Electrification for energy savings
The imperative to save energy is pervading all aspects of modern technology, from items such as domestic appliances and personal transport to large industrial equipment and utility infrastructure like pumping stations. Efficiency gains can be achieved through a combination of converting bulky hydraulic or mechanical systems to smaller, lighter electrical alternatives, and by upgrading existing motorized equipment using more efficient drives featuring energy-saving controls and high-efficiency power semiconductors.
Consumer and e-Transport
Initiatives such as minimum energy-performance standards and energy-rating labels are encouraging consumers to choose more energy-efficient appliances such as refrigerators and washing machines. Refrigerators that utilize sophisticated variable-speed control of the compressor motor, in place of simple on/off operation, have been shown to reduce energy consumption significantly, while also improving temperature regulation and reducing audible noise. In washing machines, similarly, smaller and more efficient motors with variable-speed operation permit advanced wash programs that save both heating energy and water usage.
As far as personal transport is concerned, electrical actuators are relentlessly replacing numerous mechanisms to boost the fuel efficiency of today’s cars. The market for motor drives used in the automotive and transportation industry is estimated to reach around $1.8 billion by 2018, according to a report by Research & Markets (Servo Motor & Drives Market: 2013 – 2018).
Subsystems such as power steering and water pumps, which are traditionally driven mechanically from the engine, are being replaced by smaller, lighter and more efficient electrical alternatives. Electric traction is also on the increase, evidenced by the growing number of hybrid-electric vehicles (HEVs) in manufacturers’ ranges. The role of the internal combustion engine as a range extender for the HEV is expected to diminish as batteries, battery-management systems and electronic drives continue to improve, ultimately leading to widespread use of vehicles driven purely by electric motors.
Other forms of electric transport such as electric bicycles (ebikes) can further offset the use of conventional gasoline and diesel vehicles. The rising popularity of ebikes is driving double-digit growth of this market in Germany. Additionally, ebikes already account for 15% of all new bicycle sales in the Netherlands, as reported by online journal Bike Europe.
Industrial drives demand efficiency
Many industrial applications incorporating electric motors, such as process automation equipment, HVAC systems, and pumps ranging from small, low-power applications to large units employed in irrigation systems or oil and gas extraction, are moving to more sophisticated electronic drives. Drive vendors are reporting growing customer interest in variable-speed controls for energy savings, particularly in high-power applications.
Reducing the speed of an AC motor by 20% can reduce the input power consumption by as much as 50%. In an application such as a high-power water pump, changing the motor speed to adjust the pump capacity according to demand can be far more energy efficient than using other means such as valves to restrict the output of a fixed-speed system that draws the rated power continuously.
Motors and control
Figure 1 shows the main functional blocks of a generic electronic motor controller, comprising a microcontroller, gate driver and power-transistor bridge, and a means of detecting the rotor position to provide data for the motor-control algorithm hosted on the microcontroller’s CPU. Host-interface and power-management blocks are also shown.
Figure 1: Typical electronic motor-drive architecture.
With greater reliance on electronic motor controls across diverse markets, and the rapid development in emerging applications such as electric vehicles, a variety of motor types are in use. The Brushless DC (BLDC) motor is widely used in applications such as fans and blowers, and requires the electronic controller to coordinate commutation according to the rotor position. High-power industrial drives are predominantly based on AC motors such as permanent magnet or induction motors, controlled by algorithms such as field-oriented control or scalar-slip control.
A BLDC controller can be hosted on a microcontroller, such as the Atmel ATmega48, which has enough processing capability to execute the commutation algorithm and perform other system-level functions while also integrating essential peripherals such as an Analog-to-Digital converter (ADC) and a timer with PWM output.
Field-oriented control of an AC motor ensures that the rotor and stator flux act at 90° to ensure maximum torque. The speed of the motor is determined by the number of poles and the frequency of the applied voltage. Since the number of poles is essentially fixed for a given motor, a Variable-Frequency Drive (VFD) is needed to adjust the rotor speed. The VFD comprises an input converter, a DC link and an output inverter that constructs an output of variable frequency from a PWM waveform.
Power semiconductor advances
While variable-speed control, as calculated by the microcontroller, has an important influence on the overall energy consumption of the application, the efficiency of the electronic drive as a unit is related to the performance of the power devices in the output inverter supplying the motor windings. The power stage may comprise a discrete gate driver connected to an H-bridge or three-phase power-transistor bridge. Alternatively, an integrated power module may be used, combining both functions as a single unit.
The latest field-stop trench-gate Insulated Gate Bipolar Transistors (IGBTs) are ideal for high-voltage applications. IGBT families from manufacturers such as International Rectifier, Infineon, Fairchild or STMicroelectronics offer voltage ratings up to 1200 V or higher. By achieving lower switching losses compared to earlier IGBT technologies, these latest devices can maintain high efficiency at switching frequencies above 20 kHz.
Various MOSFET technologies are available for use in systems operating at lower voltages. Advanced silicon-based families such as Infineon’s latest OptiMOS™ or ST’s STripFET™ devices are available in various voltage ratings suitable for automotive or industrial applications, and deliver a combination of low on-state resistance (RDS(ON)) and excellent RDS(ON) x gate charge (Qg) figure of merit (FOM) for enhanced switching and conduction efficiency. Innovative packages such as CanPak™, which is an option for Infineon OptiMOS devices, allow two-sided cooling for reliable operation at high operating current.
Promising even more favorable conduction and switching performance in small device sizes, emerging generations of MOSFETs featuring technologies such as silicon carbide (SiC) or gallium nitride (GaN) present new opportunities to increase efficiency even further. SiC and GaN are both wide bandgap materials, requiring energies of more than 2 electron volts (2eV) for electrons to transition from the top of the valence band to the bottom of the conduction band.
