Using RC motors on E-bikes [Archive]

[Miles]

Piotrek,

How much difference do you think sine wave commutation would make to reducing the motor noise?

 

[eP]

How much difference do you think sine wave commutation would make to reducing the motor noise?

To telling the truth i don't know more than you Miles i'm afraid
If mechanical sources generating most of the noise (bearings, belts, gears) than elimination of higher harmonics couldn't help much for significantly lowering noise i suppose. If magnetostriction effect is main source of noise than maybe sine could help. Zero magnetostriction effect stator material could help maybe even more for that case.

I think noise issue is a myth: when i ride downhill the street i hear only winds noise in my ears or vehicle's noise. Uphill i hear my rusty chain. If somebody is very sensitive to noise than maybe noise could be an issue for him.

 

[Miles]

Maybe it's not so much the level of the noise as the quality of it... I don't mind the sound of aircraft with propellers but the sound of jet engines at the same SPL is horrible.

Would a sine wave drive be much more complicated? Would it be more efficient?

 

[Miles]

I found this:

Ref: http://www.semicon.toshiba.co.jp/eng/application/homeappliance/airconditioner/selection/1178215_3199.html

 

[eP]

Would a sine wave drive be much more complicated? Would it be more efficient?

Look at that article:
http://www.edn.com/blog/1470000147/post/1700012370.html

Which raises the question, why doesn’t everyone use PMSMs instead of BLDC motors? Because until recently programming the complex sinusoidal waveform control algorithms drove the development costs too high and required a more powerful (and more expensive) processor than what’s needed for the relatively simple trapezoidal control of a BLDC motor. With the motor control development platforms recently introduced by companies such as IR, Microchip, Freescale, and STMicro, the algorithm development is already done for you, and you don’t have to be a algorithm jock to use a higher-torque, low-noise, energy efficient PMSM.

So its seems to be more complicated but is also more efficient as they said.

 

[Miles]

Thanks. Interesting article.

I guess direct drive hub motors would benefit the most, if they went PMSM?

 

[safe]

If somebody is very sensitive to noise than maybe noise could be an issue for him.

Yeah it always fascinates me about people that complain about noise. I use the engine noise as a way to know when to shift gears. Without the motor noise I would need some kind of tachometer to know where I was in the powerband because with the standard PWM powerband the torque is better down low than up high and that makes riding the powerband by feel difficult sometimes.

The motor noise assists in knowing when to shift.

So I don't see it as a negative... and it sounds kind of cool too... my motor makes this smooth humming sound all the way up to the no load area when it starts to break up... I like the sound... it feels like a high tech electric version of a road racing engine.

I'm hoping that the RC motors make a high pitched squeal so that it will also make music when you ride...

Huuuuummmmmmmmm.... eeeeeeeeee...... click (shift)
Huuuuummmmmmmmm.... eeeeeeeeee...... click (shift)
Huuuuummmmmmmmm.... eeeeeeeeee...... click (shift)
Huuuuummmmmmmmm.... eeeeeeeeee...... click (shift)

 

[Kraeuterbutter]

maybe it would be less problem to know when you need to shift, when the power of the electric motor was not that great..
then you would feel it ?!???

some 200m from here is an computer shop
the guy there is about 60years old, going to work every day with an electric scooter...

also its not loud, its so that when i have the window open (4th floor) i hear him go to work and drive home every day..
i would prefer it, when it was totaly quiet

 

[eP]

I guess direct drive hub motors would benefit the most, if they went PMSM?

I don't know - i'm not an expert yet
I think the less efficient could gain more than their more efficient BLDC counterparts IMHO.

If we talking about noise and PMSM than maybe we should investigate the transverse flux motors too ? :wink:

They could have the best torq to weight ratios and lowest working rpm range. So they are best candidates to be salient champions

They are not used in RC world these days however, but who know what will be tomorrow ? They are very efficient for massive torque applications. So they could fit well to the greatests propellers.

 

[safe]

maybe it would be less problem to know when you need to shift, when the power of the electric motor was not that great..
then you would feel it ?!???

