Crystal Detector
A crystal detector is a radio-frequency detector in which a mineral crystal, together with a metal contact, rectifies the incoming signal so that its audio can be heard. The family includes galena (lead sulfide), silicon, carborundum (silicon carbide), iron pyrites (iron sulfide, fool's gold) and combinations such as zincite with chalcopyrite. This entry looks at the materials and their role in radio, while the companion cat-whisker entry looks at the point contact itself and its link to the transistor.
Each mineral had its own personality, and radio amateurs chose among them by sensitivity and stability. Galena was the most sensitive and the cheapest, with lead-gray cubic crystals mined as lead ore, but its surface had only a few responsive spots and the fine wire needed a feather-light touch. Iron pyrites worked in a similar fashion and was a common cheap alternative. Silicon, pressed against a metal contact following Greenleaf Whittier Pickard's 1906 patent, was stabler though less sensitive. Carborundum, an artificial silicon carbide first made industrially in the 1890s, was proposed as a detector by Henry Dunwoody in 1906; it was hard and robust, handled firm contact pressure, and usually needed a small bias battery, but it tolerated the strong signals of a nearby spark transmitter. Pickard's Perikon detector of about 1907 paired zincite with chalcopyrite and used a clamped contact, a more stable arrangement than a roaming wire.
The reason for sensitive spots, and for their unreliability, was that natural crystals are not uniform. What the detector needed was a region where the surface had the right balance of impurity atoms and defects, so that a rectifying barrier formed under the metal point. Natural galena varied in composition from one lump to another, a trace of impurities that nobody could measure then. A given spot could drift with moisture, oxidation or vibration, and the user restored it by tapping the mineral or moving the contact. Because the barrier was a few atoms thick, a tiny surface change meant a large change in sensitivity. Understanding this required the semiconductor physics that arrived in the 1930s, and the concept of deliberate doping was introduced to make the result repeatable rather than a lottery.
The crystals were doing real semiconductor physics a generation before the theory existed. Pickard and others noted that some crystals were good, some were not, and that the quality of the result changed with purity, which was a form of empirical materials science. In 1907 H. J. Round, a Marconi engineer, saw yellowish light emitted from a carborundum detector when current was passed through it, the earliest observation of electroluminescence, the principle of the light-emitting diode. In the 1920s the Russian engineer Oleg Losev found that a zincite contact biased into a particular region showed negative resistance and could amplify and oscillate, building a receiver he called a crystadyne; although it was abandoned in favor of tubes, it resembled a solid-state amplifier in spirit.
Crystal detectors reappeared when new frequencies defeated the vacuum tube. Radar developed in World War II worked at microwave frequencies where tube capacitances and transit times made detection and mixing difficult, and point-contact silicon diodes served as mixers. Research programs in the United States, notably at Bell Labs, MIT's Radiation Laboratory and Purdue, learned to purify silicon and germanium and to control their doping, which put the engineers into a position to build transistors afterwards. The knowledge that the unhelpful quirks of galena had originated in impurity was one of the most valuable by-products of the war.
A common misconception is that crystal detectors failed because crystals are inherently inferior. In fact the physical mechanism was sound and survives in the modern Schottky diode and in silicon carbide power electronics, where the same carborundum material is now grown as controlled wafers and used in electric-vehicle inverters and fast chargers. What failed was reproducibility: nobody could manufacture a good contact on demand. Another misunderstanding is that a detector amplifies; it only rectifies, and so the received signal can never exceed the power collected by the antenna.
Read crystal detector as a catalog of minerals that behaved as semiconductors before the word had a theory: each crystal was a lesson about purity, surface and contact that the transistor builders later turned from luck into engineering.