The AC Induction Motor Ebike Project

I guess the word "electrocuted" means "being killed by electricity". Anyway, you knew what I meant, I've been "shocked" before and survived.

Back on topic... any comments about AC Induction motors compared to Permanent Magnet motors within a constrained power environment of 1000 watts? Unless I'm mistaken the AC Induction motor offers a far better solution and even though it is slightly less ideal when considering power density (power-to-weight) it does seem the better choice for a racing motor if the "game" is making the most of 1000 watts.

attachment.php
 
Last edited:
More Realistic Comparison

People are bound to ask:

"At what point can the permanent magnet motor out perform the AC Induction motor in the area of low end torque?"

...in order to discover that you need to use a motor that has a lower motor constant so that it builds it's power at lower rpm. In the chart the permanent magnet motor reaches it's no load speed at 2988 rpm.

attachment.php


...you can see that the permanent magnet motor does slightly produce more torque at the lower rpms compared to the AC Induction motor, but then it hits it's no load speed and stops. The AC Induction motors powerband is very wide and it continues to produce power (at reducing torque) up to 4050 rpm and beyond. From what I've read the AC Induction motor can be run at three to four times it's natural frequency in the field weakening area. This would translate to:

1800 rpm * 3 = 5400 rpm

1800 rpm * 4 = 7200 rpm

...so the powerband is very wide.
 

Attachments

  • Comparison 003.jpg
    Comparison 003.jpg
    85 KB · Views: 19,507
Last edited:
Resistance Is Not Futile...

There is a rating system for AC Induction motors called "NEMA" (National Electrical Manufacturers Association) where they classify how motors behave. At first you think:

"Wow, that has to be some kind of complicated thing or something." :geek:

...but when you dig deeper you realize that most all of what causes performance is the amount of resistance that the rotor has. Higher or lower resistance changes the way the motor behaves.

That's pretty simple.

Class A - Least resistance (usually using copper in the rotor)

Class B - More resistance (most standard iron rotors are Class B)

Class C - Yet more resistance

Class D - Like swimming through syrup electronically

------------------------------

This all makes perfect sense because motor "slip" is the magnetic field within the rotor chasing after the magnetic field that the stator is creating. A rotor that "chases" with less resistance requires more magnetic forces to get itself fully charged because lower resistance means less activity in the rotor. A high resistance rotor tends to spike up in it's reaction very quickly, but then saturates earlier.

What's really cool about this is that you can control WHEN the powerband is going to peak... if you wanted to you could design a motor that really screams at high power levels, but that motor will have less starting torque than is possible.

Class B is not a bad compromise for an ebike... the starting torque is "good enough" and there is a nice spike up in the higher power areas. Using simple iron plates for the rotor this is the cheapest thing imaginable. The vast majority of AC Induction motors are of this type because they are so cheap.

Class A has the highest efficiency and the best peak power, so there would be those that would be tempted to use this design as well. It is also more expensive (copper rotor) but will also run cooler. You pay more for these motors.

Class C and D "might" make some sense if you wanted to have really strong low end torque... maybe for an ebike that was limited to 20 mph this might be of value. The torque drops off really fast with these motors. If you look at it as a comparison to permanent magnet motors then the highest resistance cores behave a lot like a permanent magnet motor. (it's almost a straight line drop of torque from the low end to the top of the powerband)
 

Attachments

  • NEMA Class A.jpg
    NEMA Class A.jpg
    47.7 KB · Views: 552
  • NEMA Class B.jpg
    NEMA Class B.jpg
    44.7 KB · Views: 560
  • NEMA Class C.jpg
    NEMA Class C.jpg
    39.5 KB · Views: 571
  • NEMA Class D.jpg
    NEMA Class D.jpg
    47.5 KB · Views: 513
  • Permanent Magnet Torque.jpg
    Permanent Magnet Torque.jpg
    51.8 KB · Views: 529
Last edited:
The Kv Trap

http://en.wikipedia.org/wiki/Brushless_DC_electric_motor

The Kv rating of a design of brushless motor is the constant relating the motors unloaded RPM to the (peak, not RMS) voltage on the wires connected to the coils (the "back-EMF"). For example, a 5700 Kv motor, supplied with 11.1 volts, will run at a nominal 63270 rpm. By Lenz's law, a running motor will create a back-EMF proportional to the RPM. Most ESCs do not boost the battery voltage. Once a motor is spinning so fast that the back-EMF is at or above the battery voltage, it is impossible for those ESCs to "speed up" that motor, even with no load.

I was looking again at the essential differences between the synchronous motors and the asynchronous... the sticking point for widening the motor powerband is the Kv rating for things like the RC Brushless motor. Like it's less efficient counterpart, the Brushed motor, the synchronous Brushless motor still has a built in speed limit.

The AC Induction motor has no limit... you can just keep increasing the frequency and the motor just keeps accelerating, though at reducing levels of torque. This is because for the AC Induction motor voltage has absolutely no relationship to motor speed. (bizarre and hard to imagine, but true) Depending on how the motor is put together you can accomodate just about any voltage. More voltage tends to be better than less from an efficiency standpoint. (saves copper to have higher voltage)

However, one thing does come to mind... other than the Wye - Delta switch concept (which is good) you could also develop an ESC that could have a voltage doubler built into it. Ideally you could run "direct drive" voltage for as long as you could before you hit the no load speed, then electronically upshift through a voltage doubler and then just continue on even higher. That's getting really complicated to do, but it's the only way I can think of to get the wide powerband that I'm looking for using a Brushless motor and stay under the legal power limitations.

