November 2004

Power on a Chip

Continued from page 2

By David H. Freedman

In Hand
Earlier this year, Epstein and his coworkers finished making  engines in which each of the individual parts functions: the combustion chamber burns fuel, and the turbine blades spin. The resulting device is sealed all around, with holes on the top and bottom for air intake, fuel intake and exhaust. One shortcoming: it doesnt run continuously. The obstacle, says Epstein, is imperfections that imbalance the blades and cause them to wobble. We think we know what to do to correct it, he says. The problem is that it takes three months to get new parts when you make an adjustment, so were just waiting for the new parts. Epstein predicts the chip will be functioning within monthsa little ahead of schedule. Spearing estimates a version capable of putting out enough power to run devices would take two to three years more, with another year or two beyond that to produce a marketable version.

That means conceding an early lead in the power MEMS race to fuel cells, which are already hitting the market. Albany, NYbased MTI Micro Fuel Cells is preparing to launch one the size of a deck of cards for use in handheld industrial devices such as radio-frequency ID-tag readers and has plans to roll out a slightly smaller version for cell phones, PDAs, and digital cameras. Medis Technologies of New York City intends to sell a $20 disposable micro fuel cell next year.

Our competition is fuel cells, absolutely, says Epstein. But he insists that turbine chips can make up any lost ground. Up to now a few million dollars has been invested in microturbines, compared to the billions invested in fuel cells, he points out. Epsteins faith is fueled by the inherent advantages he sees in turbines. Even micro fuel cells are larger, and theyre much more finicky about fuel than a turbine engine. But in the end, it all comes down to power. Most micro fuel cells struggle to put out a watt or two, while Epsteins prototypes could provide 15 to 20 watts, more than enough to keep a power-hungry handheld device going. Laptop computers can require 50 watts, but a few turbines working together could easily pump that much power out. Likewise, Epstein envisions that a cluster of tiny engines, each capable of producing up to a hundred watts, could supply a home with an efficient and reliable source of electricity.

That switchover will surely take time. But Epstein sees it as the natural extension of the remarkable progress jet engines have made throughout the second half of the 20th century, from the novel fighter planes that appeared in World War II to the behemoth engines that power todays jumbo jets. And though Epstein predicts that, from an engineers point of view, his tiny chip-based turbines will initially perform more like the pioneering jets of the 1940s than like todays superefficient gas turbines, he is fully confident in the technologys vast potential to evolve. Indeed, the aging engines in his labs museum are an ever present reminder of the gas turbines awesome power.