Science, Physics, Math, & Myth

[Chalo]

"High Torque" - If this is the case, what is a more accurate way to describe a High-turn-count motor? "Slow-speed" motor?

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

 

[zener]

"High Torque" - If this is the case, what is a more accurate way to describe a High-turn-count motor? "Slow-speed" motor?

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

That makes sense.

 

[liveforphysics]

MXUS is so brainwashed by the myth that they are resisting winding a motor with 3 turns instead of 4, when a 3 turn would be right in the sweet spot with just enough inductance that it would serve everyone's needs with fast or slow determined by voltage. As you approach 100V, speed and power that blows what's commonly available out of the water is easy and cheap, just low hanging fruit ripe for the pickins, and those who want better low speed efficiency can simply run a lower voltage and still get the same torque they would with the slow wind motor by running the proper current.

im with you John, but we do not know the nominal or max RPM of the MXUS 3000 motor (before eddy current losses starts to add big amounts of heat).
by the upgraded e4bikes V2 with 0,33mm lams things are better, but with the stock one and its 0,5mm lams?

In Justin's own hubmotor dyno testing work, he said that eddy currents in hubmotors as a portion of losses is so low it's insignificant across the whole practical speed range he measures. Hysteresis core losses are a factor that climbs linearly with RPM, but eddy losses measure to be insignificant.

 

[miuan]

"High Torque" - If this is the case, what is a more accurate way to describe a High-turn-count motor? "Slow-speed" motor?

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

That makes sense.

Only at first sight. Actually the reason why the slower motor produces more torque is because the controller is more efficient when driving the slower motor. Once you use the faster motor with battery AND a controller with proportionally lower voltage and higher current rating, the total efficiency will be the same for both systems.

 

[zener]

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

That makes sense.

Only at first sight. Actually the reason why the slower motor produces more torque is because the controller is more efficient when driving the slower motor. Once you use the faster motor with battery AND a controller with proportionally lower voltage and higher current rating, the total efficiency will be the same for both systems.

So optimal is a 60V battery with a high A controller and some beefy pure copper cables with a 3T motor ?
That saves alot of serial wiring complexy for the battery. 8)

 

[liveforphysics]

Yup, running either 12S or 20S (to leverage either 60Vdc MOSFETs or 100Vdc MOSFETS) is where to be if your interest is in ultimate efficiency and performance (or for hot-rodding.) Though again, it's not because the motor cares, it's because running 20S lets the most power to be sent to the motor for a given number of FETs in a controller.

Bucking through the motors winding is extremely efficient, the motor really doesn't care if you start from a 1,000V battery or 10V battery, it just changes what duty cycle% the controller uses to hold whatever phase current you selected.

Remember, no amount of voltage, higher or lower, makes torque in the motor. The motor only 'reacts' to whatever the amount of amp-turns happen to be on the tooth. From a stall (and hence no BEMF being generated) it may only take <1v across the phase leads to generate full torque on that motor (it makes that <1V from bucking down pack voltage).

Pick a motor kV that enables the motor to spin up to the desired peak RPM from 12S or 20S packs, and you will have chosen the highest performance option (due entirely to best leveraging available MOSFET tech, the motor itself doesn't care of course).

 

[Punx0r]

The point about the stock phase wires being inadequate for a low-turn motor is possibly valid (it would be good to have empirical evidence to confirm it).

We all know the axle design on most low-end hub motors is dire and the phase wiring is a big part of that. We can only hub the manufacturers get the message at some point...

 

[cal3thousand]

I'm liking this thread for all of the effort that is being thrown at explaining motor design, and ultimately, performance in all these different ways. It helps a semi-noob like myself.

So how about this conundrum. Say a person purchased a low turn count, 'fast'-wind motor that is small (rated at 250W; whatever that means ) and mistakenly decided to lace it to a large wheel (say a 700c). He finds out that 12S and 30A is horribly tuned for this setup. He doesn't want a smaller wheel because his bike would look stupid with a 700c front and a 20" rear and the geometry would be awful. Should he try to parallel his packs down to 6S (and modify controller LVC to suit) and turn the battery amps up to 60A or even 100A?

 

[Allex]

Alright, so on Low-turn (21X3T) motors, how would you deal with the small phase wires to get enough current to the windings without overheating the phases?

