The development of computing applications has sparked widespread attention and discussion about how the electronics industry can reduce energy consumption. According to research by the Electric Power Research Institute (EPRI), mechanical applications, including motor drives, consumer white goods, and industrial machinery, account for more than 50% of global electricity consumption, making this area a primary target for new low-energy designs.
Besides energy management ICs, the solution for this type of mechanical drive control application lies in using intelligent chips in the form of MCUs. MCUs can manage motors with higher efficiency and lower cost, accelerating the transition from electromechanical control to electronic control and realizing the control of variable speed motors (VSMs).
Compared to traditional DC motors, MCU-controlled brushless DC (BLDC) motors offer higher efficiency, a higher torque-to-inertia ratio, higher speed performance, lower noise, better thermal efficiency, and lower EMI. The efficiency of these intelligent motors can exceed 95%, while induction motors only reach 85%. Furthermore, in many applications, MCU-controlled variable-speed BLDC motors can save 25% to 40% of energy compared to constant-speed motors.
Energy efficiency of BLDC motors
The emergence of cost-effective MCUs for motor control has also attracted the attention of traditional motor manufacturers and application vendors, prompting them to re-evaluate and reconsider some of the special control technologies for VSMs. Implementing VSMs using traditional scalar control methods requires manufacturers to increase the size of the motor to accommodate large transient or peak voltages.
To avoid the increased costs associated with larger motors, manufacturers have begun to explore field-guided control (FOC) (also known as vector control) technology to reduce motor size. FOC technology offers superior dynamic response characteristics, higher power density, and lower torque ripple, all of which contribute to improved system efficiency. Furthermore, FOC eliminates the need for position sensors, requiring only a shunt resistor, thereby reducing manufacturing costs and enhancing reliability.
Field-guided control technology maintains a 90° angle between the stator and rotor magnetic fields in the motor by changing the current in the stator windings. Although the angle of the stator magnetic field in the system is known, the angle of the rotor magnetic field must still be measured or estimated in order to calculate the angle difference between the two.
Once the angle of the rotor magnetic field is determined, the vector control algorithm can calculate the optimal timing and magnitude for applying voltage to the stator phase windings. Since these vector control algorithms are data-intensive, common sensorless FOC implementations currently employ 16-bit or 32-bit MCUs, DSPs, or DSC processors to handle complex trigonometric equations.
Furthermore, to ensure the required accuracy, a lookup table needs to be built into the system. This necessitates the use of large-capacity flash memory and complex software algorithms to handle issues such as current calculation, vector rotation, space vector modulation, and proportional-integral control. These factors all increase the cost of the control system.
The newly launched 8-bit microcontrollers have been specially enhanced in terms of architecture. For example, Infineon Technologies' XC800 series has the hardware functions required to achieve a more cost-effective FOC system (as shown in Figure 1).
Low-cost 8-bit FOC solution
The MCU with FOC functionality integrates an 8051-compatible processor core and a powerful on-chip processing unit—a vector calculator. This calculator can simultaneously perform multiple calculations on vector data (one-dimensional arrays) (as shown in Figure 2). The vector calculator consists of multiple processing units, including a coordinate rotation digital calculation (CORDIC) unit and a multiplication/division unit. When used in conjunction with a 16-bit acquisition/comparison unit and a fast on-chip A/D converter, it can perform 16-bit mathematical operations.
The CORDIC algorithm incorporates a small lookup table and utilizes addition, subtraction, and shift operations to perform iterative calculations of various complex mathematical and trigonometric functions, such as the Clarke and Park algorithms. CORDIC outputs results with up to 16 bits of precision, and its functionality is largely independent of the CPU core, thus saving resources for other control tasks.
The multiplication and division unit can perform 16-bit and 32-bit mathematical operations and can be used to replace the standard 8051 MUL/DIV instructions. To further reduce flash memory capacity and improve access speed, we can also add a library of fixed-point and floating-point mathematical operations to the bootstrap ROM.
As mentioned earlier, the primary goal of the FOC algorithm is to ensure that the stator's magnetic field remains perpendicular to the magnetic field of the permanent magnets within the rotor. This relationship is estimated through a single shunt current measurement, which requires rapid triggering of the A/D converter using a corresponding PWM mode. This is achieved by employing an event-based hardware trigger between the intelligent PWM unit CapCom6E and the A/D converter. This event-based trigger eliminates interrupt latency, enabling fast and accurate current measurement.
Integrating the on-chip computing unit and peripherals to implement FOC saves significant resources for other system control functions using low-cost 8-bit MCUs. For example, at a PWM frequency of 15kHz and a current measurement speed of 133μs, the FOC control function only utilizes 58% of the CPU performance, leaving ample margin for other dedicated functions. Unlike hard-coded FOC implementations, MCUs with integrated vector calculators are reprogrammable. This capability can be used to optimize the motor startup process by constructing a programmable ramp or employing methods to reduce the magnetic field (e.g., weakening the ID component of the FOC algorithm).
Sensorless FOC Evaluation
The FOC Driver Application Toolkit can be used to evaluate sensorless FOCs based on an 8-bit MCU. This toolkit includes an MCU with an integrated vector calculator, a three-phase power conversion board, a 24V BLDC motor, a plug-in power supply, and the complete FOC source code.
Furthermore, hexadecimal code can be downloaded via the CAN-USB bridge, allowing modification of motor parameters such as speed and current control during motor operation, thus achieving real-time control. Infineon also provides users with a complete development environment, including a free toolchain, helping them utilize the same toolkit to achieve the next stage of application development and customization.
