Power Surge For A Monster Chip

For two decades, Victor A.K. Temple has labored in relative obscurity, pursuing a semiconductor technology so arcane and unglamorous that it even turns off most silicon engineers. No wonder. The chips that fascinate him are far from the cutting edge and account for just 0.1% of all chip sales. Their name--monster power ICs--is catchy but misleading. They aren't superstar number-crunchers but integrated circuits designed to modify, switch, and otherwise handle heavy-duty electrical currents--up to 1,200 volts--that would fry ordinary chips.

Finally, diligence is paying off for Temple, who heads Harris Corp.'s Power Research & Development operation in Latham, N.Y. The radically different power chip he patented last year has been cited by R&D Magazine's panel of experts as one of 1993's 100 most significant new products. And a financial dividend seems guaranteed, too. Market researcher Dataquest Inc. is predicting a surge in demand for monster chips, from sales of roughly $70 million worldwide in 1992 to $335 million by 1997.

That increase will be sparked partly by recent energy-saving rules and incentives from government agencies. By incorporating power chips that use less electricity, appliance makers are already developing more efficient air conditioners, refrigerators, and washers. Factory equipment, aircraft, ships, and submarines may benefit next. Electric utilities count on power chips to make existing facilities more efficient and thus help meet growing demand. And sales of the chips could soar in 1998, when California will require that 2% of the cars sold there be pollution-free. They'll probably have electric motors, and Harris thinks its chips will help produce snappier acceleration, longer range, and faster battery recharges.

SAVING GRACE. What these uses have in common are electric motors. The Edison Electric Institute says the U.S. has 1 billion motors, which gulp half of America's electricity. Much of that is wasted, because most motors now run at one speed. The solution is to vary the motor speeds with power chips. Motors that run only as fast as they need to "can reduce the electric bill by up to 40%," says Daniel Artusi, director of power products at Motorola Inc.'s Semiconductor Products Sector in Phoenix. If all single-speed motors were variable-speed models, he adds, "the savings would be equal to nearly 1% of gross domestic product."

The design that accounts for most current monster chip sales is IGBT (insulated-gate bipolar transistor). Although the chip was developed in the U.S. in the 1980s, the market has been dominated by Toshiba and Hitachi in Japan and Thomson and Siemens in Europe, countries where electricity costs are so high that demand for energy-efficient products is brisk. While U.S. chipmakers focused on big-ticket industrial motors, says Harris' Temple, "the Japanese used IGBTs to go after mass-market things, like air conditioners." Their production volumes took off, and their power-chip prices fell.

Then late last year, Harris introduced a new power chip called the MOS-controlled thyristor, or MCT. It will put the U.S. back on top, Temple asserts, because of its higher current-handling capacity--more than double that of the same size IGBT--and half the electrical loss. A different approach to the internal structure of the chip accounts for these gains. The first MCTs don't switch on and off as fast as IGBTs, but a new model will, Temple claims, perhaps by yearend. Switching speed is crucial for precisely regulating variable-speed motors.

Motorola's Artusi retorts that the MCT's advantages aren't as big as Temple maintains. "We've looked carefully at MCT technology," he says, and decided to stick with IGBT. Still, Harris has powerful allies. One is the Electric Power Research Institute in Palo Alto, Calif., which does research for the utility industry. EPRI figures that MCT chips could replace practical every type of switch used by its member companies. "All of our power-saving R&D programs for the next five years are based on these devices," says Harshad Mehta, EPRI's manager of power systems technology. Moreover, he says that ultrahigh-power MCTs--now in the research stage at Harris, with funding from EPRI--will simplify the conversion of high-tension lines from a.c. to d.c., which lets the same lines carry twice as much power. These chips could also help avert blackouts by shifting electricity from one regional grid to another.

Even the U.S. Air Force likes MCT chips. Jet fighters are stuffed with so much electronics that heat buildup is a big problem. The heat can't be vented without affecting performance and increasing the plane's signature on enemy radar, "so it is dumped into the fuel," says William U. Borger, chief of aerospace power research at Wright Patterson Air Force Base in Dayton. The drawback is that "if you want to keep the electronics cool enough to land the plane, you often come in with hundreds of pounds of fuel" that otherwise could have been used for a longer flight. With their lower heat loss, MCT chips may alleviate that problem.

TOTAL CONVERSION. The Air Force sees another advantage. Using small, MCT-equipped electric motors instead of hydraulic systems to power such things as landing gears and wing flaps may help trim a plane's weight by up to 15%. That could boost the speed and maneuverability of both fighters and civilian aircraft--which explains why AlliedSignal Inc.'s aerospace division has "converted completely to MCT" for airplane power systems, says chief scientist Robert C. Eckenfelder. And he expects the entire aerospace market to follow suit.

What the chips do for planes, they may also do for electric cars. Variable-speed motors under the hood might obviate the need for heavy transmissions. MCT chips could also replace the old, solid-state controller on electric motors. In theory, that could extend the car's range by saving up to half the 15% of an electric car's battery power that is lost as heat because of resistance from the controller. John R. Wallace, director of electric vehicle development at Ford Motor Co., is skeptical about how big the mileage increase might be. But his group recently got its first MCT chips. And in lab tests, he says, "they appear to do what they're supposed to do." If that holds up in vehicle tests, he adds, "they'll become our devices of choice."

Temple isn't resting on his laurels in the meantime. He notes that current MCT chips are suitable for only about 20% of the markets Harris is eyeing. He has plenty of new ideas for improving his baby's power-handling capabilities. So even after 20 years, Temple's work may be just beginning.

      OF MCTs
      A new breed of chip shows  promise in the transportation and utility 
        -- May make it possible to replace bulky hydraulic systems on landing gears 
      with more compact electronic systems
        -- May boost mileage 
        -- Could speed up and prolong battery charges
        -- Could instantly reroute power to avert blackouts
        -- Double the capacity of high-voltage distribution lines
Before it's here, it's on the Bloomberg Terminal.