Electronic Design

Barrie Gilbert: From Crystal Sets To Submicron Silicon

Gilbert (2003)

In 1945, an eight-year-old schoolboy in England, who earlier had lost his father in the war, fiddled with the position of the cat's whisker on the lump of galena in his crystal set. He wondered why only certain spots lit up his headphones. This was Barrie Gilbert's first encounter with the magic of semiconductors.

Later, using pennies earned from his paper route, he became the beneficiary of war-surplus radar. One day in 1948, brandishing a VCR97 6-in. cathode ray tube, he announced to his mother, "I'm going to make a TV set with this!" From the kitchen sink she absently replied, "That's nice, dear." He did make that set, with help from Practical Wireless magazine. The neighbors watched the Coronation of Queen Elizabeth on the small, green screen, the only TV on the street.

His can-do spirit and a lively sense of humor opened the door to opportunities in the U.S. In 1964, with plane ticket in hand, he was ready to head to Hewlett-Packard in Colorado, when he got a call from a vice president at Tektronix. He confessed he hadn't treated Tek's job application form seriously. For instance, he addressed the question "Do you have any physical limitations?" with "Yes. I rely on a constant supply of oxygen, and can't run the 4-minute mile (yet)." Asked to suggest a starting salary, he wrote "$100,000 will be okay, for a start." Tek's offer of $10,400 was less than HP's, but his atlas showed that Portland, Oregon was a quarter-inch from the Pacific while Loveland was several inches inland. With salt water in his veins, Gilbert switched to Tek.

It was a fortuitous choice. "I had incredible latitude to devise monolithic circuits, using our new on-site wafer fab. Many of today's basic cells had not yet been developed, so it was easy to invent. It was an exciting time," he says.

His concept of superintegration, a dense implementation of logic functions, having dozens of transistor fragments merged into a single region, was the forerunner of I2L (Integrated Injection Logic). He also explored the potential of superintegrated structures for analog purposes using carrier-domain devices. Among those devices were the "steerable localized injection multiplier," a radically new approach to analog multiplication, and a magnetometer that converted magnetic-field strength directly to frequency as a domain raced around a circular track. These and other novel concepts were presented at the ISSCC in the 1970s, at a rate of a paper per year, with five receiving the Outstanding Paper Award.

While designing the Tektronix 7000 series of oscilloscopes and many of its custom ICs, he formulated what he later identified as translinear (TL) circuits. He exploited the reliable logarithmic properties of the bipolar transistor, inventing all of the common cells. He also promoted the value of current-mode (CM) signal-processing. Today, TL and CM are in everyday use. However, in a recent paper, he cautions about the dangers of "design myopia." Gilbert has about 65 patents ("I've lost count"), including many of a seminal nature.

Noting that street-hype about digital technology is well-deserved ("It has profoundly and irreversibly changed life on Earth"), he stresses that analog techniques are far from obsolete. "Analog has always been, and always will be, indispensable. Keep in mind that digital computers are only incidentally electronic. They exploit the extremely small size, high speed, and low cost of silicon technologies. It's possible to build a computer using just NOR gates, but there isn't the remotest hope of using one cell type for the immensely diverse analog functions. Analog invention remains as essential and as vibrant as it ever was. Product demand is stronger today than at any other time in history," he says.

Invention, Innovation, and Discovery
He muses over the terms invention, innovation, and discovery. "Invention is an instant," he says. "It's all about thought, and ideas happen anywhere—in the bathtub, on a train, even while dozing. But innovation is a process. It's about things and doing. It requires time, equipment, and energy to turn thoughts into things. Discovery happens when an unexpected event fortuitously encounters a prepared mind. There's a lot of overlap. It's often hard to know whether one is creating or discovering."

He notes that invention is often a solitary pursuit, but the innovation needed to breathe life into ideas is always a team effort. "I've had the good fortune to be surrounded by outstandingly talented and dedicated people," Gilbert says. "But any large organization has its share of doubters and naysayers, who should be politely but firmly ignored. The creation of something truly novel requires swimming against powerful undercurrents of ingrained opinion."

In his home lab, also a composing hideaway with racks of MIDI synthesizers and surrounded by a tantalizing assortment of vintage items—including the old 'ray tube that once lit with the promise of a new Queen's face, but now drowned by the ingress of air—he hands you the crystal set, solid, four-square on a mahogany base. It's that set which first invited him to ask "Why?" to which the response has always been "Why not?"

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish