Data Acquisition KEYTHLEY

A
GREATER
MEASURE
OF
CONFIDENCE
Data Acquisition
and Control
Handbook
A Guide to Hardware and Software for

Computer-Based Measurement and Control
1st
Edition
www.keithley.com
Data Acquisition
and Control
Handbook
A Guide to Hardware and Software for
Computer-Based Measurement and Control
Preface
The Data Acquisition and Control Handbook is a comprehensive
overview of issues that influence the selection and use of equipment
for computerized data acquisition and control. The handbook is primarily a guide to building test and measurement systems that use a
personal computer as a controller and a variety of plug-in boards and
external instruments to gather data and control external processes.
These processes cover multiple industries and markets, including the
fields of factory automation, semiconductors, optoelectonics, telecommunications, automotive, medical, computers, peripherals, aerospace,
research, and education. The goals of this handbook include:
Identifying basic electrical theory that applies to measurement
and control, regardless of the application or selected instrumentation.
Identifying the fundamental building blocks, sensors, and
processes of data acquisition, in order to help the reader form a
sound approach to system design.
Discussing common data acquisition applications that can
serve as models for developing similar systems.
The history behind some current practices is also noted.
Journals, texts, and the World Wide Web are good sources for more
information on specific applications and topics. Search engines can be
invaluable for identifying vendors specializing in many types of sensors, along with recommendations on instrumentation suitable for
specific applications.
For background on making more accurate measurements of lowlevel signals, request a copy of Keithleys Low Level Measurements,
which provides a thorough grounding in the field of sensitive electronic
measurement.
TA B L E
OF
CONTENTS
Section 1: Data Acquisition and Control Overview

1.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1-2
1.2 Data Acquisition and Control Hardware . . . . . . . . . . . . . . . . . . . . .1-2
1.2.1 Plug-in Data Acquisition Boards . . . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.22 External Data Acquisition Systems . . . . . . . . . . . . . . . . . . . . . . . . .1-3
1.2.2.1 Real-Time Data Acquisition and Control . . . . . . . . . . . .1-4
1.2.2.2 Discrete (Bench/Rack) Instruments . . . . . . . . . . . . . . . .1-5
1.2.2.3 Hybrid Data Acquisition Systems . . . . . . . . . . . . . . . . . . .1-5
Section 2: Communication Buses and Protocols

2.1 Computer Hardware Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-2
2.2 Processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-2
2.3 Bus Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-3
2.3.1 ISA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-3
2.3.2 PCI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-4
2.3.3 PCMCIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-5
2.4 Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-5

2.4.1 Serial Ports (RS-232, RS-422, RS-485) . . . . . . . . . . . . . . . . . . . . . . .2-5
2.4.2 Parallel Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-8
2.4.3 IEEE-488 (GPIB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-8
2.4.4 Universal Serial Bus (USB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-9
2.4.5 IEEE-1394 FireWire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-10
2.4.6 Distributed Measurements and Ethernet . . . . . . . . . . . . . . . . . .2-11
2.4.7 Converters, Extenders, Repeaters . . . . . . . . . . . . . . . . . . . . . . . . .2-12
Section 3: Software Overview

3.1 System Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-2
3.2 Open vs. Closed Programming Environments . . . . . . . . . . . . . . . .3-2
3.2.1 History of Open vs. Closed Programming Environments . . . . . .3-2
3.2.2 Open vs. Closed Programming Environments Today . . . . . . . . . .3-3
3.2.2.1 Learning Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-4
3.2.2.2 Text-Based vs. Graphical Syntax . . . . . . . . . . . . . . . . . . . .3-4
3.2.2.3 Programming Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . .3-5

3.2.2.4 Windows Messaging and Event Management . . . . . . . .3-6
3.2.2.5 Debugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-7
3.2.2.6 Application Deployment . . . . . . . . . . . . . . . . . . . . . . . . . .3-7
3.3 Software Development under Windows . . . . . . . . . . . . . . . . . . . . . .3-7

3.3.1 Software Structure Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-7
3.3.2 Device Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-7
3.3.2.1 Driver Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-8
3.3.2.2 The Evolution from DOS to Windows . . . . . . . . . . . . . . .3-9
3.3.3 The Application Programming Interface . . . . . . . . . . . . . . . . . . .3-10
3.3.3.1 The Role of an API . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-10
3.3.3.2 Hardware Independence . . . . . . . . . . . . . . . . . . . . . . . . .3-10
3.3.3.3 ActiveX Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-11
3.3.3.4 The Benefit of ActiveX . . . . . . . . . . . . . . . . . . . . . . . . . . .3-11
3.4 FIFO and Buffer Overrun Issues . . . . . . . . . . . . . . . . . . . . . . . . . . .3-12

3.4.1 The Role of DMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-13
3.4.2 Polled vs. Event-Driven Control . . . . . . . . . . . . . . . . . . . . . . . . . .3-13
3.4.3 Tight Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3-14
3.4.4 Managing Speed and Accuracy Tradeoffs . . . . . . . . . . . . . . . . . .3-15
Section 4: Basic Component Theory

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2
4.2 Passive Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2
4.2.1 Resistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-2
4.2.1.1 Resistor Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-3
4.2.2 Capacitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-6
4.2.3 Inductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-10
4.3 Op Amp Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-11

4.3.1 Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-12
4.3.2 Power Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-12
4.3.3 Input and Output Impedance . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-12
4.3.4 Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-13
4.3.5 Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-13
4.3.6 Inverting vs. Non-Inverting Operation . . . . . . . . . . . . . . . . . . . . .4-14
4.3.7 Normal Mode and Common Mode Voltages . . . . . . . . . . . . . . . .4-15
4.3.8 Single-Ended vs. Differential, Bipolar vs. Unipolar . . . . . . . . . .4-17

4.3.9 Single-Ended vs. Differential Inputs for
Signal Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-19
4.4 Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-19

4.5 Digital I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-21
4.5.1 Digital Logic Types and Logic Levels . . . . . . . . . . . . . . . . . . . . . .4-22
4.5.2 TTL Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-23
4.5.3 CMOS Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-24
Section 5: Basic Analog and Digital I/O

5.1 A/D Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-2
5.1.1 A/D Resolution and Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-2
5.1.2 Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-3
5.1.3 Input Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-5
5.1.4 Maximum A/D Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-5
5.1.5 A/D Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-6
5.1.5.1 Successive Approximation A/D . . . . . . . . . . . . . . . . . . . .5-7
5.1.5.2 Integrating A/D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-7
5.1.5.3 Flash Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-7
5.1.5.4 Sigma-Delta Conversion . . . . . . . . . . . . . . . . . . . . . . . . . .5-8
5.1.6 Aliasing and Anti-Aliasing Filters . . . . . . . . . . . . . . . . . . . . . . . . . .5-8
5.2 D/A Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-9

