The AC Induction Motor Ebike Project

[safe]

The Low Voltage Surprise

I was just browsing the internet looking at AC Induction motor projects and came upon this one:

http://www.speedace.info/solar_van.htm

...it seems that they had found some little company in Italy that makes LOW VOLTAGE AC Induction motors:

http://www.bestmotor.it/frameset_ei.html

Apparently you can get what should be fairly high efficiency AC Induction motors that run at voltages that would be appropriate for an ebike. So maybe the trick is in the motor winding. Just like with the brushed and brushless motors, if you increase the thickness of the copper winding wires you can run less voltage and get the same result.

Maybe the trick might be as simple as drilling out the stator somewhat to make room for more copper and rewind the motor with thicker wires? (the extra copper adds some weight, but not that much)

That would be great...

 

[safe]

Curtis Low Voltage AC Induction Motor Controllers

http://curtisinst.com/index.cfm?fuseaction=cProducts.dspProductCategory&catID=8

http://curtisinst.com/index.cfm?fus...6924-E7F3-0B66-A1A1352AFC5AC92F&SearchString=

Q: Can I use an industrial motor work with the Curtis AC controller?

A: No, the industrial motors are wound for high voltage operation. The AC induction motor controllers designed for material handling applications are wound for low voltage, typically in the 15 - 50 V rms range. We work closely with motor suppliers and can help you to locate a suitable motor.

...so apparently the focus should be on the motor and not the controller.

Curtis would not create a controller for an idea that was implausible. Obviously low voltage AC Induction motors are in fact possible and are in actual use for DC battery powered systems like forklifts.

It might not be necessary to increase the base voltage above 48 volts.

 

[safe]

The Electric MotorSport Example

When I first saw this I didn't look closely at how they did it. Apparently they custom rewind the motor for lower voltage:

http://www.electricmotorsport.com/store/ems_ev_parts_motors_ac-induction.php

These custom wound AC induction motors can produce an amazing amount of power and torque. Some of the benefits of an AC induction motor are a longer power band, higher RPM, regenerative braking, and easy reversing. This is the motor used in the E1 project. This motor can hot rod any electric motorcycle, ATV, neighborhood vehicle, go cart, or trike.

...as it turns out the Curtis Model 1236 AC Induction motor controller works at from 36 to 48 volts.

So again... it's all about custom rewinding the motor...

 

[spad4me]

This is what I liked about this concept in the beginning.
A 48 to 60 volt three phase motor.
Rewound to perform well at lower than factory voltages with an inexpensive controller .
That can probably run all day. on a sip of power.
That requires a 1000 watts .or less

 

[safe]

Higher voltage is always better, but if you are restricted to a lower voltage it does seem possible to adapt. I guess the efficiency losses due to stepping up the voltage would basically erase any efficiency gains you might make with the motor at the higher voltage. So if it's essentially the same either way then all it's going to take is to get the correct rewind. Unfortunately I have no real baseline to start with, so I'll have to just make a blind guess and then expect the probable need to rewind again once I know where things are going. It can take a while to discover all the right things to do.

 

[safe]

The Two Chip Solution?

Simplification is always nice.

If I can get a chip to do AC PWM control logic and then attach that to a Bridge Driver to actually deliver the current then aside from some peripheral issues that is all that is needed.

The only worry I have with this Bridge Driver is that it's rated as 600 volts but only 10 phase amps. I wonder if the 20 amps that is drawn from the battery is too much for a 10 amp phase current? It seems like this might be a close fit... but heck... it's simple and cheap enough to want to try it.

 

[safe]

Two Chips Might Be For Real...

This idea of only needing two chips to pull off the controller for an AC Induction motor is gathering steam for me. I'm finding all kinds of integrated chips of differing capabilities that exist for the task. It makes sense, something as common as Three Phase motor drive is going to have a market for custom parts that simplify development. And since the original market is for things like air conditioner units the marketplace is large. (much like the RC market being a cross over for ebikes)

Here's one that looks about right... it will allow 20 amp currents and would allow me to scale up to 100 volts if needed.

This sort of thing is very good when you imagine making an actual product out of this idea. If the controller can be assembled quickly and cheaply and with few parts and the AC Induction motor is cheap to build and durable it does provide a more realistic base than permanent magnet motors, especially of the brushed variety. You could make millions of these and there is no need for precious magnet materials. The ultimate WalMart product.

RC Brushless Motor People!

