Re: [CAD_CAM_EDM_DRO] Xylotex or Gecko step driver for my Mill/Drill?

Roy J. Tellason wrote:

>On Saturday 15 May 2004 11:05 pm, Jon Elson wrote:
>
>
>
>>>but I ruled that one out because it's a unipolar only board.
>>>
>>>
>
>
>
>>Good move, as most unipolar drives are very poor performers at higher
>>speeds
>>
>>
>
>Why is that?
>
>Would it have something to do with a bipolar setup energizing coils in the
>reverse direction rather than just turning them off?
>
>
>

Yup, that is basically the problem. With the typical unipolar drive,
the power
source is connected to the center tap, and transistors ground one end or the
other of the winding. When you ground one end, there is still circulating
current in the other half, fighting the field you are trying to build in the
newly powered winding. This energy has to go somewhere, and it ends
up being burned off in some diode-resistor arrangement. It is a hard
topology to harvest this energy constructively. If there is no resistor,
the current in the shut-off winding dies very slowly. If there is a
resistor,
then it gets quite hot at high step rates.

With a bipolar drive, the power source is applied across the whole winding,
and is reversed to reverse the field. The field already in the winding
opposes the power source at the moment the transistors are switched,
and the reversed voltage across the winding collapses the existing field
and quickly builds it again in the opposite direction. The energy mostly
ends up back in the power supply, so there are no big resistors to burn
off the left-over field.

There are ways around this, but they are generally more complicated than
the bipolar drive scheme. The only advantage to unipolar is it saves a
4 transistors per motor.

Jon