You are on page 1of 14


1.1 Supervisory Computer for Analog Loops 17

Introduction 4 Computer-Supervised DCS Systems 18
An Overview of Signals and Signal Processing 4 Production Monitoring and Control 18
Online Information System 19
Continuous and Digital Signals 5
Supervisory Control Techniques 19
Digital Signals 5
Supervisory Control Algorithms 19
Analog Instruments 6
Advanced Control Strategies 22
Analog Signal Processing 6
Computer Interface with DCS 23
Application and Examples of Analog
Hardware 23
Instruments 8
Software 24
Digital Instruments 9
Conclusions 24
Signal Conversion 9 References 24
Intelligent and IC Sensors 10 Bibliography 25
Basic Hardware 11
Inputs and Outputs 12
Communications and Networks 12 1.3
Virtual Instruments and Software 13 COMPUTERS — SIMULATION BY ANALOG
Application and Examples of Digital AND HYBRID SYSTEMS 26
Instruments 13 The Analog Computer 27
Comparison of Analog and Digital Instruments 14 Operational Amplifier 27
Bibliography 14 The Inverter 27
The Summer 27
Frequency Response of the OP-AMP 28
Analog Circuits for Differential Equations 28
Magnitude and Time Scaling 30
Nonlinear Components 32
Introduction 15 The Analog Computer Unit 32
History of Supervisory Control 15 Hybrid Computers 33
Generic Features of DCS 16 Hardware 33
Computer-Supervised Analog Controllers 17 Operation 34
Functions 17 Software 34
Tasks 17 Process Simulation 35

© 2006 by Béla Lipták

2 General

Laplace Transforms 35 Alarm Basics 59

Lag Functions 36 A.M. Costs and Designs 60
Control System Simulation 38 A.M. and Safety Instrumentation 60
Hybrid Simulation 38 Alarm Set Points 60
References 40 Alarm Presentation 60
Bibliography 40 Alarm Processing 61
Improving the Plant’s A.M. 61
Symptoms of A.M. Problems 61
The Tools and Elements of A.M. 61
A.M. Improvement and Rationalization
Factors to Be Considered 41 Project 62
The Old Debate 41 References 63
HVAC Damper Actuators 41
Pneumatic Actuators 41
Hysteresis 42 1.7
Direct Coupled Electronic Actuator 42 The Nature of Sound 64
Pneumatic vs. Electronic 43 Introduction 64
Control Signals 43 Speech Synthesis 64
Converters 43 Adaptive Differential Pulse-Code Modulation
Electrical Safety 43 (ADPCM) 65
The Modern Debate 44 Formant Synthesis 65
References 44 Text Synthesis 66
Bibliography 44 Voice Recognition 66
Word Boundary Detection 66
Feature Extraction 66
Pattern Standardization and Normalization 68
HUMAN ENGINEERING 46 Word Matching 68
The Control Room of the 21st Century 46 High-Level Linguistic Components 69
Introduction 46 Practical Implementation 69
Man–Machine System 47 Bibliography 70
Information Theory 47
Characteristics of Man 47
Body Dimensions 48
Information Capability 48
Performance (Efficiency and Fatigue) 49 Introduction 71
Application of Human Engineering 51 Electric Noise 71
Statistics of Operator Population 52 Reduction of Noise 71
Setting Priorities 52 Noise Types 72
Hardware Characteristics 54 Noise Sources 72
Information Coding 55 Grounding, Wiring, Filtering 74
Operator Effectiveness 55 Ground Systems 74
Operator Load 55 Wiring 75
Environment 56 Filtering 78
Summary 58 Applications 78
Bibliography 58 Transducer Signals 78
Common-Mode Rejection Measurement 80
Thermocouple Signals 81
1.6 Low-Level Signal Multiplexing 82
Introduction 59 Conclusions 86
History of Alarm Systems 59 References 87
Definitions 59 Bibliography 87

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments
W. P. DURDEN (1995) H. EREN (2005)

Types: Analog and digital instruments

Costs: A few dollars to thousands of dollars depending on functionality and complexity

Sensors: Common sensors for both types; additionally, IC (integrated circuitry) and intelligent
sensors for digital instruments

Components: Purely analog components for analog instruments; analog and digital components and
microprocessors for digital instruments

Displays: Needle indicator for analog instruments, numeric light emitting diode (LED) or liquid
crystal display (LCD) for digital instruments

Communications: Wired (e.g., 20 mA) for analog instruments; wired, RF, microwave, optical, sonic, etc.
for digital instruments

Networking: Limited network capabilities for analog instruments; unlimited capabilities for digital

Vendors (partial list): Analytical Measurements, Inc. (

Athena Controls, Inc. (
Brighton Electronics, Inc. (
Crompton Instruments (
Devar, Inc. (
Dresser Instruments (
Dunham Instruments (
EDL, Inc. (
Encore Electronics, Inc. (
Hioki Corp. (
Industrial Instruments & Supplies, Inc. (
Kahn Instruments, Inc. (
Morehouse Instrument Co. (
Ono Sokki Technology, Inc. (
Precision Devices, Inc. (
PTC Instruments (
R & R Instrumentation, Inc. (
Taylor Precision Products LP (
Weschler Instruments (
Weston Aerospace (

Digital Instruments
Alldos Eichler GmbH. (
AMETEK Instruments (
A&S Company, Inc. (
D&D Security Products, Inc. (
Davis Inotek Instruments (
Draeger Safety (
Dranetz Technologies, Inc. (

