RE: [CAD_CAM_EDM_DRO] Re: unipolar or bipolar?

Hi Mariss,

thanks for the wonderful tutorial on steppers.

I was recently designing drivers for my motors, using SGS L297 & 298 and
incorporated the "half-step with shaping" method you mention, where the peak
current rises to 1.41 x the nominal. Now I think I will limit my maximum
current to the nominal value...

However , I think the advantage of using this method lies in its ability to
be smoother than the standard half-step method, because the output currents
are more sinusoidal. Could you confirm this?

I would then just run lower currents but still keep the shaping on the
pulses.

Thanks
Guy Sirois

-----Original Message-----
From: mariss92705 [mailto:mariss92705@...]
Sent: Thursday, February 21, 2002 11:41 PM
To: CAD_CAM_EDM_DRO@yahoogroups.com
Subject: [CAD_CAM_EDM_DRO] Re: unipolar or bipolar?

Snip---
There is a common misperception some users make regarding current set
with microstep drives. It has to do with the fact microstepping
drives the motor with sine and cosine weighed currents. When the
current in one winding reaches its maximum the other winding current
is at zero. From this some surmise that it should be possible to
increase the set current to 141% of rated since the motor power
dissipation would then be identical to a full step drive. That is
correct as far as power dissipation goes but it neglects a very
important point most are not aware of.

The magnetic flux path is not shared by the windings. This means the
current in one winding does not contribute by adding or subtracting
to the to the magnetic flux of the other winding. Though the
dissipation may equal a full step drive by running the current at
141% of rated, the iron magnetic saturation is dependent on the
current of one winding entirely.

The result is the iron in the motor saturates at the higher current
and degrades the motor's linearity. Consequently the microstep
placement is not optimal anymore and the motor will exhibit greater
low speed resonance than it otherwise would have.
Snip---