The methods used to bias amplifiers in integrated circuits are substantially different from those used in discrete circuits. To understand why this is the case and to appreciate the constraints imposed by integrated circuit technology on circuit design in general, it is necessary to learn some fundamental facts about how integrated circuits are constructed. Applications and Classifications of Integrated Circuits We have in several previous discussions that diodes and transistors are found in both discrete- and integrated-circuit form. Discrete components are packaged in individual enclosures and have leads that allow electrical connections to be made between component terminals, such as emitter, base, and collector. Discrete circuits are composed of discrete components, including resistors, capacitors, and/or inductors, interconnected by wires or through conducting paths on a
printed circuit board. Figure 6-13(a) shows a typical discrete transistor circuit. In contrast, the components of an integrated circuit are all constructed on a single, ‘tiny piece of semiconductor crystal, called a chip, which may contain hundreds of diodes, transistors, resistors, and/or capacitors. (Inductors are not constructed in integrated-circuit form.) The conducting paths that interconnect the components (a) (b) Figure 6-/3 (II) Discrete (/1/(1 (/I) integrutrd circuits. Tlu: (1//1 holds in its mouth (/11 integrated-circui: chill tluu could contain all th •. circ uitry discrete circuit «b): courtesy of Philips Science and industry Division) .
of an integrated circuit are contained entirely within the device, and the only leads that are brought out are those necessary for power supply connections. grounds, and circuit inputs and outputs. Figure 6-13(b) shows a typical integrated-circuit chip of the size that might well contain all of the discrete circuitry shown in Figure 6-13(a). An integrated circuit (IC) fabricated entirely on a single silicon chip is
called a monolithic l’C, A hybrid integrated circuit contains one or more monolithic circuits interconnected with external resistors and capacitors using thin-film or thickfilm techniques. , One obvious advantage of integrated circuits over their discrete counterparts is the fact that ‘very complex circuits can be constructed in very small packages, with attendant savings in wiring, assembly, cooling requirements, and material costs. Besides the economic benefits derived from miniaturization, integrated circuits are inherently more reliable than discrete circuits. Because the Ie components are contained in a single rigid structure, an Ie is not so susceptible as a discrete circuit to the kinds of mechanical failures that afflict the latter: ruptures and shorts in interconnecting authenticated by shock and vibration; connector alignments; solder-joint failures; and so forth. Furthermore, a good deal of back-up or redundant circuitry can be included in an Ie at very little additional cost, thus ensuring
satisfactory performance in the event that some components do fail. Back-up circuitry is feasible because circuit complexity is not so significant a cost factor as it is in discrete circuits, due tQlthe nature of the manufacturing process. As we shall see in later discussions, the number of components that can be fabricated on a single chip is limited more by the need for maintaining isolation between components than by their cost. Although numerous steps are required to manufacture an lC, a large number of identical devices can be fabricated simultaneously, contributing to cost reduction. t
–. Finally, integratedcircuits are advantageous in high-fr-equency applications and in high-speed computer circuits because of the small distances that electrical signals travel between individual components. Long signal paths create delays and phase shifts that limit the frequency at which such circuits can operate.
The starting point in the manufacturing procedure for both discrete components and integrated circuits is the production of a single symmetric crystal, most often a silicon crystal. Silicon is obtained from certain chemical compounds, but is not initially in a form suitable for semiconductor devices. In its natural form, silicon is said to be poly crystalline, because it is composed of a large number of crystals having different orientations. To obtain a single crystal of uniform orientation. it is necessary to melt the poly crystalline silicon and then allow it to cool and solidify under certain closely controlled conditions. The location in the “melt” where cooling takes place and the rate at which cooling occurs are particularly critical. In one manufacturing process. a crucible of molten silicon is pulled slowly through a furnace and the crystal forms where the cooling melt emerges. The development of a single crystal by controlled cooling of molten material is called crystal growth. A seed 1 is often used as a “starter” upon which the crystal is grown. In the process most often used for production. called the technique, the seed crystal is brought into contact with the molten crystal formed using the technique is in the shape of a cylinder up to 5 in. in diameter and several feet long, This cylindrical ingot is then sliced, using a diamond cutter. into thin wafers, about 0.5 mm thick. Figure 6-14(a) shows typical cylindrical ingot