Obviously you run heavy gauge wire from the controller as close as you can to the motor/Axle, but at best you can get what about 10AWG through the bearing? - Even with a very short 10AWG conductor through the bearings, if your phase amps are up in the 200A-300A range, it's going to be cooking that 10AWG wire, no?

And then if you solve the phase wire issue, you still need a controller capable of outputting that much current. Everything has to get bigger with the higher current needs of a low-turn motor.

I'm looking for practical solutions, not something I have to Mod the heck out of.

I have 10AWG Turnigy wires in my motor and feed them with 400Phase Amps, they dont even get warm.
Yeah I know 400 is much, I know that the motor heats up quick, but it is damn fun to ride...

 

[Thud]

So how about this conundrum. Say a person purchased a low turn count, 'fast'-wind motor that is small (rated at 250W; whatever that means ) and mistakenly decided to lace it to a large wheel (say a 700c). He finds out that 12S and 30A is horribly tuned for this setup. He doesn't want a smaller wheel because his bike would look stupid with a 700c front and a 20" rear and the geometry would be awful. Should he try to parallel his packs down to 6S (and modify controller LVC to suit) and turn the battery amps up to 60A or even 100A?

You can't polish a turd....& to quote JohninCR "watts are watts". I would recomend he just crank down the controller to limit it from over feeding the poorly selected motor....& start looking for other options. reducing voltage is one way, but that affects the top speed more & runing higher amperages Creates a tippyer canoe of potential meltyness.

Mosfet's,diodes, heck, all the ellectronics perfer voltage over amperage.

 

[teslanv]

Alright, so on Low-turn (21X3T) motors, how would you deal with the small phase wires to get enough current to the windings without overheating the phases?

Obviously you run heavy gauge wire from the controller as close as you can to the motor/Axle, but at best you can get what about 10AWG through the bearing? - Even with a very short 10AWG conductor through the bearings, if your phase amps are up in the 200A-300A range, it's going to be cooking that 10AWG wire, no?

And then if you solve the phase wire issue, you still need a controller capable of outputting that much current. Everything has to get bigger with the higher current needs of a low-turn motor.

I'm looking for practical solutions, not something I have to Mod the heck out of.

I have 10AWG Turnigy wires in my motor and feed them with 400Phase Amps, they dont even get warm.
Yeah I know 400 is much, I know that the motor heats up quick, but it is damn fun to ride...

Is this because the total Phase current is divided up evenly between the three phase conductors? If that is the case, then each phase conductor would only see 133.3 Amps of a 400A Total phase load, and it seems likely that 10AWG wires would not be overly stressed in this instance.

 

[liveforphysics]

I'm liking this thread for all of the effort that is being thrown at explaining motor design, and ultimately, performance in all these different ways. It helps a semi-noob like myself.

So how about this conundrum. Say a person purchased a low turn count, 'fast'-wind motor that is small (rated at 250W; whatever that means ) and mistakenly decided to lace it to a large wheel (say a 700c). He finds out that 12S and 30A is horribly tuned for this setup. He doesn't want a smaller wheel because his bike would look stupid with a 700c front and a 20" rear and the geometry would be awful. Should he try to parallel his packs down to 6S (and modify controller LVC to suit) and turn the battery amps up to 60A or even 100A?

There would be no need to reduce pack voltage. Just increasing phase current is all it would take. Note that making this increased torque wouldn't consume any more power from the battery over starting from a higher turn-count motor and making the same torque.


It seems we've got a group of folks now who have grasped the concept! Woot!! If you are someone who gets it now, please help guide/correct others when folks start spewing motor related BS. JohninCR my friend, thank you for so often being the solitary light of wisdom amidst a sea of delusion in motor related threads.

Does everyone also get why it's no additional power draw from the battery to reduce the turn-count and increase the phase current proportionately?

 

[liveforphysics]

So how about this conundrum. Say a person purchased a low turn count, 'fast'-wind motor that is small (rated at 250W; whatever that means ) and mistakenly decided to lace it to a large wheel (say a 700c). He finds out that 12S and 30A is horribly tuned for this setup. He doesn't want a smaller wheel because his bike would look stupid with a 700c front and a 20" rear and the geometry would be awful. Should he try to parallel his packs down to 6S (and modify controller LVC to suit) and turn the battery amps up to 60A or even 100A?

You can't polish a turd....& to quote JohninCR "watts are watts". I would recomend he just crank down the controller to limit it from over feeding the poorly selected motor....& start looking for other options. reducing voltage is one way, but that affects the top speed more & runing higher amperages Creates a tippyer canoe of potential meltyness.

