Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Jun 04, 2012, 02:32 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
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.
http://www.rcgroups.com/forums/showthread.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.
prodriverex is offline Find More Posts by prodriverex
Last edited by prodriverex; Jun 07, 2012 at 07:35 AM.
Reply With Quote
Sign up now
to remove ads between posts
Old Jun 04, 2012, 08:58 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
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..
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 04, 2012, 09:10 PM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
sorry the armature pole should be 45.. divisible by 3..
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 12, 2012, 12:52 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
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?
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 12, 2012, 05:54 PM
Jack
USA, ME, Ellsworth
Joined May 2008
16,699 Posts
Quote:
Originally Posted by prodriverex View Post
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).

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
jackerbes is offline Find More Posts by jackerbes
Last edited by jackerbes; Jun 12, 2012 at 06:06 PM.
Reply With Quote
Old Jun 13, 2012, 07:38 AM
Suspended Account
United States, FL, Pompano Beach
Joined Oct 2011
1,425 Posts
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
zeroback is offline Find More Posts by zeroback
Reply With Quote  (Disabled)
Old Jun 13, 2012, 10:01 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
Quote:
Originally Posted by jackerbes View Post
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?
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 13, 2012, 10:39 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
Quote:
Originally Posted by zeroback View Post
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?
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 13, 2012, 11:46 AM
Jack
USA, ME, Ellsworth
Joined May 2008
16,699 Posts
Quote:
Originally Posted by prodriverex View Post
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
jackerbes is offline Find More Posts by jackerbes
Reply With Quote
Old Jun 13, 2012, 09:42 PM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
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.
prodriverex is offline Find More Posts by prodriverex
Reply With Quote
Old Jun 14, 2012, 09:00 AM
Jack
USA, ME, Ellsworth
Joined May 2008
16,699 Posts
You already know a lot more about motor theory than I do. And more than I want to know, I am more just the guy that just wants to use them more than to fully understand them.

The LRK wind has proven to be one of the best for some uses but I don't know if the presence of the arms that are not wound contributes to that or not.

From the viewpoint of winding, we usually start with the primary goal of attaining a Kv that is best for the prop we want to use so that leads us to select a wind type, turn count, and termination method. Then we want to get the turns on with the largest possible wire to have the best current capacity.

As we wind, the limiting factor for wire size on DLRK and ABC winds is that the winds on two adjacent arms will crowd each other. The LRK wind eliminates the crowding and gives you more room on the arms that are wound for larger wire. So with LRK you have a motor with the shortest length and largest diameter wire as compared to the other winds.

Here is a thread you may enjoy reading:

http://www.rcgroups.com/forums/showthread.php?t=1356379

In post #6 Christian Lucas (he is the Lucas of LRK fame) gives his general rules for when to use the LRK and DLRK winds. He says:

"..Best 12 slot and 14 P with LRK winding, always and yet and in the future.

Only if you have a very short stator use dLRK winding.

ABC winding only with 8P and 16P on 12N stator, LRK and dLRK only with 10P and 14P for best result.

To get high rpm/volt use 8P or 10P and for high torque use 14P or 16P with the correct winding..."

I think that is probably very good advice. His remark about "short stators" refers to the radial length of the arms, not the height. So I think your stator in the image you posted would be a stator with shorter arms.

Other words of wisdom I have read here and put in my notes are that:

"..The ideal Magnet Coverage for LRK motors is between 70% and 75%. 79% will give you more power but efficiency will start to deteriorate. 56% will not give you the power you want...."

On the magnet coverage for DLRK and ABC winds, I've read that those can use a much higher percentages of coverage, like even up into the 90% or 95% or so range.

So there are a lot of variables and considerations and also many tools and calculators that that we use to give us things to try. Those give us things to consider more than clearly defined "best answers" to our questions. And your question brings in the additional consideration in that your interest is more in generating power than in using it.

The Bewicklungsrechner will not, as far as I know, offer you a LRK wind in response to you entering a N and P count. But in the Advanced mode you can enter a LRK winding scheme (like one copied from the Kombinationstabelle ) to get the cogging and winding factor info for the LRK winds. But the table does not list info for the higher arm and pole counts like you have.

Go to the Advanced mode and enter the A-b-C-a-B-c LRK winding scheme for a 12N14P motor and you'll see a result of 84 cogging steps per turn and a winding factor of 0.96593. I think that is the highest WF of any and all winds and it is possibly also what makes the LRK wind such a good choice.

I read Brian Mulders motor tutorial some time in the past and it overwhelmed because it was too complex for me. I need to read that again now as I will understand it better. So much good stuff, so little time...

Jack
jackerbes is offline Find More Posts by jackerbes
Reply With Quote
Old Jun 15, 2012, 05:57 AM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
Thanks for sharing and advise.
I spent about a day each reading BM from part 1 to 5.. It gets heavier and really straining.. I have to print it out and read in depth..

