[Joe Ross]

Originally Posted by Joe Johnson

As to claims of linearity, that depends on what you mean. The output may actually be very linear, but the system need not respond in a linear way.

My experience in automotive systems is that when you go from low PWM freq (100-400Hz) to higher PWM freq (10-20kHz), the motors do not respond the same, especially at low duty cycles.

For example, if, in order to get a system moving, we needed say 20% duty cycle at 100Hz, we would need to bump the duty cycle up to 30-40% at 10kHz PWM freq.

This was very puzzling but we eventually attributed this our mechanical system's time constant. At 100Hz, the mechanical system's time constant was such that the system had time to react during a single PWM pulse -- These "full on" pulses provided a kick to get the motors and gears turning. At 10kHz, the system can only react to the average not the individual pulses, so it required more on time to get things rolling.

Through the private beta test channels, I asked these types of questions. They just released some test data publicly, which show a very nice linear response without any slow starting when driving a CIM motor. http://forums.usfirst.org/showthread.php?t=10182