Increased signal-to-noise ratio of sensorless control using current oversampling

This paper presents a novel approach based on current oversampling with low computational effort that significantly increases the signal-to-noise ratio with regard to sensorless control of permanent magnet synchronous machines at low and zero speed. The novel method is based on calculating arithmetic averages of all current samples taken during a period of pulse width modulation (PWM). This algorithm is implemented in a field programmable gate array (FPGA). Compared to a conventional current measurement with single sampling, the arithmetic average of the current samples has a much higher signal-to-noise ratio than the current samples themselves. The current derivative, which is used for sensorless control, is then calculated from the difference between two consecutive current averages. Experimental results validate the functionality of the novel approach.