General question about hub motors...

[RoughRider]

i would like to undestand, why the rotation speed of a hubmotor is so low?

I know, that you dont need 3000rpm on a bicycle direct drive hubmotor...

but generaly, what makes them rotate so slow?

Is it the controller(frequency) or is it the wireing(number of pairs of poles)?

 

[OneWayTraffic]

Because they are often directly linked to the rotation speed of the wheel. (Direct drive.) This has advantages (simple design) and disadvantages, chiefly low power density and hence high weight.

Some internal motors have internal gears to run the hub at a RPM 4 or 5 times higher than the wheel. They have lower weight for a given power output, higher lowdown torque per weight/voltage.

 

[johnrobholmes]

The relatively large diameter, and large number of stator and magnet poles help with the slow speed. The winding determines the final KV within the power capabilities of the motor.

 

[Drunkskunk]

-Large Diameter
-Lots of polls
-high winding count
All of these contribute to a low RPM motor. Gearless Hub motors used in bikes generaly run at 200 to 450 RPM. Although the coil might be passing the magnet at the same speed as a 3000 RPM motor, if there are 10X as many, then it goes 10X slower. Thats the theory anyway. in reality, there are only 4 or 5 times more magnets and coils, but that combined with the winding count and diameter slow the motor down without losing power.

 

[RoughRider]

comon the diameter is not so big at all...

the D of a X5 is around 200mm of the airgap...thats not big...

e.g. the UQM motors D is 280mm(of the airgap maybe 220mm) and they spin to 8000rpm...


OK...many magnets many poles...i understand...but all you need is just more frequency in the controller to let it spin faster right?

 

[johnrobholmes]

Right, the higher the pole and slot count, the faster the controller must switch phases to run the motor.


Diameter is relative. Take a 200mm motor and a 100mm motor and the 200mm motor will run 4x slower if all else is the same.


Comparing a UQM to an X5 is apples to oranges. The UQM 75 works on 250 to 400v, the x5 is 36 to 75v. Is the UQM slotless, two pole, four pole, etc? Double the size of the 280mm UQM and it will be 4x slower. Double the slot and magnet count and it will halve in speed. Double the length and it will halve in speed. All relative.

 

[RoughRider]

wait a second...

we are talikng about the unloaded speed..

what have voltage, diameter and length to do with the rpm of the motor??

as far as i understand...they have nothing to do with it...its just about the poles and the frequency, right?

 

[johnrobholmes]

Wind, size, and many aspects of construction affect the rpm per volt of the motor. Stronger magnets or smaller airgap will slow it, as well as larger construction.

 

[chet]

The hub is actually an AC motor. Isn't the rotational speed maxed by the phase, controlled by the controller?

 

[johnrobholmes]

Since the driving voltage is DC, technically it is a three phase DC motor. The construction is the same as a three phase AC though, so it is merely semantics.

The controller chops the battery voltage, so at 1/2 throttle you have 50% duty cycle. The corresponding motor speed is roughly the Kv* battery voltage* duty cycle before load is applied.

With newer controllers the motor speed is not limited by the speed of phase switching, but on older controllers they may not be able to switch fast enough to get full speed out.

 

[rkosiorek]

Since the driving voltage is DC, technically it is a three phase DC motor. The construction is the same as a three phase AC though, so it is merely semantics.

The controller chops the battery voltage, so at 1/2 throttle you have 50% duty cycle. The corresponding motor speed is roughly the Kv* battery voltage* duty cycle before load is applied.

With newer controllers the motor speed is not limited by the speed of phase switching, but on older controllers they may not be able to switch fast enough to get full speed out.

the "H" bridge design of the controller converts it to AC. it alternatetly connects each motor wire to the (+) and (-). that word ALTERNATETLY is the key. like in ALTERNATING CURRENT or AC.

so technically it is a 3 phase AC MOTOR driven by a H bridge power converter that converts DC to AC.

rick

 

[johnrobholmes]

I guess most controllers do use 6 step commutation, but I have always been taught in my CIM classes to call them 3 phase DC since the supply is DC. Probably varies from teacher to teacher :lol:

 

[dscline]

the "H" bridge design of the controller converts it to AC. it alternatetly connects each motor wire to the (+) and (-). that word ALTERNATETLY is the key. like in ALTERNATING CURRENT or AC.

The commutator does that in a simple brushed DC motor, and it's still considered a DC motor. :wink:

 

[rkosiorek]

maybe so. but without modificatio a brushless motor could be connected to a 3 phase AC source and run. well maybe the peak voltage may have to be reduced. the same could not be said for a brushed DC motor. it will not run on AC at all. unless the comutator was removed and replaced with slip rings.

rick

 

[JCG]

without modificatio a brushless motor could be connected to a 3 phase AC source and run.

I find this very intriguing. So, would this be feasible:

12 V battery, large capacity --> power inverter --> 120 VAC --> 3-phase motor controller with 120 V input --> BLDC hub motor?

