Sensor History

The National Semiconductor Days

In the early seventies the only commercial silicon based sensor work in the Valley was Art's work (just starting at National Semiconductor), Don Lynam at IC Transducers (also just starting) and the effort at Fairchild (coming to a close through Cognition.) Elsewhere in the world there were other commercial efforts with the work continuing at Honeywell Microswitch, Hans Keller at Keller in Switzerland, Kulite in New Jersey and at Phillips in Europe. The largest research efforts were at the Universities in the United States. There were significant efforts at Case Western, under Dr. When Ko, Stanford under Dr. Kendall Wise, and by John Gragg at Carnegie Mellon. National Semiconductor wanted to get into the business because of the potential automotive applications. Art was asked by National's management to explain "transduction." Art told them he would put it into the simplest terms, and one they could relate to best. He described transduction as the ability to take silicon and convert it to money. That appealed to National, Art was hired and the ten year transducer effort at National was to begin. Art has more accurately described the circumstances, as the events surrounding this statement also provide an insight into the character of National Semiconductor at the time.

In the mid seventies I assumed responsibilities for marketing of all National's hybrid products, including Transducers. At the time there was a major effort by the car companies to develop an automotive MAP sensor. At National we worked with Delco Electronics and Ford to codevelop two types of MAP sensors. The Delco version had a sensor die similar to the Honeywell sensor die of the time and was packaged in a housing similar to the Fairchild ignition module. This product and versions of it are still manufactured by Delco and other aftermarket suppliers. Similar versions of this sensor were developed for all other major car companies in the world and its specification is the defacto industry standard. The Ford version was a silicon variable capacitance pressure sensor. It is still manufactured today by both Ford and Motorola. However, it is not used by any other car manufacturer. It is more costly than the piezoresitive version. There are other thick film hybrid MAP sensors also serving this market.

In 1977 I was participating in a Transducer Range Commanders conference in Seattle along with Joe Mallon, vice president of engineering at Kulite. At this time Joe had a ton of patents for silicon piezoresistive pressure sensor processing. (In 1983 Joe, Kurt Peterson and Janusz Bryzek were to become cofounders of Novasensors; more on them later.) I got to know Joe from this meeting and found him to be the most knowledgeable person at that time concerning temperature effects due to semiconductor processing for pressure sensors in silicon. Even today most companies reference his original work for determining appropriate concentration levels for dopants in silicon to set the temperature coefficient of resistance and sensitivity.

In 1978 I was attending Semicon West in San Francisco and was at "Herr Doktor" Janusz Bryzek's presentation on discrete temperature compensation of silicon pressure sensors. Janusz presented a circuit that had at least twenty amplifiers, and several hundred resistors and many potentiometers. The most elaborate scheme for temperature compensation I had ever seen, truly a technical wonder to behold. He was asked by a member of the audience "Doktor Bryzek what is the error in such a compensation with so many components?" Without hesitation Janusz replied "there is no error, it.... is perfect!" At the time I felt he could quite possibly be right but wondered how he would test it. (An engineers mentality, not a marketing mentality.) From this conference I got to know Janusz.

During this same time period American Hospital Supply approached National for a $5.00 disposable blood pressure sensor. The first work started then for what is today, most probably, the second largest pressure sensor application in the world behind the automotive MAP sensor.

The 1977 National Semiconductors Transducer Handbook became the reference book in the transducer business. Most sections of this book are still reprinted with each reprint of the Sensym handbook. This handbook is still the reference book of choice for pressure sensors. The 1977 handbook was unique because each section had an unusual introductory title and preface including, "The pig who squealed Dixie," a section concerning acoustic measurements and "Samson and Delightful," a section on signal conditioning. This book was the result of a years work by Art, Ray Pitts a Ph.D. in physics who was consulting and rewriting the bible at this time, and myself. Ray was the major contributor. He had an unhappy ending to his story at National and a tragic ending to his life shortly thereafter.

In 1980 I became Director of Operations for Transducer Products at National and Art Zias reported to me. It was a truly challenging and joyful time period for us all as Art kept everyone on their toes and entertained. In addition to engineering, Art performed as master of ceremony at the National Semiconductor annual sales meeting and would use me as a sounding board for many of his anecdotes. One of his more memorable ones was about Charley Sporck, founder and CEO of National Semiconductor. In reaction to financial analysts criticism of him at the time, Charley made the statement "he would chomp on groins and spit testicles." Art, in reference to this statement told an audience of several hundred National employees and sales Reps "that it just goes to show you that angry rich men can develop strange gourmet fetish's." The charter for the transducer business was to determine what was needed to grow the business to $100 million in a short time period. At the time I didn't know it, but the only other alternative was to exit the business.

Most notable was Mike Scott, director of operations for hybrid and transducer products from 1974 to roughly 1979. I reported to Mike during this time period and he was the most knowledgeable marketing, and possibly the brightest person I have ever met. Mike, who previously worked for Mike Markula at Fairchild, left National to become the first president of Apple Computer and left Apple in 1983 when he and Steve Jobs did not see things the same. Markula, Chairman of the Board for Apple, opted for Steve instead of Mike. Mike left Apple at this time with his eight million shares of Apple stock and has enjoyed himself ever since. He did have one fling in the satellite launching business with the failed launch of their first satellite, and company, Starstruck. The concept for their satellite had considerable technical appeal. But in the end they had an expensive boat ride that saw their satellite and investment forever disappear into the depths of the Pacific Ocean. Floyd Kvamme was Vice President of marketing and sales for National. Pierre Lamond headed R&D. Both Floyd and Pierre are very well known today in the venture capital community. At National our semiconductor fab processing was done in the linear group, headed by Bob Swanson. Bob left National with a team of National linear people and started Linear Technology, a very profitable semiconductor company today. Their leaving National caused Charley Sporck to file legal action against the group as it "appeared" unusual to see them talking daily in the cafeteria for some time prior to leaving. Especially since they had never socialized before this time. One of the cofounders of Linear, Brent Welling for whom I worked for a brief time period at National, would later join us at Sensym as Vice President of marketing and sales.

At National I was a member of their eleemosynary committee and had the opportunity to visit a sensor group at Stanford University. Kendall Wise was head of the effort and was just leaving to form a sensor research effort at the University of Michigan. I met Jim Knutti who was involved with an implanted integrated injection logic, combination pressure and temperature sensor used to monitor bodily functions of sheep, funded by the National Institute of Health. Jim worked with Dr. Henry Allen and the two of them later (1984?) started a company to manufacture force sensors using silicon for what Art had earlier published as "the fingertips of the robot." This company, Transensory Devices Inc , was acquired by IC Sensors and Jim and Henry stayed with IC Sensors. Jim left a short time after the acquisition and started a sensor operation in Switzerland with Ascom. Ascom was eventually sold, there was a fire in their fab, and finally they exited this business. Henry joined Jim and the work done with Ascom allowed the two of them to start Silicon Microstructures in 1992. SMI was acquired by Exar in 1995 and Jim and Henry continue to be active at SMI/Exar. They have done some very good work with silicon stress concentration, modeling thereof, and silicon structures for low pressure devices and accelerometers.

Next