5.2.1 Four-Wire Remote Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-10
5.3 Interfacing Digital I/O to Applications . . . . . . . . . . . . . . . . . . . . . .5-10

5.3.1 Interfacing with Mechanical Switches . . . . . . . . . . . . . . . . . . . . .5-11
5.3.2 Contact Debouncing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-12
5.3.3 Dry Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-13
5.3.4 Slow Moving Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-13
5.3.5 Dealing with TTL-Incompatible Signal Levels . . . . . . . . . . . . . .5-14
5.3.6 Digital High Current or High Voltage . . . . . . . . . . . . . . . . . . . . . .5-14
5.4 Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-15

5.4.1 Digital Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-16
5.4.2 Analog Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-17
5.5 Ground Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-18
Section 6: Temperature Measurement

6.1 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-2
6.2 Thermocouples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-2
6.2.1 Features and Operating Principle of the Thermocouple . . . . . . .6-2
6.2.1.1 Advantages and Disadvantages of Thermocouples . . . .6-2
6.2.1.2 Operating Principle of the Thermocouple . . . . . . . . . . .6-3
6.2.1.3 Simplifying the Measurement System . . . . . . . . . . . . . . .6-4
6.2.1.4 Linearization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-7
6.2.1.5 Thermocouple Alloys, Extensions, Terminal
Pins, and Other Interconnects . . . . . . . . . . . . . . . . . . . . .6-8
6.2.2 Physical Construction of Commercial Thermocouples . . . . . . . .6-8
6.2.3 Thermocouple Types and Applications . . . . . . . . . . . . . . . . . . . .6-10
6.2.3.1 Base Metal Thermocouples . . . . . . . . . . . . . . . . . . . . . . .6-10
6.2.3.2 Noble Metal Thermocouples . . . . . . . . . . . . . . . . . . . . .6-10
6.2.3.3 Other Types of Thermocouples . . . . . . . . . . . . . . . . . . .6-12
6.3 Resistive Temperature Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . .6-12

6.3.1 Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-13
6.3.2 Principle of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-13
6.3.3 Application of RTDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-14
6.3.4 Three-Wire Bridge Configuration . . . . . . . . . . . . . . . . . . . . . . . . .6-16
6.3.5 Four-Wire RTD Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-17
6.3.6 Converting RTD Resistance to Temperature . . . . . . . . . . . . . . . .6-17
6.3.7 Excitation Current and Joule Heating . . . . . . . . . . . . . . . . . . . . . .6-18
6.4 Thermistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-19

6.4.1 Thermistor Circuit Configuration . . . . . . . . . . . . . . . . . . . . . . . . .6-19
6.4.2 Converting Thermistor Resistance to Temperature . . . . . . . . . .6-20
6.5 Semiconductor Linear Temperature Sensors . . . . . . . . . . . . . . . .6-21

6.6 Thermal Shunting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-21
Section 7: Strain Measurement

7.1 Strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-2
7.2 Poissons Strain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-2
7.3 Strain Gauges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-2
7.4 Gauge Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-6

7.5 Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-6
7.6 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-7
7.6.1 Full Bridge Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-7
7.6.2 Half Bridge Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-8
7.6.3 Quarter Bridge Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-9
7.7 Strain Gauge Signal Conditioning . . . . . . . . . . . . . . . . . . . . . . . . . .7-10

7.8 Shunt Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-10
7.9 Load Cells, Pressure Sensors, and Flow Sensors . . . . . . . . . . . . . .7-11
7.10 Acceleration, Shock, and Vibration . . . . . . . . . . . . . . . . . . . . . . . . .7-12
7.10.1 Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-12
7.10.2 Shock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-13
7.10.3 Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-13
7.10.4 Resonance and Q . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-13
7.10.5 Accelerometer Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-14
7.10.5.1 Strain Gauge Accelerometer . . . . . . . . . . . . . . . . . . . . .7-15
7.10.5.2 Piezoelectric Accelerometer . . . . . . . . . . . . . . . . . . . . .7-16
7.10.5.3 Spring-Resistive Accelerometer . . . . . . . . . . . . . . . . . .7-16
7.10.6 Instrumentation Requirements . . . . . . . . . . . . . . . . . . . . . . . . . .7-16
Section 8: Related Topics of Interest

8.1 Current Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-2
8.1.1 Voltage Burden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-2
8.1.2 DMM vs. A/D Board for Current Measurements . . . . . . . . . . . . .8-3
8.2 Connection Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-4

8.2.1 Cable Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-4
8.2.2 Conductors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-5
8.2.3 Shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-5
8.2.4 Cable Capacitance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-7
Section 9: Application Examples

9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-2
9.2 OEM/Factory Automation and Data Acquisition . . . . . . . . . . . . . .9-2
9.2.1 Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-3
9.2.2 Measurement Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-4
9.3 Semiconductor CVD Application . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-5

9.4 Process Monitoring in a Nuclear Power Plant . . . . . . . . . . . . . . . . .9-6
9.5 Tensile Test Stand Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-8
9.6 Burn-In and Stress Testing of Electronic Devices . . . . . . . . . . . . . .9-9
9.7 Performance Characterization of Shock Absorbers . . . . . . . . . . .9-11
9.8 Instrument-Grade, Low Cost Analog Output Control . . . . . . . . .9-12
9.8.1 Four-Wire Remote Sense Test Description . . . . . . . . . . . . . . . . .9-12
9.8.2 Four-Wire Remote Sense Test Procedure . . . . . . . . . . . . . . . . . . .9-12
9.8.3 Constant Current Source Test Description . . . . . . . . . . . . . . . . .9-14
9.8.4 Quadrant I, Resistive Load Test Procedure . . . . . . . . . . . . . . . . .9-15
9.8.5 Quadrant I and III, Battery Charge/Discharge Test Procedure . .9-16
9.8.6 Source/Sink Test Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-17
9.8.7 Quadrant I and II, Battery Charge/Discharge Test Procedure . . .9-17
Appendix A: Selection Guide for Plug-in Boards vs.

