1956 · Germanium Phototransistor · Work 828 of 1,871
A tiny phototransistor of 1956 on the workbench of a radio amateur in England in the evening, a hand in a grey jumper sleeve holding with tweezers a small glass tube one centimetre long, two thirds of it lacquered a plain smooth sky blue, its head clear glass with a rounded lens tip, a narrow brown-red ring at its base with three thin silvery wires, the clear tip held into the cone of light of a desk lamp, below it green felt, a small perforated board with a tiny glowing bulb, a soldering iron in its wire stand, a pocket knife, lamp light from the right.
What it was used for
The phototransistor is a transistor controlled by light rather than by the current of the emitter. It was invented by John N. Shive at Bell Telephone Laboratories; the Bell Laboratories Record of May 1950 describes the device as a cylinder about as big as a .22 calibre rifle cartridge, with a germanium wafer only 3 thousandths of an inch thick at the contact point. The same article names toll dialling in the Bell System as a possible application, alongside the existing photocell uses: automatic doors, television, sound motion pictures, wirephotos, industry. It was not Bell that brought it to market but the Radio Receptor Company of Brooklyn: the Transistor Museum lists the RR66 as a germanium PNP alloy junction photo transistor, available from mid-1954, with the stated purpose of light sensing and switching; an August 1954 advertisement in the Proceedings of the IRE named industrial controls, punch card and perforated tape readers, street light control, sound-on-film, burglar alarms, automatic door openers and counters among its applications; a metal base with a clear glass top, 3/8 inch in diameter.
What was new
Against the photocell, everything that means vacuum technology fell away: no envelope, no grid, no heated cathode. The Bell Laboratories Record of May 1950 gives as an advantage the high power output for a photo-electric device, in some cases enough to operate a switch directly without the preliminary amplification usually required. Added to that: good response to rapidly fluctuating light, particular sensitivity in the spectrum of ordinary incandescent lamps, low impedance and small size. The source stays cautious: the device is still in the experimental stage and is not expected to displace existing photocells, just as the transistor is expected to supplement rather than supplant tubes. Between that caution and the first marketed type lay 4 years, and it came from a small firm in Brooklyn, not from Bell or Western Electric.
What the sources leave open: The year 1956 could not be substantiated. What is documented is May 1950 (Shive, Bell Laboratories Record) and mid-1954 (RR66, first photo transistor marketed, per the Transistor Museum). For the Mullard OCP71 only a self-published page claims 1956, and that page is demonstrably wrong on neighbouring dates; the Museum of Technology dates the OCP71 to the 1960s. Sensitivity figures, prices and unit numbers are not documented.