Multitooth motor winding idea

I have this idea...

All you really smart folks out there this is your big chance to blind us with your intellect.

Please feel free to point out the flaws in my logic or where I flat out missed things, made mistakes or you know something I don't. That's the whole idea.

1. Make a custom tooth count stator. So that there is enough teeth to wind 2 or 3 adjacent teeth in the same direction and have 3 or 6 phases.

2. I want to remove the factory magnets. They are typically the same width as a single stator tooth. I want to replace them with magnets that are as wide as 2 or 3 stator teeth. That depends on the stator tooth count and if you are winding 2 or 3 teeth.

3. Start from an empty stator.

4. Depending on the stator tooth count, adjacent 2 or 3 stator teeth would be wound in the same direction. This effectively makes them into a single stator tooth.

5. Typically you wind 3 adjacent stator teeth forward/reverse/forward. I want to wind 2 or 3 adjacent teeth all forward, then 2 or 3 teeth all reverse and then 2 or 3 teeth all forward again. I also want to have 3 or 6 phases.

6. There is an implication which to my limited experience sounds absolutely fantastic. KV is based on turns per tooth. Since you are winding wire in the same direction around 2 or 3 teeth, you've effectively dropped Kv by 2 or 3X (I think). Well that's great if you want a really slow motor such as a hub, but this in an outrunner and I want something like 4000 RPMs. So then put 1/3 (ish) as many turns on the 3 adjacent teeth to get back to the original Kv. Now for the cool part. Turns of wire takes up space on the stator teeth. Since each tooth now has 1/3 as many turns on it, it's about 66% less full. Use 66% more copper per turn. This means each phase is now 66% more capable of handling current. A 10kw motor ideally becomes a 30kw motor. Right?

7. The kv is going to be higher...probably 2X or 3X higher depending on if you wind 2 or 3 adjacent teeth the same direction. This is because turns per tooth define Kv and you've reduced the number of teeth quite a lot. This sounds like it means the opposite of number 6. I frankly don't understand the implications in this area yet.

8. If there's 6 phases, then there's now half as many teeth per motor on top of the 2 or 3 tooth adjacent winds. I think this too will increase Kv by 2X.

9. Pole counts will go down significantly so eRPM constraints will matter less despite the higher Kv...hopefully.

10. Adjacently wound teeth in the same direction and lower pole counts also means fewer transitions per rotation and therefore less iron losses.

Conclusion: Wont this make for a motor that has 2-3X more current handling, probably has a somewhat higher Kv and has a much lower overall stator and magnet count so fewer transitions per rotation which reduces iron losses.

Like I said earlier tear it apart, point out the logic flaws or missteps I made, add additional ideas, corrections and so on. I'm certainly no expert on this. I've wound 2 outrunners. I just found out about 6 phase motors a few weeks ago. What do I know? LOL!

More thought on Kv and multi tooth winds.

Assume I am winding a 6 phase motor that at 3 phases was 50Kv.
Assume 10 turns per tooth = 50 Kv on the original motor.
Assume that the stator teeth are 100% full of copper.
Assume each original phase can handle 100 amps.

1. With half as many stator teeth per sets of 3 phases, that's going to double the Kv or 100kv at 10 turns per tooth.

2. If 2 adjacent teeth are wound the same direction with the same number of turns (10 each) as done originally, due to number 1, Kv should remain at 100. See number 4 in my original post. IE: The turns per "effective" stator tooth are now 20.

3. To get back to 50Kv per sets of 3 phases would require 20 turns per adjacent same direction wound stator teeth.

4. To get more current handling, means fewer turns per adjacent same direction wound teeth to make room for more copper per turn.

5. Split the difference between higher Kv and more current handling. Go for 125kv and have 25% more space per tooth for more copper or 125 amps per phase and there are now 6 phases so that really means the motor doubled it's wattage.

Again...pure speculation...like I really know what I'm talking about. Find the flaws, make corrections and so on. This is going to hopefully go from theoretical to an actual motor wound this way.

ElectricGod said:

I have this idea...

All you really smart folks out there this is your big chance to blind us with your intellect.

Please feel free to point out the flaws in my logic or where I flat out missed things, made mistakes or you know something I don't. That's the whole idea.