© 2006 by Béla Lipták

4 General

Dwyer Instruments, Inc. (

Dynalco Controls (
Entran Devices, Inc. (
Fluke Electronics (
Garmin, Ltd. (
Hanna Instruments, Inc. (
Hoyt Electrical Instrument Works, Inc. (
Intercomp (/
Keithley Instruments, Inc. (
Koehler Instrument Co., Inc. (
Leader Instruments Corp. (
MC Miller Co. (
NEC San-ei Instruments, Ltd. (
NeTech Corporation (
Nihon Kohden (
Polar Instruments, Inc. (
Protek T&M (
SE International, Inc. (
Sensotec, Inc. (
Sierra Instruments, Inc. (
SunTech Medical Instruments (
Texas Instruments (
Turner BioSystems (
Valhalla Scientific, Inc. (
Wagner Instruments (
Warren-Knight Instrument Co. (
Winters Instruments, Inc. (
Yokogawa Corp. (

INTRODUCTION considerable progress in digital instruments has taken place.

Therefore, this section will concentrate primarily on digital
Instruments are manmade devices that measure the parame- instruments. However, signals and signal processing will be
ters of physical variables. They may be analog, digital, or a introduced first since they are common to both types of
combination of the two. Nowadays, most instruments are instruments. In the following sections, the underlying prin-
digital because of their advantages over analog counterparts. ciples of analog and digital instruments will be discussed
However, the front ends of many instruments are still analog; separately, some examples will be given, and a comparison
that is, most signals from sensors and transducers and the between the two will be provided at the end of the chapter.
first stage of signal processing are still analog. Nevertheless,
it is important to mention that in recent years digital instru-
ments operating purely on digital principles have been devel-
oping fast. Today’s smart sensors based on digital principles AN OVERVIEW OF SIGNALS AND SIGNAL PROCESSING
contain the complete signal condition circuits and the sensors
in a single chip. The outputs of smart sensors can directly be All instruments operate on signals detected by sensors from
interfaced with other digital devices. physical phenomena. One of the main differentiators between
Sensors and transducers are the most basic and primary analog and digital instruments is the signal processing, which
elements of all instruments. They respond to physical varia- requires different theoretical approaches and hardware. The
tions to produce continuous outputs that convey useful infor- signals generated by sensors can be processed in three dif-
mation. The outputs may be in the forms of amplitudes, ferent ways:
frequencies, or phase differences of currents, voltages, power,
or energy. As in the case of all signal-bearing systems, in
both analog and digital instruments, there are useful signals 1. By analog techniques that directly deal with analog
that respond to the physical phenomena and unwanted signals signals
that are imposed as various forms of noise. 2. By converting analog signals into digital forms and
In the last 10 to 15 years or so, due to rapid progress in implementing the systems as digital instruments
integrated circuit (IC) technology and the availability of low- 3. By dealing with the signals purely in digital forms as
cost analog and digital components and microprocessors, digital-to-digital inputs and outputs

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments 5

A signal is defined as “any physical quantity that varies Digital Signals

with time, space, or any other independent variable.” Most
signals occurring in real world are analog in nature, i.e., they Digital signals are arrays of numbers that are used in com-
provide a continuous stream of information about the physical putational analysis of systems and signals. Digital signals can
quantity. Common electrical signals include variations in volt- either be generated or directly be derived from analog signals
ages, currents, frequencies, electric and magnetic properties, using A/D (analog to digital) converters.
phase relations, etc. On the other hand, the physical phenom-
Analog-to-Digital In the A/D conversion process, analog sig-
ena may be temperature, position, sound, vibration, chemical
nals are first sampled at discrete time intervals to obtain a
reactions, biological activity, light intensity, and so on.
sequence of samples that are spaced uniformly in time. This
Analog signals are processed by analog signal processing
is achieved by multiplying a continuous signal by a periodic
techniques, which can be described as a body of theoretical
train of Dirac delta functions spaced T seconds apart. T repre-
and practical methods that can be implemented. These tech-
sents the sampling period; its reciprocal is the sampling rate
niques include but not restricted to amplifying, filtering,
( fs = 1/T ). After the analog signals have been sampled, quan-
transmitting, estimating, detecting, modulating, demodulat-
tization is necessary to put the discrete-time signal into discrete
ing, analyzing, displaying, and reproduction of signals. On
numerical values. This is followed by coding to convert the
the other hand, due to easy and cost-effective availability of
quantized values to binary sequences.
advanced microprocessors and the supporting components,
For example, a discrete-time sinusoidal signal may be
the majority of modern instruments are digital and require
expressed in terms of sequences as:
digital signal processing. Digital systems provide powerful
processing and data handling capabilities simply by software x (nT ) = A sin(ω nT + θ ), 1.1(3)
and/or firmware implementations. Also, once the signals are
converted to digital formats, they can be managed by any where n is the integer variable termed the sample number,
computer or digital system. This provides a wide range of and T is the time interval between samples known as the
possibilities for data processing, communications, storage, sampling period.
and visual displays. The Fourier series representation of a periodic discrete-
time signal x(n) sequence is:
N −1
k =0
j 2π kn / N