How about up to 600 volts and 200 amps for an RC motor? (160 hp)

http://www.mskennedy.com/client_images/catalog19680/pages/files/4851re.pdf

 

[safe]

Apparently the idea of using a single MOSFET bridge as an integrated circuit has been discussed over at Endless-Sphere a couple of months ago. Their observation was that when it comes to increasing power levels (stressing the MOSFETS) that the individual MOSFET has better cooling behavior. That does sound logical.

However...

I'm not involved in the "arms race" of increasing power levels and am desigining around a fixed 1000 watt limit, so I don't think that those issues apply to my case.

They priced out some of the products they could find and found the prices are about the same either way. The big advantage of the integrated unit is simplicity and ease of assembly, so it's more of an issue of how one manufactures the product efficiently than anything else.

If I can find the right Three Phase Bridge IC at the right price I'll go with it, but if not, I can always go manual with six individual MOSFETS and build it from scratch. (for the prototype)

http://endless-sphere.com/forums/viewtopic.php?f=2&t=11246&sid=9da9588af3ee969e52fdb934704e60ed

 

[safe]

"Slip" (slop) should mean better "Start"

The Brushless DC Permanent Magnet motor is in many ways more similiar to a Switched Reluctance or Stepper motor than to the AC Induction motor. With the Brushless DC Permanent Magnet motor the starting needs to be done very carefully because the timing of the pulses is dependant on feedback from the motor. (backEMF signal)

AC Induction motors can be run "blind" using the open loop VF (Voltage vs Frequency) control scheme. When the AC Induction motor starts up it just starts spinning the magnetic field around the iron rotor and "eventually" enough "slip" develops so that you get a reaction. Assuming you start with a low enough frequency and are careful to limit current using PWM (which is how it's normally done) the motor should start with fewer problems. There is no synchronization to get right. (this is why they are asynchronous motors)

Basically the AC Induction motor is more "Idiot Proof" by design compared to the other motors that rely on feedback.

"Idiot Proof" is a good thing... I prefer things that can be allowed to be a little sloppy and lower quality in the design because that means they can be made more cheaply and survive more abuse. (the AK-47 machine gun was designed using these principles) This design feature of the AC Induction motor adds to the attractiveness for a low end product. (if it's good enough for your dishwasher it's probably good enough for a cheap ebike)

The AC Induction motor maintains much of the design sloppiness that the DC Brushed motor employs. One does not need to worry about synchronization with the brushed motor because it's mechanically built into it's design. This makes the AC Induction motor in some ways similiar to the brushed motor because of it's lower tech synchronization methods. (in other ways the AC Induction motor is very unique)

One "can" synchronize the AC Induction motor and if you do you can control things like "slip" very carefully, so there is an "upgrade path" with the design as one might want to improve things. I like the fact that you can be at the low end and also scale up to the higher end and it all works with the same hardware.

 

[safe]

First Steps Into The Math...

I'm only at the beginning of creating mathematical models of an AC Induction motor for ebikes so bare with me as I am learning as I go. For Permanent Magnet motors I've gotten the math down 100% and as most recent proof of that I just got done assembling a new motor/gearing combination on my old bike and it delivered the exact performance as expected. But the AC Induction motor uses different mathematics... so I'm just getting started...

Here is my first stab at comparing a Permanent Magnet (PM) motor to an AC Induction (ACIM) motor that is designed with 4 poles so that it's peak torque is at only 1800 rpm. The Permanent Magnet motor has a no load speed a little above 6000 rpm. (which is high for a brushed motor, but probably about right for a brushless motor) Obviously you can reduce the peak rpm for the Permanent Magnet motor and that will improve the low end torque, but then you lose top end speed. This type of a tradeoff is less of a problem with the AC Induction motor because of it's wider powerband.

...the torque is just sort thrown up there as a reference point, so just sort of go with it as something to ponder at this stage. (don't obsess on the details at this point)

The basic idea is that the AC Induction motor builds it's power earlier because of the 4 pole design... it then continues to deliver power for a long time until it runs out of torque. The comparable Permanent Magnet motor is actually stronger on a power-to-weight ratio basis, but given that we are (suppose to be) dealing with a FIXED power limitation our concern is less about "peak" than "average" power.

So as long as you sort of twist your mind towards the idea of "best overall legal power" (as opposed to a world without limits) then the idea makes sense.

Repeating once again... for raw and illegal power above the 1 hp limit you want to use Permanent Magnet motors because the math shows they can deliver more for less weight.