P1: FJU/FFV Revised Pages P2: FJU Qu: 00, 00, 00, 00

Encyclopedia of Physical Science and Technology EN001H-24 May 26, 2001 14:28

Analog-Signal Electronic Circuits

Donald T. Comer
David J. Comer
Brigham Young University

I. Introduction to Analog- and Mixed-Signal

Electronic Circuits
II. The Analog Amplifier
III. Digital-to-Analog and Analog-to-Digital
Converters
IV. Mixed-Signal System Applications

GLOSSARY construction of circuits that perform mathematical op-

erations such as summing or integration of input analog
Analog-to-digital conversion The conversion of an input signals.
analog voltage or current to a digital output code that Sample-and-hold circuit An electronic circuit that sam-
represents the value of the input analog signal at the ples a time-varying input analog signal under control
time of conversion. of a digital sample gate and generates a constant out-
Analog signal A voltage or current that is continuous in put signal equal to the input signal at the time the gate
time and carries information that is related to the signal. signal is asserted. When the gate signal is not asserted,
Binary counter An electronic circuit that registers the the output voltage holds constant at its previous value.
occurrence of pulses on an input line and generates an VLSI circuits An abbreviation for very large scale in-
output binary code representing the number of pulses tegrated circuits. These circuits are fabricated on or
that have occurred since a particular starting time. within a silicon substrate and may include 10,000 to
Comparator circuit An electronic circuit that compares several million transistor devices.
two input signals and generates one of two possible
output levels, depending on which input signal has the
largest magnitude. ELECTRONIC CIRCUITS may be classified as either
Digital-to-analog conversion The conversion of an input analog or digital circuits on the basis of their intended
signal, coded as an N -bit digital code, to an output application. For many years, the design of analog circuits
analog voltage or current that has a value determined was treated as a separate area from that of digital circuit de-
by the numeric value of the input code. sign. Today, it is becoming common practice to design cir-
Op amp (operational amplifier) An electronic ampli- cuits that include both analog and digital features merged
fier that has a very high voltage gain and allows the into a single product. This trend has been accelerated

531
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

532 Analog-Signal Electronic Circuits

by the availability of integrated circuit processes that al- analog and digital circuits and to fabricate both on a com-
low the fabrication of both kinds of circuits on a common mon substrate. This became known as mixed-signal VLSI
substrate. Circuits that combine both analog and digital processing and is utilized today by a high percentage of
features have come to be known as mixed-signal circuits. commercial integrated circuit companies.
The most widely used substrate material is silicon. Early
integrated processes concentrated on the fabrication of
I. INTRODUCTION TO ANALOG- AND bipolar junction transistor (BJT) circuits. In recent years,
MIXED-SIGNAL ELECTRONIC CIRCUITS processes involving the metal-oxide-semiconductor field-
effect transistor (MOSFET) have become very popular.
Analog electronic circuits are designed to perform pro- The design of digital circuits on silicon can typi-
cessing functions on electronic signals. Analog signals cally take advantage of computer automation and expert
are typically low-level, nonquantized, continuously vari- computer-automated design systems. This results from the
able signals that are produced by analog transmitters or fact that digital designs are generally very structured and
transducers. For example, a radio station may transmit involve the use of a limited number of common build-
through its antenna a signal which consists of a sinusoidal ing blocks. However, the design of the analog portions
signal that varies in amplitude in accordance with an im- of a mixed-signal circuit often uses unique cells that do
posed modulating signal. Several stations in a geographic not readily lend themselves to design automation and may
area transmit simultaneously, each using a slightly differ- produce unpredictable and often harmful interactions with
ent range of frequencies. The function of a radio receiver is the digital circuitry. Substrate noise, instability, and tem-
to select the range of frequencies that includes the desired perature drift are factors that often cause undesirable in-
signal, amplify signals in this range, remove any unwanted teractions between analog and digital sections of a mixed-
components, and drive an audio speaker. The peak ampli- signal integrated circuit. Therefore, special techniques are
tude of the received signal may be about 50 µV at the required in the design of these circuits.
output of the receiving antenna. This signal must be am-
plified to a voltage level sufficient to drive a speaker. The
functions of selecting the desired frequency range (called II. THE ANALOG AMPLIFIER
filtering) and amplifying the signal are processes carried
out by analog electronic circuits. A. Amplifier Specifications
The electronic circuits of the early 1900s were imple-
The electronic amplifier is considered to be the fundamen-
mented by using vacuum tubes, and their use was almost
tal building block of analog circuit design. To construct
entirely centered around radio and communication appli-
an amplifier, one must utilize a device that is capable of
cations. In the 1930s, amplifiers were also used in the con-
power gain; that is, this device must be capable of being
struction of the transatlantic cable that was developed to
activated by a low-power input signal to produce a higher-
provide telephone service between the United States and
power output signal. The increased power is provided by
Europe. In the 1940s, analog electronic circuits were used
a dc power supply.
in radar, sonar, and some weapon-control applications as
The primary function of an amplifier is to generate an
well as in an expanded communication role.
output signal that is a replica of the input, but at a higher
With the invention of the bipolar junction transistor in
signal strength. Amplifiers may be designed specifically
1951 and its rapid development in succeeding years, the
for voltage gain, power gain, or current gain. The most
vacuum tube began to be replaced by transistors in am-
common use of an amplifier is as a voltage-gain device.
plifier applications. By the 1960s, engineers had replaced
Figure 1 shows a typical amplifier diagram where an input
vacuum tubes with transistors in most commercial analog
signal, Vi , is applied to its input terminal and an output
electronic applications.
voltage, Vo , is generated at its output. The voltage gain,
In 1964, Fairchild Semiconductor Corporation intro-
G v , of the amplifier may be expressed as
duced the industry’s first analog amplifier that was con-
structed by using integrated circuit technology. In this Vo
technology, all the resistors, capacitors, and transistors Gv = . (1)
Vi
required to produce the amplifier were fabricated simulta-
neously on a common silicon substrate. This technology, For example, an amplifier with a voltage gain of 100 would
originally known as integrated circuit technology, eventu- convert a 50-µV input signal to a 5-mV signal.
ally evolved into VLSI (very large scale integration), in In many applications the dc level of the input or output
which thousands of transistors can be fabricated on a sil- is not important. Rather, it is the variational or alternating
icon substrate. In the 1980s, it became feasible to merge portion of the signal that contains the information to be
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