It's the nature of the PWM power delivery. In it's stock form the energy is held constant from the battery being regulated by the controller. This translates (through the inductance of the motor) to a situation where low rpms actually delivers more torque (and current) than the rpms really warrant. In order to bring back the linear power curve (easier to ride) you need to switch the controller logic to "armature current limiting" which will flatten it out.

Using "armature current limiting" with the RC motor is an important idea also because it will resolve the problem that Recumpence has had which is excessive torque when you don't want it. By smoothing out the power and making it more linear you solve several problems at once.

Gears and "armature current limiting" are a natural combination because they complement each other.

...but this is a more advanced topic. Before you get the RC motor tuned like this you have to solve the basic issues of getting the gears sorted out. "Armature current limiting" is something you do later when you are tweeking the bike to make it more refined.

 

[dirty_d]

you cant really do sinusoidal commutation with a regular BLDC motor, it needs to have an absolute position sensor on the shaft to work because you need to know exactly where in the commutation step the rotor is to determine the AC output amplitude. and i think for it to be of any benefit the backemf has to be sinusoidal also. you might be able to do it without position sensors though, if you know the time since the last commutation you can calculate in software how fast the motor is rotating, and using that speed you can guess what amplitude to make the output through the commutation step, if the motor doesn't change speed since you measured it it will be right. its likely not worth the extra trouble though. and it will take a whole lot more processing power to create the right waveform and still do everything else.

 

[Miles]

The new Toshiba chip is a sensorless design:

http://www.embedded-control-europe.com/prodnews?cat=2&pid=302

http://www.toshiba-components.com/prpdf/5807E.pdf

 

[dirty_d]

yea thats another way you can do it, if the motor backemf is sinusoidal then you can measured the backemf of the unpowered coil and use the amplitude to determine your output amplitude.

 

[dirty_d]

i just came to a realization here, im not sure if this is already blatantly obvious to everyone, but it just clearly occurred to me. you cant make two motors of equal size and of equal power output and of equal voltage where one is for example 100 Kv and the other is 50 Kv. the high Kv one requires few windings of thick wire with low resistance say for example 0.030ohms. the low Kv motor requires twice as many winding wraps of wire half the size(maybe not exactly half i dunno) with the wire being twice as long. that increases the resistance by 4. so now we have 2 equal motors only with the Kv and resistance different ill use 48V for the rest of this stuff, the max power output from the high Kv motor is (48.0 / 0.030) * 48.0 = 19200W, from the low Kv motor (48.0 / 0.120) * 48.0 / 4.0 = 4800W! 4 times less power than the low resistance high Kv motor, by doubling the voltage to 96V it will give you the same power and performance as the high Kv motor, this means that if you want a small light motor that is also powerful, while being high torque and low speed instead of high speed and low torque like rc motors you need to run it at high voltage and low amperage.

this explains why direct drive hub motors are so big, in order to get high power and torque with low voltage, you need both low resistance and low Kv this takes many turns of thick wire, or a larger number of poles, these both take up space and mass. geared motors are smaller because they have higher Kv and just use gears to lower the rpm and increase torque. another good example is the PMG-132 motor its 24 pounds and has almost the same power output as that HXT motor, the HXT being 130Kv and the PMG being 50Kv and both operating at similar voltages

besides the chance of electrocuting yourself, higher voltage and lower amperage seems to be an advantage, losses in the controller are mostly due to I^2*R right?, i don't think increasing voltage increases the losses much, but reducing current by 2 reduces resistive losses by 4. the battery pack would require a larger number of smaller Ah cells for the same capacity, it would probably complicate the bms, but if a cell goes bad its cheaper to replace.

EDIT: wait actually halving Kv and doubling voltage would give you the same torque/speed wouldn't it? damn you universe. so what is it just inescapable? if you want more torque you need a bigger motor?

 

[eP]

EDIT: wait actually halving Kv and doubling voltage would give you the same torque/speed wouldn't it? damn you universe. so what is it just inescapable? if you want more torque you need a bigger motor?

