Cat-Whisker Detector
A cat-whisker detector is a radio detector made by touching a fine, springy metal wire, the whisker, to the surface of a semiconducting crystal such as galena (lead sulfide). The small contact area passes current much more easily in one direction than the other, which allowed early radio receivers to rectify an incoming signal and recover its sound without a battery or vacuum tube. It is the artifact in which a semiconductor first did useful electronic work, decades before anyone could explain why.
Rectification at a metal-crystal contact was first reported in 1874 by Karl Ferdinand Braun. Braun, a German physicist who later shared the 1909 Nobel Prize in Physics with Guglielmo Marconi, noticed that the current through certain metal sulfides depended on the polarity and the pressure of the contact. This was an unexplained curiosity in the laboratory, and it contradicted Ohm's law in the form physicists knew it. Braun's observation lay almost unused until wireless telegraphy needed a way to detect weak signals, when the coherer, a tube of metal filings, proved slow, fussy and noisy.
The use for radio was developed in the 1890s and 1900s by several people in parallel. Jagadish Chandra Bose in Calcutta worked on millimeter-wave radio in the 1890s and built detectors using metal-semiconductor contacts, including galena, receiving a US patent in 1904 for a detector of electrical disturbances. In 1906 Greenleaf Whittier Pickard, an American engineer working with the American Wireless Telephone and Telegraph company, found that a point contact on a piece of silicon gave a workable detector and patented it; he went on to test thousands of minerals and later founded a company marketing detector crystals. In the same year Henry Dunwoody patented carborundum (silicon carbide) as a detector crystal. The device that resulted, a wire probing a mineral, became the standard receiver detector until the vacuum tube took over in the 1920s.
A galena detector was adjusted by hand, because the working spots were scarce and fragile. The operator stroked the whisker across the crystal face until the headphones suddenly gave a clear signal, a sensitive spot, then protected it from jarring; a slammed table could ruin the adjustment. The spots were regions where impurities or crystal defects happened to make a good barrier, a lottery that no one could control without knowing about doping. This lack of reproducibility is exactly why the vacuum tube, with its steady and predictable behavior, replaced the detector in commercial radio, despite the tube's cost and battery demands.
Physics eventually caught up, and then the whisker came back in wartime. Alan Wilson's 1931 band theory of semiconductors explained conduction in impure crystals, Walter Schottky and Nevill Mott described the barrier at a metal-semiconductor contact in 1938 and 1939, and the work of Russell Ohl at Bell Labs on silicon in 1940 uncovered the pn junction. Microwave radar, in which vacuum tubes failed at high frequency, relied on silicon point-contact diodes, with the 1N21 and 1N23 types produced by the thousands during World War II, and germanium diodes were developed in parallel at Purdue and elsewhere. Teams that had studied these purified crystals built the knowledge that led to the transistor.
The point-contact transistor of December 1947 is, quite literally, a cat-whisker detector with two whiskers. John Bardeen and Walter Brattain at Bell Labs placed two gold contacts very close together on a block of germanium and found that current at one could control current at the other, with gain. The device was described as a development of the old rectifying contact, a fact that surprised later observers who assumed the transistor had come from a wholly new apparatus. Point-contact transistors were then displaced by junction transistors because they were noisy and difficult to make consistently, the same weakness that had sunk the cat whisker.
One misconception is that the whisker detector was a pn junction, which it was not. Its barrier was formed between a metal and the semiconductor surface, closer to a modern Schottky diode, and it worked without any external power. Another is that the name refers to an animal: the thin wire simply resembles a cat's whisker. Echoes survive in current technology; the probe tips used to test chips on a wafer are fine wires that touch pads, and Schottky diodes are made in large numbers inside radio-frequency circuits and power converters, but now with controlled metal films rather than a hunted spot.
Read cat-whisker detector as the semiconductor's first job description: a barely understood, hand-tuned contact that taught radio what rectification could do, and left a puzzle about crystals that physics and the transistor eventually solved.