Data Acquisition Techniques Using Personal Computers Howard Austerlitz CYBEX A Division of Lumex, Inc. Ronkonkoma, New York Academic Press, Inc. Harcourt Brace Jovanovich, Publishers San Diego New York Boston London Sydney Tokyo Toronto Front cover photograph courtesy of International Business Machines Corporation. This book is printed on acid-free paper. @ Copyright © 1991 by A C A D E M I C PRESS, INC. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. IBM PC, PC/XT, PC/AT, PS/2, PC DOS and Micro Channel are trademarks of IBM Corporation. MS DOS, Microsoft Windows and Microsoft C are trademarks of Microsoft Corporation. OS/2 is a trademark of Microsoft and IBM Corporation. Unix is a trademark of AT&T Information Systems. Apple Macintosh is a trademark of Apple Computer, Inc. Nu Bus is a trademark of Texas Instruments. Intel is a trademark of Intel Corporation. Motorola is a trademark of Motorola Corporation. Lotus 1-2-3 and Symphony are trademarks of Lotus Corporation. ASYST is a trademark of Keithley Asyst. LABTECH NOTEBOOK is a trademark of Laboratory Technologies Corporation. Academic Press, Inc. San Diego, California 92101 United Kingdom Edition published by Academic Press Limited 24-28 Oval Road, London NW1 7DX Library of Congress Cataloging-in-Publication Data Austerlitz, Howard. Data acquisition techniques using personal computers / Howard Austerlitz. p. cm. Includes bibliographical references and index. ISBN 0-12-068370-9 1. Microcomputers. 2. Automatic data collection systems. 3. Computer interfaces. I. Title. TK7888.3.A872 1991 04.6' 16-dc20 91-2754 CIP PRINTED IN THE UNITED STATES OF AMERICA 91 92 93 94 9 8 7 6 5 4 3 2 1 To my wife Kiel, whose guidance and understanding made it all possible Preface In recent years personal computers (PCs) have become common fixtures in most laboratories due to their low cost and wide range of hardware and software support . They have replaced minicomputers as de facto plat- forms for data acquisit ion sys tems. Data Acquisition Techniques Using Personal Computers is intended to be a tutorial and reference for engi- neers , scientists , s tudents , and technicians interested in using personal computers for data acquisition and analysis. It is assumed that the reader knows the basic workings of personal computers and electronic hardware , although these aspects will be re- viewed briefly in this work. Sources listed in the bibliography are good introductions to many of these topics . Only the family of IBM PCs and compatible systems (PC/XT/AT computers) will be covered in any great detail here , since they represent the largest hardware and software support base for scientific and engi- neering applicat ions. Howeve r , I B M ' s PS/2 systems (based on Micro Channel) and Apple ' s Macintosh II computers (based on NuBus) will be covered briefly. This book stresses " r e a l " applications and includes specific exam- ples as well as a survey of commercial ly available hardware and software products . It is intended to provide all the information you need to set up a data acquisit ion system based on a personal computer . In addition, it will serve as a useful reference on personal computer technology. The area of software is as important as hardware , if not more so. Software topics , such as programming languages, interfacing to a PC ' s software environment , and data analysis techniques , are covered in de- tail, along with a survey of commercial data acquisition application pro- grams. Throughout this work, the term personal computer will refer to a generic machine . It can be an Apple Macintosh or an IBM PS/2 system. The abbreviated term PC will imply an IBM P C / X T / A T system or com- patible, based on an Intel 80x86 family microprocessor and running MS-DOS (or IBM DOS) software. xi xi Preface I wish to acknowledge the many people who helped me with this undertaking. My thanks to Academic Press for getting the project started and seeing it through to its conclusion. I am grateful for the assistance I received from manufacturers in the data acquisition field, including Ved Vasconcelos from Keithley Metrabyte , Kate Kressman from Keithley Asyst , Shari Worthington from Labora tory Technologies, and Iris Polaski from Burr-Brown/Intel l igent Instrumentat ion. I also wish to thank every- one at C Y B E X who helped me, especially Jim Smith. Finally, I want to acknowledge Orndorff, the lap-top editor who kept me company during all those late nights at my PC. Howard Austerli tz C H A P T E R m Introduction to Data Acquisition Data acquisit ion, in the general sense , is the process