1960 · Planar Integrated Circuit · Work 843 of 1,871
A tiny experimental chip under a microscope on a laboratory bench in Palo Alto at night, a round glossy gold metal base six millimetres across with a broad rim and a small flat tab, filled with smooth black resin, in its centre a small grey rectangular silicon chip with a fine pattern of pale lines, around it eight tiny gold contact points, several joined to the chip by short gold wires, the base held in tweezers by a hand in a white shirt sleeve, beside it a broken grey silicon disc and a small plain glass vial of amber resin, an oscilloscope behind with two green square waves.
What it was used for
Micrologic was Fairchild's first line of integrated circuits. The first element, the "F" flip-flop, combined 4 transistors and 5 resistors - 9 components - on one silicon die. Robert Norman designed the circuit as a modified Direct Coupled Transistor Logic (DCTL), because that suited the early planar process; Jay Last led development from August 1959. The first monolithic devices worked on 26 May 1960, and circuits using the isolation method later preferred on 27 September 1960. The element was publicly introduced in March 1961 at a press conference at the IRE Show in New York, with a photograph in LIFE.
What was new
What was new was not the circuit in one piece - Jack Kilby built a working version with fine gold "flying-wires" at Texas Instruments on 12 September 1958 - but the connection: Jean Hoerni's planar process and Robert Noyce's patent let the conductors be formed as deposited metal, with no flying wires. Production depended on that; only a few dozen of Kilby's type went out for evaluation, and TI's first commercial device, the Type 502 flip-flop of March 1960, cost $450 each. Isolating the devices in the crystal was at first unsettled: first channels etched from the back and filled with epoxy, then the p-n isolation that Isy Haas and Lionel Kattner developed with a boron diffusion.
What the sources leave open: All dates, component counts and prices from the Computer History Museum (Silicon Engine). Its page on the planar circuit does not mention Kilby; the date 12 September 1958 and the $450 price appear on its page about the "solid circuit". Who contributed what to the p-n isolation is only briefly described there, and the source mentions no dispute between Last and Noyce.