1. Overview
This paper uses the TI 240XDSP as the core of the fully digital control system for a brushless DC motor, resulting in a servo system requiring very few system components. The TMS320F240X is a high-performance 16-bit digital signal processor (DSP) from TI, specifically designed for the digital control of motors. This DSP includes a fixed-point DSP core and a series of microcontroller peripheral circuits, integrating digital signal processing capabilities with efficient motor control. It allows for software replacement of analog devices, easy modification of control strategies and parameters, and also features fault detection, self-diagnosis, and communication with a host computer.
2. Hardware Introduction
The TMS320LF2407A's core is the TMS320C2XX (Figure 1), which uses a Harvard architecture and pipelined operation. At a clock frequency of 20MHz, the instruction cycle is only 50ns, and most instructions can be completed within a single instruction cycle. Its central arithmetic logic unit (CALU) includes a 32-bit CALU, a 32-bit accumulator, CALU input/output data scale shifters, a 16-bit × 16-bit multiplier, a product scale shifter, data address generation logic (including 8 auxiliary registers and 1 auxiliary register arithmetic unit (ARAU), and program address generation logic). When the processor is working continuously, it can simultaneously execute: a) data reception and transmission via the serial port; b) internal timer decrementing; c) generating three-phase pulse width modulation (PWM) waveforms; d) acquiring 4 analog signals; e) watchdog timer decrementing. It also includes features not found in other DSP series chips, such as dual 10-bit analog-to-digital converters and a PWM-based manager (6 comparator units, 12 PWM outputs, and 2 photoelectric encoder interfaces). Its PWM waveform generation unit includes programmable dead-time control, capable of outputting asymmetric, symmetric, and space vector PWM waveforms. The LF2407 is the only x240x series DSP capable of expanding external memory and is also the most powerful and feature-rich model in terms of control capabilities, widely used in code development, system simulation, and practical systems.
Figure 1 Overall structure of TMS320Lx240x series DSP controller
3. DSP Control Flow
Figure 2 shows the control and drive circuit for speed regulation of a three-phase brushless DC motor implemented using the TMS320LF2407A. This design uses the TMS320LF2407A microcontroller as the core of the system controller, with power MOSFETs as the power conversion elements. Only two phases of the motor are conducting at any given time to control the commutation elements, and PWM is used to control the motor's torque and speed. Here, three Hall sensors H1, H2, and H3, spaced 120° apart, are connected to the three capture pins CAP1, CAP2, and CAP3 of the TMS320LF2407A after being shaped and isolated by a circuit. Capture interrupts are generated to provide the commutation time and position information.
Figure 2. Speed control and drive of a three-phase brushless DC motor using TMS320LF2407A.
Theoretically, the speed and torque control of a brushless DC motor are mainly based on the following engineering calculation equations for torque and back electromotive force.
Where N is the number of stator coils per phase of the brushless DC motor, is the length of the rotor, r is the inner diameter of the rotor, B is the magnetic flux density of the rotor, is the angular velocity of the motor, i is the phase current, L is the phase reactance, is the position of the rotor, and R is the phase impedance.
As can be seen from the equations, the back electromotive force is proportional to the motor speed, and the torque is almost proportional to the phase current. Based on these characteristics, the control strategy shown in Figure 3 is adopted in the scheme. A deviation between the given speed and the speed feedback is generated, which, after speed regulation, produces a current reference quantity. The deviation between this reference quantity and the current feedback quantity, after current regulation, forms the control quantity for the PWM duty cycle, thus realizing the speed control of the motor. Current feedback is achieved by detecting the voltage drop across the resistor. Speed feedback is obtained through calculation using the position quantity output from the Hall sensor. The position quantity output from the position sensor is also used to control commutation.
Figure 3 Speed and current control of a three-phase brushless DC motor
4. Software control
The system employs a PWM control algorithm. The motor receives DC current, and only two power transistors are active at any given time. The PWM control signal from the 2407 is directly connected to the driver, and the driver's output is then connected to the control electrode of the power MOSFET. The 2407's CPU clock frequency is 20MHz, and the PWM frequency is 20kHz.
4.1 Phase Current Detection
After receiving the amplified voltage drop signal across the resistor, the TMS320LF2407A converts it to a current signal via an A/D converter. At the end of the conversion, the A/D module sends an interrupt request signal to the CPU, awaiting CPU processing. Every 50µs, the DSP controller samples the phase current, thus implementing a 20kHz current regulation loop. Based on the current error, the PID controller adjusts the duty cycle of the PWM pulse at the beginning of each PWM cycle.
4.2 Rotor position and speed detection
Mastering the appropriate commutation timing can reduce torque ripple. Position detection is used not only for commutation control but also for generating speed control inputs.
The position signal is obtained through three Hall effect sensors. Their output signals are 1200 ohms out of phase. Each mechanical rotation has six commutations. The correct commutation timing can be obtained by setting the DSP to dual-edge triggered capture interrupt function. The specific capture interrupt is obtained by setting the DSP's capture ports CAP1~CAP3 as I/O ports and detecting the level status of these ports.
Position signals can also be used to generate speed control quantities. Once the time interval between two commutations is measured, the average angular velocity during the two commutation times can be calculated using the following formula.
The time interval between two commutations can be obtained by capturing the value of the T2CNT register of Timer 2 when an interrupt occurs.
4.3 Current and speed regulation
The phase current can be adjusted by changing the pulse width of the PWM signal with a carrier frequency of 20kHz.
Ierror="Iref"-Imea
cyclenew = cycleold + IerrorK
If cyclenew >= Timer_period, then cyclenew = Timer_period.
If cyclenew > Timer_period, then cyclenew = 0
Where Iref is the user-desired reference current;
Imea—The actual measured phase current;
Ierror—The phase current error to be adjusted;
Speed adjustment uses a PI algorithm to achieve optimal dynamic performance. The calculation formula is as follows:
In the formula, Iref represents the speed regulation output.
ek - the velocity deviation of the kth iteration;
Kp - Speed proportionality coefficient;
Ki - Velocity integral coefficient;
T - Speed adjustment cycle;
Experiments have shown that it can generate excellent three-phase PWM control waveforms. Figure 5 shows the generated PWM waveforms.
Figure 4. PWM waveform generated by DSP control of a three-phase brushless DC motor
The software modules described above, including system initialization, position signal detection, and PWM signal output, can implement a basic speed control system for a three-phase brushless DC motor with a position sensor. However, to build a more complete system, additional functional modules are needed, such as a control module for adjusting the motor speed and a data logging module for storing system operation data. The communication between the TMS320LF2407 and the PC uses a half-duplex RS-485 interface circuit. Since the PC provides an RS233 interface, an interface conversion between RS-32 and RS-85 is required.
5. Conclusion
The innovation of this article lies in implementing PWM waveform output on a DSP using the PI algorithm commonly used in industrial control. Due to the limitations of the performance of microcontrollers, it is difficult to meet the requirements of high-speed and high-precision motor control. However, using a DSP can effectively realize PWM waveform output for the control of brushless DC motors.