The picture is the plan view layout of the magnets along the axial. the stator arms will be on top of it.
As for the stator arm design, I am exploring the following option:
1. To use LRK/dLRK winding on soft iron that will look like (side view)
|[] | //////////// |[]|
|[] |-----------|[]|
|[] |///////////// |[]|
|--------|--------|
|--------|--------|
|--------|--------|
|--------|--------|
|--------|--------|
===============

where ==== is the static rotor;
[] on the left and right is the dual halbach array arrangement. dual so it line up on both sides, and this is the side view;
| /////////// |
|-----------| is the winding on the soft iron laminate (well still no idea how to do it yet)
|/////////// |

When you said the radial length of the stator arm, are you referring to the
| ┐
| } length or width of this?
| ┘

2. Wave Winding w/o iron..


Well that will probably save lotsa weight without the soft iron and keep the axial gen very flat like a pancake. But how much sacrifice on the power output remain to be seen (or calculated).

Maybe even can be wound as if to be like LRK if we can get the dimension accurately correct. Is that even doable?
prodriverex is offline Find More Posts by prodriverex
Last edited by prodriverex; Jun 15, 2012 at 06:05 AM.
Reply With Quote
Old Jun 15, 2012, 06:22 AM
Jack
USA, ME, Ellsworth
Joined May 2008
16,699 Posts
Looks like you have a plan, you asked:

"..When you said the radial length of the stator arm, are you referring to the
| ┐
| } length or width of this?
| ┘..."

Lucas's statement left me confused too so I asked about in (maybe in another thread) and when I say axial I mean how far out the stator arms extend from the center of the rotor, perpendicular to the shaft it you will. So basically the more length you have for turns the better it is for the LRK motors.

I think of the dimensions of a stator arm as being length perpendicular to or away from the shaft and height as being top to bottom of the stator parallel to the shaft. More length gives room for more turns around, more height gives room for turns with a longer length of wire in each coil. I'd love to see one of the experts chime in and say if I have that right...

LRK is sort of strange but really needs to be tried if you are curious. It looks like half of the wire is missing and many think of it as inferior to a dLRK wind because it is older and not as common. But it is not inferior and is actually superior in some instances.

I have never been unhappy with the ones I have done. And some of those did not have the recommended and lower percentages (75% or so) of magnet coverage because they were originally cheap dLRK motors. I think of the LRK as a good choice when you like to fly more on torque and mid-throttle power (thereby getting better flight duration out of a battery). If you are looking for speed and high RPM it is not as good a choice (as Christian Lucas clearly said).

Maybe some day I'll get around to optimizing the magnets for a LRK motor but I have not done that yet. I have not moved or removed any magnets for that matter, just worked with the original configurations.

Jack
jackerbes is offline Find More Posts by jackerbes
Last edited by jackerbes; Jun 15, 2012 at 06:31 AM.
Reply With Quote
Old Jun 15, 2012, 09:27 PM
The Engineer
prodriverex's Avatar
Singapore
Joined Dec 2009
610 Posts
I spent the night punching thru some possible permutations


So most likely permutation 3 will save lotsa weight..
while 15 will have 6 less cogging if that is any diff at 4000 to 5500rpm which is not very high right?
But interestingly 27, raised the WF to 0.93301. If I am correct, higher WF will result in better V-rms throughput?

The winding was: AabBCcaABbcCAabBCcaABbcCAabBCcaABbcCAabBCcaABbcC
And for LRK, it read this? A-b-C-a-B-c-A-b-C-a-B-c-A-b-C-a-B-c-A-b-C-a-B-c-

And the result: Cogging remain at 240, but the WF increased to 0.96593!
prodriverex is offline Find More Posts by prodriverex
Last edited by prodriverex; Jun 15, 2012 at 09:59 PM.
Reply With Quote
Old Jun 16, 2012, 06:40 AM
Jack
USA, ME, Ellsworth
Joined May 2008
16,699 Posts
I think you're right about the effect of the higher winding factor. It really strains the limits of my understanding of motors when the discussion gets to things like that. That is when I just want to try it and see if it works better.

If you have not seen it, the winding factor is discussed here and that is the first explanation of it that I could find:

http://en.wikipedia.org/wiki/Winding_factor

The is a link on that wikik page to this "winding editor" in Sweden:

http://www.eme.ee.kth.se/emetor/emetor.php

And that is interesting but it *really* makes my head hurt!

Jack
jackerbes is offline Find More Posts by jackerbes
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion How thick are wings meant to be ???? vengeance111 Electric Plane Talk 15 May 15, 2012 10:29 AM
Question How tight should my wheels be? Taildragger Zoidman Beginner Training Area (Fuel) 5 Nov 21, 2011 01:15 PM
Help! How thick should be a wing profile? jorgecameo Foamies (Scratchbuilt) 7 May 15, 2009 01:17 PM
How thick should a trailing be? Vito Sailplane Talk 22 Dec 08, 2004 10:11 AM
How thick should my copper braid wire be?? ferradas Electric Plane Talk 1 Feb 22, 2002 07:43 AM