The AC Induction Motor Ebike Project

Discussion in 'Electric Bicycles' started by safe, Aug 29, 2009.

  1. safe

    safe Active Member

    The AC Induction Motor Ebike Project

    After much searching for an AC Induction motor that would be suitable for an ebike I arrived at the older model Frigidaire Dishwasher motor. Apparently they made a newer one that was a little different, but looked to be bigger. I like this one because of the smaller dimensions.

    This thread will track the progress of this project. (the motor arrived today)

    The dimensions are more of a "pancake" than the Unite motors I've been using and the width is about 3.75" and the diameter of the shell is about 6" on the inside and 5" on the outside.

    I don't know the weight exactly, but it feels about the same as a Unite motor... somewhere around 10 lbs. A 10 lbs motor is okay (in my opinion) if it can deliver the perfect flat powerband over a large rpm range.

    The central goal here is to create the "ideal" LEGAL powered motor that delivers a flat 750 watts (or slightly above) across the widest possible range of rpms while the motor is under full load. It's effectively a "race motor" for an ebike racing class that I want to invent and promote someday.

    This motor will be EBRR "Electric Bicycle Road Racing" Legal. (which means the power input will be strictly restricted to a maximum of 1000 watts of power with no exceptions)

    The racing class will be called "Formula 1K".

    And so the project begins...

    Attached Files:

    Last edited: Aug 29, 2009

  2. safe

    safe Active Member


    This older motor was a Single Phase and not a Three Phase motor. Unfortunately they didn't make all the grooves the same size as they used what appears to be a capacitor starting technique and some smaller starter wires to get the thing going. Not sure how I'll adapt to that... otherwise the wires came out really easily and it looks like a pretty good motor to start with. The rotor is made of pressed iron plates like the stator and there is a slight rotational skew to them in order to reduce ripple currents. (all to be expected) With the very open design the airflow looks to be very good, so cooling should happen easily.

    The wire thickness is slightly thicker than 18 AWG wire and the larger grooves held 33 turns of that wire.


    The Single Phase motor is listed as 3450 rpm for 60 Hz.

    Doing the math we would get:

    Single Phase (2 poles, 4 grooves needed) : 3450 / 1 = 3450 rpm

    Two Phase (4 poles, 8 grooves needed) : 3450 / 2 = 1725 rpm

    Three Phase (6 poles, 12 grooves needed) : 3450 / 3 = 1150 rpm

    Four Phase (8 poles, 16 grooves needed) : 3450 / 4 = 863 rpm

    Five Phase (10 poles, 20 grooves needed) : 3450 / 5 = 690 rpm

    Six Phase (12 poles, 24 grooves needed): : 3450 / 6 = 575 rpm

    ...and there are exactly 24 grooves available. If you wire the motor as a Three Phase then you double up on the grooves with one overlapping the other on both sides.

    The advantage of a Six Phase controller is that you can eliminate the need for a geardown unit. You could take 575 rpm and a smaller sprocket like about an 11 tooth and then match that with a front chainring of about 55 teeth and you get the proper final gear relationship.

    Hmmmm.... more complexity in the circuits to do Six Phase, but it might produce a really nice motor that way. Will have to research more the positives and negatives of higher phase motors. Three Phase is clearly the easier path to take.

    Attached Files:

    Last edited: Aug 29, 2009
  3. safe

    safe Active Member

    Notes About "Multi-Phase" AC Induction

    Due to the potential benefits resulting from the use of a phase order higher than three in transmission, some interest has also grown in the area of multi-phase machine. For machine drive applications, multi-phase system could potentially meet the demand for high power electric drive systems, which are both rugged and energy-efficient. High phase number drives possess several advantages over conventional three-phase drives such as: reducing the amplitude and increasing the frequency of torque pulsation, reducing the rotor harmonic currents, reducing the current per phase without increasing the voltage per phase, lowering the dc link current harmonics, higher reliability and increased power in the same frame. The high phase order drive is likely to remain limited to specialized applications where high reliability is demanded such as electric/hybrid vehicles, aerospace applications, ship propulsion, and high power application where a combination of several solid state devices form one leg of the drive. The research has been underway for the last two decades to investigate the various issues related to the use of multi-phase machine as a potential alternative to the conventional three-phase machine. This paper, therefore, reviews the progress made in multi-phase induction machine drive research and development since its inception. Attempts are made to highlight the current and future issues involved for the development of multi-phase induction machine drive technology for future application. it appears that if I could pull off Six Phase power then I would have less to worry about as far as things like heat in the motor because less current is required per wire to get the job done.