Mosfet's,diodes, heck, all the ellectronics perfer voltage over amperage.

There is no need for him to reduce voltage Thud. The controller bucks down whatever it needs. Until the point you physically change out the FETs in the controller to a lower voltage version, you gain nothing in choosing a lower pack voltage.

 

[cal3thousand]

So how about this conundrum. Say a person purchased a low turn count, 'fast'-wind motor that is small (rated at 250W; whatever that means ) and mistakenly decided to lace it to a large wheel (say a 700c). He finds out that 12S and 30A is horribly tuned for this setup. He doesn't want a smaller wheel because his bike would look stupid with a 700c front and a 20" rear and the geometry would be awful. Should he try to parallel his packs down to 6S (and modify controller LVC to suit) and turn the battery amps up to 60A or even 100A?

You can't polish a turd....& to quote JohninCR "watts are watts". I would recomend he just crank down the controller to limit it from over feeding the poorly selected motor....& start looking for other options. reducing voltage is one way, but that affects the top speed more & runing higher amperages Creates a tippyer canoe of potential meltyness.

Mosfet's,diodes, heck, all the ellectronics perfer voltage over amperage.

There is no need for him to reduce voltage Thud. The controller bucks down whatever it needs. Until the point you physically change out the FETs in the controller to a lower voltage version, you gain nothing in choosing a lower pack voltage.

Except for target speed right? Turn up the amps and only put as many watts into the motor as you dare, using the throttle hand (or other control measure like CA)

Also, wouldn't the battery like it better to be paralleled down in this case?

 

[liveforphysics]

Also, wouldn't the battery like it better to be paralleled down in this case?

The battery has no configuration preference for having some amount of power drawn from it. It's just as silly to say "high torque battery configuration" as it is to say "high torque motor wind".

If you have a 2t motor and a 4t motor, and you've got a controller to feed them the matched amounts of power (meaning the 2t would be getting fed 2x the average phase current at 1/2 the average phase voltage). Both end up drawing exactly the same amount of load from the battery to make the same amount of torque or power etc.

 

[Allex]

So you do not touch the DC current, only phase to make same torque?

 

[zener]

So you do not touch the DC current, only phase to make same torque?

I guess the controller does it automically becouse of lower resistance from motor.

 

[liveforphysics]

So you do not touch the DC current, only phase to make same torque?

Yes. The motor has no way to know or care what your battery current may happen to be. If your controller is drawing 1Amp off a 1,000V battery, or drawing 10A off a 100V battery, your motor could never know or care, it only experiences the resultant phase current, which after the controller bucks it down would be identical between the two.

 

[cal3thousand]

So you do not touch the DC current, only phase to make same torque?

Yes. The motor has no way to know or care what your battery current may happen to be. If your controller is drawing 1Amp off a 1,000V battery, or drawing 10A off a 100V battery, your motor could never know or care, it only experiences the resultant phase current, which after the controller bucks it down would be identical between the two.

Ahh... In this case, one would change the batteryhase ratio in the controller, not the overall Amp draw. I'm seeing more clearly now

 

[John in CR]

"High Torque" - If this is the case, what is a more accurate way to describe a High-turn-count motor? "Slow-speed" motor?

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

To make more torque at any rpm it makes more heat.

 

[Chalo]

"High Torque" - If this is the case, what is a more accurate way to describe a High-turn-count motor? "Slow-speed" motor?

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

To make more torque at any rpm it makes more heat.

Not every application is constrained by heat. Most, in fact, are not. Most are constrained first by cost, then by weight or physical size. If your motor will never overheat, then more stall torque is better-- if getting it doesn't require bigger, more expensive components.

Miuan's observation that a low turn count motor can be accommodated by a higher current controller and battery ignores the fact that high current controllers and batteries and wiring harnesses cost and weigh more, and are physically larger, than lower current versions. That makes them intrinsically less suitable for most applications.

 

[John in CR]

Alright, so on Low-turn (21X3T) motors, how would you deal with the small phase wires to get enough current to the windings without overheating the phases?

Obviously you run heavy gauge wire from the controller as close as you can to the motor/Axle, but at best you can get what about 10AWG through the bearing? - Even with a very short 10AWG conductor through the bearings, if your phase amps are up in the 200A-300A range, it's going to be cooking that 10AWG wire, no?

