This is a good question. I understand that Unipolar and Bipolar Half Coil uses the least number of coils active so it doesn't give great low speed torque. Because of the short(er) length of wire involved there's relatively low inductance so the torque lasts at higher speeds because the motor is able to turn faster, quicker for a given power.
Bipolar Series uses two coils in series (also the full coil depending on how many wires you've got) and so has very good low speed torque, higher than unipolar or bipolar half coil. Bipolar Series, two coils in series (or the full coil) has higher inductance than unipolar or bipolar half coil. High inductance means the torque drops off rapidly as speed increases.
Bipolar Parallel uses two coils in parallel ( or the full coil - in parallel !) and so has good low speed torque like Bipolar Series. But the inductance is roughly the same as Unipolar or Bipolar Series because the effective coil length is one coil, the same as Unipolar and half that of Bipolar Series. That means the torque is also lasts through to higher speeds.
The current per phase can be confusing and you need to consider how many/much coils are active in each drive method.
Higher inductance basically means the motor is slower to energise each coil. As the step rate increases the motor is required to energise coils faster, the coils inductance increasingly works against this so the average power used over the time the coil is energised is less and so the torque is reduced. Increasing voltages works to counter inductance and increase the average power used by a coil by quickening the time taken to get to the rated amperage. This is why drivers using resistors to limit current, as opposed to choppers are slower than choppers for the same volts and amps.
As a quantitative measure my machine runs far better with bipolar parallel than the other methods, to the extent that microstepping unipolar lost steps badly. Bipolar series ran well on microstepping but only at slower speeds, lost steps at higher step rates. Bipolar parallel works great with microstepping and with about 1.8 times faster g00's.
As you used a Pacsci motor as an example I'll mention I got best response with mine by having as high a voltage as the driver would stand (48v) and not necessarily the all the rated Amps. Not providing the full amps means theoretically that you're not getting the full torque, but in real terms I found that wasn't much of an issue as having the motor respond faster and not loose steps over a wider range made the machine run better. Bipolar parallel gave a much wider 'sweet spot' operating range. Hope that makes sense. YMMV.  |