Inside every plug-in vehicle there’s a black box the size of a six-pack cooler that connects the battery to the electric motor. It’s called the power inverter. This crucial, but often overlooked component converts the battery pack’s high-voltage direct current (DC) into alternating current (AC) pulses that control the traction motor.

A DC-to-AC inverter, basically a fast-acting silicon semiconductor switch, functions something like an Engine Management System does in a internal-combustion power plant. It feeds the driver’s commands to the traction motor in the form of pulse-width-modulated drive signals at frequencies that can range from 10 Hz to 10 kHz.

Because all electric traction power passes through the inverter, any efficiency losses that occur within cut directly into a plug-in vehicle’s battery-only driving range.

In fact, more efficient inverter technology ranks second in importance only to more power-dense batteries for extending battery-only range in next-generation plug-ins.

Improved electric and hybrid vehicles are not alone in their need for better inverters. High-efficiency inverter technology would also greatly benefit industrial motors, consumer electronics, appliances and data centers as well as photovoltaic and wind energy systems.

It’s no surprise then that electronics and materials researchers worldwide are working to develop improved semiconductors that could deliver inverter performance that is superior to conventional silicon—including fewer switching losses, greater thermal efficiency and importantly, reduced system costs. Even Google is working on this issue, having established last year a prize competition—The Little Box Challenge—that will award $1 million to the developer of the best inverter design for "green energy" applications; see https://www.littleboxchallenge.com/.

The goal of this research is to develop what are called wide bandgap (WBG) semiconductors. To physicists, WBG materials exhibit a relatively large quantum energy range in which no electron states can exist—a bigger electron energy gap compared to silicon between the top of the valence band and the bottom of the conduction band. In practice, it refers to the amount of energy that is needed to release electrons from a particular semiconductor material for conduction.

Semiconductors with wider bandgaps can, for example, withstand higher applied electric fields, or voltages, as well as operate at higher temperatures, power densities and frequencies.

In the automotive sector, the U.S. Department of Energy recently awarded research grants to General Motors ($3.99 million) and Delphi ($2.46 million) to support three-year, cost-shared projects to develop high-efficiency, cost-competitive integrated power inverter modules based on WGB semiconductors for plug-in vehicles.

Automotive Engineering previously reported on Toyota’s continuing research efforts to develop more efficient automotive power electronics modules using silicon carbide; see http://articles.sae.org/13244/.

Smaller, lighter inverters

Inverters in current plug-ins rely on silicon-based power transistor technology that was developed for industrial applications over the last 25 years, said Pete Savagian, GM's General Director for Electric Drives and Systems Engineering and a veteran of the company’s pioneering EV-1 program. These insulated-gate bipolar transistors (IGBTs), often tuned for automotive use, combine good efficiency and fast switching, he explained, but expanding plug-ins’ battery-only driving range means moving beyond silicon.

Two emerging WBG semiconductors, silicon carbide and gallium nitride, are expected to fill that role, Savagian continued, because they “can bring three to ten times better energy efficiency when they're turned on and especially, when they're turned off. And when you’re switching at rates of 10,000 Hz, reducing losses becomes important.”

Wide-bandgap inverter technology "plays upon the ability of the transistor material to run at higher temperature and with fewer losses than silicon-based power electronics," explained A.J. Lasley, Director of Electronic Controls Advanced Engineering at Delphi in Indianapolis. “The improved efficiency can directly translate into longer range.”

He noted that wide-bandgap materials, particularly silicon carbide semiconductors, have been trying to push into industry for many years, with the recent DoE-supported projects aiming to "push the limit" in plug-in inverter technology.

"The new materials offer great potential for allowing us to reduce the size of inverters by as much as 30% and cut energy losses by 20% to 30%,” Lasley said.

According to GM's Savagian, the new WBG semiconductors would allow “using less semiconductor material in inverters than we do now. The resulting smaller footprint means that everything else can shrink as well, including all the support equipment—electrical connectors, cooling system, heat exchanger, and the housing and chassis structures.”

Such physical and operational downsizing should in addition yield significantly cheaper power inverter units, Savagian predicted. He noted that the inverter typically accounts for about 40% of the total cost of an electric drive train, which includes an electric motor and a gear reduction system.

Both Savagian and Lasley stressed that one of the principal benefits of WBG semiconductors to plug-in vehicles is that they would enable engineers to integrate power inverters directly into the transmission systems.

“Their smaller size means that the mounting and packaging can be more rigid and robust," Savagian observed. "It also would enable engineers to incorporate the devices into the transmission units, saving space and weight. You could, for instance, get rid of the electrical cables, which makes assembly easier.”

Beyond silicon

Experts note that gallium nitride has similar bandgap characteristics to silicon carbide, which is a more mature technology. But silicon carbide chip fabrication "is very expensive, while gallium nitride offers the possibility of lower-cost manufacturing because of it is more compatible with the underlying substrate materials,” said Jayant Baliga, Director of the Power Semiconductor Research Center at North Carolina State University. Baliga, a pioneer in power electronics, invented and commercialized IGBT devices when he worked at General Electric.

Baliga’s NCSU center is taking the lead in the Power America program, also known as the Next Generation Power Electronics National Manufacturing Innovation Institute. This is a five-year, $140-million R&D effort established in January 2015 by the DoE “to drive WBG semiconductor costs to make them more competitive with silicon materials.” In the case of silicon carbide, the researchers are attempting to adapt existing silicon chip foundries to silicon carbide chip fabrication, Baliga noted.

Anant Agarwal, the senior WBG expert at the DoE’s Advanced Manufacturing Office, has said he expects that highly efficient power electronic devices using the new semiconductors will be able to achieve price parity with traditional silicon-based devices within about five years.

Power America’s members comprise a dozen companies as well as seven universities and laboratories, including ABB, Arkansas Power Electronics International, Cree, Delphi, John Deere, Monolith Semiconductor, Qorvo, Toshiba, Transphorm, United Silicon Carbide, VACON and X-FAB.

Besides NCSU, the program’s academic and lab partners are Arizona State University, Florida State University, the National Renewable Energy Laboratory, the U.S. Naval Research Laboratory, the University of California, Santa Barbara and Virginia Polytechnic Institute and State University.