One of the perks of this job is that I get to meet some really interesting people. Last week was no exception, as I got to meet two engineers who pioneered the research for the electret condenser microphone (ECM) when they worked for Bell Labs in the early 1980s. They are James West and Gerhard Sessler. The pair was in town as they were being inducted into the New Jersey Inventors Hall of Fame (who knew New Jersey had an Inventors Hall of Fame?). Because the team was also instrumental in the development of the MEMS-based microphone, Ray Stata, the founder and chairman of Analog Devices, joined the meeting. Stata pioneered some of the early MEMS work at Analog Devices more than 15 years ago. This made the meeting that much more fun.

Sessler joined Bell Labs in 1959, a time when the primary driver for microphone development was for telephones. His development work on MEMS-based microphones occurred in the 1980s at the University of Darmstadt, where he designed the first condenser microphones based on silicon micromachining. West was simultaneously working on a similar concept at Bell Labs. In the mid 1970s, MEMS were used mainly for pressure sensing. But according to Sessler, "microphones simply measure pressure. They're just a lot more sensitive."

According to West, the evolution of microphones started with a carbon version, which was invented by Thomas Edison (also in New Jersey). However, the story goes that Edison never got the credit (or financial reward) for the carbon microphone because it was held up in the patent process by the Bell Labs legal team. The carbon mike was followed by the ECM, then a silicon version, and finally the MEMS model.

Left to right: James West, Ray Stata, and Gerhard Sessler after West and Sessler were inducted into the New Jersey Inventors Hall of Fame.

The first patent for a MEMS-based microphone was granted in 1983. At the time, there was a lot of resistance to the MEMS mike, as it cost about 10X compared with the carbon existing technology, and were not as reliable, as they were expected to have a life of only about one year. In fact, Sessler's bosses told him not to bother even trying to get a patent because it wouldn't be worth the effort. But Sessler thankfully forged ahead anyway.

I'm told that MEMS microphones already account for about 20% of the cell phone market. The downside to the MEMS microphone is its cost--it's about 2X, as compared with an ECM microphone, which costs about 25 cents. Analog Devices happened to launch a family of MEMS microphones last month.

The biggest current market MEMS devices are crash sensors in automobiles. But they really hit the microphone scene in a big way with the introduction of Motorola's extremely popular RAZR phone. In that application, the mike was designed as a surface-mount device, one that could withstand high temperatures. Since that time, the industry is producing about 200 million MEMS-based microphones per year, compared with about 2 billion ECM microphones in the same time period.

The current MEMS microphone consists of two ICs in a package, the sensor and the pre-amp. Analog Devices, who continues to develop MEMS-based microphones, believes it won't be long before further integration occurs. This includes a model with a digital output.

The tradeoff between the microphones is that the MEMS mikes can withstand higher temperatures, but they're more expensive than the ECM counterparts. But like most technologies, as volumes increase, costs will come down. In fact, some of the latest cell-phone designs have up to four microphones, as OEMs try to get the best recording experience.

Richard Nass is editor in chief of Embedded Systems Design magazine and editorial director of TechInsights' Embedded Group. He can be reached at rnass@techinsights.com.