Continuous signals, also known as analog signals, are defined where N is the fundamental period.
for every value of time from −∞ to +∞. Continuous signals The coefficient ck is given by:
can be periodic or nonperiodic. In periodic signals, the signal
repeats itself in an oscillatory manner, e.g., the sinusoidal N −1
 − j 2π kn 
∑ x(n) exp 
waveform (as in Figure 1.1m), which can be expressed as: ck =  1.1(5)
N n= 0
N 
x(t) = Xm sin(ω t) −∞ < t < ∞ 1.1(1)
The frequency-domain of discrete-time signals that are
where x(t) is a time-dependent signal, ω = 2πft is the angular not necessarily periodic can be expressed in a discrete Fourier
frequency, and Xm is the maximum value. transform (DFT) as:
The signals can be periodic but not necessarily sinusoi- ∞
dal; for example, they may be triangular, sawtooth, or rect-
angular waveforms. The periodic signals can be expressed as
X (ω T ) = ∑ x(n) exp[− jω nT ]
n =−∞
a combination of pure sinusoidal waveforms known as the
Fourier series. That is: and

x(t) = X0 + X1 sin(ω1t + φ1) + X2sin(ω2t + φ2) + … 2π

+ Xn sin(ωn t + φn) x (n) = X (ω T ) exp[− jω Tn]d (ω T ) 1.1(7)
1.1(2) 2π 0

where ω1, ω2, …, ωn are the frequencies (rad/s); X0, X1, … , These equations are essential in signal processing for
Xn are the maximum amplitudes of respective frequencies; digital instruments where analog-to-digital signal conversion
and φ1, φ2, … , φn are the phase angles. takes place.
In Equation 1.1(2), the number of terms may be infinite, In all digital instruments, the Nyquist sampling theorem
and the higher the number of elements the better the approx- must be observed; that is, “the number of samples per second
imation. Nevertheless, high-frequency contents can have sig- must be at least twice the highest frequency present in the
nificant effect on digital signal processing, as will be explained continuous signal,” Equation 1.1(2). As a rule of thumb,
next. depending on the significance of the high frequencies, the

© 2006 by Béla Lipták

6 General

Sensors Pre- Transmission

Physical Filters Outputs
and conditioning and
quantity Amplifiers displays
transducer circuit conditioning

FIG. 1.1a
Functional blocks of an analog instrument.

sampling must be about 5 to 10 times the highest frequency • High common mode rejection ratio (CMRR)
component in the signal. • Low sensitivity to changes in power voltage
• A broad operating frequency range
• High environmental stability
Operational amplifiers can be configured as inverting or nonin-
Analog instruments are characterized by their continuous sig- verting amplifiers. In addition, they can perform many other
nals. A purely analog instrument measures, transmits, displays, functions, for example as multipliers, adders, limiters, and filters.
and stores data in analog form. The signal processing is real- Another version of op amps is the instrumentation ampli-
ized by analog components that are integrating together as fiers, which are essentially high-performance differential
functional blocks, as illustrated in Figure 1.1a. Some examples amplifiers that consist of several closed loops within the chip.
of functional blocks are bridges, amplifiers, filters, oscillators, Instrumentation amplifiers are used extensively in applications
modulators, offset circuits, level converters, and buffers. where sensor signals are extremely weak and noisy. An instru-
Three basic components are common in all types of ana- mentation amplifier has improved CMRR (up to 160 dB), high
log signal processing: resistors, capacitors, and inductors. input impedances (e.g., 500 MΩ), low output impedance, low
The main function of a resistor is to limit the current. The offset currents and voltages, and better temperature character-
current will flow through a capacitor only if the voltage istics than common op amps do.
changes across it. In the case of inductors, voltage is estab-
lished only as a result of a change in the current across it.
Other analog components, including semiconductor devices
such as diodes, transistors, operational amplifiers, and recti-
A small number of the most essential functional blocks nec-
fiers, are based on these three basic elements.
essary for analog signal processing, such as amplifiers, adders,
Two basic types of semiconductor devices exist: bipolar
integrators, filters, modulators, and demodulators, will be
and metal oxide semiconductors (MOS). Many modern ana-
explained next.
log components and circuits are manufactured in integrated
circuit (ICs). ICs contain all basic circuit components such
Amplifiers An amplifier is a device that provides amplifi-
as resistors, transistors, and diodes densely packed on silicon
cation of the input signal in terms of voltage, current, or
chips. The operational amplifier (op amp) is an example of
power. Many different types of amplifiers exist, including
an IC.
voltage, power, current, transconductance, and transresistance
Operational amplifiers are fundamental building blocks amplifiers, and, in terms of frequency range, audio (15 Hz to
of analog signal processing circuits. They are made either as 20 kHz), radio frequency (RF = 20 kHz to 0.1 GHz), and
integrated (monolithic) circuits or hybrid circuits (combina- video (10 Hz to 6 MHz) amplifiers. As an example, Figure
tion of monolithic and discrete parts). An op amp is made 1.1b illustrates two single-transistor amplifiers, bipolar and
from hundreds of transistors, resistors, and capacitors in a MOSFET (metal oxide semiconductor field effect transistor).
single chip. It has two inputs (inverting and noninverting) See Figure 1.1b.
and a single output. A good operational amplifier has the
following properties:
Adders An adder (summer) is a circuit, based on op amps,
whose output is the weighted algebraic sum of all its inputs.
• High input resistance, hundreds of megaohms or few In the case of the voltage adder, in Figure 1.1c, inputs V1, V2,
gigaohms and V3 are added to give an inverted output, that is:
• Low output resistance, less than few ohms or fraction
of one ohm R0 R R
• Low input offset voltage, a few millivolts or micro- V0 = − V1 − 0 V2 − 0 V3 1.1(8)
R1 R1 R1
• Low input bias current, a few picoamps More inputs can be connected without affecting the response
4 6
• Very high open loop gain, 10 to 10 of the existing inputs since the inverting input is held at a
• Low intrinsic noise virtual ground by the feedback mechanism.