1. Make a 26, 32 or 52 tooth stator. 26 teeth will work for 3 phases and 3 adjacent same direction wound teeth. 32 or 52 teeth allows for 6 phases. 32 teeth allows 2 same direction adjacent teeth and 52 teeth allows 3 same direction adjacent teeth.

Click to expand...

No blinding flash of brilliance here but you should know that a three phase motor must have a tooth count that is divisable by three so at the first hurdle your plan fails. Secondly the multitude of different winding options for 3 phase electric motors have been studied and analysed to death so you might want to do some research. I have quite a few IEEE papers on the subject floating about, I'll see if I can put my hands on a few and post for you.

kiwifiat said:

ElectricGod said:

I have this idea...

All you really smart folks out there this is your big chance to blind us with your intellect.

Please feel free to point out the flaws in my logic or where I flat out missed things, made mistakes or you know something I don't. That's the whole idea.

1. Make a 26, 32 or 52 tooth stator. 26 teeth will work for 3 phases and 3 adjacent same direction wound teeth. 32 or 52 teeth allows for 6 phases. 32 teeth allows 2 same direction adjacent teeth and 52 teeth allows 3 same direction adjacent teeth.

Click to expand...

No blinding flash of brilliance here but you should know that a three phase motor must have a tooth count that is divisable by three so at the first hurdle your plan fails. Secondly the multitude of different winding options for 3 phase electric motors have been studied and analysed to death so you might want to do some research. I have quite a few IEEE papers on the subject floating about, I'll see if I can put my hands on a few and post for you.

Click to expand...

In context of everything else in this thread, I meant to list tooth counts so that in groups of 2 or 3 wound as a single tooth you would still have 3 phases or 6 phases. That's stated multiple times. Look for logical errors or conceptual errors revolving around Kv, current handling, winding and related details. I'm familiar with the various ways that a 3 phase motor is typically wound. This is a bit different. I want to wind 2 or 3 teeth all in the same direction as if they are a single tooth and then explore the implications of this. LRK vs dLRK or whatever is not relevant. Same for WYE vs Delta. Not relevant.

So looking at what you said again...lol...all my tooth counts are way wrong. I made number 1 open for discussion by NOT listing actual tooth counts. I looked at them several times and flat missed that. There's probably a few more obvious things here that I missed too. I only read this 5 times and found plenty of other things to correct before posting.

number 5 was also wonko..fixed.

Just a brain fart there. LOL

ElectricGod said:

6. There is an implication which to my limited experience sounds absolutely fantastic. KV is based on turns per tooth. Since you are winding wire in the same direction around 2 or 3 teeth, you've effectively dropped Kv by 2 or 3X (I think). Well that's great if you want a really slow motor such as a hub, but this in an outrunner and I want something like 4000 RPMs. So then put 1/3 (ish) as many turns on the 3 adjacent teeth to get back to the original Kv. Now for the cool part. Turns of wire takes up space on the stator teeth. Since each tooth now has 1/3 as many turns on it, it's about 66% less full. Use 66% more copper per turn. This means each phase is now 66% more capable of handling current. A 10kw motor ideally becomes a 30kw motor. Right?

Click to expand...

If I understand what you're describing, you'll have more copper on the ends of the teeth rather than thru them, than if it were wound normally. As I understand it, only the copper that's between the teeth is "active" in moving the motor--the rest is wasted in "end turns".

So you end up with a motor that has less torque than the same one wound normally.

I don't know how that affects it's speed or power handling capability, but if more of the power is "wasted" at the ends (end-turn losses) then I expect it'll actually only have the same basic power handling as the other method, as far as effective power output is concerned.


I'm not a motor expert by any means, so I could be wrong in any of the above.

Perhaps Major or Miles or one of the other motor experts can chime in on that.

I'm struggling to understand what you propose (a simple diagram would have been very useful) but it sounds like you want to wrap a winding around several adjacent teeth, bridging the gaps between the teeth. My initial impression would be that this would most likely screw up the magnetics somewhat and decrease torque/efficiency.

As a sanity check, consider that the only determinate of torque capability of a motor is the amount of copper wound around its teeth. It doesn't matter how you rearrange it, if you don't increase the amount of active copper you won't get any more torque capability. So changing star to delta or KV or 3 to 6 phases doesn't really matter. It might make the motor better suited to the limitations of your controller, but not inherently a better motor. As amberwolf said, end turn losses are a real consideration though, if you can reduce these somehow for a given tooth geometry (without incurring additional losses) then you could be on to something.