Analog-Signal Electronic Circuits 533

processed. Amplifiers intended only to transmit the vari-

ational portions of an input signal are referred to as ac
coupled amplifiers while amplifiers that are intended to
amplify and transmit dc as well as variational levels are
referred to as direct coupled amplifiers.
In the typical amplifier circuit of Fig. 1, the resistance,
R L , represents the amplifier load. The power, Po , delivered
to the load can be expressed as
Vo2 (r ms)
Po = , (2)
RL
where Vo (rms) is the root-mean-square value of the output
voltage. The power delivered to the input of the amplifier FIGURE 2 Typical amplifier step response.
is
Vi2 (rms) decreases to 0.707 of its midband value. The frequency
Pi = , (3) where this occurs is approximately equal to the bandwidth
Ri
of the amplifier.
where Vi (rms) is the root-mean-square value of the input
Some amplifiers are designed to have a high gain over
voltage and Ri is the input resistance of the amplifier.
only a very narrow band of frequencies. These stages are
The power gain of the amplifier of Fig. 1 may be ex-
called narrowband, or tuned, amplifiers and are indispens-
pressed as
able in communication applications.
Po G 2 Ri The efficiency of an amplifier may be expressed as the
GP = = v .(4) ratio of the power delivered to the load to that drawn from
Pi RL
the power supply. For the amplifier of Fig. 1, the efficiency,
Other important parameters of an amplifier are its equiv-
η, may be expressed as
alent input noise and its bandwidth. The equivalent in- 
put noise is determined by measuring the noise voltage Vo2 (rms) R L
produced at the output, Vo (noise), and dividing it by the η= , (6)
VI SS
voltage gain to give
where V is the power-supply voltage and I SS is the average
Vo (noise) current drawn from the power supply.
Vi (noise) = . (5)
Gv Other amplifier parameters that are important to some
The equivalent input noise represents a lower limit on the applications are the rise time, the settling time, and over-
input-signal voltage that can be amplified. If the input shoot. Figure 2 shows an amplifier output waveform in
signal is smaller than the noise voltage, the signal will be which rise time, overshoot, and settling time are defined.
obscured in noise, even after amplification. In this case, the input to the amplifier is assumed to be a
The bandwidth of a wideband amplifier specifies the step function.
frequency range over which maximum amplification oc-
curs. This maximum, or midband, voltage gain is normally B. Types of Amplifiers
measured at a single moderate frequency where the am-
plifier behaves close to ideal. The frequency of the input Several important types of amplifiers are the narrowband
signal is then increased to a point where the amplifier gain amplifier, the wideband amplifier, the power amplifier, and
the operational amplifier, or op amp.