    It's kind of like deciding if I want a three cylinder motor or a six cylinder. We all know that as you increase the number of pistons that the overall system becomes smoother, however, the complexity also goes up. Each piston becomes less stressed while the overall power is still higher.

    This might solve the smaller groove problem because I could just wind all the grooves to the smaller size and leave extra space in the bigger ones.
    Last edited: Aug 29, 2009
  4. safe

    safe Active Member

    Fitting To The Bike

    It does look like the width is good and the size can be handled within the space I've already built into the frame.

    I'm planning on using this motor on the new bike (as in my avatar) and not my old bike which will still get a brushed motor for now.

    Attached Files:

    Last edited: Aug 29, 2009
  5. safe

    safe Active Member

    Actual Information

    Emerson Electric Company

    Part Number: S447301

    Model: S62NXHMP-7198

    HP: 1/3

    Volts: 115

    Single Phase

    Frequency: 60 Hz

    RPM: 3450

    Amps: 6.2

    Temperature Rating: 40C

    Manufacture Date: 5/1988


    So this is an old Single Phase motor that seems to be about the right size as a foundation, but there's no way you could run it stock for anything. Here's another for $29:

    ...had I to do it over again I might have taken a look at the newer version that is a true Three Phase from the start because then the stator grooves would already be the same size. (a problem I have to deal with on this one)

    It's also possible that the Frigidaire Dishwasher motor was originally a Single Phase and the replacement is still Single Phase and I got my information wrong somewhere along the way. But I will say that upon taking it apart it does look like the situation is workable. I knew it was going to involve a rewind before I started.
    Last edited: Aug 29, 2009
  6. spad4me

    spad4me Member

  7. safe

    safe Active Member

    It's probably a good motor, but 17 lbs is a "non-starter" in my opinion for actual ebike use. That's about double the weight of the motor I'm using. Are you going to rewind it to run better at lower voltage or are you going to step up the voltage? Rewinding is better I think because you lose some efficiency every time you step up the voltage and 240 VAC would require some serious stepping up.

    However, I'm glad you are doing a parallel development effort and have a good knowledge of this stuff because we will be able to share information about our struggles.

    I'm going to give the Six Phase concept a try because I want the ability to reduce the motor rpm by half so that I can eliminate any geardown unit. Being able to drive the chain directly will reduce complexity and improve efficiency because that's one less source of friction.

    Since I'll be building my own controller it will make for a good winter project. I'm going to use a generic PIC and then I can program the algorithm myself. There are some techniques that you can get away with because of Six Phase power since the many phases allow more overlap. You can do what they call "Flat top" algorithms where you only open the MOSFET's when the curve is nearing the top. All the midrange (0.5 or less amplitude) is essentially wasted energy because the angle is bad so you can leave the MOSFET's closed during those areas. This would allow me to run a leaner motor which should run cooler. I've been reading a bunch of AC motor theory stuff... there's a lot out there.

    When it comes to getting the motor rpm matched to human pedal power rpm your choices are geardown units (which are complex) or more sophisticated controller algorithms. I'm taking the software approach this time so that I can avoid the mechanical approach. (my home built geardown unit took six or seven rebuilds before I got it to work reliably)

    With the Induction motor I can abandon multispeed gears without sacrificing anything, so that's an another reduction in complexity.

    So are you going with the "custom" Three Phase chip that does the algorithm?

    If I knew of a Six Phase "custom" chip I'd buy it, but I think I'm probably on my own having to program it myself. I used to be a computer programmer, so it shouldn't be too hard. I already know "C" and did assembly language programming back in college.

