This paper, authored by Yu Bin of Hunan Institute of Technology, presents a fully digital design for an intelligent control system of a brushless DC motor using the TMS320LF2407 DSP controller, a dedicated DSP for motor control. The paper presents the hardware circuit design scheme for system implementation and introduces the design of the main functional modules of the data processing board and power drive board. Mixed programming using C and assembly languages ​​was employed, and simulation debugging was successfully performed. 1 Introduction Permanent magnet brushless DC motors are a new type of motor. Their control systems offer numerous advantages, including small size, high efficiency, and good economy, and have enormous development potential. Digital signal processors (DSPs) produced by TI (Technology, Instruments) offer advantages such as high precision and strong reliability. The TMS320LF 2407 DSP can realize complete brushless DC motor control functions, significantly simplifying the control circuit, reducing costs, and increasing system reliability. The hardware design of a motor control system using a DSP as the main control unit is simple. This paper introduces an intelligent control system for a permanent magnet brushless DC motor based on the TMS320LF2407 DSP. 2 Overall Scheme Design This system utilizes a DSP to realize fully digital dual-closed-loop control of a permanent magnet brushless DC motor. A deviation is formed between the given rotational speed and the speed feedback quantity. After speed regulation, a current reference quantity is generated. The deviation between this current reference quantity and the current feedback quantity is then regulated to form the control quantity of the PWM duty cycle, thereby realizing the speed control of the motor. The current feedback is achieved by detecting the voltage drop across the resistor R. The speed feedback is obtained by calculating the position quantity output by the Hall position sensor. The position quantity output by the position sensor is also used for commutation control. The system principle block diagram is shown in Figure 1. [IMG=Figure 1 System Principle Block Diagram]/uploadpic/THESIS/2007/11/2007111614451190877R.jpg[/IMG] Figure 1 System Principle Block Diagram 2.1 TMS320LF2407 DSP Controller The TMS320LF240x series from TI is a DSP controller specifically designed for control applications and digital motion control. The TMS320 LF2407 chip, a new member of the TMS320LF24x series of DSP controllers, is a fixed-point DSP chip under the TMS320C2000 platform. Several advanced peripherals are integrated into the chip, forming a true single-chip controller. It is a low-cost, low-power, and powerful upgrade product for digital control of motor motion. Its CPU core is 16-bit with a processing speed of 30 MIPS. The TMS320LF2407 chip contains a 10-bit unipolar AID conversion module, with a total of 16 AID sampling channels and a minimum conversion time of 500ns, enabling three-phase voltage sampling of the motor without phase compensation. However, due to its unipolar nature, a boost circuit is required for AC sampling to ensure the AC signal voltage range is between 0.3V and 3V. Six pulse width modulation (PWM) outputs can be used for motor control in variable frequency speed control systems. Three capture units (CAPs) can be used to record pulse widths, providing a time reference for FFT algorithms and an interface with a grating encoder for speed measurement. 2.2 Position Detection: Photoelectric encoders can be used to measure the position of a rotating device. They are fixed to a rotating shaft, with a light source on one side of the code disk and a photodetector on the other. When the encoder rotates with the rotating machinery, it generates A-phase (generated by the inner ring) and B-phase (generated by the outer ring) pulse waves on the photodetector. Based on the position and timing information of these two pulse waves, the speed, acceleration, and direction of rotation of the rotating mechanism can be determined. When the code disk rotates clockwise, the leading edge of the B-phase pulse leads the A-phase pulse; when rotating counterclockwise, the leading edge of the A-phase pulse leads the B-phase pulse. The A-phase and B-phase signals are captured and input via the CAP/QEP pins of the DSP. 2.3 Speed ​​Detection: The motor speed is calculated by the DSP using the detection signal from the photoelectric encoder. The speed value can be easily calculated based on the A-phase and B-phase signals captured by the DSP's CAP/QEP pins, using the time values ​​corresponding to the rising and falling edges of any one phase signal. Alternatively, the motor speed value can be obtained by dividing the number of pulses detected within a fixed time period by the fixed time period. The parameters used for speed calculation and speed adjustment are stored in six cells starting from 300H in the data area, with AR2 serving as the data address pointer. The variables stored in each cell are listed in Table 1. [IMG=Table 1 Variables stored in the six cells starting from 300H]/uploadpic/THESIS/2007/11/2007111614465098359L.jpg[/IMG] Table 1 Variables stored in the six cells starting from 300H 2.4 Current Detection The current sensor is an important component in the servo system. Its accuracy and dynamic performance directly affect the system's low-speed and high-speed performance. Current detection methods include resistance detection, optocoupler detection, and various other methods. This system uses a Hall element current detection method based on the magnetic balance principle. The device used is a Hall effect magnetic field compensated current sensor. It is internationally