1. The Narrowband, or Tuned, Amplifier

The narrowband circuit is required in communication ap-
plications wherein signals contained in a small range of
frequencies are to be amplified. Several radio and televi-
sion stations may simultaneously transmit signals to a geo-
graphic area. Each signal has a different channel or carrier
frequency and transmits information over a small range
of frequencies centered about this carrier frequency. The
FIGURE 1 Typical voltage amplifier. receiving circuitry includes amplifiers that can be tuned to
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

534 Analog-Signal Electronic Circuits

the carrier frequency and amplify only the small range of amplifiers and generates an output voltage that repre-
frequencies near the carrier frequency. For example, the sents the dependent variable in the differential equation.
first amplifier stage in an FM radio receiver might have A system allowing such simulations is called an analog
maximum amplification at a frequency of 100.3 MHz. The computer.
range of frequency over which near-maximum amplifica- The op amp is also a building block in circuits that con-
tion occurs may extend from 100.2 to 100.4 MHz. vert between an analog signal and a corresponding digital
code. A circuit that accepts an analog input signal and
converts this signal to a corresponding N -bit output digi-
2. The Wideband Amplifier
tal code is called an analog-to-digital converter, or ADC.
The wideband stage amplifies signals falling within a cer- A circuit that accepts an N -bit digital code and converts it
tain broad range of frequencies. The highest frequency to a corresponding analog output signal is called a digital-
at which the circuit provides maximum amplification is to-analog converter, or DAC.
several orders of magnitude above the lowest frequency Prior to the 1960s, op amps were very expensive preci-
of maximum amplification. Amplification is constant be- sion circuits costing hundreds of dollars. In 1964, the first
tween these two frequencies. An example of a wideband op amp was fabricated on an integrated circuit chip. Since
amplifier is the video amplifier in television receivers, that time, the integrated circuit op amp has evolved into a
which amplifies signals from around 30 Hz to 4 MHz. Be- high-performance amplifier costing only a few cents.
low 30 Hz and above 4 MHz, the amplification decreases. Although the frequency performance of the op amp is
The ratio of upper-corner frequency to lower-corner fre- somewhat limited, it is a very important component in
quency is 133,333 in this case. In a narrowband circuit, electronic systems. It is used in many analog circuits as
this ratio is near unity. well as in circuits that combine analog and digital signals
such as the ADC and DAC.
3. The Power Amplifier
C. Op Amp Behavior
Many amplifying stages are designed to have high values
of voltage gain or current gain, but they produce an output Figure 3 shows a common symbol for the op amp. This
signal with very small power. The first few stages of ra- amplifier is designed to amplify the difference between the
dio or television receivers generally produce output power two input-signal voltages, V1 and V2 . The output voltage
levels in the milliwatt range. for the circuit of Fig. 3 is given by
Some applications require a high-power signal. Com- Vo = A(V1 − V2 ), (7)
mon examples of this are the signals that drive the audio
speakers in a public address system, a high-fidelity radio where A is the voltage gain of the op amp. Since the circuit
receiver, or a compact disc (CD) player. Smaller audio amplifies the difference between the two input signals,
speakers may require tens of watts while larger ones are it is referred to as a differential amplifier. Typical low-
driven by amplifier stages that produce hundreds of watts. frequency voltage gains for a general-purpose op amp are
These stages are called power amplifiers. 200,000–300,000 V/V.
If the voltage gain, A, is very large and the input re-
sistance is also large, the circuit of Fig. 3 approaches an
4. The Operational Amplifier, or Op Amp ideal op amp. When negative feedback is applied around
The op amp was so named as a result of the mathematical an ideal op amp, a voltage node designated a virtual ground
operations this device can perform on electronic signals. can be produced. Figure 4 shows one useful application
The op amp, along with a small number of resistors, can of an op amp and illustrates the idea of a virtual ground.
generate an output signal or voltage that is proportional In the circuit of Fig. 4, a signal that is proportional to
to the sum of several input signals. With a slightly dif- the output voltage is returned, or fed back, to the inverting
ferent configuration of resistors, the op amp can generate input of the op amp through the resistance R2 . This is
an output signal that is proportional to the difference of
two input signals. The addition of a capacitor can create a
circuit having an output signal that is proportional to the
integral of the input signal.
Several op amps, resistors, and capacitors can be in-
terconnected or configured in a manner that allows dif-
ferential equations to be simulated. The excitation func-
tion or independent variable is applied to the system of FIGURE 3 The operational amplifier (op amp).
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