of collecting informa- tion from the real world. For most engineers and scientists these data are mostly numerical and usually collected, stored, and analyzed using a computer . The use of a compute r automates the data acquisition process , enabling the collection of more data in less time with fewer errors . This book deals solely with automated data acquisition using personal com- puters . An illustrative example of the utility of automated data acquisition is measuring the tempera ture of a heated object versus time. Human observ- ers are limited in how fast they can record readings (say, every second, at best) and how much data can be recorded before errors due to fatigue occur (perhaps after 5 minutes or 300 readings). An automated data acqui- sition system can easily record readings for very small time intervals (i .e. , much less than a millisecond), continuing for arbitrarily long time periods (limited mainly by the amount of storage media available). In fact, it is easy to acquire too much data , which can complicate the subsequent analysis. Once the data are stored in a computer , they can be displayed graphically, analyzed, or otherwise manipulated. Most real-world data are not in a form that can be directly recorded by a computer . These quantit ies typically include tempera ture , pressure , dis tance, velocity, mass , and energy output (such as optical, acoust ic , and electrical energy) . Very often these quantit ies are measured versus time or posit ion. A physical quanti ty must first be converted to an electri- cal quanti ty (voltage, current , or resistance) using a sensor or transducer. This enables the data to be condit ioned by electronic instrumentat ion, 1 2 CHAPTER 1 Introduction to Data Acquisition which operates on analog signals or waveforms (a signal or waveform is an electrical parameter , most often a voltage, that varies with time). This analog signal is continuous and monotonie, that is, its values can vary over a specified range (for example , somewhere between - 5 . 0 volts and + 3.2 volts). The values can change an arbitrarily small amount within an arbitrarily small time interval. To be recorded (and understood) by a computer , data must be in digital form. Digital waveforms have discrete values (only certain values are allowed) and have a specified (usually constant) time interval between values. This gives them a " s t e p p e d " (noncontinuous) appearance , as shown by the digitized sawtooth in Figure 1-1. When this time interval becomes small enough, the digital waveform becomes a good approxima- tion of the analog waveform. If the transfer function of the t ransducer and the analog instrumentat ion is known, the digital waveform can be an accurate representat ion of the time-varying quantity to be measured. The process of convert ing an analog signal to a digital one is called analog-to-digital convers ion, and the device that does this is an analog-to- digital conver ter (ADC). The resulting digital signal is usually an array of digital values of known range (scale factor) separated by a fixed time interval (or sampling interval). If the values are sampled at irregular time intervals, the acquired data will contain both value and time information. The reverse process of convert ing digital data to an analog signal is called digital-to-analog convers ion, and the device that does this is called a digital-to-analog conver ter (DAC). Some common applications for DACs include control sys tems, waveform generators , and speech synthe- sizers. A general purpose laboratory data acquisition system typically con- sists of A D C s , D A C s , and simple digital inputs and outputs . Figure 1-2 is (a) Analog Waveform (b) Digitized Waveform Figure 1-1 Comparison of analog and digitized waveforms: (a) sawtooth analog waveform and (b) a coarse digitized representation. Introduction to Data Acquisition 3 Mass Keyboard Display Storage I τ I COMPUTER I Digital Digital Inputs ADC DAC Outputs zu ^ Multiplexer | | ^ u r r j ^ e x e r Analog Inputs Analog Outputs TTTT ΊΤΤΤΤ Inputs from Sensors Outputs to Controls Figure 1-2 Simplified block diagram of a data acquisition system. a simplified block diagram of such a system. Note that additional channels are often added to an A D C or D A C via a multiplexer (or mux), used to select which one of the several analog input signals to convert at any given t ime. This is an economical approach when all the analog signals do not need to be simultaneously monitored. Economics is a major rationale behind using personal computers for data acquisit ion sys tems. The