recommended as a key current detection device in power electronic circuits. It integrates the concepts of current transformer, magnetic amplifier, Hall element, and electronic circuitry into one unit. It has measurement, feedback, and protection functions. [IMG=Figure 2 Hardware System Block Diagram]/uploadpic/THESIS/2007/11/2007111614484315524X.jpg[/IMG] Figure 2 Hardware System Block Diagram It is actually an active current transformer. Its advantage is "magnetic field compensation." The measured primary magnetic field and the measured magnetomotive force in the measuring winding are compensated to zero in real time. That is to say, there is actually no magnetic flux in the iron core, so its size can be made very small without fear of iron core saturation or frequency and harmonic effects. The reason why the magnetomotive forces of the two can be fully compensated is due to the Hall effect. Once the two are unbalanced, a Hall electromotive force will be generated on the Hall element. It serves as the input signal of a differential amplifier powered by ±15V. The output current of the amplifier is the measured current of the sensor, automatically and quickly restoring the magnetomotive force balance, that is, the Hall output always remains zero. In this way, the waveform of the current faithfully reflects the waveform of the measured current on the primary side, only with a turns ratio relationship. 3. Hardware System Design According to the system design scheme, the control hardware of the system can be composed of two printed circuit boards: a data processing board and a power drive board, as shown in Figure 2. The data processing board performs AID data reading, serial port data reading, control algorithm calculation, and PWM waveform output. The power drive board samples the three-phase terminal voltage and current of the motor, performs filtering and voltage division processing, and amplifies the six-channel PWM output signals to drive the power devices. The six PWM ports of the DSP chip output three-phase square wave currents with a 120° phase difference. The PWM output signals are transmitted to the power MOSFET drive circuit through a precision linear optical isolator to drive the power MOSFETs, and the six power MOSFETs drive the motor. The two printed circuit boards are connected by a double-row 34-core cable to transmit the AID sampling signal and the PWM output signal, respectively. The data processing board is based on the TI TMS320LF2407 DSP controller. This DSP controller is equipped with two 8-channel 10-bit AID converters with sample-and-hold, and its minimum conversion time is 500ns, which can sample the three-phase terminal voltage and current in a timely manner. The board provides a 20MHz active crystal oscillator. The JTAG port on the data processing board connects to the XDS510 board plugged into the computer motherboard, allowing the computer to download and debug programs online, greatly improving work efficiency and shortening the development cycle. An RS232 interface is also designed on the data board, allowing the computer to remotely control the motor via a serial data cable. The board's SRAM is a static RAM chip—IS61C1024-15, 128K×8 bits, used to store programs and immediate operands. The SRAM chip is from Silicon Integrated Systems, Inc. (ISSI), with an access time of 15ns, and can be used in conjunction with a DSP. The GAL is a programmable logic array device from AMD, used for addressing and chip selecting of the external SRAM. For future improvements or upgrades, some expansion interfaces are designed, such as data and address line interfaces. The power driver board includes a power inverter circuit, a MOSFET pre-stage driver circuit, an optocoupler circuit, and an AID sampling circuit for terminal voltage and current. The brushless DC motor drive circuit adopts a three-phase bridge fully controlled circuit. The stator winding of the BLDC motor is Y-connected, and the number of pole pairs of the motor is 2. The inverter uses a bridge-type drive circuit composed of six N-channel MOSFETs—IRFZ44N. 4. Software System Design The entire control system software adopts a modular design concept, conforming to the current top-down mainstream design approach. In this system, the CPU clock frequency is 20MHz, and the PWM frequency is 20kHz. The ADC conversion is initiated by a timer 1-cycle matching event, ensuring that the current is sampled once per PWM cycle. The current is adjusted in the A/D conversion interrupt handler to control the PWM output. A capture interrupt is triggered every 60° mechanical angle rotation of the rotor to perform commutation operation and speed calculation. The main program flowchart of the system is shown in Figure 3. [IMG=Figure 3 System Main Program Flowchart]/uploadpic/THESIS/2007/11/2007111615012666263Q.jpg[/IMG] Figure 3 System Main Program Flowchart 5 Simulation Results This paper presents a fuzzy control simulation of the system. The speed setpoint jump response curves are shown in Figures 4 and 5. [IMG=Figure 4 Speed ​​Jump Response Curve]/uploadpic/THESIS/2007/11/2007111614513259826J.jpg[/IMG] Figure 4 Speed ​​Jump Response Curve [IMG=Figure 5 Torque Response Curve During Speed ​​Jump]/uploadpic/THESIS/2007/11/2007111614514027257E.jpg[/IMG] Figure 5 Torque Response Curve During Speed ​​Jump 6 Conclusion The simulation results show that the system can respond quickly when a given speed jumps. Torque and current fluctuations are small, and the system can quickly reach stability when the speed jumps; the static error is small, exhibiting excellent static and dynamic performance. Proceedings of the 2nd Servo and Motion Control Forum Proceedings of the 3rd Servo and Motion Control Forum