External Data Acquisition Instruments
Appendix B: Glossary
Appendix C: Diameter and Resistance of Various Wire Gauges
Appendix D: Safety Considerations
Index
SECTION 1
Data Acquisition and

Control Overview
1.1
Definition
Although concepts like data acquisition and test and measurement can
be surprisingly difficult to define completely, most computer users,
engineers, and scientists agree there are several common elements:
A personal computer (PC) is used to program test equipment
and manipulate or store data. The term PC is used in a general sense to include any computer running any operating system
and software that supports the desired result. The PC may also
be used for supporting functions, such as real-time graphing or
report generation. The PC may not necessarily be in constant
control of the data acquisition equipment or even remain connected to the data acquisition equipment at all times.
Test equipment can consist of data acquisition plug-in boards
for PCs, external board chassis, or discrete instruments. External
chassis and discrete instruments typically can be connected to a
PC using either standard communication ports or a proprietary
interface board in the PC.
The test equipment can perform one or more measurement and
control processes using various combinations of analog input,
analog output, digital I/O, or other specialized functions.
The difficulty involved in differentiating between terms such as
data acquisition, test and measurement, and measurement and control
stems from the blurred boundaries that separate the different types of
instrumentation in terms of operation, features, and performance. For
example, some stand-alone instruments now contain card slots and
microprocessors, use operating system software, and operate more like
computers than like traditional instruments. Some external instruments now make it possible to construct test systems with high
channel counts that gather data and log it to a controlling computer.
Plug-in boards can transform computers into multi-range digital
multimeters, oscilloscopes, or other instruments, complete with userfriendly, on-screen virtual front panels.
For the sake of simplicity, this handbook uses the term data acquisition and control broadly to refer to a variety of hardware and software solutions capable of making measurements and controlling
external processes. The term computer is also defined rather broadly; however, for most applications, a computer means an IBMcompatible PC running Microsoft Windows 95 or later, unless otherwise noted.
1.2
Data Acquisition and Control Hardware

Data acquisition and control hardware is available in a number of
forms, which offer varying levels of functionality, channel count,
speed, resolution, accuracy, and cost. This section summarizes the fea-
SECTION 1
tures and benefits generally associated with the various categories,

based on a broad cross-section of products. Refer to Appendix A for a
comparison of plug-in boards and external instruments.
1.2.1
Plug-in Data Acquisition Boards

Like display adapters, modems, and other types of expansion boards,
plug-in data acquisition boards are designed for mounting in board
slots on a computer motherboard. Today, most data acquisition boards
are designed for the current PCI (Peripheral Component Interconnect)
or earlier ISA (Industry Standard Architecture) buses. Data acquisition
plug-in boards and interfaces have been developed for other buses
(EISA, IBM Micro Channel, and various Apple buses), but these are no
longer considered mainstream products.
As a category, plug-in boards offer a variety of test functions, high
channel counts, high speed, and adequate sensitivity to measure moderately low signal levels, at relatively low cost.
Table 1-1.
Features of plug-in data acquisition boards
Least expensive method of computerized measurement and control.

High speed available (100kHz to 1GHz and higher).
Available in multi-function versions that combine A/D, D/A, digital I/O,
counting, timing, and specialized functions.
Good for tasks involving low-to-moderate channel counts.
Performance adequate to excellent for most tasks, but electrical noise inside
the PC can limit ability to perform sensitive measurements.
Input voltage range is limited to approximately 10V.
Use of PC expansion slots and internal resources can limit expansion
potential and consume PC resources.
Making or changing connections to boards I/O terminals can be
inconvenient.
1.2.2
External Data Acquisition Systems

The original implementation of an external data acquisition system
was a self-powered system that communicated with a computer
through a standard or proprietary interface. As a boxed alternative to
plug-in boards, this type of system usually offered more I/O channels,
a quieter electrical environment, and greater versatility and speed in
adapting to different applications.
Today, external data acquisition systems often take the form of a
stand-alone test and measurement solution oriented toward industrial
applications. The applications for which they are used typically
demand more than a system based on a PC with plug-in boards can
provide or this type of architecture is simply inappropriate for the
application. Modern external data acquisition systems offer:
DATA ACQUISITION AND CONTROL OVERVIEW
High sensitivity to low-level voltage signals, i.e., approximately

1mV or lower.
Applications involving many types of sensors, high channel
counts, or the need for stand-alone operation.
Applications requiring tight, real-time process control.
Like the plug-in board based system, these external systems
require the use of a computer for operation and data storage. However,
the computer can be built up on boards, just as the instruments are,
and incorporated into the board rack. There are several architectures
for external industrial data acquisition systems, including VME, VXI,
MXI, Compact PCI, and PXI. These systems use mechanically robust,
standardized board racks and plug-in instrument modules that offer a
full range of test and measurement functions. Some external system
designs include microprocessor modules that support all the standard
PC user interface elements, including keyboard, monitor, mouse, and
standard communication ports. Frequently, these systems can also run
Microsoft Windows and other PC applications. In this case, a conventional PC may only be needed to develop programs or off-load data for
manipulation or analysis.
Table 1-2.
Features of external data acquisition chassis
Multiple board slots permit mixing-and-matching boards to support

specialized acquisition and control tasks and higher channel counts.
Chassis offers an electrically quieter environment than a PC, allowing for
more sensitive measurements.
Use of standard interfaces (IEEE-488, RS-232, USB, FireWire, Ethernet) can
facilitate daisy chaining, networking, long distance acquisition, and use with
non-PC computers.
Dedicated processor and memory can support critical real-time control
applications or stand-alone acquisition independent of a PC.
Standardized modular architectures are mechanically robust, easy to
configure, and provide for a variety of measurement and control functions.
Required chassis, modules, and accessories are cost-effective for high
channel counts.
Some architectures have minimal vendor support, limiting the sources of
equipment and accessories available.
1.2.2.1 Real-Time Data Acquisition and Control

