Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Motor windings

  1. Nov 17, 2016 #1
    I learned that brushless DC motor windings have an algorithm for winding depending on several variables....
    When I google image search BLDC stator, I expect to see this :
    images?q=tbn:ANd9GcRJWHTCyyHyJCTH3fqAlYhYRU6gEkEWE5qXUYs9dOjZwCh_OhaD.jpg
    But I see a lot more of this :
    BLDC-Motor-Winding-Line-Stator-Winding-Machine.jpg

    Is this second one still a brushless DC? If so, do they just lose efficiency by winding it this way instead of putting 1 coil across multiple stator teeth? The first picture, with the coils spanning multiple teeth... this is the way I learned it. But the second way definitly looks easier to manufacture.

    So yeah, is picture 2 still a brushless DC?
    How can I figure out the difference in efficiency between the different winding methods?
     
  2. jcsd
  3. Nov 18, 2016 #2
    You are correct - there are many factors. I am going out on a limb, but I am thinking that the top on is more for continuous rotation applications ( like speed control & high efficiency) and the bottom like a servo (positioning control)
     
  4. Nov 18, 2016 #3

    Hesch

    User Avatar
    Gold Member

    Yes it is:

    In the first picture, the coils are distributed, perhaps to form a sinusoidal magnetic field along the periphery.
    In the second, the coils are forming a trapezoidal magnetic field, which is intended as for brushless dc-motors, because the controller will be more simple.

    maxresdefault.jpg

    In the above picture you see that the position of the rotor is measured by simple hall sensors ( 6 states ).
    If you had a sinusoidal commutation, an encoder with much higher resolution would be needed, to follow the sinusoidal shape and keep track of cos φ, which implies a more complex controller with sin, cos tables. Otherwise the motor will not run stable.

    With a trapezoidal commutation, you can switch the phase within a range ( 60° ), where in the torque will be constant.

    The trapezoidal commutation were preferred at a time where processors were much slower ( Intel 8080 ).
    Today the speed of processors is not a problem, even with cos/sin calculations.
     
  5. Nov 18, 2016 #4
    Thanks for replies. Yeah, the algorithm I have winds them over multiple teeth for trapezoidal or sinusoidal control. I think with the intention that if you want a sinusoidal magnetic field you need a sinusoidal shaped magnet (it is curved at its edges as opposed to straight out radially)

    I was wondering if I could find a way to put numbers on either of these winding systems. What would the math behind them be... anyone know even where to look for this information?
     
  6. Nov 18, 2016 #5

    Hesch

    User Avatar
    Gold Member

    No, it's much easier ( much more robust and precise ) to distribute the windings.
    When you move a winding from around one tooth to another, you are moving a "step" on the sine wave. In this way you can form/build a nearly perfect sine wave.

    These steps ( higher harmonics ) are removed by placing the magnets skew on the rotor ( not in parallel with the axis ), smoothing the steps.

    A motor with sinusoidal commutation is a synchronous motor, and if you have one with permanent magnets, try to place a nail on the rotor. It will find a skew position.
     
  7. Nov 18, 2016 #6

    Hesch

    User Avatar
    Gold Member

    You will need a numerical simulation program. The manufacturer of such motors has one, I promise.

    Having coming up with a suggested number of windings, the manufacturer will test the motor ( measuring torque(ω), presence of harmonics in the back emf, etc. ). Then move some windings, replace the magnets, test again.

    But you may drive the motor mechanically by another motor ( include a lot of inertia, so that the motor will be running smooth ) with some widings in the stator, measure/analyze the back emf, change the windings, test again, until the back emf is satisfactory.

    At last, replace the test wire with wire with as much cross section area as possible.
     
  8. Nov 19, 2016 #7
    The most common PM BLDC motor is optimized for a trapezoidal drive and that's why you see more examples as shown in the second photo. There are sinusouidal BLDC motor controllers used with motors optimized for it, but electronically controllers are much more complex and expensive. Still there are some applications that call for sinusoidal BLDC motor drives.

    The math involved in calculating motor constants is unwieldy. It can also be hard to find the specifications you need to do it. If you're just doing a project at the hobby level you can get the Kv constant as indicated already, by spinning the motor with a known rpm and measuring the back EMF. Lacking a milli-ohmmeter you can get winding resistance by doing a four wire resistance measurement also known as a Kelvin measurement. Once you have those two constants, you have pretty much what you need to optimize motor setup in terms of stator wind. You can use a ratio of known turns to known Kv to find a new number of turns for a new Kv.

    In the past I've taken an unwound motor and just put a quick two phase wind on it with light magnet wire to get a Kv measurement. From there I can get a turns to Kv ratio in order to arrive at a desired Kv with a specific number of turns. In some cases you can hit the Kv you want more closely by opting for a Delta or Wye configuration. The two are pretty much identical in terms of motor output and efficiency, but Delta has a little bit less copper loss. Resulting Kv values are different for a given number of turns by a factor of root three so it can be a way get a Kv value that would otherwise be more difficult to achieve.

    The physical size required of a BLDC motor depends on how much power output you need. The limiting factors are flux saturation of the stator core and copper loss which results from winding resistance. Heating is a primary limitation and copper loss is a big part of that. Lacking simulation software, which is more often the case than not, you kind of have to wing it there. Copper loss can provide a gauge. The bigger the motor the lower the copper loss since heavier wire can be used in windings. Core saturation puts a cap on power output which goes up similarly with increasing motor size as copper loss goes down.

    The most helpful site I've found for doing PM BLDC motor winds is here; http://www.bavaria-direct.co.za/constants/
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted



Similar Discussions: Motor windings
  1. Motor winding? (Replies: 2)

Loading...