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments 7


300 kΩ 22 kΩ 2.2 MΩ 22 kΩ
C2 C2

2 kΩ C1 2 kΩ C1

100 kΩ vo 100 kΩ vo
vs + vs +
− −
2 kΩ 1 kΩ

160 kΩ 1.5 MΩ

10 kΩ C3 C3
10 kΩ

(a) (b)

FIG. 1.1b
Single-transistor amplifiers: (a) bipolar common-emitter, (b) MOSFET common source.

R1 R0 Differentiators can be obtained by changing the positions

of R and C.

Filters Analog filters are essential parts of both analog and
V1 digital instruments. They are used to extract the useful infor-
V2 mation from the sensors, as most sensors generate broadband
V3 Vo
signals containing various forms of noise. Filters are useful
for the elimination of external and internal noise that inter-
FIG. 1.1c feres with operation of an instrument.
An adder with three inputs. There are two basic types of analog filters, passive and
active. Passive filters are made from purely passive compo-
nents, such as resistors, capacitors, and inductors. Active fil-
Integrators An integrator can be defined as a circuit whose ters contain active elements such as op amps in addition to
output rate of change is proportional to the applied input. the passive components. The filters can be configured to be
The output of an integrator for an input Vi is: low-pass, high-pass, band-pass and band-stop, or notch filters.
The subject of filters is an important but vast topic; therefore,

dV0 some limited detail is provided below.
= kVi or V0 = k Vi dt 1.1(9)
This equation shows that the output Vo is obtained by inte- Low-Pass Active Filters A typical example of a low-pass
gration of the input Vi . Figure 1.1d illustrates an integrator filter is given in Figure 1.1e. In the process of an active low-
based on an op amp. The integrator is an important building pass filter design, the cutoff frequency of the filter must be
block in the design and implementation of active filters. determined first. The cutoff frequency is the frequency at
which the gain of the filter is about −3 dB.

1 C
V0 = −
RC ∫Vi dt
Vi +
R1 R2 Vo

Vi Vo C2

FIG. 1.1d FIG. 1.1e

An integrator. A low-pass filter.

© 2006 by Béla Lipták

8 General


C2 Vo
− R7 F
− R2
Vi + D −
+ E
R4 R6 R5 +

FIG. 1.1f
A three-amplifier biquad filter.

The choice of resistors in filters is critical. As a rule of increases. A typical complex active filter is illustrated in
thumb, resistors R1 and R2 are selected as 25 kΩ for a 10 kHz Figure1.1f.
or 250 kΩ for 100 Hz filter. The values of capacitors can be
calculated for the selected frequency by substituting the selected Modulators and Demodulators In many electronic sys-
resistance values. The calculations are repeated to obtain the tems, modulations and demodulations are essential for allow-
nearest available capacitors by changing the resistor values. ing convenient and efficient transmission and reception of
information. Basically, modulation is the process of encoding
High-Pass Active Filters High-pass active filters are analo- information on a carrier signal such that its amplitude, fre-
gous to low-pass filters and pass only high frequencies. quency, or phase varies with the message signal, known as
the modulating signal. There are three types of modulation:
Band-Pass or Band-Stop Active Filters A straightforward way amplitude modulation (AM), frequency modulation (FM),
of building a band-pass or a band-stop filter is by combining and phase modulation. Examples of AM and FM are illus-
low-pass and high-pass filters in cascaded form. trated in Figure 1.1g. Demodulation is the process by which
information is extracted from the carrier signal.
Notch Filters These filters are usually used to suppress sin-
gle sharp frequencies such as 50 Hz. The best performance
is obtained by adjusting component values while monitoring APPLICATION AND EXAMPLES OF ANALOG INSTRUMENTS
the circuit performance on an oscilloscope.
Various design techniques are used to configure analog Analog instruments are used in many applications where
filters. Some of these design techniques include Butterworth, digital readouts are required. Examples of such applications
Chebyshev, and Bessel–Thomson filters. The complexity of are control panels, testing systems, manufacturing and pro-
filters increases as the order of the filter (number of poles) duction equipment, dashboards in all types of transportation

Carrier signal Carrier signal

Modulating signal Modulating signal

AM Modulated signal FM Modulated signal

(a) (b)

FIG. 1.1g
Two types of modulation, (a) AM and (b) FM.

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments 9

signal x(t) Analog signal
Sample and

T 2T nT Time
Sampling period T (t)
FIG. 1.1j
Analog-to-digital conversion process.