Torque don't depend on Kv.
Torque depend on motor's dimetions/geometry, poles count, air gap length, magnets, core's permeability.

Kv depend on all above + turns serial/parallel config. So for any given motor geometry you can get Kv you want by right choice of winding configuration.

 

[Miles]

if you want more torque you need a bigger motor?

Yup!

 

[recumpence]

I have huge torque with my reduction setup. Reduction complicates the drive system, though. But, for a bike I plan to keep long term, the complication is worth it.

Matt

 

[eP]

I have huge torque with my reduction setup. Reduction complicates the drive system, though. But, for a bike I plan to keep long term, the complication is worth it.

Matt

More narrow magnetic poles and different winding scheme could also act the same way like reduction.

 

[dirty_d]

Torque don't depend on Kv.
Torque depend on motor's dimetions/geometry, poles count, air gap length, magnets, core's permeability.

but Kv depends on motor's dimensions/geometry, poles count, air gap length, magnets, core's permeability. + windings so its the same deal

torque does depend on Kv when you are talking about motors that operate at the same voltage and output the same power. for a 100Kv motor that has a max power of 5000W and a 50Kv motor that is also 5000W that operate on 48V the first will produce 5000W at 2400rpm@19.9Nm, and the second at 1200rpm@39.8Nm. and the second motor will have to be bigger. i hope im wrong though, if someone can tell how to make two motors the same size and weight that output the same power except one being slow and torquey and the other fast and little torque without gears let us know!

 

[eP]

Torque don't depend on Kv.
Torque depend on motor's dimetions/geometry, poles count, air gap length, magnets, core's permeability.

but Kv depends on motor's dimensions/geometry, poles count, air gap length, magnets, core's permeability. + windings so its the same deal

torque does depend on Kv when you are talking about motors that operate at the same voltage and output the same power. for a 100Kv motor that has a max power of 5000W and a 50Kv motor that is also 5000W that operate on 48V the first will produce 5000W at 2400rpm@19.9Nm, and the second at 1200rpm@39.8Nm. and the second motor will have to be bigger. i hope im wrong though, if someone can tell how to make two motors the same size and weight that output the same power except one being slow and torquey and the other fast and little torque without gears let us know!

Torque don't depend on Kv as you are able to wind the same motor for 2 different Kv versions but at the same torque.
These motors cannot have the same max power for the same voltage of course, but could have the same torque max.
So Kv don't depend directly on torque max motor capability, Kv depend mostly on winding scheme.

 

[dirty_d]

ep, right i agree with you there, i was just using the case where the voltage of each motor is the same. if the voltage is different for two different motors of different Kv then the torque can be the same.

 

[recumpence]

I have huge torque with my reduction setup. Reduction complicates the drive system, though. But, for a bike I plan to keep long term, the complication is worth it.

Matt

More narrow magnetic poles and different winding scheme could also act the same way like reduction.

Hmm, well, my AXI is a 14 pole motor. My new Plettenberg is a 20 pole motor. So. it should have more torque and it should run smoother at low RPM with high ploe count.

Am I correct, or am I missing something?

Matt

 

[eP]

Hmm, well, my AXI is a 14 pole motor. My new Plettenberg is a 20 pole motor. So. it should have more torque and it should run smoother at low RPM with high ploe count.

Am I correct, or am I missing something?

If all others dimentions are the same, magnets strength, etc it should be the true (more torque for more poles)

 

[dirty_d]

the Kv for the plattenberg is 215 rpm/V that translates to 0.044 Nm/A, the AXI is 225 rpm/V which translates to 0.042 Nm/A, the terminator will have a little more torque if the current limit is kept the same, and i would think it will be smoother since it has more poles, the AXI has 42 commutation steps per mechanical rotation, and the terminator has 60.

does it mention the winding resistance of the terminator? from the data on the site i calculated 80 milliohms, which seems high.

 

[recumpence]

Oh,the Terminator is larger diameter as well. The AXI is 63mm diameterm while the Plettenberg is 77.5mm. So, the increased diameter should increase torque also.

For what this thing costs, it had better be phenominal!

Matt