1965 · Ceramic Dual In-Line Package · Work 855 of 1,871
An assembly bench in a computer works in California, a green circuit board drilled with rows of small holes lying in a plain wooden frame, along it a row of small oblong ceramic packages two centimetres long, each with a plain smooth unmarked dark grey lid and two rows of seven short silvery legs set into the holes, a hand in a white shirt sleeve pressing one more package into its row with the thumb, beside the board a further package still held in a flat silver stamped frame, its legs joined along a continuous edge strip, a reel of solder wire, small side cutters, warm work light from above.
What it was used for
In December 1965 Fairchild advertised the dual in-line package, a plug-in housing for integrated circuits, expressly for makers of commercial equipment: its leads had dimensional tolerances to fit holes drilled to commercial standards, one could use standard circuit boards, flow-solder it and handle it the way one would any discrete component. In August 1965 Fairchild, as reported in Electronics, counted among its major customers the Advanced Scientific Instruments division of EMR Corp. and Burroughs, both in computers, and Hughes Aircraft. The difference in assembly practice was the decisive point: flat packs were usually assembled under a microscope and joined one at a time, whereas a board filled with plug-in packages could be flow-soldered in a single pass. The insertion tool built by Technical Devices of Culver City placed more than 10 circuits a minute.
What was new
What was new was size and pitch, not miniaturisation. George Sideris, manufacturing editor of Electronics, wrote in August 1965 that next to the flat packs the package seemed big, crude and ungainly, but that it was nearly 4 times the size and therefore easier and cheaper to handle; it had 14 leads in 2 rows of 7, the rows 0.3 inch apart, on a 0.1 inch pitch, twice the flat-pack spacing, and printed wiring could run under the package. Fairchild was not alone: in May 1965 Texas Instruments had announced its own plug-in package with 2 rows of 8 round pins, the rows 0.2 inch apart, not interchangeable with Fairchild's, and Sylvania stated that its similar package resulted from an independent analysis of the market. The barb: because equipment assemblers did not want a single source, they pressured their suppliers into the same package style - the design became the industry standard and thereby common property.
What the sources leave open: No patent on the package could be found, including by inventor and assignee search; that is not proof of absence. The design year is open: the Computer History Museum says 1965, while the contemporary source dates only the introduction, in March 1965. The figures on assembly cost savings come from Fairchild itself, as the journalist marks them. Unit numbers and a package price are missing.
Connections
Every connection names its kind. “Forerunner” on its own would be a claim.