typical data acquisition system of 10-15 years ago, based on a minicomputer , cost about 20 times as much as today ' s sys tems , based on personal computers , at around the same per- formance levels. This is largely due to the continuing decrease of elec- tronic component costs along with increased functionality (more .logic elements in the same package) . Since personal computers have become commonplace in most labs, the cost of implementing a data acquisition system is often jus t the price of an add-in board and support software, which is usually a modera te expense . There are , of course , applications where a data acquisition system based on a personal compute r is not appropriate and a more expensive, dedicated system should be used. The important system parameters for making such a decision include sampling speed, accuracy, resolution, 4 CHAPTER 1 Introduction to Data Acquisition amount of data , multitasking capabilities, and the required data process- ing and display. Personal computer-based systems have certain limitations in these areas , especially regarding sampling speed and handling large amounts of data. However , newer , high-performance personal computers keep "pushing the edge of the e n v e l o p e " ; they can out-perform dedicated data acquisition sys tems. The evolution of the PCs based on the Intel 8 0 x 8 6 microprocessor (or CPU) family, the IBM P C / X T / A T , PS/2 , and compat- ible sys tems, is demonst ra ted in Table 1-1, showing bus width and the amount of available memory space. In recent years , Apple ' s Macintosh computer line has gained popu- larity as a platform for data acquisition, now that a nonproprietary inter- face, N u B u s , is used. These machines , based on the Motorola 68000 family of microprocessors , have certain advantages , including a graphi- cal, consistent operating environment (using icons) and a linear memory addressing space. (The segmented addressing space of the Intel 8 0 x 8 6 family will be discussed in Chapter 5.) Software is as important to data acquisition systems as hardware capabilities. Inefficient software can waste the usefulness of the most able data acquisition hardware system. Conversely, well-written software can squeeze the maximum performance out of mediocre hardware . Software selection is at least as important as hardware selection and often more complex. Data acquisition software controls not only the collection of data but also its analysis and eventual display. Ease of data analysis and presenta- tion are the major reasons behind using computers for data acquisition in the first place. With the appropriate software, computers can process the TABLE 1-1 INTEL 80x86 CPU Family Bus Width Characteristics CPU DATA BUS SIZE ADDRESS BUS SIZE MEMORY SPACE (bits) (bits) (Mbytes) 8086 16 20 1 8088 8 20 1 80286 16 24 16 80386 32 32 4096 80486 32 32 4096 Introduction to Data Acquisition 5 acquired data and produce outputs in the form of tables or plots . Without these capabili t ies, the equipment is not much more than a sophisticated (and expensive) data recorder . An additional area of software use is that of control . Computer outputs may control some aspects of the system that is being measured, as in au tomated industrial process controls . The software must be able to measure system paramete rs , make decisions based on those measure- ments , and vary the compute r outputs accordingly. For example , in a tempera ture regulation system, the input would be a temperature sensor and the output would control a heater . In control applications, software reliability and response time are paramount . Slow or erroneous software responses could cause physical damage. A plethora of commercial ly available PC-based software packages can collect, analyze , and display data graphically, using little or no pro- gramming (see Chapte r 11). This software allows the user to concentra te on the application instead of worrying about the mechanics of getting data from point A to point Β or how to plot a set of Cartesian coordinates . Many commercial software packages contain all three capabilities of data acquisit ion, analysis , and display (the so-called integrated packages) , while others are optimized for only one or two of these areas . The important point is that you do not have to be a computer expert or even a programmer to implement an entire personal computer-based data acquisit ion sys tem. Best of all, you do not have to be rich, either. The next chapter examines the world of analog signals and their t ransducers , the "front e n d " of any data acquisition system.