Critical real-time control is an important issue in data acquisition and
control systems. Applications that demand real-time control are typically better suited to external systems than to systems based on PC
plug-in boards. Although Microsoft Windows has become the standard
operating system for PC applications, it is a non-deterministic operating system that cant provide predictable response times in critical
SECTION 1
measurement and control applications. Therefore, the solution is to

link the PC to a system that can operate autonomously and provide
rapid, predictable responses to external stimuli.
1.2.2.2 Discrete (Bench/Rack) Instruments
Originally, discrete electronic test instruments consisted mostly of
single-channel meters, sources, and related instrumentation intended
for general-purpose test applications. Over the years, the addition of
communication interfaces and advances in instrument design, manufacturing, and measurement technology have extended the range and
functionality of these instruments. New products such as scanners,
multiplexers, SourceMeter instruments, counter/timers, nanovoltmeters, micro-ohmmeters, and other specialized instrumentation
have made it possible to create computer-controlled test and measurement systems that offer exceptional sensitivity and resolution. Some
systems of this type can service only one channel or just a few channels, so their cost per channel is high. However, the addition of switch
matrices and multiplexers can lower the cost per channel by allowing
one set of instruments to service many channels while preserving high
signal integrity. These instruments can also be combined with computers that contain plug-in data acquisition boards.
Table 1-3.
Features of discrete instruments for data acquisition
Support measurement ranges and sensitivities generally beyond the limits of

standard plug-in boards and eternal data acquisition systems.
Use standard interfaces (e.g., IEEE-488, RS-232, FireWire, USB) that support
long-distance acquisition, compatibility with non-IBM-compatible computers, or use with computers without available expansion slots.
Most suitable for measurement of voltage, current, resistance, capacitance,
inductance, temperature, etc. May not be effective solutions for some types
of specialized sensors or signal conditioning requirements.
Generally slower than plug-in boards or external data acquisition systems.
More expensive than standard data acquisition systems on a per-channel
basis.
1.2.2.3 Hybrid Data Acquisition Systems

Hybrid systems are a relatively recent development in external data
acquisition systems. A typical hybrid system combines a DMM-type
user interface with several standard data acquisition functions and
expansion capabilities in a compact, instrument-like package. Typical
functions include AC and DC voltage and current measurements, temperature and frequency measurements, event counting, timing, triggering, and process control. Keithleys Integra Series, which includes
the Model 2700 and Model 2750 and their associated plug-in modules,
provides multiple board slots for expanding the systems measurement
capabilities and channel capacity (Figure 1-1).
DATA ACQUISITION AND CONTROL OVERVIEW
Figure 1-1.
Keithley Model 2700 hybrid data acquisition system
Table 1-4.
Features of a hybrid data acquisition system
Delivers accuracy, measurement range, and sensitivity typical of bench

DMMs, and superior to standard data acquisition equipment.
DMM front end with digital display and front panel controls provides resolution equivalent to a DMM (18- to 22-bit A/D or better).
Built-in data and program storage memory for stand-alone data logging and
process control.
Uses standard interfaces (IEEE-488) that support long-distance acquisition
and provide compatibility with non-PC computers.
Cost-effective on a per-channel basis.
Limited expansion capacity (less critical because base test capability is
already complete).
Generally slower than plug-in boards or external data acquisition systems.
SECTION 1
SECTION 2
Communication Buses
and Protocols
2.1
Computer Hardware Overview

The overall performance, suitability, and long-term usefulness of a PCbased data acquisition system often depends as much on the PC used
as on the measurement hardware. As a general rule, a faster processor,
more memory, and more disk storage will improve system performance. It is also important to follow the recommendations on hardware and software given by manufacturers concerning compatible
computers and peripherals.
2.2
Processor
PC manufacturers introduce faster, more powerful PCs continually by
making improvements in architecture, processor speed, disk storage,
memory, peripherals, etc. Often, these improvements are incremental,
so they may not provide sufficient incentive to the user to replace an
existing PC. Therefore, the installed base of PCs used in test and measurement will contain several product generations, each with varying
degrees of suitability for a desired application.
A survey of computing hardware recommended for use with various data acquisition products will show that few require the use of a
cutting-edge PC to function. While the minimum system requirements
for installing and running most new software applications continues to
escalate, these requirements are typically a few generations behind the
current state of PC technology. For example, when a typical high-end
PC included a Pentium III processor and 64128MB of RAM, the minimum workable system for many data acquisition hardware and software products was still an 80486 processor with just 832MB of RAM.
A recommended system would fall somewhere between the extremes;
for example, a Pentium-class PC with a 233MHz processor, 64MB RAM,
510GB of fixed disk storage, and a VGA display adapter capable of 256
colors at 800600 resolution.
Be aware that while a minimum PC configuration may offer sufficient resources to manage the test and measurement task, it probably
cant handle much more. Spreading resources too thin by running multiple programs during data acquisition can result in sluggish program
response, lock-ups, or other problems.
Increasing the amount of on-board RAM can boost system performance dramatically, especially with Windows operating systems.
For large or concurrent applications, adding RAM can reduce the need
for the system to swap information to and from disk drives, which is far
more time-consuming than retrieving the same information from
RAM. Compared to upgrading a processor, motherboard, or an entire
system, installing more RAM is a relatively easy and inexpensive way to
extend the useful life of an older machine.
SECTION 2
In 2001, most newer PCs contain PCI expansion slots, but few (if
any) ISA slots. Although the PCI architecture offers some advantages in
terms of speed of interrupt handling, some users may find it easier to
continue using ISA if they have a sizeable investment in ISA-based test
and measurement systems platforms, software, and maintenance
resources. Even as PCI-based machines assume architectural leadership, PC manufacturers have begun to discuss producing slotless PCs
that will require the use of external test and measurement hardware.
2.3
Bus Architecture
Over the PCs history, a number of internal PC buses have been developed. The original ISA (Industry Standard Architecture) bus and the
current PCI (Peripheral Component Interconnect) bus are the two
most common architectures employed for data acquisition products.
A few other architectures, notably IBMs Micro Channel
Architecture (MCA), the Enhanced Industry Standard Architecture
(EISA), and the Video Electronics Standards Association (VESA) local
bus, were introduced prior to the PCI. All these buses were designed to
replace or augment the ISA bus with advanced features that are now
common to PCI. These features included higher speed, 32-bit operation, Plug-and-Play operation, and bus mastering. The simple but wellestablished ISA bus has actually outlasted all of them. While a few
Micro Channel and Apple data acquisition products were developed,
neither architecture took lasting root in the data acquisition market.
Today, neither is regarded as a mainstream data acquisition platform.
2.3.1
ISA
The ISA bus was a core element of the original IBM PC (ca. 1981),
although the term ISA was not adopted until the bus was well established and other buses were introduced by IBM and other computer
manufacturers. Initially, the ISA bus was an 8-bit, 4.77MHz bus
designed to satisfy the speed and data path requirements of the PCs
8088 processor. In 1984, the ISA bus was upgraded to 16 bits and 8MHz
for the 80286 processor used in the IBM PC/AT, and the card-edge connector was physically extended to provide a 16-bit data path. Both the
8-bit and 16-bit variations are included under the general term ISA.
ISA plug-in data acquisition boards and communication interfaces still
exist today. Some have only the primary 8-bit card edge connector and
will fit an 8- or 16-bit ISA slot. Sixteen-bit boards include the secondary
card edge extension and will only fit a 16-bit ISA slot.
Direct Memory Access (DMA) and Plug-and-Play (PnP) are two
noteworthy features of the ISA bus. DMA improves data transfer
between peripherals and memory by enabling an expansion board to
write to system memory directly, without intervention of the microprocessor. Strictly speaking, Plug-and-Play was introduced on later ISA
COMMUNICATION BUSES AND PROTOCOLS
boards to coincide with the introduction of Windows 95. PnP enabled