Analog-Signal Electronic Circuits 535

(device M8 ). Other devices in this circuit provide dc bias

current and active loads for the gain stages.
Actual op amps such as that shown in Fig. 5 differ from
the ideal op amp defined by Eq. (7) and Fig. 3 in several
ways. A practical op amp has a finite, rather than an infi-
nite, input impedance and a finite, rather than zero, output
impedance. Actual op amps also exhibit imbalances that
cause the output to depend on input voltages and an input
offset voltage, VO S . This dependency can be expressed as
FIGURE 4 Inverting op amp. Vo = A(V1 − V2 − VO S ). (11)
The input offset voltage, VO S , is typically caused by mis-
called negative feedback and has the effect of reducing matches in the input stage, for example, in devices M1 and
the voltage gain of the circuit. It also causes the node x M2 in the circuit of Fig. 5, and it may be positive or neg-
to appear as a virtual ground. A virtual ground means that ative. This circuit may also function over a limited range
only a small voltage appears at this node and the resistance of input voltages, sometimes called the common-mode
to ground is very small, approaching zero. input-voltage range. The offset voltage, VO S , will also typ-
From Eq. (7), the voltage Vx may be determined as ically depend on both the common-mode input range and
Vo the power-supply voltage, +V . Thus, a practical op amp
.Vx = (8) will include specifications on CMR (common-mode in-
A
put range), CMRR (common-mode rejection ratio), and
If the output voltage, Vo , is finite, which is guaranteed
PSRR (power-supply rejection ratio). The bandwidth and
by the presence of negative feedback, then the voltage
gain of the op amp are usually specified, often as a gain–
Vx will approach a value of zero volts as A approaches a
bandwidth product, which is the product of the low fre-
large number. Thus, the voltage Vx is always a very small
quency gain and the bandwidth of the amplifier measured
voltage. While it will not be shown here, it also has a
in an open loop (nonfeedback) condition.
very small resistance to ground and, therefore, satisfies
the requirements of a virtual ground.
The virtual-ground node is sometimes referred to as
III. DIGITAL-TO-ANALOG AND
a summing junction since it allows the currents flowing
ANALOG-TO-DIGITAL CONVERTERS
through any connected resistances to sum together and
flow through the feedback resistor, R2 . For example, the
The digital-to-analog converter, sometimes called a D/A
current flowing into the node x through R1 is Vi /R1 . The
converter or a DAC, is one of the most commonly used
current flowing from node x through R2 is −Vo /R2 . As-
suming that no current flows into the input terminal of
the op amp, the satisfaction of Kirchhoff’s current law
requires that
Vi −Vo
= . (9)
R1 R2
This establishes the voltage gain of the circuit at
Vo −R2
G= = . (10)
Vi R1
Thus, for an ideal op amp, the gain is determined by the
resistor ratio R2 /R1 and is independent of the op amp gain
A as long as it is large.
Practical op amps are typically constructed from tran-
sistor devices arranged in a manner such as that shown
in Fig. 5. This typical op amp circuit is constructed from
MOS (metal-oxide semiconductor) transistor devices and
consists of an input differential stage (devices M1 and FIGURE 5 Typical MOS (metal-oxide semiconductor) op amp
M2 ), a second gain stage (device M9 ), and an output stage schematic.
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

536 Analog-Signal Electronic Circuits

binary value of one or zero. The signal B1 is assumed to be

the control for the first-bit, or most-significant-bit (MSB),
current, while B N is the control for the least-significant-bit
(LSB) current for an N -bit DAC. Although other types of
coding are available, the most common is binary coding.
The following equations are based on the assumption that
the DAC is binary coded.
In Fig. 7, a current designated the full-scale current,
FIGURE 6 The digital-to-analog converter (DAC). I F S , may be defined as
I F S = KI R E F , (12)
mixed-signal circuits; that is, it processes both analog and
digital signals in the same circuit. Its input signals are where K is a scaling factor determined by the ratio of R0
digital, often derived from a microprocessor or a computer to R.
bus, while its output is an analog signal. Figure 6 shows The current I1 , or MSB current, is chosen to be exactly
a block diagram of the DAC. The input to this converter one-half the full-scale current. Thus I1 is expressed as
is a digital word of a prescribed number of bits, while the KI R E F IF S
output is an analog signal uniquely related through coding I1 = = . (13)
2 2
to the input digital word.
A large number of D/A conversion techniques have been If all bit currents are assumed to be binarily scaled in
developed over the years. However, the DAC circuits most descending order, then the output current can be expressed
suitable for integrated circuit design may be classified into as
 