Signal Conversion

Analog-to-Digital Conversion Analog-to-digital (A/D) and

FIG. 1.1h digital-to-analog (D/A) converters are used to interface the
An analog multimeter. (Courtesy of Hioki Corporation, www.hiok- analog environment (the real world) and the digital environ- ment. Analog signals obtained from a myriad of real-world
sources, including temperature, motion, sound, and electrical
signals, are converted to digital forms by A/D converters.
equipment, laboratories, explosion-proof areas, consumer After the conversion, the signals are processed using many
products, and machinery. A typical example of analog instru- different algorithms such as digital filters and statistical tech-
ments is the multimeter, as illustrated in Figure 1.1h. niques. When necessary, the processed signals then can be
converted back to analog forms to drive devices requiring
analog inputs such as mechanical loads and sound speakers.
DIGITAL INSTRUMENTS A/D conversion involves three stages: sampling, quanti-
zation, and encoding. Once the analog signals are sampled,
In most modern instruments, the original analog information quantization takes place to determine the resolution of the
acquired from a physical variable is converted to digital form sampled signals. In coding, the quantized values are con-
to obtain a digital instrument. Analog-to-digital (A/D) con- verted to binary numbers. Figure 1.1j illustrates a typical A/D
verters are used for signal conversion together with appropri- conversion process of an analog signal.
ate multiplexers and sample-and-hold devices. The multiplex- All modern A/D converters produce digital output values
ers enable the connection of many sensors and transducers to in the integer format. These can be in binary codes or the
the same signal-processing media. The typical building blocks twos-complementary form. The length of the output word
of a digital instrument are illustrated in Figure 1.1i. defines theoretical resolution of the A/D converter and the
Digital systems are particularly useful in performing ranges that an A/D converter can produce. The smallest change
mathematical operations, numeric displays, and storing and in the input voltage (VLSB) that A/D can convert is:
transmitting information. Digital instruments consist of three
essential subsystems: converters, general-purpose processors
Vmax − Vmin
together with mass storage and communication peripherals, VLSB = 1.1(10)
and application-specific software. In this section, some essen- 2n
tial features of digital instruments will be discussed, includ-
ing signal conversion, intelligent and IC sensors, basic hard- where n is the number of bits produced by the A/D converter
ware, inputs and outputs, communications and networks, and Vmax and Vmin are the maximum and minimum input
virtual instruments, and software. voltages that an A/D converter can handle correctly.

Sensor Analog Multiplexer Microprocessor

Physical A/D Digital systems
or signal
signal converter and processing
transducer conditioning

FIG. 1.1i
Basic building blocks of a digital instrument.

© 2006 by Béla Lipták

10 General

Due to a diverse range of application requirements, many Dn –1 is the Most Significant Bit (MSB), and D0 is the Least
different types of A/D converters exist. The most popular are Significant Bit (LSB).
counter ramp, successive approximation (serial), flash (par- However, the output code of an A/D converter is normally
allel), and delta–sigma (∆/Σ) converters. interpreted as a fraction of its Full Scale (FS) value, which
is considered as the unit. So the binary codes from an A/D
Quantization A number of quantization methods exist, one of conversion process must be interpreted in fractional numbers,
which is amplitude quantization, defined as the process by which is equivalent to dividing the value of each bit by 2 .
which a signal with a continuous range of amplitudes is trans-
formed into a signal with discrete amplitudes. In this method, Digital-to-Analog Converters Analog signal generation from
the process is usually memoryless; that is, the quantizer does the digital world may be part of the measurement process or
not depend on previous sample values to predict the next value. may be required for control or sensor stimulation purposes.
Although this is not an ideal solution, it is commonly used in Digital signals are converted to analog waveforms by D/A
practice. converters and must be filtered and amplified before being
The precision of the quantizer is limited by the number applied to the appropriate output transducers. However, some
of discrete amplitude levels available to approximate the con- new types of instrumentation now deal exclusively with dig-
tinuous signal amplitudes. These levels are referred to as rep- ital signals. These include logic and computer network ana-
resentation or reconstruction levels, and the spacing between lyzers, logic pattern generators, and multimedia and digital
two representation levels is called the step size. Quantization, audio devices. Such instruments usually include high-speed
like sampling, introduces some degree of uncertainty in the digital interfaces to move data in and out of the instrument
value of the analog signal. For A/D converters with large at high speeds.
number of bits (8 bits or more) this effect can modeled as an Digital information requires digital-to-analog converters
additive noise. The amount of quantization noise relative to (D/A) for the conversion of the binary words into appropri-
the signal, or Signal-to-Noise Ratio (SNR), depends on the ately proportional analog voltages. A typical D/A converter
nature and magnitude of the input signal. Quantization noise consists of three major elements:
can be both nonlinear and signal dependent. However, pro-
vided that the step size is sufficiently small, it is reasonable 1. Resistive or capacitive networks
to assume that the quantization error is a uniformly distributed 2. Switching circuitry, operational amplifiers, and latches
random variable and its effect is similar to that of thermal 3. A reference voltage
In addition to quantization noise, A/D converters have Also, an important requirement is the switching of the
other sources of noise, such as internal voltage reference reference voltage from positive to negative for bipolar signals.
noise and thermal amplifier noise. In practice, most convert-
ers rarely achieve their theoretical noise floor. Therefore, Intelligent and IC Sensors
when selecting an A/D converter, one uses the number of bits
of converter as a rough estimate of its resolution. Neverthe- In recent years, progress in electronic instruments has been
less, in practical applications, a close examination of the making a major turnaround due to the availability of IC
converter specifications should reveal its true performance. sensors in the form of micro- and nanosensors. Most of these
sensors are smart sensors that have intelligence due to inte-
Coding Coding is a process of representing the finite output gration of complex digital processors into the same chip. A
states of the quantizer by a sequence of n bits. The codes general scheme of a smart sensor is illustrated in Figure 1.1k.
formed using bits are normally known as binary codes and In this particular example, the sensor is under a microproces-
can be of two types: unipolar and bipolar. sor control. Excitation is produced and modified depending
Unipolar codes are used to represent unipolar quantities, that on the required operational range. The processor may contain
is, quantities with a predefined sign (positive or negative). The
most common unipolar codes are natural binary, BCD (Binary-
Coded Decimal), and the Gray Code. In order to illustrate the Physical A/D
Sensor Analogue
coding process, the natural binary code is described below. conditioner converter
In the natural binary code, each bit of a number represents
a successive power of two according to the position that the bit Gain control, Control
occupies. Given a number of n-digits (represented by Dn −1 , D/A, etc.
Sensor Data in
Dn – 2, … D1 and D0 ), the value of the number is determined control Digital
Data out
by the following expression: processor
Control in
n −1 n− 2
Dn−1Dn− 2 K D1D0 = Dn−1 × 2 + Dn− 2 × 2 +L
FIG. 1.1k
+ D1 × 2 + D0 × 2
1 0
1.1(11) General form of smart sensors.