Windows 95 to identify the boards present in the PCs expansion slots.
However, the concept of automatic board identification was a feature
of MCA and EISA architectures years before Windows 95. Unfortunately, the prevailing DOS operating system did not support PnP, so few PCs
could take advantage of this feature. Plug-and-Play operation requires
a combination of PnP-compatible peripherals, system hardware, system BIOS, and operating system software.
Plug-and-Play capability allows a computer to recognize a new
internal or external peripheral plugged into the system automatically,
and to configure system resources (interrupts, memory addresses, etc.)
to operate these devices. Without PnP, installing an ISA board typically
involved setting physical jumpers or switches, or configuring the board
through software to establish IRQ, I/O addresses, or DMA channels for
the peripheral. In a heavily populated system, users could find it difficult to install all the desired peripherals without running out of
resources or causing conflicts.
The ISA bus is a well-entrenched standard that will persist despite
dwindling support from Intel, Microsoft, and computer manufacturers. Even in the 8-bit version, ISAs speed is more than adequate for
many applications. A wide variety of ISA data acquisition boards
remain available.
2.3.2
PCI
Intel introduced the 32-bit PCI (Peripheral Component Interconnect)
bus in 1993. It appeared first in Pentium and later-generation 486
machines. PCI is the current PC bus standard and was the only internal
PC bus approved by Intel and Microsoft as of January 2001.
The PCI bus offers a number of performance advantages over the
ISA bus. The PCI bus runs at 33MHz compared to 8MHz for the ISA bus.
PCIs improved PnP capability allows jumperless installation of PCI
boards and automatic allocation of interrupts, I/O addresses, and DMA
channels. Interrupts can be shared by different PCI devices. PCI also
features bus mastering, which, like DMA, allows PCI devices to take
control of the bus and perform data transfers directly to memory, without intervention of the processor. PCIs improved design allows simultaneous bus mastering of multiple devices.
Since the introduction of the PCI bus, the total number of slots on
motherboards has dwindled, and there are few (if any) ISA slots.
However, a PCI-based machine with three or four PCI slots still maintains a high level of expansion capability. Hard disk controllers, serial
and printer ports, sound, video, and other system resources that usually required separate interface boards in an ISA platform are now populated on PCI motherboards directly, leaving the few existing slots free
for other peripherals. PCI data acquisition boards are available for
SECTION 2
most applications implemented with ISA boards and provide for easier
installation and better performance than ISA.
2.3.3
PCMCIA
Most laptop computers now include at least one Personal Computer
Memory Card International Association (PCMCIA) slot, which enables
the computer to accept data acquisition boards, interfaces, and other
PCMCIA peripherals. In terms of architecture and performance, the
PCMCIA bus most closely resembles the ISA bus with added Plug-andPlay and hot-swapping features. PCMCIA data acquisition boards are
readily available, although the selection is limited in comparison with
ISA and PCI versions. The PCMCIA bus also has less to offer in terms of
channel counts, high resolution A/D, and A/D speeds greater than 100
kilosamples/second.
PCMCIA boards are roughly the size of a credit card, which requires
cabling to an external connector pod or terminal block. This tiny form
factor precludes some useful features typical of full-size boards, such
as power supply regulation, advanced signal conditioning, and maximum tolerance of electrostatic discharges.
2.4
Connectivity
The topic of connectivity includes a variety of external buses and ports
that are available on PCs, either as standard equipment or as plug-in
interface options. These include parallel printer ports, the traditional
RS-232 serial port, RS-422 and RS-485 serial ports, IEEE-488, and
newer high speed buses such as the Universal Serial Bus (USB), IEEE1394 FireWire bus, and Ethernet. These buses are summarized in
Table 2-1.
2.4.1
Serial Ports (RS-232, RS-422, RS-485)

The Electronic Industries Associations (EIA) RS-232 serial standard is
one of the most common communication protocols. Introduced in
1962, RS-232 became an early common feature on mainframe and mini
computers, data terminals, printers, modems, and other data equipment long before PCs were developed.
RS-232 was originally specified for data communications between
a single driver device and a single receiver over distances up to 15
meters (50 feet), at speeds up to 19,200 bits per second. These distance
and speed limits have proven to be conservative; RS-232 is now routinely used at speeds of 115230 kbits/second. RS-232 has also been
operated at distances greater than 25 meters. The ability to achieve
these operating parameters depends on the environment and quality
of cabling used. Generally speaking, a single system cant provide both
maximum speed and maximum distance.
Table 2-1.
External PC buses
Generic
Name
Industry
Designation
Typical
Max.
Distance
Typical
Max.
Speed
Serial
RS-232
15m (50ft)
115kbits/s
Standard serial interface.

One device per port.
Serial
RS-422
1220m
(4000ft)
115kbits/s
One transmitter may drive

up to 10 receivers. Normally
requires a plug-in RS-422
adapter.
Serial
RS-485
1220m
(4000ft)
115kbits/s
Supports up to 32 transmit or
receive devices on one bus.
Normally requires a plug-in
RS-485 adapter.
Parallel
(Printer
Port)
SPP, EPP,
ECP
~15m
(50ft)
100+ kB/s
Popular for printers,

scanners, disk drives, other
peripherals. Not a mainstream DAQ interface.
General
IEEE-488
Purpose
Interface
Bus (GPIB)
~2m (6ft)
Can be
extended
1MB/s
A standard fixture on many

scientific instruments and
peripherals. Supports up to
15 devices on one bus.
Normally requires a plug-in
GPIB adapter.
Universal USB
Serial Bus
5m (16.5ft)
per cable
drop; 15m
(50ft) total
12Mbits/s
(480Mbits/s
planned)
Supports hot plugging, PnP,

and connection of up to 127
devices using USB hubs.
USB is standard on many
new computers. Driver support in W95 OSR 2.1, Win 98,
Win 2000.
FireWire
IEEE-1394 4.5m (15ft)
Ethernet
10BaseT,
100BaseT
Features/Advantages
100400Mbits/s Supports hot plugging, PnP,

(1+ Gbits/s
and daisy chaining of up to
planned)
63 devices. FireWire is not
yet widely implemented.
Driver support in Win 98,
Win 2000.
~925m
10Mbits/s,
(3000ft);
100Mbits/s*
farther on
LANs or
over Internet
Standard high speed office/

industrial network. Interface
required for computer.
*Raw bit rates. Actual throughput is application dependent.
SECTION 2
The traditional RS-232 connector was a 25-pin D-shell design.