Punx0r said:

I'm struggling to understand what you propose (a simple diagram would have been very useful) but it sounds like you want to wrap a winding around several adjacent teeth, bridging the gaps between the teeth. My initial impression would be that this would most likely screw up the magnetics somewhat and decrease torque/efficiency.

As a sanity check, consider that the only determinate of torque capability of a motor is the amount of copper wound around its teeth. It doesn't matter how you rearrange it, if you don't increase the amount of active copper you won't get any more torque capability. So changing star to delta or KV or 3 to 6 phases doesn't really matter. It might make the motor better suited to the limitations of your controller, but not inherently a better motor. As amberwolf said, end turn losses are a real consideration though, if you can reduce these somehow for a given tooth geometry (without incurring additional losses) then you could be on to something.

Click to expand...

No...I want to wind 2 or 3 adjacent teeth. Each tooth will have same direction winds on them. It's more or less the same thing as winding around all 3 teeth at once, but there's much better copper fill.

Everytime you wind a tooth you end up where you started so how do you plan on reducing end turns and increasing the tooth fill ?,
By filling the tooth next door the magnet ring will need to be changed as you know and the only thing achieved is a differing layout the strands on each tooth will have the same number and diameter facing the same way so the field strength will be identical and the resistance will not change much at all so the same current will flow.

What you need to have a look at is an inductors charicteristics with an iron core,

It's the inductance we manipulate to change the kv this is achieved by altering the resistance of the static coil with less turns and or increased strands per turn etc , so to answer your question the only way it would alter kv is if you change the static resistance of the coil or reduce end turns giving more field magnetics or the third way is increasing the neodymium field strength that would lower the kv to, over time as magnets age the kv on a motor will rise if you damage a motor by overheating it watch when it cools the kv will have increased slightly due to the neodymium field strength being weaker.

There tons of formula for working out how fast a dc or ac motor spins I'll have a bash through my old books I forgot a lot of it but I can remember it being a very intesting topic I went as far as eddy currents in stator's and skin effect etc I covered it for my nvq3 sparky qualification about 8 years ago so I could do with a freshen up.

My thought is to reduce iron losses. Typical hub motors have so many teeth and switching as a result is quite high. Same for large outrunners. This increases iron losses significantly. Winding teeth side by side the same direction AND also using larger magnets that span that same number of teeth will greatly reduce the over all tooth count and poles. The hope is that reduced iron losses will make for a significantly more efficient motor. 6 or 9 phases will help that a little too...or should.

Usually those kinds of losses are more to be worried about in higher speed motors, from what I recall.

End-turn losses and less active copper area might have more of an impact than the lower core losses.

If the stator is wider it might make up for the end-turn losses, as that increases the active area of copper (that's parallel to the magnets, running thru the teeth).

Maybe look up some of Miles' and others' posts about FEMM and see about simulating it to see what happens?

amberwolf said:

...

Perhaps Major or Miles or one of the other motor experts can chime in on that.

Click to expand...

Thanks amberwolf,

I tend to look at the copper in the slots as opposed to "around teeth". Of course I have to go back and consider end-turn copper loss, but it's the slot copper producing torque, or really the net current in the slot.

Look at the OP suggestion to wind 3 adjacent teeth in the same direction. Assume those 3 coils are series connected. Assume, as most are, 2 coil sides per slot. The 3 coils = 6 coil sides placed in 4 adjacent slots. The 2 outer slots have a single coil side from this group of 3 and a coil side from a neighboring group. The 2 inside slots have 2 coil sides per slot all from the same group. In each of those 2 slots, the current in the 2 coil sides flows in opposite directions, cancelling each other.

The net result is 2/3 of the slot copper is wasted.

So not a good winding pattern.

Regards,

major

Major...

Is that why side by side teeth are always wound forward/reverse/forward so that you don't waste copper?

How about those end turns? How can we take advantage of them?

IIUC, (simplification follows) since they're not parallel with the magnets (i.e., parallel with the slots), you can't, really--the force created from that non-parallel flow in them doesn't act on the magnets the way the force from the parallel ones (in the slots) does.

If there was any way to only have current flow thru the copper in the slots themselves, it'd be a more efficient motor (less weight, etc), but there isn't (it has to get *to* the slots somehow).

