Shop OBEX P1 Docs P2 Docs Learn Events
Propeller as a brushless motor controller — Parallax Forums

Propeller as a brushless motor controller

James LongJames Long Posts: 1,181
edited 2006-12-12 18:24 in Propeller 1
Ok....so I want to use the Propeller as a Brushless motor controller. I know that a stepper is a type of brushless motor.....but I'm referring to a Brushless motor like would be used in an RC car.

The topic I'm looking for......

1.What are my options when it comes to a true sine wave.....would it be advantageous?

2. What problems will I occur with Phasing? Considering real 3 phase power is 90deg out of phase.

3. What speed can I expect with a 3 phase motor? (Spin and Assembly)

4. Any tips you can think of.

Thanks in advance for the input,

James Long

Comments

  • Bryan K.Bryan K. Posts: 47
    edited 2006-12-11 07:35
    I have done some research into brushless DC motors. I actually have some 1/2 hp ones on a few of my projects, one being a barstool racer. I have though about the idea of hacking the current speed controller, and found out the speed controller sends out a square wave output to the motor windng, not a sine wave similar to a three phase AC motor, on which I also have knowledge of. Running a brushless DC motor, similar to the ones used on RC aircraft connect the wires in a certain sequence. For instance, let 's let A, B, and C be the three legs of the motor, you have to sequence the windings. AB then BC then CA...and so on. Your speed is dependent on the switching frequency, if you can sequence the phase windings at a higher rate, then the motor rotates faster. In order to reverse the motor direction, just switch two of the phases. Ill see what else I can pull up from previous research.
  • GavinGavin Posts: 134
    edited 2006-12-11 11:32
    To get a sine wave type drive use pwm on the 3 low side drivers?
    Speed is determined by how fast the 3 phase change, torque by the pwm width?
    I suspect asm might be needed but you could start with spin for low speed.
    Back EMF for detecting position using the A/D conversion methods.
    Nice thing with the Prop is you can dedicate cogs for phase and A/D and dont have to worry about interupt code like normal micros. In theory the code should be easy[noparse]:)[/noparse] It is on my list of things to do with the prop.

    Gavin
  • James LongJames Long Posts: 1,181
    edited 2006-12-11 15:12
    Let me know.....if you ever make the code.



    I'm interested.



    James L
  • LawsonLawson Posts: 870
    edited 2006-12-11 18:08
    short answer: yes the Prop should be able to drive a brushless DC motor

    Long answer: But the devil is in the details!· specifically driving the coils at the correct phase relation to the rotor under variying loads and speed conditions is not easy.· having a seperate rotor position sensor greatly simplifiys the software and circuit, but is harder to install and a bit more expensive than extracting the needed position information from the back emf of each motor coil.· If you want good low speed performance and reliable reversing a sensor based motor controller is the only way to go.· (the back emf signal that a "sensorless" controller relies on is speed sensative, and disapears when the motor stops)

    good luck!
    Marty
  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-12-11 18:37
    Jianwen Shao does a good job discussing the technique of back-emf for brushless motors in his master's thesis: http://scholar.lib.vt.edu/theses/available/etd-09152003-171904/unrestricted/T.pdf

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • LawsonLawson Posts: 870
    edited 2006-12-12 01:48
    WOW that's an awsome source of info!·

    heh, I propose a further improvement on this scheme.· The·good points·off-time sensing and On-time sensing methods described in this paper can be combined.· Since all the outputs of this sensing scheme are digital, opto-isolators or equivelent can be used to decouple the analog signal processing·from the ground referance of the pwm driver.· This would allow letting the analog portions of the sensing circuit to float·and referance to·Vmotor/2 instead of ground allowing directly senseing, with no attenuation, of·the back-emf of the undriven coil during the on-time of the pwm cycle.·

    This is similar to what this => http://www.4qdtec.com/srd.html·circuit uses to controll the mosfets.· the key part is that while the circuit can run off of several hundred volts, with the right mosfet and diode, the portion of the circuit that drives the mosfet never sees more than ~9v.·



    my 2 cents,
    Marty
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2006-12-12 18:02
    Hmmm...
    Seems to me that you could get the Propeller to drive two [noparse][[/noparse]or more brushless motors].
    That would mean you could have two wheels steer or run a twin engine airplane or a twin prop boat.

    I suspect that the sine wave might be easy to synthesize and shift into three phases. After all, it can be done on much less. Generally it is a DAC from one lookup table, but each output starts 120 degrees apart.

    CD brushless 3-phase motors have hall sensors to use for precise speed control. That seems to offer more control than back EMF and many motors can be used as is. The Propeller has the pins and capacity to handle the feedback, so why bother trying to measure back EMF at the zero crossing?

    Either way, the back EMF or hall sensors are only required for variable speed [noparse][[/noparse]not used in airplane motors]

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "If you want more fiber, eat the package.· Not enough?· Eat the manual."········
    ···················· Tropical regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-12-12 18:10
    Kramer said...
    The Propeller has the pins and capacity to handle the feedback, so why bother trying to measure back EMF at the zero crossing?
    Because for certain applications it is pretty much required. Thier current main use is high powered engines for R/C aircraft. Encoders are typically bulky and thier added weight will add to the total weight and throw off the center of gravity. By using back EMF you reduce the weight and are not limited to where it is physically placed so you can place it where it helps maintain the proper center of gravity.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • Phillip Y.Phillip Y. Posts: 62
    edited 2006-12-12 18:24
    One thing that is not defined in Jianwen Shao's Thesis is what the low speed is
    that controlling the motor becomes difficult, probably around 60rpm.

    below this speed the motor behaves more like a stepping motor
    with only 6 steps per turn instead of 48 for a 7.5 degree stepper.

    So speed control is great, but position control or slow smooth motion
    is a problem that usually requires encoder feed back.

    So for the drive motor of a fast RC car great , but not·something like the·Scribbler™ robot .



Sign In or Register to comment.