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments 11

full details of the transducer characteristics in ROM (read only and low cost, and power efficiency, micro-controllers are
memory) enabling the correct excitation, gain, and so on. commonly used in electronic instruments and the associated
instrumentation systems. Therefore, in this section, detailed
Basic Hardware information on micro-controllers will be provided.

Microprocessors and micro-controllers constitute the heart of Micro-Controllers Many micro-controllers are single-chip
almost all types of modern digital instruments and instrumenta- devices that have memory for storing information and are able
tion systems. They play an important role in data acquisitions, to control read/write functions and manipulating data. The
data processing, and control. The applications of microproces- performance of a micro-controller is classified by its size, that
sors and micro-controllers in instruments and instrumentation is, the number of bits that the CPU (central processing unit)
systems can be categorized according to the following roles: can handle at a time. Many types or families of micro-control-
lers are offered by a diverse range of manufacturers. Therefore,
1. Data handling functions that include data acquisi-
in order to select appropriate micro-controllers for a specific
tion, data processing, information extraction, data
task, it is vital to understand the differences in the character-
compression, interpretation, recording, storage, and
istics of different families.
The software for micro-controllers to perform specific
2. Instrumentation control, which includes sensors, actu- tasks can be written in either assembly codes or high-level
ators, system resources, and process controls languages such as Basic or C. The program written in other
3. Human-machine interface, one of the significant roles digital devices can be downloaded into the system memory
of digital systems, to provide a meaningful and flexible through the serial communication port of the micro-controller.
user interface to the operators for ergonomic informa- All high-level programs are compiled into machine language
tion and control display for execution. Significant numbers of compilers are available
4. Experimentation and procedural development, con- on the market for the different families of micro-controllers.
sisting of commissioning, testing, and general proto-
typing of the targeted system under investigation
Digital Signal Processors Digital signal processors (DSPs)
Micro-controllers and digital signal processors are spe- are specialized microprocessors that involve partial computer
cial microprocessors that have built-in memory and interface architecture and fulfill fast operational needs. A simplified block
circuits within the same IC. Due to smaller sizes, simplicity diagram of a typical DSP processor is given in Figure 1.1l.

Program address bus

Data address bus

address bus

X-data X-data Y-data Y-data

C Program Program RAM & Address
Address RAM &
sequencers ROM generator ROM ROM generator

Program bus
R X-data bus
Y-data bus
data bus Timer Serial I/O
interface Multiplier port

FIG. 1.1l
Block diagram of a typical DSP.

© 2006 by Béla Lipták

12 General

DSPs incorporate special hardware features that are capa- 5. Controller Area Network (CAN) is a multiplexed wir-
ble of speeding up calculation of digital filters, fast Fourier ing scheme that was developed jointly by Bosch and
transforms (FFTs), and other frequency-based algorithms. Intel for wiring in automobiles. The CAN specifica-
Modern general-purpose microcomputers can also address tion is being used in industrial control in both North
the needs of digital signal processing if some of the necessary America and Europe.
hardware and special instructions are added. As a result, the
distinction between the DSP processor and microprocessors, Data-Acquisition Boards A data-acquisition board is a
in general, is in the degree of specialization. This particular plug-in I/O board that allows the treatment of a set of analog
DSP has two data memories, denoted as X and Y (Figure 1.1l). signals by a computer. Thus, these boards are the key ele-
For the implementation of signal processing, say in a finite ments to connect a computer to a process in order to measure
impulse response (FIR) filter, X-memory can be used to store or control it. Data-acquisition boards normally operate on
the samples of input signals and Y-memory to store the conditioned signals, that is, signals which have already been
impulse response. filtered and amplified by analog circuits. However, some
One of the main characteristics of the DSP processors is data-acquisition boards can deal with the signals directly
the capability of handling integers for which they are generated by sensors. These boards include all the necessary
designed. Most of the existing DSP processors are either 16- signal-conditioning circuits.
or 32-bit devices. For example, the Motorola DSP56300 pro- The number of analog inputs normally managed by a data-
cessors are 24-bit integer processors that offer a compromise acquisition board is eight or 16. This system is made up of
between inexpensive 16-bit and powerful 32-bit devices. an analog multiplexer, an amplifier with programmable gain,
a sample-and-hold circuit, and an A/D. All the elements of
Inputs and Outputs this system are shared by all the analog input channels. Some
data-acquisition boards also offer analog outputs to carry out
The data communication between micro-controller and out- control operations. When available, the normal number of
side world is realized through input and output (I/O) pins. analog outputs is two.
The format of data transmission of a micro-controller can be In addition to analog inputs and outputs, data-acquisition
universal asynchronous receiver transmitter (UART) — a port boards also supply digital I/O lines and counters. Digital lines
adapter for asynchronous serial communications. A diverse are used for process control and communication with peripheral
range of I/O hardware and software is available. Some of devices. Counters are used for applications such as counting
these I/Os suitable for instruments and instrumentation sys- the number of times an event occurs or generating a time base.
tems are explained below:
Communications and Networks
1. Universal Synchronous/Asynchronous Receiver Trans-
mitter (USART) — a serial port adapter for either asyn- In many applications, instruments are networked by computers
chronous or synchronous serial communications. to measure many variables of a physical process. The resulting
Communications using a USART are typically much arrangement for performing the overall measurement function
faster than those using UARTs. is called the measurement system. In measurement systems,
2. Synchronous serial port — Synchronous serial ports instruments operate autonomously but in a coordinated man-
are used to communicate with high-speed devices such ner. The information generated by each device is communi-
as memory servers, display drivers, additional A/D cated between the instruments themselves or between the
ports, etc. In addition, they are suitable to implement instrument and other devices such as recorders, display units,
simple micro-controller networks. base stations, or a host computer. Currently, the coordination
3. Serial Peripheral Interface (SPI) (Motorola) — A syn- of instruments is largely realized by digital techniques.
chronous serial port functioning as an enhanced In digital instrumentation systems, the transmission of
UART. data between devices is realized relatively easily by using
4. I C bus — Inter-Integrated Circuit bus (Philips) — is serial or parallel transmission techniques. However, as the
a simple two-wire serial interface that was originally measurement system becomes large by the inclusion of many
developed for 8-bit applications. The I C bus is a two- instruments, communication can become complex. To avoid
line, multi-master, multi-slave network interface with this complexity, message interchange standards are used,
collision detection. Up to 128 devices can be con- such as RS-232, USB, EIA-485, and IEEE-488.
nected on the network, and they can be spread out over For long-range data transmission, modem, microwave, or
10 m. The two lines of the network consist of the serial radio frequency (RF) media are selected. In these cases, var-
data line and the serial clock line. Each node on the ious modulation techniques are used to convert digital signals
network has a unique address, which accompanies any to suitable formats. For example, most modems use fre-
message passed between nodes. Since only two wires quency-shift keyed (FSK) modulation. The digital interface
are needed, it is easy to interconnect a number of with modems uses various protocols such as MIL-STD-188C
devices. to transmit signals in simplex, half-duplex, or full duplex