    Attached Files:

    Last edited: Aug 30, 2009
  8. spad4me

    spad4me Member

    I have to see where this takes me.
    I have since found many different motors some better some worse on weight .
    I believe that I will use a ballbearing 1 hp 3000 + rpm 48 to 60 volt three phase motor.
    The final weight is expected to be around 9 pounds.

    I am going for sunset tech not sunrise tech all of the surprises have been removed for me allready
  9. safe

    safe Active Member

    More Poles, Less Phases?

    I was reading some more about the windings possible and came to realize that 2 poles per phase is not the only way to do it. You can actually go to 4 poles per phase in the windings and that way I could use simple Three Phase control (that custom chip) and save myself a lot of trouble.

    The more poles you use the better because the formula is:

    Motor Speed = 120 * Frequency / Pole Count


    It's funny when you learn this stuff over the internet because multiple solutions can be found and each one has it's plusses and minuses.

    The goal is to get the lowest rpm motor with the least effort.


    PHASES: As the name implies, three-phase motor stator windings will consist of three separate groups of coils called phases. Phases must be displaced from each other by 120 electrical degrees. Phases must be electrically balanced (contain the same number of coils) and connected for the same number of poles. Phases will always be designated as A, B, and C.

    POLES: Poles refer to a coil or group of coils wound and connected to produce a unit of magnetic polarity. Poles are referred to as either north or south. The number of poles a stator is wound for will always be an even number and refers to the total number of north and south poles. For example, a four-pole motor will have two north and two south poles alternately spaced around the stator.
    Last edited: Aug 30, 2009
  10. spad4me

    spad4me Member

    Your emerson motor could be considered a two phase motor .
    The single phase power is shifted by the capacitor by ninety degrees to start the motor.

    See what I mean by sunset tech a motor re winder has long ago solved your problem.
    Last edited: Aug 30, 2009
  11. safe

    safe Active Member

    It was a 2 pole Single Phase motor before I removed the wires.

    This process will be like learning to rewind the DC brushed motors... I will eventually figure out how the windings need to be, the thickness of the wires, the number of turns and eventually I'll focus on better and better results.

    It's a "process" that's for sure.


    Here's the "enhanced" way to figure out the true speed. You actually divide the phases from the pole count. Interesting.

    Attached Files:

    Last edited: Aug 30, 2009
  12. Proudest Monkey

    Proudest Monkey New Member

    Ok I gotta admit...I'm impressed..

    But how come noone's talking about miniature reactors?
    If they can put em in submarines....
  13. spad4me

    spad4me Member

    The lead shielding would weigh too much.
  14. Proudest Monkey

    Proudest Monkey New Member

    No I was thinking more along the lines of a Russian design....less shielding...but could probably only be ridden for 1-2 minutes at a time before it rads you out..

    I just put a CB with 5' antenna on my's possible...
  15. spad4me

    spad4me Member

    Safe, I just found a battery thread that shows your battery setup.
    How long does it take to charge the twelve sets of 20 nicads.
    Have any failed yet?
  16. safe

    safe Active Member

    The batteries have worked flawlessly. They charge up in about an hour depending on how drained they were. My first set on the older bike went for a time before I got the random and expected flawed cell. No problem, pop open the battery and like a flashlight I just replaced the problem cell.

    It's so good that it's kind of puzzling that it's not done by others yet.

    The idea doesn't work for anything but NiCad or NiMh though... so since everyone is all excited about LiFePO4 these days you can't use it for that.

    NiCads are indestructable and can last for a decade...
  17. safe

    safe Active Member

    Taking A Closer Look...


    This is very important and worth a second look. Most of the documentation about AC motors tends to use the term "poles" very loosely and I think I got the wrong idea about the relationship between "phases" and "poles" at first. The main thing is to realize that it's the number of subdivisions within the total poles available that defines the motor rpm.

