Figure 1. Advanced algorithms, such as Field Oriented Control (FOC), improve the dynamic control of the motor torque by keeping it constant within the rated speed range. 
Figure 1. Advanced algorithms, such as Field Oriented Control (FOC), improve the dynamic control of the motor torque by keeping it constant within the rated speed range.
Designers are always on the lookout for semiconductors and algorithms that help to boost the efficiency of appliances with a minimum addition to the overall system cost. Motor-control systems need to compensate for system input changes and can use control algorithms to ensure the efficient operation of the motor. Using advanced algorithms, such as a field oriented control (FOC), motor torque can be controlled dynamically, keeping it constant within the rated speed range (see Figure 1). Toward this, the most commonly used motor-control loop is the Proportional Integral Derivative (PID) controller, which comprises error calculation (reference minus measured variable), compensator (controller), and output generation to the system.

Motor-Control Design Issues in Appliances

Given the increasing efficiency and environmental standards requirements, designers can reasonably expect to see more control loops in their system designs. In advanced motor-control systems, it is normal to have a minimum of three control loops. This can be understood as follows: while an outer system control loop keeps motor speed constant under varying load conditions; an inner loop controls the motor current by constantly adjusting the duty cycle of the pulse width modulator (PWM) peripheral, under an analog-to-digital converter’s (ADC’s) interrupt. An extra inner loop monitors current in the same way, but is aligned with a different current vector.

The right kind of controller is required to execute advanced control algorithms. It would be a bonus if the same controller could also handle the system design portion. Consider the latest washers, air conditioners and automotive applications that use permanent magnet synchronous motors (PMSMs). These motors, while conforming to the latest energy standards, require efficient motor-control algorithms involving current control at very high-speed operation. FOC algorithms meet this requirement, but necessitate two current-control loops running between 8 kHz to 20 kHz rates, and a third speed controller loop executed at rates of 1 kHz or less.

Some vendors offer fixed-hardware based PID controllers supported by an MCU, which limit the flexibility of the controller section. These “hard-wired” motor-control algorithms -- besides being proprietary and expensive -- are limited in their ability to support advanced motor-control algorithms. At the high end, expensive digital signal processors (DSPs) are overkill for many motor-control applications. In addition, some standard DSPs lack motor-control peripherals. While they do speed up FOC algorithm execution, DSP-based motor-control schemes get very expensive because they require a number of external peripheral chips.

DSCs Make Motor Control Efficient

DSCs offer a clear-cut design advantage over the previous two approaches. Featuring motor-control specific peripherals, such as PWMs, ADCs and comparators, DSCs support most motor-control schemes in an economical manner. The DSP core inside of a DSC speeds up computation, while the on-chip peripherals accommodate the multiple control loops required for motor control. And, programming DSCs is easy -- because they maintain the same look and feel as an MCU.

Appliances with variable-speed motor drives are a good fit for DSC-based controls. Since these applications cannot afford major hardware changes to accommodate the motor control, FOC algorithm control using DSCs makes practical sense. A three-phase inverter is used as the power stage to drive the motor windings in these appliances. With the addition of a DSC and current-sensing circuitry based on two low-cost resistors, a complete FOC design is obtained that improves the motor efficiency.

Sensorless FOC on a PMSM-based Appliance

Figure 2. A DSC-based FOC execution block. Click to enlarge 
Figure 2. A DSC-based FOC execution block.
The sensorless control technique implements an FOC algorithm by estimating the position of the rotor without using position sensors. The FOC algorithm (see Figure 2) is configured by ensuring that the PWM triggers ADC conversions for the two windings, using two shunt resistors. The reference speed of the motor is set and the ADC interrupts are enabled to execute the algorithm.

The position estimator, which is based on the currents and voltages of the motor, uses a motor model to measure the motor position indirectly. A current-observer model assists in the indirect measurement of back EMF, by feeding the motor and its model with the same input. A closed-loop observer, which is modeled after the motor, ensures that the estimated value matches the measured value.

DSCs help to implement FOC by ensuring that the stator magnetic field stays 90 degrees ahead of the rotor at all times. Using a three-phase voltage, the FOC algorithm generates a voltage vector to control the three-phase stator current. The amplitude, frequency and phase of this voltage vector are all controlled by the FOC algorithm. Converting the physical current into a rotational vector using transforms makes the torque and flux components time-invariant. We can exploit this time invariance to handle the control with conventional PI controllers, as with a DC motor.

The FOC process begins by measuring the three-phase motor currents. From circuit theory, we know that the instantaneous sum of the three-phase current values will be zero. Therefore, in practice, by measuring only two of the three currents, the value of the third current can be determined. There is a reduction in hardware cost, because only two current sensors are required. The FOC technique is universal because it works with both synchronous and asynchronous motors.

On-Chip Analog Peripherals
DSC’s fast and flexible on-chip ADCs support current sensing and offer useful triggering options. For example, an ADC conversion triggered by the on-chip PWM module enables a cheap current-sensing circuit, by sensing inputs at specific times where switching transistors allow current to flow through sense resistors. Crucially, some DSCs feature ADCs that have the capability to capture multiple signals simultaneously. This feature helps to eliminate any delay in motor-current measurements between two phase samples.

Specialized, on-chip peripherals aboard DSCs enable implementation of the control algorithms and the entire appliance system design comfortably. Figure 3 shows an example of a washing machine interface under DSC control. The DSC’s ADC channels can be used to measure the motor current, the motor temperature and the temperature on the heat-sink connected to the power switches. User-interface elements, such as LCD or key switches, can be interfaced using the I/O lines of a DSC. The appliance can also be diagnosed and calibrated using the DSC’s serial ports.

DSCs also provide fault interfaces that include input lines, which can be used to shut down the PWMs in case of catastrophic system faults. In addition, by designing a DSC-based active power factor correction (PFC) block, the appliance can be readied for European markets that have stringent energy regulations. The PFC block comprises an inductor, a power switch and a diode. ADCs on the DSC measure the current and voltage values from the DC bus. Based on these inputs, the DSC controls the power switch through a PWM module, by executing a PID loop, which keeps the PF value close to unity. Alternative, non-DSC-based PFC implementations require an extra ASIC.

Figure 3. A DSC in action as the system controller in a washing machine
Figure 3. A DSC in action as the system controller in a washing machine


DSCs, with their motor-control friendly peripherals, are natural choices for the role of signal controllers in appliance designs that need advanced algorithms such as FOC. Some vendors, such as Microchip Technology, even offer FOC algorithm libraries for free with full source code. The role of FOC algorithms in being able to control the stator current leads to a reduction in torque ripple and results in quieter appliances. DSC-based FOC implementations save energy and offer better response to dynamic load changes. Being programmable devices, DSCs enable faster production lines by allowing the rapid customization of newer appliance models to address multiple geographies and markets.

Note. dsPIC is a registered trademark of Microchip Technology Inc. in the U.S.A. and other countries. All other trademarks mentioned herein are property of their respective companies.

Jorge Zambada is Senior Applications Engineer in Microchip Technology’s Digital Signal Controller Division. For more information, contact Microchip Technology, 2355 West Chandler Blvd., Chandler, AZ 85224; (480) 792-7200;