one of three basic categories: (1) current-scaling circuits, B1 B2 BN
IO = IF S 1 + 2 + · · · + N (14)
(2) voltage-scaling circuits, and (3) charge-scaling cir- 2 2 2
cuits. These approaches scale and sum precisely coded or
currents, voltages, or charge to generate the output analog
signal. The first category is generally best suited for bipo- N
I F S Bi
IO = , (15)
lar technology while the latter two are more widely used in i=1
2i
CMOS (complementary metal-oxide–silicon) technology.
where [B1 · · · B N ] is the binary input code.

A. The Current-Scaling DAC

The first monolithic DAC to become widely used in in-
dustry was the DAC-08, originally designed by Precision
Monolithics and Motorola. The DAC-08 uses the current-
scaling approach.
Figure 7 illustrates a current-scaling DAC that uses pre-
cision current sources set up by a reference amplifier,
which works with an R − 2R ladder and scaled n– p–n
bipolar junction transistors to create binary-weighted cur-
rent sources.
The reference amplifier in Fig. 7 consists of an op amp,
A1 , working in conjunction with transistor Q 0 and resistor
R0 to produce a circuit that forces the current I R E F into
the collector of Q 0 . The devices Q 1 through Q N can be
designed to produce currents that are equal to, or precisely
related to, I R E F by proper choice of R and the dimensions
of the transistors. The circuit of Fig. 7 allows for switching
between a true input, I O , and its complement, I O∗ . This may
be accomplished by a differential switch as illustrated in
Fig. 8.
The DAC of Fig. 7 can be characterized in terms of FIGURE 7 Current-scaling DAC: (a) block diagram; (b)
the following equations, assuming that all inputs have a schematic.
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

Analog-Signal Electronic Circuits 537

analog output between any two adjacent codes. It is

expressed as a percentage of full scale or in terms of an
LSB value.

B. The ADC
Analog-to-digital, or A/D, conversion is the reverse of the
FIGURE 8 Differential DAC switch. operation previously described as D/A conversion. The
function of an A/D converter, also called an ADC, is to
measure an analog input signal and provide a quantized
An alternative notation for this equation is representation of the signal in the form of a digital word.
I F S (n) Just as the DAC is considered a decoding device, the ADC
IO = , (16) can be considered an encoding device in the translation of
2N
signals between the analog and digital worlds. Figure 9
where n is the decimal equivalent of the binary input code
shows a block diagram representing an ADC.
[B1 · · · B N ].
The analog input signals consist of the input signal,
Vi , as well as a precision reference, VR E F , assumed to
1. DAC Specifications be a dc voltage. The output consists of a digital word of
N bits. In practice, the output may be generated either
Digital-to-analog converters are normally chosen for ap-
as a serial bit stream or in parallel as shown. However,
plication within a given system on the basis of their resolu-
for the purpose of definition, a parallel output with binary
tion and accuracy, which, in turn, is generally determined
coding will be assumed, with B1 representing the MSB and
by N , the number of input bits of the converter. The fol-
B N representing the LSB. In the block diagram of Fig. 9,
lowing is a listing and definitions of key DAC terms and
assuming signals are referenced to a common ground, the
specifications:
function of the ADC can be defined as that of generating
binary words [B] = [B1 , B2 , . . . , B N ] according to the
Resolution The resolution of a DAC is given by the num-
equation
ber of bits, N . The resolution is the smallest increment
of output that the DAC can produce. An 8-bit, DAC has Vi 2 N
n= , (17)
a resolution of 8 bits, or one part in 28 . This results in VF S
a percentage of 0.39%. where n is an integer and the decimal equivalent of [B],
Linearity Linearity is the maximum allowable deviation and VF S is the nominal full-scale voltage. The nominal
from an ideal straight line drawn between the zero-scale full-scale voltage of the ADC is related to the reference
and full-scale outputs. It is often given as a percentage voltage by the equation
or as a fraction of an LSB. For example, ( 12 )LSB lin-
earity for an 8-bit DAC corresponds to 0.195%. Some VF S = kVR E F , (18)
manufacturers call this parameter the nonlinearity. where k is a constant scale factor. Alternatively, [B] may
Monotonicity A DAC is said to be monotonic if each be calculated from the equation
digital-code increment produces an output equal to or
larger than that of the preceding code. A DAC is gener-
ally expected to be monotonic to increments as small as
an LSB, but in general the smallest increment for which
the DAC remains monotonic determines its monoto-
nicity.
Settling time The settling time is measured from the in-
stant of a digital-input-code change to the time that
the analog output reaches its corresponding new value
to within a specified error band. Usually the output is
expected to settle to within an error band of ( 12 )LSB.
Worst-case settling time is typically measured between
the zero-scale and full-scale codes.
Differential nonlinearity The differential nonlinearity
of a DAC is the maximum deviation from ideal of the FIGURE 9 The analog-to-digital converter (ADC).
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

