How thick should my core for the coil to be? - RC Groups
Jun 04, 2012, 03:32 AM
The Engineer
Discussion

# How thick should my core for the coil to be?

Hi there, i am referring to this old thread:
Designs on Axial Flux motors, curious about coil types.
https://www.rcgroups.com/forums/show....php?t=1326848

I am planning something on paper and not ready to start any mechanical work yet.

I am planning an axial layout gen in halbach arrangement basis. Well not as elaborate as Launchpoint's using 5 magnets to a pair of pole. Will start with simple 3 magnet to a NS pole which I believe is common layout.
So lets say we have space for 84 magnets, we get 42 poles.

If the winding is to get 3 phase (star or delta better?) for easy workability with say available ESC, and I need 150V output to produce 2.5kW, that's almost 21A. Add 15% buffer we derive 25A that is flowing thru the coil.

Say we work out the N as 480 turns, do I need to wind using a wire capable to take 25A? That's insane 8AWG! I haven't got to understand this part so I am asking this here.

Or should it be 25A/480 and the closest is AWG34 at 0.16mm: http://www.powerstream.com/Wire_Size.htm)

Thanks.
Last edited by prodriverex; Jun 07, 2012 at 08:35 AM.
 Jun 04, 2012, 09:58 AM The Engineer I think i derived the number of armature to be 46. i read around some describe their motor to be 12N10P. What does it mean? 12 coils, 10 poles? What design call for more poles than coils and vice versa? Here's a pic of the magnet layout..
 Jun 04, 2012, 10:10 PM The Engineer sorry the armature pole should be 45.. divisible by 3..
 Jun 12, 2012, 01:52 AM The Engineer Maybe no one understand my question? Maybe not? Anyway, too thin a wire will cost heat to build up if high voltage is pumped through. Need to go back to drawing board to calculate using less turns then with thicker wire. But then again, if each turn carries 0.3V, x480 turns then it sounds about right at 144V?
Jun 12, 2012, 06:54 PM
Jack
Quote:
 Originally Posted by prodriverex Maybe no one understand my question? Maybe not? Anyway, too thin a wire will cost heat to build up if high voltage is pumped through. Need to go back to drawing board to calculate using less turns then with thicker wire. But then again, if each turn carries 0.3V, x480 turns then it sounds about right at 144V?
12N10P means 12 Nuten (stator arms) and 10 Poles (magnets).

http://www.powerditto.de/bewicklungsrechner.html

Use it in German or English, as an example, enter Nuten 45 and poles 30 and it will show you your winding and magnets and how to wind them.

Look at the winding factor for that, it is 0.86603 which is pretty good. A perfect winding factor is 1.0 but that is impossible to attain, so the closer you are to 1.0 the better the winding factor is.

Look at the cogging steps too, the more cogging steps you have the more permutations the ESC has to make to run the motor.

I'm not a motor expert but you want the winding factor to be high and the cogging steps to not be too high or the ESC cannot keep up with it. As an example, the common 12N14P motor has 84 cogging steps per turn. LCM(12,14)
and its winding factor is: 0.93301. That is a good motor.

Jack
Last edited by jackerbes; Jun 12, 2012 at 07:06 PM.
 Jun 13, 2012, 08:38 AM Suspended Account The individule wire wound on each phase never see's the total current that we see when measuring the amp consumption with a watt meter between the ESC and the battery
Jun 13, 2012, 11:01 AM
The Engineer
Quote:
 Originally Posted by jackerbes 12N10P means 12 Nuten (stator arms) and 10 Poles (magnets). Go to this link: http://www.powerditto.de/bewicklungsrechner.html Use it in German or English, as an example, enter Nuten 45 and poles 30 and it will show you your winding and magnets and how to wind them. Look at the winding factor for that, it is 0.86603 which is pretty good. A perfect winding factor is 1.0 but that is impossible to attain, so the closer you are to 1.0 the better the winding factor is. Look at the cogging steps too, the more cogging steps you have the more permutations the ESC has to make to run the motor. I'm not a motor expert but you want the winding factor to be high and the cogging steps to not be too high or the ESC cannot keep up with it. As an example, the common 12N14P motor has 84 cogging steps per turn. LCM(12,14) and its winding factor is: 0.93301. That is a good motor. Jack
Thanks Jack. I derived 42 poles mainly because limited by magnet size and trying to cover up any gap in between magnets, without resorting to customize fan-shape.
Was studying how to get efficient number of stators and see if we could employ LRK winding.. thanks for the link!

Now i tried 39,42 and the result for winding factor comes out to 0.93301, as you said - good! But cogging is 252.. hmm.. is that good or bad? Will LRK winding helps?
Jun 13, 2012, 11:39 AM
The Engineer
Quote:
 Originally Posted by zeroback The individule wire wound on each phase never see's the total current that we see when measuring the amp consumption with a watt meter between the ESC and the battery
So glad to know that now. But hand winding 480 turns, OMG.. this is only prototype.. unfortunately.. no automation.

Did I get it right from the start?
Aim: A generator to push out 1.5kW continously from rpm of at least 4000rpm and peak at 5500rpm for 2.5kW. Voltage preferred, 120 to 150V..

So I use Faraday's Law of Induction..
E=-N[(B.A/t)]

The magnet is rated at 12,500 Gauss or 1.25T but we derated it to 1T.
Magnet area is 1"x0.188", converted to sq-m.
t=0.01091s

N is 13,491. Since there are 42 poles so N should be 321? (Sorry not 480, i took 28 poles previously in my other calculation..)

Is that the right start?
Jun 13, 2012, 12:46 PM
Jack
Quote:
 Originally Posted by prodriverex Thanks Jack. I derived 42 poles mainly because limited by magnet size and trying to cover up any gap in between magnets, without resorting to customize fan-shape. Was studying how to get efficient number of stators and see if we could employ LRK winding.. thanks for the link! Now i tried 39,42 and the result for winding factor comes out to 0.93301, as you said - good! But cogging is 252.. hmm.. is that good or bad? Will LRK winding helps?
I only know bits and pieces and only for a few motors. Don't know how this works for generators. But know that you have said that, maybe someone can help with that.

On the LRK question, you would wind every other pole so you need an even number of poles that is also divisible by 3. If you take the 42N28P and change it to 21N28P so it would like a 42 Nuten motor with only 21 Nutens wound it gives you the same ABC wind and the same numbers for the cogging steps and winding factor.

The LRK motor is the one that has the traits of having the best torque while, at the same time, being a little more efficient with the input power drawn from a battery. But again that is motors, not generators...

So I guess we need to await the arrival of the real motor experts!

Jack
 Jun 13, 2012, 10:42 PM The Engineer Hi Jack, I should have study harder when i was in polytechnic and go for a graduate degree.. now i have to go back to basic.. and basic tells me that motor and generator almost similar .. but uses right hand fleming's rule.. current is opposite direction.. My project focus on the generator essential.. I will pass the filtering and voltage linearing to my partner who's more than qualified than me. I will try out the LRK formula.. But, pardon my ignorance.. the real reason to alternate the tooth's winding? Is there any real effect from the laminated iron which is unwound after all when passing thru the magnets? Looking at http://www.southernsoaringclub.org.z...-motors-3.html working principle being explained, it still did not explain why skipping alternate tooth.