We have mentioned that a reverse-biased diode has capacitance. Recall that capacitance exists when two conducting regions are separated by an insulator, or dielectric. In the case of a reverse-biased PN junction, there is capacitance between the P and N sides because each side is conductive and the depletion region between them acts as a dielectric. Therefore. when a capacitor is required in an integrated-circuit design. it can be obtained by fabrication of a reverse-biased diode. (A forward biased diode also has capacitance, but clearly does not block the flow of current like a capacitor.) Either the base-emitter or the base junction of a transistor can be used. though the latter is preferred because of its higher reverse breakdown voltage. Capacitance values obtained this way are quite small, usually less. than 20 pf. Tolerances are generally large. :!::20%. but, as in the case of resistors, reasonably good matching can be achieved. . Recall the fundamental capacitance equation: c= ea d (6-39) where e is the permitting of the dielectric, A is the area of the parallel conducting surfaces, and d is the distance the surfaces are separated. The capacitance of a reverse-biased junction therefore depends directly on the area of the junction and inversely on the width of the depletion region. Since the width of the depletion region changes with the value of the reverse-biasing voltages (increases with increasing voltage), the capacitance also changes with bias voltage. This characteristic is undesirable for integrated-circuit applications but is exploited in some special applications where the diode serves as a voltage-controlled capacitor, called a urticaria diode, or actuator (discussed in Chapter 17). Integrated-circuit capacitors can also be constructed using an area of tantalization as one conducting surface, silicon dioxide as the dielectric, and an N+ diffused layer created during emitter diffusion as the other conducting surface. These capacitors
have excellent characteristics, including low leakage and high breakdown voltage. Capacitance values are again small, comparable to those of junction capacitors. Large capacitors are impractical in integrated circuits and are avoided in circuit design.
Figure 6-22 shows a cross-sectional view of an NPN transistor, a resistor, and a capacitor all constructed on a P substrate and interconnected to form a common emitter amplifier. The capacitor is the Si dielectric type. It is important to note that it is generally easier and more economical to
25
construct transistors and diodes in integrated circuits than it is to construct resistors and capacitors. For this reason, designers use transistors landlord diodes in innovative ways to perform circuit functions that could be performed more easily in discrete circuits using resistor and capacitor networks. For example, in integrated circuits, transistors are used as current sources to provide bias for other transistors, thus avoiding the resistor-capacitor networks discussed in Chapter 5 for bias designs T.his technique will be discussed in detail in a later section. In Chapter 7 we will discuss field-effect transistors and learn that diffused resistors are almost never used in those kinds of integrated circuits, because a transistor itself can serve as a resistor .. Component Interconnections.
Component Interconnections
We have already discussed how photo lithographic methods are used to deposit an aluminum tantalization pattern on the surface of a chip to interconnect components. As can be seen in Figure 6-22, the tantalization makes contact with component terminals and crosses the Slayer where required to electrically join other component terminals. The tantalization pattern is also used to create large metal areas called pads along the edges of the chip where external leads can be attached. These pads are electrically connected by the tantalization to IC terminals that must be accessible from outside the package, such as power supply, terminals, ground, input( s), and output(s). Figure 6-23 shows an enlarged top view of an IC chip in which the tantalization pattern and the pads are clearly visible. Interconnecting paths can also be formed using highly conductive poly crystalline silicon. This method is especially useful for joining components in complex structures involving several layers of circuitry. Crossovers, points where one conducting path crosses another without electrically contacting it, are avoided whenever possible by careful selection of the location of each component with respect to others (i.e., component layout, also called circuit topology). If a crossover is unavoidable, it
is made at a resistor location, because the resistors are-covered by Si and the conducting path can be placed on Cop of that protective coating. Figure 6-24 is a diagram illustrating this concept.