538 Analog-Signal Electronic Circuits

 
B1 B2 BN Vi Although there are many practical methods of con-
1
+ 2 + ··· + N = , (19) structing an ADC, the type of converter described next
2 2 2 kVR E F
serves as an instructive example.
where the binary coefficients of [B] have a value of either
one or zero as required to make the equality true.
Because an A/D conversion using any real circuit will 1. The Counting ADC
require a finite time, it is customary to assume that the Figure 10 shows a simplified diagram for a counting type
input signal is not changing during the conversion cycle. of ADC which can be constructed from the following
If Vi is likely to change, it is common practice to employ a building blocks:
sample-and-hold (S/H) circuit to capture and hold the in-
put during the A/D conversion. In such a case, the digital 1. N -bit binary counter
output represented by Eq. (17) or Eq. (19) will correspond 2. N -bit current-scaling DAC
to the value of input that existed at the start of the con- 3. S/H circuit
version cycle. To clarify the point in time for which the 4. Comparator circuit
conversion is valid, most ADCs include a start-conversion 5. D flip-flop and AND gate
digital control signal and a digital status signal that indi-
cates when the conversion is complete. These optional The circuit also requires a clock signal.
signals are shown as dashed lines on the block diagram of The operation of the circuit can be described from the
Fig. 9. point in time that a start-conversion gate is applied. This

FIGURE 10 Staircase (counting) ADC. S/H, sample and hold.

P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

Analog-Signal Electronic Circuits 539

signal causes the D flip-flop to set, which immediately TABLE I Typical VLSI ADC Parameters
switches the S/H circuit to a hold mode and gates the Parameter Specification
clock signal into the binary counter. The counter, which
was previously reset to an all-zero count, begins a binary Resolution 12 bits
count-up sequence. Its output is applied to the DAC and Linearity ±( 12 )LSB
will cause a staircase buildup of an output current, I O . Differential nonlinearity 13 bits
The node voltage, Vx , will initially set positive due to an Full-scale error Adjustable to zero at 25◦ C
assumed positive value of input, but Vx will decrease as Full-scale temperature coefficient 20 ppm/◦ C
I O increases in binary LSB increments. The voltage Vx Analog input range 0–5 V or −2.5 to +2.5 V
may be expressed as Output code Binary
Digital interface TTLa
Vx = V A − I O R. (20) Conversion time 2 µsec
When I O R equals or exceeds V A , the comparator control a TTL, transistor–transistor logic.
node VX will swing negative, which will drive the com-
parator output to a low state, which, in turn, will reset
the flip-flop and cause the conversion-complete line to go the same definitions as for the DAC. It must be empha-
high. The character stored in the binary counter at this sized, however, that if the output of an ADC is an N -bit
time will correspond to the digital conversion of Vi . code, it does not follow that the accuracy is one part in 2 N .
This fact can be verified by noting that the circuit holds Table I lists specifications for an ADC that are typical of
the binary code which makes those achieved by using VLSI technology.
I O R ≥ Vi . (21)
Using Eq. (16), which expresses I O for the DAC circuit, IV. MIXED-SIGNAL SYSTEM
and assuming that I F S is given by VR E F /R R E F , one can APPLICATIONS
write the inequality of Eq. (21) as
n K VR E F R The mixed-signal circuits described earlier may be used
≥ Vi . (22) in a variety of ways to construct electronic systems with a
R R E F 2N
range of choices between digital and analog approaches.
If R is chosen equal to R R E F , then This section illustrates a few systems that can be de-
Vi 2 N signed by using digital, analog, and mixed-signal building
n≥ , (23) blocks.
VR E F
which is the desired result. The significance of the equal-
or-greater condition of Eq. (23) is that the binary code must A. Digital/Analog Waveform Generator
advance past the point where Vx swings negative but will Complex analog waveforms are required in applications
not advance more than a fraction of an LSB past before the ranging from the synthesis of human speech to laboratory
conversion cycle is terminated. The accuracy and gain of waveform test generators. Figure 11 shows a system capa-
the comparator will be a major factor in determining the ble of generating an arbitrary waveform from data stored
accuracy of the ADC, along with the linearity of the DAC. in a digital memory and accessed sequentially in time by a
Because the digital count to the DAC always starts at a clock-driven binary counter. The digital output characters
zero count and may have to cycle through all 2 N possible are converted to analog by the DAC. Filtering is employed
count combinations before finding the analog match, it is to smooth the quantized edges of the output waveform.
inherently slow. Furthermore, its conversion time is vari-
able, with larger analog inputs requiring more time than
that required by smaller inputs. B. Digital Audio Systems
Other types of ADC circuits are used to decrease the Audio recording, since the invention of the phonograph
time required to perform a conversion. Architectures avail- by Edison in 1877, has traditionally been accomplished
able for higher conversion speeds include the successive- by the storage of analog signals on cylinders, disks, or
approximation ADC and the parallel, or flash, ADC. magnetic tape. However, in recent years, a transition to-
ward digital storage has taken place, with audio signals
2. ADC Specifications being encoded as digital samples and stored as bits on
tape or CD media. Continuous high-quality audio may be
The key specifications for an ADC are its resolution, lin- reproduced by reading the digital samples and converting
earity, and conversion time. Resolution and linearity have back to analog by using a DAC followed by a smoothing
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

