Fairchild 2n697
The Fairchild 2N697 was a silicon NPN transistor introduced by Fairchild Semiconductor in 1958 as part of the 2N696 and 2N697 pair, and it is a landmark device because it was one of the first silicon transistors that could be bought in quantity for fast switching. It was a diffused-base mesa transistor: the base region was formed by diffusing impurities into the silicon, and the surrounding silicon was etched away to leave a raised plateau, or mesa, so that the junction area was small and the capacitance low. It was not a planar device, a distinction that is often blurred, since Fairchild's planar transistors, notably the 2N1613, came in 1960.
The mesa construction gave it speed and power handling that the earlier alloy and grown devices lacked. In a mesa transistor, the collector is the bulk of the wafer, the base is a thin diffused layer, and the emitter is a smaller diffused region on top. Etching a mesa around the active region limits the collector-base junction area, which cuts capacitance and increases switching speed, and diffusion controls the base thickness precisely, so that gain can be consistent from one device to the next. Silicon added further advantages over germanium: it tolerated higher temperatures and had much lower leakage current, a feature that mattered in cramped, hot military and computer equipment.
Its early customers came from computing and defense, and that shaped the product. Fairchild supplied the 2N697 to IBM for computer logic and to Autonetics, the North American Aviation division that built guidance systems for the Minuteman missile, among others. These buyers paid prices far above those of consumer transistors, with early quotes reported at roughly 150 dollars per device, and in return they demanded uniformity and reliability. That revenue gave the young company a foothold and justified investment in better process control, and the performance targets set by these buyers influenced how Fairchild engineered later parts.
The mesa transistor had a weakness that the planar process was designed to cure. At the edge of the mesa the collector-base junction was exposed to the air, and any ionic contamination or moisture could change surface conditions, producing leakage current and drifting characteristics. Manufacturers dealt with this by cleaning the surface and sealing devices in metal cans, but the root problem remained. Jean Hoerni's planar process kept the oxide grown during diffusion in place as a protective cover over the junction, and the first planar transistors, including the 2N1613 in 1960, were more stable and could be made in larger numbers at lower cost. After that, planar silicon gradually displaced mesa devices as the standard.
The 2N697 matters historically because it helped make silicon, rather than germanium, the mainstream semiconductor. Texas Instruments had announced a commercial silicon transistor in 1954, but early silicon devices were slow and expensive. The diffused mesa design showed that silicon could also be fast, and the military and computer market was willing to pay for it. A common misconception, repeated in some histories, is that the 2N697 was the first planar transistor or the first silicon transistor; it was neither, but a stepping stone that used diffusion before the oxide-protected structure arrived. Another is to treat its high early price as typical: prices fell rapidly as yields rose and volumes grew.
The part number also illustrates the 2N naming system of the era, in which a registered 2N number identified a standard device that other manufacturers could second-source. The 2N697 stayed in catalogs for decades as a general-purpose silicon switching and amplifier transistor in metal can packages, and its descendants and its planar successors are still imitated in modern small-signal parts.
Read Fairchild 2N697 as the bridge device of the silicon transition: a fast, diffused mesa transistor that proved silicon could be both quick and manufacturable, and whose exposed junction edge made the case for the planar process that replaced it.