It still seems that the AC Induction motor is the way to go... (worth the effort to try anyway)
 
Last edited:
Power-To-Weight Ratio

Much of the fascination with RC Brushless motors of late has to do with the very favorable power-to-weight ratios they possess. However, if you look at some of the more leading edge AC Induction motors out there they deliver performance that even makes these RC motors look tame:

http://www.acpropulsion.com/tzero/index.php

Now, AC Propulsion produces its unique propulsion technology, tzero technology, for other electric cars. The AC induction motor produces up to 220 horsepower but weighs only 110 pounds. It provides all of the tractive effort, and most of the braking, too. Highly effective regenerative braking recaptures kinetic energy from deceleration, returning it to the battery for later use.

Now let's do that math on this...

220 hp = 165,000 watts (165 kW)

Dividing... 220 hp / 110 lbs = 2 hp / lb

So an equivalent RC Brushless motor would need to weigh about one pound in order to equal the power-to-weight ratio of the bigger AC Induction motors in order to get two hp. For the case of the ebike with it's one horsepower limitation you would "in theory" only need a motor weighing one half of a pound!. However... it's also true that efficiency of AC Induction motors drops quickly at about 1 hp, so it's not a direct comparison.

I keep looking at different motors, but the industrial preference in design is to use these massive cast iron frames in order to help dissipate the heat. All that cast iron acts like a huge heat sink... which makes sense on some shop floor, but not on an ebike.

So I'm still at the place of "I just don't know yet"... (will it work well or not?)

-----------------------------------

Thinking in the "larger perspective" the only real difference between the RC brushless motor and the AC Induction motor as far as the design is that the RC motor uses magnets for the rotor. Permanent magnets have plusses and minuses... the plus is that they respond instantly to magnetic fields created by the stator and that means instant and synchronous torque. The negative side is that since the response is directly related to the magnets themselves and the magnets create a backEMF it means there is a fixed top speed to the design.

Switching out the magnets and putting an AC Induction motor solid iron or copper squirrel cage as the rotor means that torque is not glued to the stator fileds, but is asynchronous and "slips" behind all the time. So the negative is that torque lags somewhat compared to the instant response you get with permanent magnets. But the plus is huge... on the positive side the lack of permanance in the rotor reaction means that you can just keep driving and driving the rotor faster and faster. The big advantage of AC Induction motor is powerband width.

So it really comes down to:

:D For peak power within a narrow powerband you use permanent magnets.

:D For the widest powerband you use a solid rotor and induction based fields.

...obviously the power legal permanent magnet motor desperately needs to use multispeed gearing and the induction motor does not. Given the laws about ebikes it kind of pushes us towards the AC Induction motor because the laws are all about peak power and not about how broad the powerband can be.

It's all chaos at this stage I guess... (design is still in a state of flux)

We've come a long way from when the only choice was a hub motor. :cool:
 
Last edited by a moderator:
Strategies For Rewinding

Just like with the brushed motors I've rewound in the beginning I'm sort of operating "without a clue" of how it will turn out. After going through six Unite motors and rewinding them in differing ways I've found out the good, the bad and the ugly about what can be done. Too much power delivered through a little 500 watt motor can end up overheating it causing failure. All the advantages of increased performance are lost to commutator failure. :sick:

For the AC Induction motor that I'm working on the motor started as a Single Phase with 33 turns of what looks like about 18 AWG wire. So let's consider the math and how one might make a guess at how to rewire to get it to work at half the voltage.

The rules for magnetic strength are:

Magnetic Field (B) = Number of Turns of Wire (N) * Number of Amps of Current (I)

6d407268645b13ac03976a66820cf979.png


...in the case of a lower voltage motor we need to reduce the resistance so that current can increase. With increased current you can compensate for a lowered turn count and make the motor work correctly.

As I see it:

33 turns * 18 AWG wire (Single Wind) -----> Maps to the 115 volt Stock design.

17 turns * 18 AWG wire (Double Wind) -----> Maps to a 115 / 2 = 57 volt design.


...now a 48 volt battery is pretty close to this, but it's not perfect.

We could also think of using thicker wire (might be hard to get a tight fit) or smaller wire and more winds. The issue of wire thickness is more about ease of rewinding, but it's the final calculation that can be the most important factor in performance.

My past history has taught me that more copper is better than less... that if you can find a configuration that uses more turns with thicker wire it's going to work better. However, there is limited amount of room in the grooves, so you can't increase the number that much.

In order to switch from 115 VAC to 48 VAC it might be about as simple as going from Single Wind to Double Wind.

Also, the overall design will go from Single Phase to Three Phase power (also from 2 Pole to 4 Pole) and from what I've read that usually increases efficiency from about 50% up to closer to 75%. To get 75% efficiency across a wide powerband is a "good enough" goal for a first pass at this. Most ebikes using brushed motors have efficiency of about 78% at best and that only is valid at a razors edge of rpm range. The AC Induction motor will give it's best performance not based on rpm, but on load... full load means best performance, no load means worst performance. The catch here is that "full load" happens at a razors edge of about 5% slip, so the trick will be in the controller logic... but that's another story for later on. :whistle:
 
Last edited:
AC Induction motors are proven on things like the Tesla Roadster, but at these lower power levels (around 1 hp) it's pretty much known that performance is much lower. My hope is that when you compare an AC Induction motor to a regular brushed motor (78% efficiency and peak and more like 65% average) that it compares well. The ultimate goal is to be able to build a one speed ebike where the low end power (0-10 mph) is mostly done by pedaling and then above that you sit behind the fairing and let the lowered wind resistance make pedaling unnecessary. If all things work (which I'm not sure they will yet) the AC Induction motor could be a very low cost and highly reliable basis for ebikes.

At this point I'm just gathering information so that I have a fighting chance for some measure of success... in no way am I certain yet...
 
Back
Top