However, many serial products now include 9-pin serial ports with (or
in place of) the 25-pin version. Both port types contain the same nine
signal lines, although pin numbers for specific signals are different. A
full RS-232 link includes Tx (transmit), Rx (receive), signal ground, and
several additional handshaking lines. Handshaking is important with
some peripherals (such as modems) to control transmission and prevent data overrun. It is possible to establish serial communication for
some applications using only Tx, Rx, and signal ground, in which case,
handshake pins on each piece of serial equipment must be terminated
with various loop-back and pull-up connections. Check the documentation for any RS-232 device to be used for pin-out and connection
information.
A second popular serial standard, RS-422, is similar to RS-232,
except that it uses differential data transmission. This method employs
two active lines to transmit a signal, rather than RS-232s signal line and
ground. Differential data transmission provides superior noise rejection, which supports higher data rates and greater operating distances.
Data transmission speeds up to 115 kbits/second and distances up to
4000 feet are possible. As with RS-232, maximum speed and maximum
distance are mutually exclusive. Connector pin designations for RS-422
differ from those for RS-232, so ports and equipment using these standards are incompatible. The RS-442s ability to handle multi-drop
applications, where one driver device can transmit to up to ten
receivers, is a significant advantage over the RS-232.
RS-485 is another popular serial standard, which combines elements of RS-422 with the ability to handle multiple receivers and
multiple transmitters on one bus. Like RS-422, RS-485 uses differential
data transmission and can operate at speeds of up to 115 kbits/second
over distances up to 4000 feet. In addition, RS-485 can support up to 32
driver devices and 32 receivers, making it possible to construct a multipoint network of transmitters and receivers using a single RS-485
cable run.
With the possible exception of Ethernet, RS-422 and RS-485 provide the best combination of distance and ease of use for data acquisition applications. As with RS-232, achieving long-distance, error-free
communication depends on cable quality and environmental conditions. At distances less than 4000 feet, allowable transmission speeds
increase, with speeds up to 10 megabits/second possible at 50 feet. The
use of high quality, shielded twisted pair cabling is critical to achieve
optimum performance.
Plug-in RS-422 and RS-485 interfaces for ISA and PCI buses are
readily available. Many bench-top instruments and some external data
acquisition products include an RS-232 or IEEE-488 port, both of
which can be converted to RS-422 or RS-485 by using the appropriate
adapter. RS-422 and RS-485 ports are more common on industrial

systems, where support for long-distance data acquisition is a prime
concern.
2.4.2
Parallel Port
A parallel printer port can be found on nearly every PC. The original
purpose of this port was to drive a Centronics-compatible printer at a
distance up to 50 feet. Over the years, other non-printer parallel
peripherals have been introduced, including scanners, disk drives, and
tape drives.
The parallel interface uses eight separate transmission lines, allowing data to be transmitted a byte at a time, and resulting in higher
speeds than RS-232. More recently, the IEEE-1284 standard was drafted to describe five different uni-directional and bi-directional hardware configurations and operating enhancements for the parallel port.
The first three modes are uni-directional and support data transfer
rates up to 100 kbytes per second. The last two, EPP (Enhanced Parallel
Port) and ECP (Enhanced Capability Port), are bi-directional, and offer
even higher transfer rates. These enhanced operating modes require a
parallel port and cabling that supports high speed, bi-directional operation. Most new PCI-based systems with communication ports populated directly on the motherboard include an enhanced parallel port.
Despite its current all-but-universal availability and high speed
capability, the parallel port hasnt been embraced as a mainstream
interface for test and measurement, although there is some data acquisition equipment with parallel interfacing. The reason for this lack of
popularity may be as simple as the fact that the parallel interfaces distance and speed capabilities are already duplicated or surpassed by
other standards, such as USB and FireWire. Furthermore, parallel ports
have been implemented in hardware in different ways, which can
sometimes make it difficult to get parallel devices operating properly.
Finally, the parallel port, like the ISA slot, may disappear from PCs altogether as printers and other parallel peripherals increasingly come
with USB or FireWire interfaces.
2.4.3
IEEE-488 (GPIB)
The IEEE-488 bus was developed by Hewlett-Packard in 1965 as the
HP Interface Bus (HP-IB) to provide a standardized communication
interface for lab instruments, recorders, and related equipment. The
interface proved to be both popular and versatile; it was ultimately
renamed GPIB (General Purpose Interface Bus) by the IEEE. GPIB is
also known generically as IEEE-488, although the complete standard
consists of the IEEE-488.1-1987 (hardware specification) and IEEE488.2-1992 (a superset of IEEE-488.1 dealing with software and communication).
SECTION 2
The GPIB has become an industry standard for an extremely wide

range of electronic instruments, including data acquisition products.
This 8-bit parallel bus transmits data at up to one megabyte per second. Up to 15 devices can be wired on one bus in either a daisy chain
or star configuration. The maximum distance per cable drop is 6.5 feet
(2m), although this distance can be extended to more than 6500 feet
(2km) with repeater hardware.
GPIB interfaces generally arent standard equipment on PCs, so a
plug-in adapter board and appropriate driver software must be installed in the PC to communicate with GPIB instruments. These boards
are available in ISA, PCI, and PCMCIA versions from most data acquisition manufacturers.
In 1992, some manufacturers of GPIB interfaces began to introduce
a higher-speed version of GPIB that was eventually designated HS488. The proposed change offers increased bus speed at the expense
of the robustness the standard IEEE-488 handshaking scheme affords.
Opponents of HS-488 cite compatibility and interoperability issues, a
marginal increase in practical throughput, and the availability of
newer, faster buses as reasons not to tamper with the existing GPIB
standard. Although HS-488 interfaces are now available, the standard
hasnt been widely adopted. Further, with more than 20 years of development behind them, IEEE-488 interfaces and instruments have the
advantage of an extremely large investment in business and academic
applications. These instruments are not likely to be replaced soon,
despite any advantages HS-488 offers.
2.4.4
Universal Serial Bus (USB)

