In Chapter 5, we referred to clipping as the undesirable result of overdriving an amplifier. We have seen that any attempt to push an output voltage beyond the limits through which it can “swing” causes the tops and/or bottoms of a waveform to be “clipped” off, resulting in distortion. However, in numerous practical applications, including waveshaping and nonlinear function generation, waveforms are intentionally clipped. how the transfer characteristic of a device is modified to reflect the fact that its output is clipped at certain levels. In each of the examples shown, note that the characteristic becomes horizontal at the output level where clipping occurs. The horizontal line means that the output remains constant regardless of the input level in that region. Outside the clipping region, the transfer characteristic is simply a line whose slope equals the gain of the device. This is the region of normal, linear operation. In these examples, the devices are assumed to have unity gain, so the slope of each line in the linear region is 1.
Figure 15-12 illustrates a somewhat different kind of clipping action. Instead of the positive or negative peaks being “chopped off,” the output follows the input when the signal is above or below a certain level. The transfer characteristics show that linear operation occurs only when certain signal levels are reached and that the output remains constant below those levels. This form of clipping can also be thought of as a special case of that shown in Figure 15-11. Imagine, for example, that the clipping level in Figure lS’:”ll(b) is raised to a positive value; then the result is the same(a).
Clipping can be accomplished using biased technique that is more efficient than overdriving an amplifier, Clipping circuits rely on the fact that diodes have very low impedances when they are forward biased and are essentially open circuits when reverse biased. If a certain point in a circuit, such as the output of an amplifier, is connected through a very small impedance to a constant voltage, then the voltage at the circuit point cannot differ significantly from the constant.
voltage. We say in this case that the point is clamped to the fixed voltage. (However we will reserve the term clamping circuit for a special application to be discussed later.) An ideal, forward-biased diode is like <.l closed switch, so if it is connected between a point in a circuit and a fixed voltage source, the diode very effectively holds the point to the fixed voltage. Diodes can be connected in operationalamplifier circuits, as well as other circuits, so that they become forward biased when a signal reaches a certain voltage. When the forward-biasing level is reached, the diode serves to hold the output to a fixed voltage and thereby establishes a clipping level.A biased diode is simply a diode connected to a fixed voltage source. The value and polarity of the voltage source determine what value of total voltage across the combination is necessary to forward bias the diode. several examples. (In practice, a series resistor would be connected in each circuit to limit current flow when the diode is forward biased.) In each part of the figure, Can write Kirchhoff’s voltage law around the loop to determine the value of input voltage Vj that is necessary to forward bias the diode. Assuming that the diodes are ideal (neglecting their forward voltage drops), we determine the value of u, necessary to forward bias each diode by determining the value of Vj necessary to make Vo >O. Whenever Vj reaches the voltage necessary to make VI> > 0, the diode becomes forward biased and the signal source is forced to, or held at, the de source voltage.
If the forward voltage drop across the ‘diode is not neglected, the clipping level is found by determining the value of Vj necessary to make V” greater than that forward drop (e.g., Vo > 0.7 Y for a silicon diode). Although these conditions can be determined through formal application of Kirchhoff’s voltage law, as shown in the figure, the reader is urged to develop a mental image of circuit behavior in each case. For example, in (a), think of the diode as being reverse biased by 6 Y, so the input must “overcome” that reverse bias by reaching +6 Y to forward bias the ‘diode.
Three examples of clipping circuits using ideal biased diodes and the waveforms that result when each is driven by a sine-wave input. In each case, note that the output equals the dc source voltage when the input reaches the value necessary to forward bias the diode. Note also that the type of clipping we showed in Figure IS-11 occurs when the fixed bias voltage tends to reverse bias the diode, and the type shown in Figure lS-12 occurs when the fixed voltage tends to forward bias the diode. When the diode is reverse biased by the input signal, it is like an open circuit that disconnects the de source, and the output follows the input. These circuits are called parallel clippers because the biased diode is in parallel withc output. Although the circuits behave the same way whether or not one side of the de voltage source is connected to the common (low) side of the input and output, the connections shown in Figure lS-14(a) and (c) are preferred to that in (b) because the latter uses a floating source.
A biased diode connected in the feedback path of an inverting operational amplifier. The diode is in parallel with the feedback resistor and forms a parallel clipping circuit like that shown in Figure IS-14(a). Since vis at virtual ground, the voltage across R, is the same as the output voltage Value Therefore, when the output voltage reaches the bias voltage E, the output is held at E volts. Figure lS-1S(b) illustrates this fact for a sinusoidal input. As long as
the diode is reverse biased, it acts like an open circuit and the amplifier behaves like a conventional inverting amplifier. Notice that output clipping occurs at input voltage -(R\/R,}E, since the amplifier inverts and has closed-loop gain magnitude R,IR1_ The resulting transfer characteristic is shown in Figure 15-15(c). This circuit is often called a limiting circuit because it limits theoutput to the de level clamped by the diode. (In this and future circuits in this section, we are omitting the bias compensation resistor, Re, for clarity; it should normally be included, following the guidelines
In practice, the biased diode shown in the feedback of Figure IS-1S(a) is often replaced by a zener diode in series with a conventional diode. This arrangement eliminates the need for a floating voltage source. We will study zener diodes in more detail in Chapter 18 (Section 18-7) and will learn that in many respects they are equivalent to biased diodes. Figure 15-16 shows two operational-amplifier clipping circuits using zener diodes. The zener diode conducts like a conventional diode when it is forward biased, so it is necessary to connect a reversed diode in series with it to prevent shorting of RI. When the reverse voltage across the zener ,diode reaches Vz, the diode breaks down and conducts heavily; while maintaining an essentially constant voltage, Vz, across it.
Under those conditions, the total voltage across equals Vz plus the forward drop, VI>, across the conventional diode. Double-ended limiting circuits, in which both positive and negative peaks of the output waveform are clipped.
Conventional parallel clipping circuit and (b) shows how double-ended limiting is accomplished in an operational-amplifier circuit. In each circuit, note that 1)0 more than one diode is forward biased at any given time, and that both diodes are reverse biased for -£1 < Vo < £2, the linear region. A double-ended limiting circuit using back-to-back zener diodes. Operation is similar to that shown in Figure 15-16, but no conventional diode is required. Note that diode D. is conducting in a forward direction when D2 conducts in its reverse breakdown (zener) region, while D2 is forward biased when D, is conducting in its reverse breakdown region. Neither diode conducts when -(VZ2 + 0.7) < u, < (VZI + 0.7), which is the region of linear amplifier operation.