Re: [CAD_CAM_EDM_DRO] brushless servos without hall commutation

bluecoast1 wrote:

>Hi,
>
>I have a number of brushless servos I've collected to build some cnc
>machines. All of them have optical encoders of some resolution, but
>none have hall-effect sensors for commutation.
>
>What I am wondering is if there is any reason (I can't think of any) why
>I couldn't divide the optical encoder output with a PIC or something
>to get a faked hall-effect commutation output I could use for a
>generic controller.
>
>

The problem is that when the drive powers up, it has to be able to determine
the position of the motor's rotor magnets without being able to move the
rotor, or at least without having to move it very far. You wouldn't want
to have to turn the motor up to a full turn to find the index mark.
Many newer
drives use a form of triple-quadrature, where all 3 tracks on the encoder
have transitions of the same, or nearly the same, frequency. A and B
are the standard quadrature scheme, and the phase relationship of the
3rd track tells the drive immediately where it is within a revolution.
That way, the drive's commutation logic can immediately tell which
transistors to turn on for each direction. Without this, you need to at
least
pulse coils somewhat at random and see which direction of motion you get
to set up the commutation.

Jon