The Universal Serial Bus (USB) was introduced in 1995 to address a
number of connectivity issues associated with existing serial communication standards. USB supports multiple devices and provides easier
installation, faster transmission speeds, and simpler cabling requirements than conventional parallel or serial ports. USB is also designed
to supply modest operating power directly to peripherals, eliminating
the need for external power supplies in some cases.
Many PCI-based Pentium PCs include one or two USB ports on the
rear panel as standard equipment. Dual USB ports can support elaborate USB configurations, because external USB hubs can be attached to
each port, allowing up to 127 USB peripherals to be connected. Hubs
can be stand-alone devices or they can be built into keyboards, printers, or other USB peripherals. They can also include power supplies to
drive USB peripherals.
USB operates at distances up to 16.5 feet (5m) per cable drop, with
a maximum of 50 feet (15m). Maximum speed for USB version 1.1 is
12Mbits/second. However, the USB 2.0 standard released in 2000 raises
the maximum USB speed to 480Mbits/secondfaster than the first-
generation FireWire interface, but slower than that anticipated for

future FireWire implementations.
USB peripherals are hot plug devices, so they can be attached or
removed from an energized PC without damage and without the need
to reboot. USB also incorporates Plug-and-Play, so a compatible PnP
operating system will automatically recognize and reconfigure PC
resources to handle the addition or removal of a USB peripheral. These
ease-of-use features are among the strongest arguments for USB.
Two initial concerns with USBoperating system support and the
availability of peripheralshave also been addressed. Successful USB
operation requires that a PC contain USB hardware ports, operating
system support, and driver support. Normally, all requirements can be
retrofitted to a machine without USB support.
USB is fully supported by Windows 98, Windows Me, and Windows
2000. Some versions of Windows 95 also support USB. Microsofts firstgeneration USB drivers are included in Windows 95 OSR 2.1, which
was bundled with OEM computer systems, but not available for retail
sale. There is no Microsoft USB support available for the original version of Win 95 or Windows NT. Although third-party USB drivers may
be available for early versions of Windows 95, the simplest solution
might be simply to upgrade the operating system to a later version that
supports USB.
The list of USB devices available includes mice, keyboards, PC
cameras, scanners, external modems, disk drives, printers, and speakers. Predictably, availability of USB data acquisition hardware has
lagged, but it will undoubtedly increase as USB becomes more popular.
Adapters are also available to mate conventional GPIB, serial, and parallel devices to USB.
2.4.5
IEEE-1394 FireWire
FireWire is a high-speed serial interface thats outwardly similar to
USB. The technology was originally developed by Apple Computer for
digital video, audio, and related high-speed applications, then proposed as a standard to the IEEE. In 1995, FireWire received the designation IEEE-1394 (sometimes shortened to 1394.) FireWire offers
high transmission speeds, simplified low-cost interconnects, and the
ability to daisy chain up to 63 devices on one bus.
The current FireWire standard describes data rates of
100400Mbits/second over distances up to 15 feet (4.5m). Speeds of a
gigabit/second or higher are planned, and future enhancements are
expected to extend the maximum transmission distance. Like USB, a
FireWire device can be hot plugged and is Plug-and-Play compatible.
Windows 98, Windows Me, and Windows 2000 contain drivers for
FireWire. Windows 95 will not support FireWire, at least not through
driver products offered by Microsoft. FireWire has replaced the SCSI
SECTION 2
port on more recent Apple machines. FireWire ports are not yet common on PCs, so FireWire support must usually be added to currentgeneration PCs by using a plug-in adapter.
The FireWire cable and connector are outwardly similar in appearance to USB, although the two standards are neither physically nor
electrically compatible. Cabling consists of dual shielded/twisted conductor pairs, plus two additional conductors to carry power to FireWire
peripherals. Sonys iLINK is a variation on FireWire that uses only
four conductors to carry signal. Power is not transmitted, but the signal
protocol is essentially the same as FireWire.
FireWire has been embraced for applications in digital video (DV)
and audio, which is not surprising, given that these are the traditional
strengths of Apple machines. The technology is a natural fit for any
application that requires moving large amounts of data at high speed,
and is adaptable to virtually any type of peripheral. In comparing the
speed of FireWire with USB, it appears that both standards will continue to leapfrog each other in order to gain temporary speed advantages.
In the end, FireWire may prevail in speed, but may offer less in the way
of distance and device support (63 vs. 127 peripherals). However, both
buses are relatively new, and both may ultimately become standard
equipment on PCs as complementary technologies.
As far as FireWire and data acquisition are concerned, there are few
FireWire-compatible data acquisition devices at this writing. However,
it is significant that FireWire apparently leads USB in being considered
as a replacement for GPIB in test and measurement applications.
Protocols have been developed for transmitting IEEE-488.1 and -488.2
messages and command/control sequences on a FireWire bus. This
encourages the use of GPIB (SCPI) commands with FireWire.
2.4.6
Distributed Measurements and Ethernet

Many test and measurement applications require the ability to acquire
data from separate stations located over a wide geographic area
larger than that is normally possible with plug-in boards or instruments that use traditional parallel, RS-232, or GPIB external buses.
Distances of more than a few feet between instrumentation and
sensors can result in signal degradation, noise pickup, signal conditioning complications, measurement speed restrictions, and a variety
of other problems. The nature of these limitations depends on noise in
the environment, type and level of the signal, and length of cabling,
among other factors.
There are two options for maintaining signal integrity where great
distances separate the computer and sensors. The first is to condition
and convert the signal at the sensor location to a form that wont be
affected by adverse factors, then transmit the signal to centrally located test and measurement hardware. Common examples of this tech-
nique include conversion of signals to a higher voltage (05V or 010V),