I suppose if the controller were in the stator itself, and the outputs of the FET stages were connected as "rays" out to the slots, it might be different, but I'm not sure what the actual results would be.

amberwolf said:

IIUC, (simplification follows) since they're not parallel with the magnets (i.e., parallel with the slots), you can't, really--the force created from that non-parallel flow in them doesn't act on the magnets the way the force from the parallel ones (in the slots) does.

If there was any way to only have current flow thru the copper in the slots themselves, it'd be a more efficient motor (less weight, etc), but there isn't (it has to get *to* the slots somehow).

I suppose if the controller were in the stator itself, and the outputs of the FET stages were connected as "rays" out to the slots, it might be different, but I'm not sure what the actual results would be.

Click to expand...

I agree, but I thought maybe a secondary magnet that was parallel with the end winds might be able to use that otherwise bit of wasted copper. Maybe a single controller can't handle magnets turned sideways on the ends of the stators?

Hmmm...so this brings up something else. I suspect that square stator teeth where the end turns and length turns are the same would be 50% efficient...or there about. A stator with long narrow teeth would maximize lengths and minimizes ends. There has to be a balance so that you get the most iron per tooth (width in the tooth), but have very narrow end turns.

A high-tooth-count wide stator (with deep/large slots for lots of copper if you need that) could probably make more active copper vs end copper?


Keep in mind hte magnets on the end probably won't end up as separate domains; they'll either integrate or interfere with the other magnets.

Plus, they'd have to be really skinny to fit in the very narrow space (if any) between the existing magnets and the side covers of the hubmotor. (unless you make new covers that allow for that)

amberwolf said:

A high-tooth-count wide stator (with deep/large slots for lots of copper if you need that) could probably make more active copper vs end copper?


Keep in mind hte magnets on the end probably won't end up as separate domains; they'll either integrate or interfere with the other magnets.

Plus, they'd have to be really skinny to fit in the very narrow space (if any) between the existing magnets and the side covers of the hubmotor. (unless you make new covers that allow for that)

Click to expand...

I just thought it might add to the motor, but since I've never seen it done, it's either not compelling or just adds complexity and of course the magnetic domains would meld together and then what's the point.

If you wanted to improve motor torque per unit of motor weight, you would do the opposite and double the number of teeth and double the number of magnets.

Is that due to reduced quantity of back iron needed? Or are there other advantages?

There is a lot of conflicting information on the 'net about the relationship between pole count and motor torque. The "traditional" EE school says poles = magnetic gearing and more poles = more torque (although never explaining why - at best just a handwave about "lateral air gap"). Then there's the hobbyist forums with professionals/semi-professionals and keen amateurs who design motors who say pole count is completely irrelevant to the torque potential for a given rotor volume.

liveforphysics said:

If you wanted to improve motor torque per unit of motor weight, you would do the opposite and double the number of teeth and double the number of magnets.

Click to expand...

Has anyone here investigated Cupro-Magnesium for stator windings?

It has only 75% of the IACS, but is stronger and can therefore be lighter for a given application according to copperalliance.org.uk

One method of reducing vehicle weight is by replacing Cu-ETP cables by copper-magnesium CW127C (Mg 0.14 to 0.26%), conductivity 75% IACS The copper-magnesium cables first introduced in 2010 are up to 25% stronger than Cu-ETP which enables smaller diameter (0.29 mm) cables to be used, resulting in a weight saving of up to 50% compared to Cu-ETP cables, whilst not jeopardising performance or reliability.

Click to expand...

liveforphysics said:

If you wanted to improve motor torque per unit of motor weight, you would do the opposite and double the number of teeth and double the number of magnets.

Click to expand...

Then that also doubles the amount of switching fields back and forth and doesn't that increase iron losses? I understood that less stator teeth was better for reducing iron losses than more teeth.

I'm with Punx0r on this one...too much conflicting information!

ElectricGod said:

liveforphysics said:

If you wanted to improve motor torque per unit of motor weight, you would do the opposite and double the number of teeth and double the number of magnets.

Click to expand...

Then that also doubles the amount of switching fields back and forth and doesn't that increase iron losses? I understood that less stator teeth was better for reducing iron losses than more teeth.

I'm with Punx0r on this one...too much conflicting information!

Click to expand...

