Control of brushless ac servo motor


Thread Starter


I currently have a brushless ac servo motor. I am wondering if I can use HCTL1100 to control the motor. HCTL1100 uses pulses to control motor but the brushless ac servo motor uses sinusiodal signals to power it. I've read that brushless ac and dc servo motors are basically the same. So, can I actually control the ac servo motor with pulses instead of sinusoids?
Yes, you can use square-wave commutation on a motor that was optimized for sinusoidal commutation, with the penalty of adding some torque ripple.

Your bigger problem is that the commutator on the HCTL-1100 is designed for switched reluctance motors, in which each phase is uni-directional and so just on or off. You need a little external logic to create the proper state machine for the bi-directional phases (ON+, OFF, ON-) of a brushless servo motor. I figured it out once back in the 80s, but I don't remember it now.

Curt Wilson
Delta Tau Data Systems

Yuri Mitnick

The difference between brushless ac and brushless dc motor is (phase-to-phase) back EMF (torque) distribution - sinusoidal vs trapezoidal (flattened).

HCTL1100 controls a motor using six step (block, or trapezoidal) technique. So your brushless ac motor will show torque ripple comprised of cut sinusiod 60 el.deg tops. This may be problematic for smooth low speed operation.

Besides that six step commutation may suffer from commutational torque "ticks" that occur every 60 el.deg (

For a brushless ac servo motor, there are 3 input wires going into the motor. Is it possible for me to operate the motor itself alone without the servo drive or amplifier circuit? If it is possible, what would be the input to each of the wires? What source can I use to get those inputs? Btw, is it always that there is feedback from the servo motor?

Basically what I'm trying to find out here is just what source can be used as inputs for servo motor in open loop control (ie. without the feedback)? I know that there's U,V and W inputs, but that applies when there's feedback. What happens if I wanted to remove the feedback?
Regarding the commutation of the motor, I thought the signals will be fed into PWM circuit to convert them into pulse modulated signals before being fed into a servomotor? So if that's the case, then for DC case the width of pulses will be fixed at 2 values only (on or off) while the width of AC pulses varies accordingly to the sin wave. Therefore the additional torque ripple is to provide this required extra width to satisfy the commutation condition. Is this correct?

Btw, is it possible to achieve zero error by applying encoder feedback to the system? Is square wave or sinusiodal wave output from the encoder is better for the position control of this system?
Instead of adding ripple torque to the signal, can I use this other method instead? Can I feed the square wave signal into a filter to obtain a sinusoidal signal waveform, then perhaps amplifies it to the desired value and then only feed this modified signal into the motor as the commutation signal?

Curt Wuollet

Making a sine wave from a square wave requires filtering out all harmonics leaving the fundimental. To do this for varying speeds would require a tracking or adaptive filter. It would be difficult to maintain phase coherence which is sort of the idea with commutation. Of course you could use a pll to fix that but you are really working hard to use a motor the wrong way.

Anything is possible, it's just a matter of how desperate you are to do it.