Fujitsu Labs Ltd. said it has developed has developed a gallium nitride (GaN) high electron mobility transistor (HEMT) amplifier capable of 174 watts at 63 volts at 40 percent efficiency. If that doesn’t mean much to you, Fujitsu claims the HEMT will enable it to make more energy efficient 3 (or wasn’t that 4?) G base stations…
The next level of high-output, high efficiency amplifiers is going to need gallium nitride-based transistors for the job, claims Fujitsu. But the company, whose Takeshi Mimura, who invented the HEMT back in 1980, believes that these tricky compound semiconductor based amps needed to bang out more than 150 watts to be useful for 3G base station requirements. Until this announcement, nobody has breached that barrier, claims the company.
Compound semiconductors such as gallium arsenide and gallium nitride can whack out power, but they are tricky compounds with a lot of issues hidden in their atomic layers that cause them to behave badly when currents are run by them. They are also a pig to handle in the labs. Take the following quote from Mimura talking about the early days.
“The process of manufacturing new semiconductor devices is much like the process of advance through a minefield, by groping in the dark. After we are figuratively blown apart by the mine, which represents the failures we experience with our processes, and make sacrifices in terms of yield loss, we finally make our way to a broader and more reliable road to the manufacturing process.”
Usually GaN HEMTs have proved unstable or blew out at about 50 volts, for example. But the Labs got over this by improving the uniformity of the gallium nitride epitaxial layer. The new HEMT performance conforms to W-CDMA 3G mobile communication base station standards. When supplied with 60 or more volts of power, the design satisfies the W-CDMA specification for adjacent channel leakage, while also achieving the record-setting drain efficiency of 40 percent.
Bottom line, better base stations. When, however, Fujitsu isn’t saying.