Mixed-signal FPGAs provide GREEN POWER

Electric motors are used in nearly everything, from elevators to home appliances. In 2005, the United States consumed 4,055 billion kilowatt-hours of electrical power. More than 50 percent of this power was used in electric motors, translating into a staggering 2,000 billion kW-hr bill. Unfortunately, many of the motors in use are inefficient and waste a substantial amount of the power they consume.

Adding electronic control can dramatically improve efficiency, but the high cost of control and power electronics has been a barrier to its implementation. With technology improvements in semiconductor processes and integration, mixed-signal field-programmable gate arrays (FPGAs) are emerging as an important alternative for motor control implementation. These highly integrated, flexible platforms offer the bulk of the resources needed for motor control on a single, low-cost device. Using FPGAs in lieu of fixed logic gives designers the flexibility to implement the most efficient design for their application and the ability to use the same device across a broad range of motor applications.

The efficiency of small ac motors can be as low as 50 percent. While motor efficiency improves to more than 90 percent as motor size increases, there is still opportunity to improve efficiency and reduce energy consumption. By adding intelligent load matching or variable speed control, the power efficiency of electric motors across the full range can be increased. With a mixed-signal FPGA and a soft optimized microprocessor, this can be accomplished with a minimal increase in motor cost. In fact, coupled with best practices, this combination can result in motor efficiencies approaching 95 percent. Implemented broadly, electronic motor control could result in savings of as much as 15 percent of the electric power used in the United States--an annual reduction in energy consumption of as much as 300 billion kW-hr, saving $15 billion and reducing greenhouse gases by more than 180 million metric tons.

The biggest factor impacting ac motor efficiency is loading mismatch. When an ac motor is operated near full load, it can reach efficiencies over 90 percent. Few motors are consistently operated at full load, however, due to natural load variations and oversizing, reducing efficiency by as much as 75 percent and wasting power. It is not unusual in the United States to find motors that are 2x to 3x larger than they need to be for the load that they are driving, which is an expensive mode of operation. Even when correctly sized to meet the maximum load, the motor is typically run at a lower, less efficient loading. For example, an escalator is sized to carry the maximum number of people. Most of the time, however, there are few people on the escalator, causing it to run at a low level of efficiency and waste power.

Electronic motor control
With electronic motor control, the load can be intelligently and continuously sensed and exactly matched with the proper input power, maximizing the efficiency of the motor over the full operating load range and minimizing power consumption and operating costs. Even small variations in the loading can be detected and power precisely applied to match it without affecting the speed of the motor. In effect, electronic control constantly sizes an ac induction motor to the job, so that it is always operating under ideal load conditions.

The conversion of ac motors doesn't necessarily require an expensive replacement of all of the motors that are currently in use. The Department of Energy estimates that the industrial sector alone uses 12.4 million motors larger than 1 horsepower. Motor replacement is ongoing, with as many as 600,000 motor failures and replacements annually. This means that over the next 20 years, most of the motors larger than 1 hp will need to be replaced. Replacing these with highly efficient electronically controlled motors can reduce ongoing industrial power requirements by as much as 18 percent, resulting in significant energy cost savings for the manufacturing sector.

Variable-frequency drives
For applications that can be operated at a constant speed, intelligent load matching via electronic control is a great solution. Unfortunately, not all applications that utilize ac motors can be operated at a constant speed. For low-cost drives that are suitable for applications with known loading, variable-frequency drives can be used to vary the motor's rotational speed to match it to current conditions. The ability to continuously vary the speed depending on conditions and maintain a constant ratio between frequency and voltage (V/f control) is an easy way to get variable-speed operation from a three-phase motor.

Compared with alternative solutions, FPGAs have historically been excluded from consideration as a solution for these applications because of their cost and the lack of the required analog peripherals for ac motor control. New cost-effective mixed-signal FPGAs offer a highly flexible single-chip solution, however, providing much of the resources needed for motor control implementation for a broad range of motors--from single-pole permanent magnet motors to large three-phase ac drives.

