine that will do justice to those magnets. If you absolutely must keep to your 250 rpm top speed then you will need to furl at about 400W, but even if you let it go up to about 700W you will not exceed about 350 rpm. If those blades are reasonable that should be quiet.

Flux

[tecker]

The JR connection can give you the  amperage you need as you parallel allowing you to use smaller wire and slightly smaller rectifiers . The small wire as 75 -90 turn coils  top out and bump the rotor around unless your dumping then it heats up . So as long as you can keep the voltage up to match the other coils (single coil out ) or phases (single phase out ) the wire is somewhat protected against over speed.  

[DanB]

"If you keep to star you will need about 6 in hand at about 40 turns. If you use about 65 turns and Jerry connect with 3 coils in series you will be similar. You may get 4 strands in.  If you are prepared to put up with 9 rectifier bridges then you have some other options. You could wind each coil with about 120 turns and connect each to a rectifier."

I would double check your figures with your test coil so that you're sure.

Flux - I'm not sure I agree with 40 turns there if he's shooting for 12V cutin at 140 rpm.  I would think more like 30 turns or maybe even a couple less with N50 grade magnets.

[Flux]

Yes Dan I agree with you.

He seemed hell bent to run it too slow and that was the absolute limit I thought he could start from. When he saw his mistake he could open the air gap.

He finally seems to have settled for 60 turns Jerry connected which is nearer.

These fellows never give any information on air gap or other details so it is always much of a guess. A few turns too many can be dealt with, too few would seem like a disaster but in my experience too few is usually better but I can't convince anyone.

It was hard going getting it down to something practical, at least we are within range now.

Flux

[s4w2099]

Hehe, I will take your advice you guys have much more experience than me. One thing that I could do is to upgrade the blade size of about 13ft for a slower alternator? what do you think about that one?

[s4w2099]

I have hooked up a 1.2 NiCd to a 24 volt alternator that I made for the same blade size. As you said even the 9' stalled badly as expected. Note this was only a test. the current started at about 5 amps with about 120 RPM. Then the RPMs and the current started to go down rapidly.

Even when wind picked up the charging current did not passed 200mA. and thats how many watts? 0.24Watts.

I just wanted to know if using PWM to make it easier for the blades to spin would make any difference.

By the way, the air gap in the new machine I think would be about 3/4 - 1" just for safety my test coil was 1/2 thick just as the magnet's thickness. Do you guys think I should increase teh field by making the airgap smaller and use less turns to lower the resistance? That would mean that I would need more strands of wire even for jerry. I know for certain that it would make a better machine though.

:-S

[Flux]

Even N50 is not going to hold a 13ft rotor with only 12 poles.

You are much nearer the region for 9ft. If you want the big blades then think 16 poles or bigger magnets.

Total stator thickness of 5/8 including glass covering suits 1" magnets fairly well. You should be able to work in about 7/8 gap. For 12v star connected I would go with Dan's 30 turns for general use, but your prop is a bit of an unknown.

For Jerry connection that would bring you up to about 52 turns. It would make sense at this point to set the air gap to about 7/8" and check a test coil with about 52 turns.

You need 14 v peak at cut in speed ( say 140 rpm) that is 10v rms for the 3 coils in series so each coil should give about 3.4v.

You will struggle much more if you try for 16 magnets with that #20 wire, it's not a very convenient starting point unless you want multiple rectifiers.

Yes your pwm idea could overcome the stall with a winding with too many turns. As I said before simple fixed ratio pwm is not going to do much other than run you into a very difficult learning programme.

If you take the thing to its logical conclusion and use high frequency pwm with an inductor you have a buck converter. If you can track the pwm of this to match the load then you have the holy grail of the mppt tracker.

You have to ask yourself whether you can do this. If you can then you can more or less choose any cut in speed for your alternator within reason. Bear in mind that if you choose a winding that is too slow and your converter fails you have a useless alternator. If you build it conventionally there is still no reason why you can't use your converter to match the load over the working range and see a lot of benefit, but you must abandon your desire for a top speed of 250 rpm, the correct load matching and 250 rpm limit on a 9ft rotor are just not compatible.

Flux

[s4w2099]

1" magnets

My magnets are 2" length, 1" width and 0.5 thick. Or are you talking about an alternator for the 13ft rotor?

For this machine (9') I am planning to use from 3/4 to 1" airgap. What do you think about that one?

[Flux]

Sorry, should have said 1/2" magnets, I was working on total magnet length.

Yes I was assuming you had 12 off 2 x 1 x 1/2 N50 per disc for the 9ft. You need at least 16 for the 13 ft.

3/4 to 1" gap should be fine, As I said that should take about 5/8 total stator thickness. Coils just over 1/2" then glass and resin.

Flux

[s4w2099]

Dont you think there would be any advantage by reducing the airgap, making the coils thiner with fewer turns and maybe 5 in hand if it fits?

[Flux]

Smaller gaps reduce the available winding space, you need safe mechanical gaps and with thin stators this % becomes significant. Even if you reduced the air gap to near zero you will still need half the turns.

As long as as you don't go ridiculously thin you can make the thickness what you fancy.

The optimum is coils 1/2 to 5/8 thick but down to 3/8 should be reasonable.

You ultimately have to make a decision, we could go on for ever. Thick coils , thin coils, many strands in hand, many circuits in parallel, star, delta,Jerry ,  5 phase. All possible.

If you want to stay conventional stick to the usual cut in speeds or you can try your pwm and let us know how it works.

Flux                    

[solarengineer]

Hey Flux, can i ask you about your comment

"You could wind each coil with about 120 turns and connect each to a rectifier."

if you rectify each coil seperatly would you then series each of 3 rectifiers to connect it all up?

and does this method help with stator heat?

Thanks

Jamie

[s4w2099]

Ok. the mill is up for testing. It is easily producing 80Watts even though the direction of the wind is not favorable. The wind is coming from the south and a wall of trees is in the way. Add to that that the mill is only 10ft of the ground. Another problem is that my rectifiers drop 1.7Volts according to the datasheet and I am charging 12V. Right now it does not have a tail so I have to point it towards the wind.

I hope I can see winds from the north. From that direction I have no obstacles near by. When I dont have a lot of wind I can get rid of the jerry connection and use star to charge the battery and it can still gives me from about 1-1.5 amps.

The blades are spinning very very controlled by the alternator. Its would not let them over speed at all. If shorted the blades will HALT in a second. If shorted across the ampmeter it gets out of range so I will not do that ever again.

I do not know much about furling but I have some ideas on how to make a nice furling system for this mill.