Question over different pole pairs in motors

[Santacruz]

Hi all,
I have searched, but not really found a simplified answer.
That is pole pairs in a motor.
I take as an example, MXUS 3k - 23 pole pairs and a QS205 3k - 16 pole pairs.
So one has 46 magnets and one has 32 magnets.
Does this have something to do with RPM or torque or even reliability?
What is the difference and what is better to have.

 

[John in CR]

Hi all,
I have searched, but not really found a simplified answer.
That is pole pairs in a motor.
I take as an example, MXUS 3k - 23 pole pairs and a QS205 3k - 16 pole pairs.
So one has 46 magnets and one has 32 magnets.
Does this have something to do with RPM or torque or even reliability?
What is the difference and what is better to have.

As long as the motor can produce the torque you need and assuming all else is equal, the lower the number of pole pairs the better. That's because the iron core losses vary directly with switching frequency, and the less magnets, the lower the number of times the polarity changes per revolution. It's done to achieve a more efficient motor. The problem is that the stator will require few teeth for a good match of poles and slots, and that results in deeper teeth and more laminating steel, which is more expensive and generally heavier. The hubmotors I use have only 10 pole pairs, which a big part of the reason they have a peak efficiency of over 96%. Too bad they're out of production due to cost.

 

[Santacruz]

Thank you,
That's fairly clear.

 

[thepronghorn]

Hi all,
I have searched, but not really found a simplified answer.
That is pole pairs in a motor.
I take as an example, MXUS 3k - 23 pole pairs and a QS205 3k - 16 pole pairs.
So one has 46 magnets and one has 32 magnets.
Does this have something to do with RPM or torque or even reliability?
What is the difference and what is better to have.

As long as the motor can produce the torque you need and assuming all else is equal, the lower the number of pole pairs the better. That's because the iron core losses vary directly with switching frequency, and the less magnets, the lower the number of times the polarity changes per revolution. It's done to achieve a more efficient motor. The problem is that the stator will require few teeth for a good match of poles and slots, and that results in deeper teeth and more laminating steel, which is more expensive and generally heavier. The hubmotors I use have only 10 pole pairs, which a big part of the reason they have a peak efficiency of over 96%. Too bad they're out of production due to cost.

All else isn't equal though. While you can decrease hysteresis losses with fewer pole pairs to lower the electrical frequency of the motor, you can also increase the torque constant of a motor by increasing the number of slots and pole pairs. Since hub motors are usually operated in conditions where torque/copper losses dominate rpm/hysteresis/eddy losses, increasing torque at the expense of rpm losses usually results in more efficient motors.

John's "hubmonsters" aren't particularly efficient at making torque with specific Km (Nm/sqrt(W)/kg) values below those of many normal bicycle hubmotors such as MXUS3000 and TC4080 based on data from the "Motor Data V4.35" motor comparison spreadsheet. This is probably largely due to their low pole/slot count. The hubmonsters do have low enough specific hysteresis drag (Nm/kg) compared to the same normal bicycle hub motors to end up being more efficient, but I would attribute that more to quality design and materials rather than slot/pole count configuration. I would argue it is cheaper to make an efficient motor with lower slot/pole count since hysteresis losses are inherently lower, and deep slots mostly just waste less material when they are stamped out.

 

[John in CR]

Hi all,
I have searched, but not really found a simplified answer.
That is pole pairs in a motor.
I take as an example, MXUS 3k - 23 pole pairs and a QS205 3k - 16 pole pairs.
So one has 46 magnets and one has 32 magnets.
Does this have something to do with RPM or torque or even reliability?
What is the difference and what is better to have.

As long as the motor can produce the torque you need and assuming all else is equal, the lower the number of pole pairs the better. That's because the iron core losses vary directly with switching frequency, and the less magnets, the lower the number of times the polarity changes per revolution. It's done to achieve a more efficient motor. The problem is that the stator will require few teeth for a good match of poles and slots, and that results in deeper teeth and more laminating steel, which is more expensive and generally heavier. The hubmotors I use have only 10 pole pairs, which a big part of the reason they have a peak efficiency of over 96%. Too bad they're out of production due to cost.

All else isn't equal though. While you can decrease hysteresis losses with fewer pole pairs to lower the electrical frequency of the motor, you can also increase the torque constant of a motor by increasing the number of slots and pole pairs. Since hub motors are usually operated in conditions where torque/copper losses dominate rpm/hysteresis/eddy losses, increasing torque at the expense of rpm losses usually results in more efficient motors.

John's "hubmonsters" aren't particularly efficient at making torque with specific Km (Nm/sqrt(W)/kg) values below those of many normal bicycle hubmotors such as MXUS3000 and TC4080 based on data from the "Motor Data V4.35" motor comparison spreadsheet. This is probably largely due to their low pole/slot count. The hubmonsters do have low enough specific hysteresis drag (Nm/kg) compared to the same normal bicycle hub motors to end up being more efficient, but I would attribute that more to quality design and materials rather than slot/pole count configuration. I would argue it is cheaper to make an efficient motor with lower slot/pole count since hysteresis losses are inherently lower, and deep slots mostly just waste less material when they are stamped out.

I'm pretty sure that the added stator weight is the only detriment to the efficiency gain. I don't know if it requires thicker back iron for the magnets too, but the QSv3 weighs more than the MXUS (more than just the extra 5mm of stator length would create). They pick up some extra efficiency and are useful to higher rpm due to the slot and pole difference.

Note that HubMonsters have quite a bit of extra weight that could be avoided if made for ebike use. They have only one AL side cover with the rest of the shell cast steel with a quite wide chunk of steel for the disc brake mount so it is outside the width of a wide 13" scooter rim and tire. I was able to trim 1.5kg off of one while still retaining a sizable disc mount, and if I modded one to accept another AL side cover I could save another kg or so. Then the only bit of wasted weight would be some of the 25mm axle required for heavy scooters. Even after all that the extra deep stator steel necessary for the low slot count may not make it match the specific Km with the aged approach of a thin ribbon of stator steel at the perimeter of high slot and pole count motors, which are made that way because it's cheaper. I'd pay the weight penalty for a truly high efficiency DD hubbie made for ebikes any day of the week. To bad they're still not made. The Leaf motor certainly isn't one despite that thread title. It just picks up a slight efficiency gain due to thinner lams, which is becoming fairly common with almost defacto standard 23 pole motors.

For me higher efficiency gained through lower iron losses and lower resistance for a given Kv is well worth a bit of weight penalty, because it makes extreme performance possible including true highway speeds and more.