Wide bandgap materials enable devices to offer enhanced high-temperature performance. In addition, SiC and GaN materials have high breakdown voltages that allow devices to have high maximum operating voltage in relation to die size, thereby enabling both RDS(ON) and Qg to be reduced, leading to better conduction and switching performance. In addition, both SiC and GaN have superior electron mobility and thermal conductivity compared to ordinary silicon, as well as a low temperature coefficient, making these technologies attractive in applications requiring high operating frequencies and high power output.
The EPC2014 and EPC2016 GaN power MOSFETs from Efficient Power Conversion Inc. (EPC) are rated up to 40 V and 100 V respectively. They provide ultra-low RDS(ON) and gate charge within a tiny footprint for applications requiring high switching frequencies and low on-state losses. EPC’s GaN FETs are supplied as passivated die with solder bars.
Detecting rotor position
The gate-drive signals generated by the microcontroller running the motor control algorithm determine the speed and torque of the motor. While speed range and responsiveness are related to the time to execute the algorithm, the accuracy and smoothness of the gate-drive signals have an important influence on aspects such as audible noise and vibration. High levels of audible noise can be unacceptable in some applications, particularly in domestic appliances and automotive drives. High levels of vibration, on the other hand, can shorten rotor-bearing life and may disrupt nearby mechanisms such as measuring devices or limit switches.
Correct execution of the motor-control algorithm, to generate the desired gate-drive signals at the right time over a wide rotor speed range, requires an accurate means of detecting the position of the rotor relative to the stator. A basic three-phase BLDC motor has six commutation phases. The control algorithm must know the position of the rotor to energize the stator coils in the correct sequence to turn the rotor in the desired direction and at the requested speed. The rotor position can be calculated from the back EMF induced in the unpowered stator coils, creating a sensorless drive.
A sensorless drive can save bill-of-materials costs and offers advantages in some applications, particularly where initial torque is relatively constant. However, the motor-control algorithm running on the microcontroller must take into account the fact that no EMF is generated on startup. Moreover, the position information is not as accurate as can be achieved with sensor-based feedback using devices such as Hall switches or a magneto-resistive sensor.
Hall sensors
Hall switches remain the most widely-used technology for rotor position sensing. In a basic implementation, three Hall sensors are positioned at 120° intervals around the rotor. The rotor magnet pairs trigger the Hall sensors in sequence, creating a switching pattern that changes every 60°. Positioning the Hall sensors to coincide with the ideal coil commutation points, the motor can be driven by energizing the coils using either constant or pulse-width modulated drive signals.
Figure 2: Internal rotor 3-phase BLDC motor with one magnet-pole pair and Hall sensors S1, S2, S3.
In many controllers, Hall sensors are placed inside the motor and monitor the rotor poles directly (Figure 2). However, one disadvantage is that the sensors can be exposed to high temperatures inside the motor. Sealing the motor can also be difficult, leaving the system vulnerable to the ingress of chemicals, moisture, or particles from the surrounding environment. To overcome these disadvantages, an additional magnetic code wheel may be placed on the end of the rotor shaft allowing the sensor assembly to be positioned on the end of the motor. This can also allow easier access and greater design flexibility.
To minimize signal delay and ensure a balanced duty cycle, suitable Hall switches should have high sensitivity, with a magnetic switching point close to zero milli-Tesla. The semiconductor process used to fabricate the sensors, however, is subject to manufacturing spread, resulting in a variance between the magnetic switching points of different sensors. To compensate for this, sensors with a Hall-Effect latch feature internal circuitry to cancel out the effects of offsets due to the Hall sensor and input amplifier. This allows the switching point to be specified within a very narrow window, thereby helping to ensure smoother and more balanced rotation.
Other features include a fast signal path that minimizes delay between input and output, and features built-in temperature compensation to stabilize the magnetic switching point over the specified temperature range.
Magneto-resistive sensing
In applications where a separate magnetic code wheel is to be used, an angular position sensor can provide an effective alternative to an array of Hall switches. Magneto-Resistive (MR) sensors are ideal for this application, taking advantage of the fact that the magneto-resistance effect is predominantly dependent on magnetic field direction rather than field strength. However, the strength of the applied magnetic field must exceed the sensor’s specified minimum saturation field strength.
Anisotropic Magneto-Resistive (AMR) sensors such as the KMA210 and KMA220 from NXP Semiconductors, allow accurate rotor position sensing using a single or pair of sensors. The sensor contains a pair of Wheatstone bridges, which produce orthogonal differential signals that are then amplified and converted into the digital domain. A CORDIC (COordinate Rotation DIgital Computer) algorithm is then applied to calculate the angle, which is then converted to a proportional analog voltage. Two independent sinusoidal output signals are generated, which are proportional to the sine and cosine of the angle between the sensor and direction of the field.
Figure 3: NXP Semiconductors’ magneto-resistive angular position sensors are available in single or dual-channel configurations.
The KMA210 and KMA220 are automotive-qualified magneto-resistive sensors that provide programmable user adjustments including the zero-angle and the angular range. In the KMA210, the sensor is housed in a separate over-molded module with integrated filters for improved Electro-Magnetic Compatibility (EMC). The KMA220 (Figure 3) integrates two sets of sensors and processing ICs for dual-redundant applications. These sensors can be used for sensing the angular position of mechanisms such as wipers, accelerator pedal or actuator valves in automotive applications, in addition to motor control applications such as electric power steering.
Conclusion
Efficient electric motor drives carry the hopes of many stakeholders as a means of reducing energy demands and ensuring the sustainability of human activity. Improvements in microcontrollers, control algorithms, power devices and sensors can each make a contribution towards these goals.