© 2006 by Béla Lipták

1.1 Analog vs. Digital Instruments 13

forms depending on the directions of the data flow. The sim- alone instruments do; consequently, virtual instruments offer
plex interface transmits data in one direction, whereas full more flexibility in data handling.
duplex transmits it in two directions simultaneously.
In industrial applications, several standards for digital Software for Virtual Instruments To develop the software
data transmission are available. These are commonly known of a virtual instrumentation system, a programming language
in engineering literature as fieldbuses. Among many others, or special software can be used. However, the option of using
some of these standards, such as the WordFIP, Profibus, and a traditional programming language (C, for example) can gen-
Foundation Fieldbus, are widely accepted and used. The erate several problems, including difficulty in programming
fieldbuses are supported by hardware and software (for exam- the graphical elements and difficulty in learning the language.
ple, National Instruments chips and boards) that allow Currently, a more utilized option is the Microsoft Visual
increases in the data rates suitable with high-speed protocols. Basic programming language, which runs under the Windows
operating system. Visual Basic has become quite popular in
Virtual Instruments and Software the development of virtual instrumentation systems because
it combines the simplicity of the Basic programming lan-
Traditional instrumentation systems are made up of multiple guage with the graphical capabilities of Windows. Another
stand-alone instruments that are interconnected to carry out important reason for its popularity is that it is an open lan-
a determined measurement or control an operation. Function- guage, meaning that it is easy for third-party vendors to
ality of all these stand-alone instruments can be implemented develop products that engineers can use with their Visual
in a digital environment by using computers, plug-in data- Basic programs. Several companies now sell Dynamic Link
acquisition boards, and supporting software implementing Libraries (DLLs), which add custom controls, displays, and
the functions of the system. The plug-in data-acquisition analysis functions to Visual Basic. The controls and displays
boards enable the interface of analog signals to computers, mimic similar functions found on stand-alone instruments,
and the software allows programming of the computer to look such as toggle switches, slide controls, meters, and LEDs.
and function as an instrument. The systems implemented in By combining various controls and displays, any virtual
this way are referred to as virtual instrumentation systems. instrument can easily be programmed.
The major advantage of virtual instrumentation is its A third option for developing virtual instrumentation sys-
flexibility; changing a function simply requires modification tems is to use a software package specifically designed for the
of supporting software. However, the same change in a tra- development of these systems, such as LabView or PC600.
ditional system may require to adding or substituting a stand- The crucial advantage of these packages is that it is not nec-
alone instrument, which is more difficult and expensive. For essary to be a Windows programmer to use them. Several
example, a new analysis function, such as a Fourier analysis, packages of this type exist, but only one, LabView by National
can be easily added to a virtual instrumentation system by Instruments, has reached great diffusion.
adding the corresponding software to the analysis program. LabView is an entirely graphical language based on two
However, to do this with a traditional instrumentation system, concepts: the front panel and the block diagram. It is extend-
a new stand-alone instrument (for example, spectrum ana- ible, so new functional modules can be added to a program.
lyzer) would have to be added to the system. The modules can also be written using general-purpose lan-
Virtual instruments also offer advantages in displaying and guages, such as C or C++. These languages provide great
storing information. Computer displays can show more colors flexibility to program functions that perform complex numer-
and allow users to quickly change the format of displaying the ical operations on the data.
data that is received by the instrument. Virtual displays can be The applications of virtual instruments are gaining
programmed to resemble familiar instrument panel compo- momentum in instrumentation systems. Nowadays, palm and
nents, including buttons and dials. In Figure 1.1m, an example laptop computers are widely available and they can easily be
of a virtual display corresponding to a two-channel oscilloscope equipped with commercial data-acquisition boards such as
is shown. Computers also have more mass storage than stand- the PCMCIA interface cards. Together with a suitable inter-
face, software, and sensors, these computers function just as
any digital instrument would.

