Sensorless positioncontrol of a stepper


Thread Starter

Clemens Zvacek

The development shows it - steppers can also be used in a PID loop. The plusses are the low price and the high reliability of the motors. But the necessary encoder drives the price up again.

I have tried to get as much information as possible about the possibilities of determining the rotor position without sensors. A student named A. Maryniak of a german university tried to solve this problem with fuzzy logic but - as his work shows - failed to find the right characteristics.

He also names the concept "TetraDrive" by Escap, but it is not possible to get any information bout that over the internet. The technique is not in their catalogs any more.

Maryniak only shortly describes it, TetraDrive means analyzing the EMF of the not used windings of the stator to determine the next step.

Does anyone have information on this or similiar ways of determining the rotor position of a stepper motor without sensors ?

You've got two mechanisms you could exploit. The first is the back EMF of the phases. In a two-phase stepper motor these are Ke*w*sin(Theta) and
Ke*w*cos(Theta). From the voltage commands/measurements, you've got to subtract out the I*R and L*dI/dt drops, then take the arctangent to get Theta.

The other possible mechanism in a hybrid motor is the varying inductance. The more that a phase is "lined up", the higher its inductance will be, and therefore the higher its L/R time constant.

Curt Wilson
Delta Tau Data Systems