You make your power by flipping the field in the iron relative to the magnets. The more often you flip the field the more power you're making for a given mass of iron, as well as heat of core loss, but using your core is why you're carrying that lump of iron.

Buk___ said:

Has anyone here investigated Cupro-Magnesium for stator windings?

It has only 75% of the IACS, but is stronger and can therefore be lighter for a given application according to copperalliance.org.uk

One method of reducing vehicle weight is by replacing Cu-ETP cables by copper-magnesium CW127C (Mg 0.14 to 0.26%), conductivity 75% IACS The copper-magnesium cables first introduced in 2010 are up to 25% stronger than Cu-ETP which enables smaller diameter (0.29 mm) cables to be used, resulting in a weight saving of up to 50% compared to Cu-ETP cables, whilst not jeopardising performance or reliability.

Click to expand...

Click to expand...

Hi Buk___,

The increase in resistance would undoubtedly cause intolerable additional heat which would significantly derate the motor. And the increase in strength of wire offers little if any benefit to the machine. Am I missing something here?

Regards,

major

major said:

Hi Buk___,

The increase in resistance would undoubtedly cause intolerable additional heat which would significantly derate the motor. And the increase in strength of wire offers little if any benefit to the machine. Am I missing something here?

Regards,

major

Click to expand...

I was reading an article about the new UK electrified train line -- High Speed 2 -- and saw mention that "new high conductivity copper alloys" are being used to "increase the maximum speed of the trains."

Having read this article, I'm confused by how the alloys are being used.

They mention that they are "high conductivity", but then list numbers lower than pure copper; and they mention saving weight (on cars not trains). Hence I thought I'd ask at the one place I know that if there are savings -- energy or weight -- to be had, someone here would likely have investigated it

Perhaps you can achieve the same field strengths by using more material, but lower overall weight?

liveforphysics said:

ElectricGod said:

liveforphysics said:

If you wanted to improve motor torque per unit of motor weight, you would do the opposite and double the number of teeth and double the number of magnets.

Click to expand...

Then that also doubles the amount of switching fields back and forth and doesn't that increase iron losses? I understood that less stator teeth was better for reducing iron losses than more teeth.

I'm with Punx0r on this one...too much conflicting information!

Click to expand...

You make your power by flipping the field in the iron relative to the magnets. The more often you flip the field the more power you're making for a given mass of iron, as well as heat of core loss, but using your core is why you're carrying that lump of iron.

Click to expand...

Right...that's what I understood too. So then which is better or perhaps that's not quite right. There's a balance to be had between flipping fields and iron losses due to flipping fields.

Large tooth count motors such as hubs have far more flipping feilds than say a C80100 outrunner. A 6kw hub is going to have far shorter stator teeth than the C80100 outrunner. Also, the hub is going to have many more teeth than the C80100. It would seem logical therefore that the hub motor will very likely be less efficient than the outrunner for no other reason than the iron losses from many more flipping feilds. Also, the C80100 will have long narrow stator teeth while the hub probably has more squarish teeth. This too is an advantage for the outrunner. It would seem that motors with long skinny stator teeth and a low tooth count are the best of all worlds. On one hand you maximize the torque bearing portion of each wind and on the other you minimize the iron losses. Then I have to ask...what is categorically a "low tooth count"? There has to be a balance here between iron loss and tooth count.

Did I get anything wrong here? Did I make sense?

What is the best balance?

They exist in industry though (just one article that jumped out at the top of a google search): http://www.ciif-expo.com/article/show.php?itemid=279

1m diameter, <1000rpm, up to 100 poles, 20,000Nm torque direct-drive PM brushless motors

Buk___ said:

. They mention that they are "high conductivity", but then list numbers lower than pure copper; and they mention saving weight (on cars not trains).

Click to expand...

I think they are talking about applications where pure copper cannot be used because its mechanical properties are insufficient. Since alloying copper usually has a detrimental effect on its conductivity, they are trying to hit a good compromise between mechanical properties and conductivity with specialist alloys. It makes sense of the train over-head cables, but I'm not sure where car wiring looms come into it, as AFAIK they, like most electric wire (and magnet wire) use pure copper!

There can be a weight (and cost) advantage to using a lower conductivity but overall lighter material like aluminium but car wiring is usually restricted by space as much as anything as modern cars have so many electrical systems the wiring loom gets bulky quickly.