    So the EXTREME would be to use a Single Phase motor and wind it as a 12 pole design. The motor speeds that you would expect would be:

    Low Throttle:
    Single Phase with 2 Poles = (120 * 10) / (2) = 600 rpm
    Single Phase with 4 Poles = (120 * 10) / (4) = 300 rpm
    Single Phase with 12 Poles = (120 * 10) / (12) = 100 rpm

    High Throttle:
    Single Phase with 2 Poles = (120 * 60) / (2) = 3600 rpm
    Single Phase with 4 Poles = (120 * 60) / (4) = 1800 rpm
    Single Phase with 12 Poles = (120 * 60) / (12) = 600 rpm

    Low Throttle:
    Three Phase with 2 Poles = (120 * 10) / (2) = 600 rpm
    Three Phase with 4 Poles = (120 * 10) / (4) = 300 rpm

    High Throttle:
    Three Phase with 2 Poles = (120 * 60) / (2) = 3600 rpm
    Three Phase with 4 Poles = (120 * 60) / (4) = 1800 rpm

    ...based on the gearing that an ebike needs the Three Phase with 4 Poles would be okay:

    Pedal Speed - From 0 to 100 rpm

    Motor Speed (11-55 sprockets) - From 60 to 360 rpm the pedal power would cover the lower rpms best and then the motor kicks in strongest to take the motor above 100 rpm. Pedal power would be your "extra boost" to get things going from a standstill.


    It's just like with rewinding DC brushed motors, depending on the way you wind them you get different behavior. The behavior we want for ebikes is a motor that develops it's power at lower rpm so that the gearing can be accommodated more easily. The Single Phase motor has a problem in that it cannot guarantee to start in the right direction and the (rare) Double Phase cannot guarantee starting either. So from a practical standpoint you need to go with a Three Phase. If you could increase the number of grooves in the stator from 24 to something like 48, then you could have more options. The standard stator uses:

    24 grooves
    12 poles

    ...which can be wound as:

    2 pole, 4 pole and for Single Phase only 12 pole.
    Last edited: Aug 31, 2009
  18. spad4me

    spad4me Member

    You have it .
    The starting problem can be overcome by shading one pole, a thick conducting wire at an edge of the rotor., or just pushing the bike then goosing the throttle.

    There is delta which is phase--- a to b, b to c, c to a. Then double delta just parallel.
    Wye which is ,a phase to ground, b to ground and, c to ground . In this case ground is a common attachment point that ensures a stable refrence . A stable ground is not needed here. There is also double wye .
    The oddest setup is called a schnidt transformer it is dangerous in that one of the legs has an unusually high voltage which by law is marked with an orange band where ever this voltage is present in a junction box.
    I recommend a delta wiring solution over all others

    I was looking at batteries and of course the iron based liithium batteries came up. two thousand dollars is just too much I believe that your nicad based solution will work just as well.
    Last edited: Aug 31, 2009
  19. safe

    safe Active Member

    Delta produces the lowest rpm right?

    Probably the main thing to have learned (for me) is that the Phase count is less critical than the way the poles are connected together. The way they usually explain the poles is not very clear in many cases. I'm glad I found that image because that cleared it up pretty quickly.

    I don't think that it's even necessary to go to Single Phase just to get the chance to go 12 Pole. I think that 4 Pole and Three Phase should cover it.

    From above:

    Pedal Speed - From 0 to 100 rpm

    Motor Speed (11-55 sprockets) - From 60 to 360 rpm
    Last edited: Aug 31, 2009
  20. safe

    safe Active Member

    Single Phase, Pedal Start, Regen?

    If you eliminated the rear freewheel and just used a single front freewheel then the chain would always be connected to the motor. You then create a "pedal first" setup where you cannot use the motor until you pedal first to get the rear wheel moving in the right direction. Since the chain is bound to the motor all the time you can now do regen on the rear wheel. If you cut the throttle to zero then there is no backEMF because induction motors only create a magnetic field if they are dealing with a preexisting field, so if you shut the throttle completely it allows the motor to freewheel. Ramp the power down and you would get the full regen until the field collapsed.

    That's all possible isn't it?

    It would allow for a very, very low rpm motor because a 12 Pole would produce:

    Single Phase with 12 Poles = (120 * 10) / (12) = 100 rpm
    Single Phase with 12 Poles = (120 * 60) / (12) = 600 rpm

    Pedal Speed - From 0 to 100 rpm

    Motor Speed (11-55 sprockets) - From 20 to 120 rpm

    ...that's almost TOO SLOW for the motor. :grin5:
    Last edited: Aug 31, 2009