If you look at it from the viewpoint of the engineers at Signetics in early 1971, you can understand why they were cool to the idea of a separate IC just as a timer. Their former colleague, Hans Camenzind, had taken a four-month leave of absence to finish his book on IC design, and he decided not to return as a full-time engineer. The semiconductor industry was in one of its miserable periods and Signetics suffered along with the others; it had laid off almost half its staff. In December 1970, Camenzind proposed to design a timer IC for Signetics as an independent contractor. The timer idea grew out of work he had done at the company on the first phase-locked loop, a dramatically successful IC. The PLL required precise oscillators, and he found it best to use a current source to charge and discharge an external capacitor directly through an external resistor. The oscillator's frequency was determined by that capacitor and resistor. Camenzind had used such oscillators in the 565, 566 and 567 ICs.

So he was fairly sure he could use the same approach to design an accurate timer/oscillator. Signetics management shot his proposal over to the engineers who used to work with Camenzind. It would be understandable if they had mixed feelings in rejecting the idea of their former colleague. But they felt that any competent engineer could build a timer with a low-cost op amp, a flip-flop and a zener diode. And it would cost less than the $5 or so that a new IC would cost initially. People worried about money in those days.

Undaunted, Camenzind pitched his idea to Art Fury, the marketing manager, a man with lots of practical engineering and marketing experience. Even before Camenzind could finish his presentation, Fury said: "I want it."

Fury was a good man to talk to. He launched three of the most popular products at Signetics. Back in 1966, he had been a market-development specialist for GE Semiconductor.

There, he was charged with getting semiconductors into toys. His first major sale was into a toy known as Tommy Turtle. A child could clap his hands; a cheap crystal microphone would pick up the sound and transmit the signal to a GE SCR, which would start a motor and cause the turtle to move about five feet until a cam interrupted the current.

This was a great sale for GE because its SCRs had a minimum voltage rating of 35 volts. Those that failed the 35-V test, though they were functionally good otherwise, were dumped into 55-gallon drums for disposal. Tommy Turtle didn't need 35 V. It ran on two D-cell batteries. Fury sold millions of them, right out of the 55-gallon drums.

There was a slight glitch the following year when the toy company, Remco, decided to add a duck, Tricky Dicky Duck. The company sold millions at the toy show in February and was prepared to start shipping in October for the Christmas season. That fall, Richard Nixon won the Republican nomination for the presidency—and Remco had to reprint millions of labels for what became Tricky Doodle Duck.

Fury and Camenzind settled on a fee of $1,200 a month for 12 months. The staff reductions at Signetics left a lot of equipment unused, so Camenzind could borrow almost any instrument he needed. He set up shop between two Chinese restaurants in Sunnyvale, Calif., and started designing.

In five months he had a breadboard that worked. He wrote a development report and passed the design review. But he suffered a nagging feeling that something wasn't quite right. He had missed something important. Two days after the design review, he decided that the current source was not necessary. He could charge the capacitor through the resistor, making the timing more accurate, and cutting the pin count from nine to eight. Nine is a lousy pin count as there is no nine-pin package. He had planned to use a 14-pin DIP with five pins unconnected. He could now use an eight-pin mini-DIP.

He was slightly behind schedule, so he didn't pursue another design review and he didn't change the development report. He just proceeded with the integration. A good switch for Camenzind. Two months before Signetics introduced the revised Camenzind timer, another company brought out the timer described in Camenzind's original report. It was the wrong circuit and was pulled off the market three months later.

When Signetics introduced the 555 timer, Fury did something crazy. Instead of pricing the new IC as was the custom (make your money as quick as you can), he set the introductory price at 75 cents. The market's reaction was what every engineer and marketer dreams about. Signetics sold half a million 555s in three months. Camenzind was inundated with phone calls from engineers who had learned that he had designed the 555. Nine out of 10 of their applications were in areas he had not considered.

The 555 was sufficiently popular that it was in short supply. Many Signetics distributors complained that they weren't getting delivery, yet an unknown distributor kept advertising immediate availability of 555s. It turned out that an inside marketing person at Signetics was assuring himself delivery by the simple process of carrying out cartons of 555s and selling them on his own behalf.

His purchase of an expensive house for cash led to his fall and imprisonment and to a freer market in 555s. When the sheriff came to take him away, he asked, "Does this mean I lose my job?" It did.

Within six months after its introduction, eight companies copied the 555. Signetics never applied for a patent and the company (now Philips) has only a minor market share. The original design, shrunk but unmodified, survives to this day — very, very successfully. The 555 timer has been the largest-volume IC every year since it was introduced in 1972.

And Camenzind is still an independent. He founded and still runs Array Design, an IC-design contractor in San Francisco.

The Century of the Engineer: Misunderstood Milestones