The transistor is the fundamental component of most analog and digital circuits. A transistor consists of a piece semiconducting crystal (usually silicon, but sometimes germanium or gallium arsenide) with which impurities such as antimony, bismuth, arsenic, and phosphorous have been mixed by a process called doping. Doping increases the semiconductor's conductivity (tendency to carry electric current) by introducing free electrons that act as negatively-charged current carriers or, conversely, by introducing positively-charged gaps in the semiconductor crystal's electron structure, holes, which act as positively-charged current carriers. The difference between the number of electrons in the outermost orbital of the added impurity (dopant) and that of the host semiconductor determines whether the dopant introduces extra electrons (producing regions of "n-type" semiconductor) or extra holes (producing regions of "p-type" semiconductor).
In 1926 Julius Edgar Lilienfield of New York filed a patent on what is now acknowledged as the blueprint for an n-p-n junction transistor. Lilienfield's title for his proposed device was "Method and Apparatus for Controlling Electric Currents," but it was never developed into a commercial application. In the mid-1940s, in order to expand and improve its telephone business, AT&T's Bell Telephone Laboratories began to develop a solid-state semiconductor switch to replace the vacuum tube. As a result, a team of scientists led by experimental physicist Walter Brattain (1902-1987), theoretical physicist John Bardeen (1908-1991), and theoretician and project team leader William Shockley (1910-1989) developed the first practical transistor in December 1947, a point-contact device made from gold foil and a chunk of germanium. Upon learning of this invention Shockley developed the bipolar junction transistor, which was sturdier and easier to manufacture than Bardeen's and Brattain's device. For these achievements the three scientists shared the 1956 Nobel Prize in physics. Today Shockley is generally noted as the initiator and director of the research program that led to the transistor's discovery, and Brattain and Bardeen are credited with the actual invention of the transistor. The word transistor was formed at this time from TRANSfer resISTOR. The company publicly announced the transisto" on June 30, 1948, but its announcement received little attention. Nonetheless, by the 1950s transistors began to rapidly replace vacuum tubes, especially in radios, and eventually became one a key ingredient of all computers.
There are two basic types of transistors, the bipolar junction transistor (BJT) and the field effect transistor (FET). Both BJTs and FETs exploit the properties of p-type and n-type doping to enable a control voltage to control a current flow, much as a faucet valve controls the flow of water through a pipe. Imagine that a strong flow of water is passing through a faucet: twisting the faucet knob rapidly back and forth will produce a corresponding series of ups and downs in the water flow, turning a fairly small physical signal into a much larger one. This is essentially how analog signal amplification using transistors works. Or, one could simply use the valve to turn the water ON and OFF; this is mode in which transistors are used to process binary signals representing 1s and 0s.
The faucet-control effect is achieved by BJTs and FETs as follows. Both n-type and p-type regions of doped silicon conduct electricity (i.e., allow charges to flow through them), but differently: An n-type region conducts current by passing negative charges--electrons--through its crystal structure, and a p-type region does so by shifting holes as if they were mobile positive charges. The central feature of a transistor is the "p-n junction"--a plane where a p-type region touches an n-type region, like cheese lying on bread. The holes in the p-type region tend to steal electrons from the parts of n-type region they are nearest to (since opposite charges attract), which deprives a thin layer of the n-type material at the p-n junction of electrons. But without free electrons, n-type material has nothing to carry current with; the electron-free layer of n-type material thus acts as an insulator, a barrier to current flow, especially to current flow that tries to make electrons move further into the n-type region, which widens the depletion layer (layer depleted of electrons). The p-n junction is thus a one-way barrier to electron flow. If a three-layer n-p-n sandwich is made, there will two such junctions, one on either side of the p-type layer. (One could also build a p-n-p sandwich, but for simplicity's sake this discussion will be restricted to the n-p-n case.) Clearly, no matter which way one tries to make current flow through two back-to-back p-n junctions, one of them will tend to block the current.
The current-valve structure of the transistor is now almost complete. Both BJTs and FETs exploit the n-p-n (or p-n-p) sandwich structure, but in different ways. (1) A BJT is arranged so that the large current to be controlled tries to flow from one n-type layer to the other. Because of the back-to-backp-n junctions in the p-n-p sandwich, however, it cannot. Therefore wires are attached to the edges of the p-type layer in the middle of the sandwich and a control voltage is used to inject electrons into it. Each electron fills in a positively-charged hole; if enough electrons are injected, the p-type layer is temporarily and reversibly transformed into a n-type layer, possessing extra electrons for conduction. As far as current flow is concerned, there is now an n-n-n sandwich, the p-n junctions have disappeared, and current can flow through freely. Varying the control voltage varies how many electrons are injected into the p-type layer, which in turn controls the amount of current that can flow through the transistor, just as a valve varies the water flowing through a pipe. (2) An FET arranges current to flow sideways through the n-p-n (or p-n-p) sandwich, through the p-type layer. Normally the current has no problem doing so, because it can slip easily between the depletion layers above and below the p-type layer. But by applying an appropriate voltage to both the n-type layers of the sandwich, one can cause the nonconducting depletion layers of the two p-n junctions to thicken inward, squeezing the conductive p-type path in the middle--perhaps shutting it off completely. Thus, as in the BJT, control of the supply of charge carriers implements a current valve.
Although FETs are somewhat slower than BJTs, they are cheaper, smaller and used less power. FETs are used as amplifiers, oscillators, and switches, being very suitable for the amplification of very small signals.
Until the invention of the transistor in 1947, digital circuits were composed of vacuum tubes, which also function as valves for current flow. (British engineering slang for vacuum tubes was "valves.") Compared to tubes, transistors are durable, small, resistant to physical shock, and inexpensive. They also use less energy. At one time, only discrete (physically separate) transistor devices existed; normally these were sealed in ceramic or plastic "cans," with wires poking out that could be connected to an electric circuit. Although discrete transistors are still used, the vast majority of transistors are built into integrated circuits as parts of microprocessors, memory chips, and the like. State-of-the-art microprocessors contain millions of tiny transistors (mostly FETs). Today, transistors are used in virtually every electronic device, including radios, televisions, computers, and communication satellites.
This is the complete article, containing 1,191 words
(approx. 4 pages at 300 words per page).
