The field-effect transistor (GET), like the bipolar. junction transistor, is a three- . terminal semiconductor device. However, the FER operates under principles completely different from those of the BET. A field-effect transistor is called a uni polar . device because the current through it results from the flow of only one of the two
kinds of charge carriers: holes or electrons. The name field effect is derived from the fact that the current flow is controlled by an electric field set up in the device by an externally applied voltage. There are two main types of Fetch: the junction field-effect transistor (JET)
and the metal-ox IDE-semiconductor GET (SAMOSET). We will study the theory and some practical applications of each. Both types are fabricated as discrete components and as components of integrated circuits. The SAMOSET is the most important component in modern digital integrated circuits, such as microprocessors and computer memories.
JUNCTION FIELD·EFFECT TRANSISTORS
Figure 7-1 shows a diagram of the structure of a FEET and identifies the three terminals to which external electrical connections are made. As shown in the figure, a bar of N-type material has regions of P material embedded in each side. The two P regions arc joined electrically and the common connection between them is called the gate (G) terminal. A terminal at one end of the N-type bar is called the drain (D). and a terminal at the other end is called the source (S). The region of N material between the two opposing P regions is called the channel. The transistor shown in the figure is therefore called an Archangel JET. the type that we will study initially. while a device constructed from a P-type bar with embedded N regions is called a Channel JET. As we develop the theory of the JFET, it may be helpful at first to think of the drain as corresponding TB the collector of a BET, the source as corresponding to the emitter, and the gate as corresponding to the base. As we shall see, the voltage applied to the gate controls the now of current between drain and source, just as the signal applied to the base of a controls the flow of current between collector and emitter.When an external voltage ;.; connected between the drain and the source of N-channel , so that the drain is positive with respect to the source, current IS establish by the flow of electrons through the N material from the source to the drain. (The source is so named because it is regarded as the origin of the electrons.) Thus, conventional current flows from drain to source and is limited by the resistance of the N material. In normal operation, an external voltage is applied between the gate and the source so that the PN junctions on each side of the channel are reverse biased. Thus, the gate is made negative with respect to the source, as illustrated in Figure 7-2. Note in the figure that the reverse bias causes a pair of depletion regions to form in the channel. The channel is more lightly
doped than the gate, so the depletion regions penetrate more deeply into the N-type channel than into the P material of the gate.
The width of the depletion regions in Figure 7-2 depends on the magnitude of the reverse-biasing voltage VI;s. The figure illustrates case where is only a few tenths of a volt, so (he depletion regions are relatively narrow. (V/l.\’ is also assumed to be relatively small; we will investigate the effect of a large Vo.!’ presently.) As V(;s is made more negative, the depletion regions expand and the width of the channel decreases, The reduction in channel width increases the resistance of the channel and thus decreases the flow of current If) from drain to source. To investigate the effect of increasing the drain current ID, let us suppose for the moment that the gate is shorted to the source (V GS = 0). As is increased slightly above 0, we find that the current If) increases in direct proportion to it.