And then if you solve the phase wire issue, you still need a controller capable of outputting that much current. Everything has to get bigger with the higher current needs of a low-turn motor.

I'm looking for practical solutions, not something I have to Mod the heck out of.

On those V2 motors, the slots are bigger, so first I would to get them to see if there's any possibility of making it 22strands by 3turn, which is only 2 strands more than the 4T.

10ga is slightly larger than the windings on a 3T, so it won't heat any more than the rest of the copper in the motor.

If you're going to run that kind of current you need cooling mods unless you're lucky enough to live where it's very flat, so redoing the harness is minor in comparison. I've found the way to get the most copper through is by using magnet wire. Having only the thin magnet wire enamel separating the phases, hall wires, and temp sensor wires from each other frees up space for more copper. Add a double layer of quality shrink over the whole thing for triple layer of insulation. Stiffer magnet wire is actually useful coming out of the axle, because it holds it's form better.

The lazy way also works well too, which is to go with large gauge wire starting just a few inches outside of the axle. Make that connection fall in your drip loop, and it gets good air flow for cooling. Copper conducts heat so well that it will sink a lot of heat out of the wire inside the motor.

 

[John in CR]

A high turn count motor, with the same battery and controller, makes more zero speed torque than a low turn count motor. All our bikes start from zero speed, so that characteristic is relevant.

To make more torque at any rpm it makes more heat.

Not every application is constrained by heat. Most, in fact, are not. Most are constrained first by cost, then by weight or physical size. If your motor will never overheat, then more stall torque is better-- if getting it doesn't require bigger, more expensive components.

Miuan's observation that a low turn count motor can be accommodated by a higher current controller and battery ignores the fact that high current controllers and batteries and wiring harnesses cost and weigh more, and are physically larger, than lower current versions. That makes them intrinsically less suitable for most applications.

If that's the power levels you're talking about, then you can get even cheaper lower voltage controllers, lower voltage BMS, and lower voltage charger, and still be able to deliver the higher current to the lower turn count motor, and save even more money. You're talking to one of the most budget minded forum members.

You would want to stay away from the 2 turn or 1.5 turn motors in that case though, because the extremely low inductance requires more overhead in terms of controller capacity and that would result in bigger controllers. Those are only found as scooter motors though, which are higher power and have plenty of room for controllers.

 

[flathill]

The phase wires running out the motor only need to be the same size as the equivalent gauge inside the motor (equivalent gauge means you have to know if the winding is bifilar, trifilar, single wind, etc...not necessarily the same size as a single strand of magnet wire). Think how long the magnet wire coils on the stator would be if unwrapped. Extending it a few inches to outside the motor with the same equivalent size conductor(s) is not going to matter much at all. You don't need to go crazy and stuff the fattest wires in there in most cases, unless you are just trying to show off, or you are racing and need that .01% to win

I've also found it is easier to fit two wires per phase (in parallel) than one fat wire per phase. Think about it...6 circles fit exactly around a circle of the same size, with no gap...the 6 circles are your 6 phase wires...the inner circle is just enough space for your temp and hall sensor wiring. The 6 phase wires coming out the motor can be magnet wire and then whole bundle (with halls and temp) can be shrink wrapped like John suggested.

Place the controller near the motor. If you can't do this for aesthetics, then just make a larger gauge extension harness for the phase wires. This is where you want a connector anyway when you need to pull the motor/wheel off the bike

 

[liveforphysics]

If I were designing scratch-built high performance/efficiency ebike drivetrains for something under ~1000w, I would use 6s at most, likely 4s.

The cost of high current switching has dropped as semiconductor tech continues to make huge improvements each year. This chip is under $20, and it's the whole power switching stage in a chip.
http://ixapps.ixys.com/Viewer.aspx?p=http%3a%2f%2fixapps.ixys.com%2fDataSheet%2fMTI200WX75GD.pdf


The additional cost of copper to go from say 14awg to 8-10awg for say 6ft of wire in the whole traction harness is about ~$2.40 of cost. The added weight is ~125g. (for 8awg) You will save far more money and complexity/cost in harnessing from having a more simple BMS alone.

Drawing 100-200A intermittently through 8awg isn't something to fear, neither is switching it (if designed for it obviously).

If I were willing to spend the effort to design a new ground up deathbike, with the goal being breaking drag racing records, I would run 20s rather than the 28s I run today.