Application and Examples of Digital Instruments

Digital instruments have all the functionality of their analog

counterparts and are used in all types of instrumentation systems.
Modern digital instruments are rapidly replacing analog instru-
ments due to advantages in communications, data handling, and
networking capailities, as will be explained in the next section.
FIG. 1.1m An example of digital instrumentation is a blood pressure
A virtual instrument display acting as a two-channel oscilloscope. recorder “Comparison of Analog and Digital Instruments,”

© 2006 by Béla Lipták

14 General

digital instruments minimize the reading error that can occur

in analog displays. Errors such as parallax and shadowing
are also eliminated. Digital instruments also reduce the mar-
gin for error due to inaccurate interpolation. Analog meters
can be very sensitive to movement or rough handling, while
digital meters are immune to such dynamic effects.
One of the distinct advantages of analog instruments is
the easy interpretation of displays. The operator can get an
intuitive feel about the variables being measured at a glance.
There is no such spatial reference with digital displays, which
require mental interpretation. This requires an additional step
in the thought process and also some familiarity with the
equipment. This additional step of interpretation may not be
tolerable in some applications, such as driving cars, where
instantaneous decisions may need to be made. In addition,
analog instruments are relatively cheaper for the same func-
tionality due to their simplicity.
FIG. 1.1n
A digital ambulatory blood pressure recorder. (Courtesy of SunTech

Brignell, J., and White, N., Intelligent Sensor Systems, Rev. ed., Philadel-
illustrated in Figure 1.1n. This wireless outdoor system oper- phia: Institute of Physics Pub., 1996.
ates on solar power. It measures temperature, humidity, pres- Cavicchi, T. J., Digital Signal Processing, New York: John Wiley & Sons,
sure, wind speed and direction, and rainfall. A wireless link 2000.
allows the transmission of data up to 48 km. Several instru- Dyer, S. A. (Ed.), Survey of Instrumentation and Measurement, New York:
John Wiley & Sons, 2001.
ments can be networked together.
Eren, H., “Analog and Discrete Input/Output, Costs, and Signal Processing,”
Instrumentation Engineers Handbook, 4th ed., B. Liptak, Ed., Boca
Raton, FL: CRC Press, 2002, pp. 123–141.
COMPARISON OF ANALOG AND DIGITAL INSTRUMENTS Fowler, K.R., Electronic Instrument Design — Architecting for the Life
Cycle, New York: Oxford University Press, 1996.
Digital hardware allows programmable operation; therefore, Garrett, P. H., Advanced Instrumentation and Computer I/O Design: Real-
Time System Computer Interface Engineering, New York: IEEE Press,
it is possible to modify the functions of an instrument through 1994.
the software, which increases flexibility and reliability. How- Grover, D., and Deller, R. J., Digital Signal Processing and the Microcon-
ever, it is worth mentioning that some signals with extremely troller, Upper Saddle River, NJ: Prentice Hall PTR, 1999.
wide bandwidths require intensive real-time processing in the Haykin, S., and Kosko, B. (Eds.), Intelligent Signal Processing, New York:
digital environment. In these cases, analog instruments may IEEE Press, 2001.
IEEE, Spectrum, Piscataway, NJ: IEEE, 1990–2002.
be the only solution.
IEEE, Transactions on Instrumentation and Measurement, Piscataway, NJ:
Many modern digital instruments have built-in intelligence; IEEE, 1990–2002.
hence, they can make self calibrations, self diagnosis, and cor- Liptak, B., Instrumentation Engineers Handbook, 4th ed., Boca Raton, FL:
rections in acquired data. A new trend is that calibrations of CRC Press, 2002.
instruments can be conducted over the Internet by using appro- Madisetti, V. K., and Williams, D. B., The Digital Signal Processing Hand-
priate sites of manufacturers and standards authorities. book, Boca Raton, FL: CRC Press, 1998.
Pallás-Areny, R., and Webster, J. G., Sensors and Signal Conditioning, 2nd
Digital instruments are suitable in applications where
ed., New York: John Wiley & Sons, 2001.
multiple measurements are necessary, as in the case of pro- Pallás-Areny, R., and Webster, J. G., Analogue Signal Processing, New York:
cess industry, for two main reasons: John Wiley & Sons, 1999.
Paton, B. E., Sensors, Transducers, and LabView, Upper Saddle River, NJ:
Prentice Hall PTR, 1999.
1. Ease in communication and networking by means of
Potter, D., “Smart Plug and Play Sensors,” IEEE Instrumentation and Mea-
remote communication methods such as RF, microwave, surement Magazine, Vol. 5(1), pp. 28–30, 2002.
the Internet, and optical techniques Predko, M., Handbook of Microcontrollers, New York: McGraw-Hill, 1999.
2. Availability of almost infinite memory for data storage Randy. F., Understanding Smart Sensors, 2nd ed., Boston: Artech House,
Soloman, S., Sensors Handbook, New York: McGraw-Hill, 1999.
In digital instruments, however, converting an analog signal
Webster, J. G. (Ed.), Wiley Encyclopedia of Electrical and Electronic Engi-
allows certain errors (for example, quantization error) to be neering, New York: John Wiley & Sons, 1999.
introduced. The extent of error depends on the sampling rates Webster, J. G. (Ed.), The Measurement, Instrumentation, and Sensors Hand-
and number of bits of the A/D converter. By the same token, book, Boca Raton, FL: CRC Press, 1999.

© 2006 by Béla Lipták