540 Analog-Signal Electronic Circuits

FIGURE 11 Digital/analog waveform generator. PROM, programmable read-only memory.

filter. Figure 12 suggests block diagrams for the recording twice the bandwidth. This rate is referred to as the Nyquist
and playback of digital audio. sampling frequency. For example, a signal with a 20-kHz
The theoretical basis of digital audio is the sampling bandwidth can be characterized by taking periodic sam-
theorem, which states that an analog waveform with a ples at a rate of 40 ksamples/sec or higher.
band-limited spectrum of B Hz may be completely char- The advantages of a digital audio recording system
acterized by taking periodic samples at a rate of at least over a traditional analog system include lower noise and

FIGURE 12 Digital audio recording and playback system: (a) recording; (b) playback.
P1: FJU/FFV Revised Pages P2: FJU
Encyclopedia of Physical Science and Technology EN001H-24 May 7, 2001 13:55

Analog-Signal Electronic Circuits 541

FIGURE 13 D/A rotary position system.

elimination of frequency modulation effects due to media SEE ALSO THE FOLLOWING ARTICLES
transport variations.
ANALOG ELECTRICAL FILTERS • CIRCUIT THEORY •
C. Computer-Controlled Robotics
DIGITAL FILTERS • DIGITAL ELECTRONIC CIRCUITS
The use of computers and microprocessors to control
mechanisms is an application that offers a very wide
choice of analog versus digital techniques. Figure 13
BIBLIOGRAPHY
shows a block diagram of a position servo system that
might be used to control the rotation of a gun turret on a
Comer, D. T. (1994). “Introduction to Mixed Signal VLSI,” Array Pub-
battleship, a cutting device on a milling machine, or the lishing Co., New York.
rotation of the head of a mechanical robot. Jaeger, R. C. (1997). “Microelectronic Circuit Design,” McGraw-Hill,
The system illustrated uses a position potentiometer New York.
with a shaft that is coupled to the rotary mechanism to gen- Johns, D. A., and Martin, K. (1997). “Analog Integrated Circuit Design,”
erate position feedback. This is an analog method. Dig- Wiley, New York.
Sedra, A. S., and Smith, K. C. (1998). “Microelectronic Circuits,” 4th
ital rotary shaft encoders are also available and the use ed., Oxford Univ. Press, New York.
of a digital feedback scheme would allow for a different Tsividis, Y. (1996). “Mixed Analog–Digital VLSI Devices and Technol-
position-system design. ogy,” McGraw-Hill, New York.