At 600k samples per second, roughly two to three times faster than is required for ac motor control, the analog-to-digital converter in a mixed-signal FPGA allows the direct measurement of the stator and rotor currents to determine rotor speed and position. Additionally, with up to 30 A/D inputs, sampling of the back electromotive force of each winding, the motor current, bus voltage and any other conditions in a motor is as easy as connecting them to the device and making the appropriate measurement.

The use of a mixed-signal FPGA with an integrated soft processor, such as the FPGA-optimized ARM Cortex-M1 microprocessor, allows motors to be built with sensorless sinusoidal current control, eliminating costly sensors and further reducing the price of the electronic controls. Together, a Cortex-M1 processor in a mixed-signal FPGA can also perform diagnostics in addition to monitoring the bus voltage, motor currents and speed. The ability to run diagnostics and respond intelligently to problems as they occur can significantly reduce damage and increase the life of the motor, further reducing the cost of ownership.

Vector control
The concept of harnessing rotating magnetic fields is still evolving with new motor technologies and electronic motor control techniques. In many applications, both the load and the speed are variable, so vector control (field-oriented) is widely used. Because it is not based on steady-state motor equations, it can deal with the varying operating conditions that are seen in many motor applications. Vector control allows responsive and accurate speed control with a changing load and offers optimum efficiency even during motor transition. It also allows full motor torque capabilities at low speed.

An advantage of using mixed-signal FPGAs for vector control is that the same device can be used to control a range of motor types, including permanent-magnet ac and brushless dc motors. By using the appropriate model for the motor type, only slight changes need to be made.

For a three-phase ac motor, the vector algorithm must be continuously calculated at a rate between 1 and 10 kHz. A substantial number of calculations have to occur in the short period of time available within each 120° phase, including trigonometry, proportional-integral-derivative, real-time flux and torque interface functions. For ac motors controlled with a vector scheme, the requirement is for a small but powerful processor that can support these calculations and the rest of the application for communication and the user interface.

In actual applications, ac motor control requires 8 to 20 Mips per control axis, which is easily achieved with an ARM Cortex-M1 processor in a mixed-signal FPGA with integrated flash memory. This combination enables the full conversion, including the measurement of the currents, to be carried out in less than 6 microseconds, allowing more than 165,000 conversions per second, which is more than adequate for most ac motor control applications.

In addition to the processor, it is important to have the right peripheral set for the application. Because a mixed-signal FPGA can support the processor, memory and peripherals within a single device, the additional components required are minimal, other than the inverter block and the motor itself. This significantly reduces the cost of the electronics and makes this solution attractive for a broad range of motor designs.

Benefits for consumers
The demand for energy savings in home appliances is putting pressure on appliance manufacturers to use more efficient motors. The largest potential savings will come from the use of electronically controlled motors in refrigerators, air conditioners, washing machines and furnaces, which together account for more than one-third of residential energy use. Most appliances in use today have an ac induction motor running at constant speed with frequent starts and stops. As a result, as much as 50 percent of the energy they consume is wasted.

A motor controller built with a mixed-signal FPGA and used with discrete trench insulated-gate bipolar transistors can achieve efficiencies approaching 95 percent across the full operating range. This means that a smaller, lower-power motor can be implemented and run continuously at lower speeds, eliminating the need to turn the motor on and off, thereby reducing power consumption, heating and motor noise.

Not only does the use of electronically controlled motors reduce power consumption, it can also result in a better product that offers consumers more features at a potentially lower price. For example, washing machines can be made with features that cannot be implemented easily with a conventional induction motor, such as balance correction, forward and reverse agitation cycles, and high-speed water extraction. Another benefit of electronic motor control is that motor speed can be continuously varied so that mechanical transmissions can be eliminated, and the resulting direct drive is less costly and more reliable. n

Mike Thompson (Mike.thompson@actel.com) brings more than 25 years of experience to Actel Corp., where he is responsible for the development and support of new microprocessor IP cores for use in the company's FPGAs. He holds a BSEE from Northern Illinois University and an MBA from Santa Clara University.

 See related image