I first assume that the hall sensors no longer need be used. A big problem - it seems that most 3-phase controllers have outputs of at least 200 V. I assume this will cause a problem for a PM BLDC motor that normally runs on 36 VDC. How could you cut the motor phase voltage down?

 

[Drunkskunk]

I find this very intriguing. So, would this be feasible:

12 V battery, large capacity --> power inverter --> 120 VAC --> 3-phase motor controller with 120 V input --> BLDC hub motor?

I first assume that the hall sensors no longer need be used. A big problem - it seems that most 3-phase controllers have outputs of at least 200 V. I assume this will cause a problem for a PM BLDC motor that normally runs on 36 VDC. How could you cut the motor phase voltage down?

feasible? Yes. Practical? Not even close. The designs are similar, but Brushless DC motors are optimized for DC Square wave pulses, were 3 phase AC is sinusoidal. 3 Phase AC works at low frequancy, were 3 pole BLDC motors work at high frequency were the rolloff of a sine wave would be problomatic.

There are too many poles, so AC would turn the motor slower, and since it wouldn't have the speed to produce the back EMF needed to choke the power coming in, the BLDC motor would get hot fast.

 

[JCG]

Good stuff, DS. I hadn't even thought about the number of poles; some of these hub motors have a ridiculous number of magnets on their periphery. Thanks for the insight.

 

[paultrafalgar]

JCG, where art thou? Missing your interesting research input. Any new developments on the ultracapacitor-battery bike front?

 

[JCG]

Paul, thanks for checking up on me... I'm currently in the self-created purgatory of:

Pedaled permanent magnet alternator --> 12 VDC storage system (incl. 16V ultracap modules) --> 12 VDC to 120 VAC inverter --> AC motor controller --> 230 V AC induction motor

all on a KMX trike, designed to be a cruiser-type thing (built for distance). With the front crank hooked to a PM alternator to charge things.

Things are too heavy. 1/2 hp wasn't enough. 1 hp isn't enough either. Now I'm forced to look into 3 hp (super heavy) to climb hills, or try something else altogether. I'll post a bunch of stuff once something works... :|

 

[Drunkskunk]

glad you're still here JCG.

but..
The inefficancy of that system is making my head hurt!
On the other hand, awesome reasearch tool!

Just currrious about 1hp not being enough... Enough for what? 1Hp, 750 watts, that ought to be able to drive the bike up a flight of stairs with propper gearing. I'm curious what you're putting it through that it isn't enough.

Your project looks like too much fun.

 

[JCG]

The inefficancy of that system is making my head hurt!

DS, you are sooooo right. The good news is that I saw the looming inefficiencies ahead of time... let me count a few:

trike = heavy frame to start with
AC motor = big current draws (low eff.) when demanding high torque at low speed!
DC to AC inverter = 1 to 2 A (at 12 VDC) current draw with no load... ugh
PMA alternator = maybe 50 W out at 12 VDC for 200 W of "pedaling effort"

It's just hard to beat the 95+% efficiency pedaling a chain connected to the wheel! But as you guessed I'm trying to make a scale model of a series hybrid. I'll likely use a mini generator instead of pedaling as the primary energy source (27 lb, 900 W Yamaha that I have on hand).

Just currrious about 1hp not being enough... Enough for what? 1Hp, 750 watts, that ought to be able to drive the bike up a flight of stairs with propper gearing.

Yeah, you'd think so. The gearing is definitely an issue. I assume that it's all about the classic torque/speed tradeoff, and I'm too far down the road on the speed side at the moment. Since I've been using mostly junk I have on hand, I have just used a 9-tooth sprocket on the motor while connecting it to the largest freewheel sprocket (28 teeth, I think) for testing.

The problem is that the motor is rated for 1725 rpm (4 poles, 60 Hz, some slip). If I buy a motor that has more poles for a slower speed, it's heavier (and more expensive). I'm going to try switching the motor controller to 50 Hz operation as a quick test, but I need to gear this thing down. I have a 45 tooth sprocket sitting around, just need a way to stick it on the wheel. I've never done that before. I would like to keep the freewheel though, since I don't have a handy way to store regen energy when I slow down (the DC bus in the controller ranges from 190 V (undervoltage) to 395 V (overvoltage) during operation).

Anyway, you get the idea. This thing will do great running fast on a flat straightaway. Acceleration, and hill climbing, aren't there yet.

 

[Drunkskunk]

You might have to drill a few holes, but it might be possable to bolt the 45 tooth ring to the back of the 9 speed 28 tooth ring. Might also need an adaptor plate made out of something.

 

[JCG]

That might be the only option. I'm going to fish for some opinions from the non-hubmotor types and see what kind of tricks there might be to fixing a large sprocket to a rear wheel - they do it all the time; should be some stuff off the shelf to buy...