A Method for Real-Time Sensorless Speed Control of Brushed DC Motors in Cost Constrained Systems

Although brushed direct current (DC) motors have given way to brushless DC motors for many high-performance applications, brushed DC motors still dominate cost-sensitive markets. For applications that require closed-loop speed control, a speed measurement method with a high frequency update rate is necessary. With cost being a significant constraint, complicated algorithms or large windowing operations are not feasible for these types of systems. This paper proposes two sensorless methods for detecting the speed of a brushed DC motor at up to 18,000 RPM. The first method utilizes a current ripple component detection method at low speeds to calibrate the motor constant. Once the motor constant is calibrated, it transitions to a back electromotive force (BEMF) measurement method for operation. The BEMF method is automatically calibrated by the current ripple method, such that no motor parameterization is required. This method is suitable for 8-bit microcontrollers running at clock frequencies of at least 8 MHz. The system was implemented and tested using a low cost ATMega328P eight-bit microcontroller. The average open-loop relative error of this speed measurement method was 3.11%. The second method utilizes the current ripple component detection exclusively. However, this method requires a 32-bit microcontroller running at 64 MHz. The average open-loop relative error of this method was 1.94%.