conversion to a current loop (420mA, 020mA), and conversion to frequency. The second method is to move the measurement instrumentation to the source of the signals, while continuing to use a centrally
located computer. Communication signals between the instrumentation and the computer travel over a standard or proprietary bus, usually in digitized form. In this second case, the arrangement comprises a
distributed data acquisition system.
There are advantages and disadvantages to each method of longdistance acquisition. Centralizing the test and measurement hardware
eliminates the expense of duplicating measurement equipment at
each sensor location. However, the hardware needed to transmit signals long distances back to the instrumentation and computer can offset some of the cost savings, and also complicates the setup.
With a true distributed system, data acquisition hardware must be
included at each node of the system, resulting in higher overall system
costs when compared to centralized test hardware. Communication
between the computer and the hardware can use one of several standard communication protocols designed for error-free communication over extended distances. These distances extend to 4000 feet for
RS-422 and RS-485 or around the globe for networked data acquisition.
Communication distances greater than 4000 feet can be achieved
by combining data acquisition equipment with a network such as
Ethernet. Network-based systems require installing a suitable interface
in the PC and configuring the operating system for network support.
These requirements are easily met using Ethernet, TCP/IP, and
Microsoft Windows. Ethernet interface boards are inexpensive and
readily available. TCP/IP is a nearly universal communication protocol
for Ethernet-based systems that is supported by Windows and UNIX.
Furthermore, most industrial data acquisition systems can be outfitted
with a processor module containing an Ethernet port and Windows
operating system. Ethernet adapters for GPIB and serial ports make it
possible to assign an address to programmable instruments so they
can communicate over a network.
A functional drawback of Ethernet data acquisition systems concerns real-time control. Like Windows, Ethernet is a non-deterministic
system that precludes reliable, real-time process control. This can
become even more of an issue when the World Wide Web is involved.
2.4.7
Converters, Extenders, Repeaters

Communication protocols include specifications of the distance over
which reliable communication can be maintained. It is often possible
to extend this distance by using external hardware to boost communication signal levels or to convert one protocol to another that can be
transmitted over a longer distance. In other cases, a converter may sim-
SECTION 2
ply allow two pieces of equipment with different interfaces to communicate with each other, with distance being a secondary consideration.
Common types of interfaces and converters include serial to IEEE488, serial to Ethernet, and IEEE-488 to Ethernet. In each case, the user
must determine whether using these devices achieves the desired goal
more cost-effectively than another method. For example, a computer
may contain RS-232 serial ports, while the application requires an
instrument with IEEE-488 interface. The user must determine whether
it is better to use a serial to GPIB adapter or to install an IEEE-488 interface board in the computer.
SECTION 3
Software
Overview
3.1
System Software
Surveys indicate roughly 80 percent of test, measurement, and data
acquisition applications are computer controlled. These automated
test systems typically use PC-based data acquisition hardware and
software. These two aspects are interdependent and equally important.
The best choice for data acquisition software for a given task
depends on any number of factors, including computer platform,
operating system, user programming skills, and application type. While
some of these items are a matter of personal preference, others are dictated by conditions beyond the users control, such as corporate policy,
financial concerns, the need for compatibility with other software or
hardware, etc. Operating systems and computing platforms other than
Microsoft Windows can be used successfully for data acquisition, but
the full burden for program development, hardware control, and support typically rests with the user. Therefore, users who strongly prefer a
non-Windows environment may still find that the economics and
greater ease of system development encourage them to use Windows
for data acquisition applications.
3.2
Open vs. Closed Programming Environments

Since the introduction of Windows 95, the general trend in PC software
has been to make computing simpler, more graphic, and more accessible to users, without the need for users to have in-depth knowledge of
application programming. However, this philosophy was associated
with computerized measurement and control software well before
Windows became popular.
3.2.1
History of Open vs. Closed Programming Environments

Prior to the introduction of the Windows environment, there were two
fundamental types of data acquisition software. First, most data acquisition manufacturers supplied drivers and/or libraries for their data
acquisition products. This software support added hardware control
capability to popular, DOS-based programming languages such as
interpreter BASIC, compiler BASIC, C, and PASCAL. Languages were
text-oriented and presented little more than a command prompt from
which the programmer had to develop all user interface, hardware control, data manipulation, and display functions. The programming environment was said to be open for several reasons:
The program was developed by the end user, who had access to
all the code.
The user was free to distribute or modify the program as needed
to provide new features or even to operate with different hardware.
SECTION 3
Commercial software developers might also develop software

tools that could coexist with the users selected programming
language and hardware.
Alternately, a user could opt for a proprietary, closed-architecture
package designed to automate hardware control, data manipulation,
graphing, analysis, and other functions. Such packages were also DOS
applications, and were normally purchased with drivers or libraries supporting a specific manufacturers data acquisition products. Other than
the ability to develop end applications, the user had little control over
how the software operated. Copyright restrictions meant that a program
developed on one computer might not be shared legally with other users
unless a runtime license was obtained for other computers.
Proprietary software packages were generally developed by third
parties who did not manufacture test and measurement hardware, so
support for new hardware could lag the introduction of the hardware.
While the libraries and drivers that were offered by hardware manufacturers were usually free, proprietary packages could be quite
expensive.
Proprietary, closed-architecture packages have classically been
associated with the concept of no programming. Users generate test
programs with tools such as pull-down menus and fill-in-the-blank
forms. Although the user might be insulated from conventional programming, there is still a learning curve associated with closed software environments, and the total effort required can equal that of
using an open programming language.
3.2.2
Open vs. Closed Programming Environments Today

Today, the dominant PC operating system is some version of Windows
(95/98/Me/2000/NT), rather than DOS. The majority of software packages are Windows applications, and the most popular data acquisition
programming languages are now Visual Basic, Visual C/C++,
C++Builder, and Delphi. Many test and measurement hardware
manufacturers provide Windows-compatible ActiveX controls,
Common Object Model (COM) objects, and other programming tools
to facilitate program development using these languages. New proprietary packages have also been introduced, such as LabVIEW, Agilent
(formerly HP) VEE, and TestPoint. While these programs are similar
in scope to DOS-based test and measurement packages, they are
Windows applications.
Despite the new data acquisition software options available, the
important question of whether to use a programming language or
select a proprietary package to develop test software remains open. In
some cases, a program development environment may not be needed
at all for relatively simple applications. The start-up software supplied
SOFTWARE OVERVIEW
with some data acquisition boards and instruments may be sufficient

for simple test needs.
Open architecture programming languages have become simpler
to use, while offering much greater versatility and power. Visual Basic
and Visual C/C++ have emerged as the leaders for development of test
and measurement programs. ActiveX technology simplifies programming without compromising performance; it also protects the investment in software development by making applications more portable.
Conversely, a proprietary development package may be suitable for
less complex applications where theres only an occasional need for
program development.
Assuming that the application is to be developed in a Windows
environment and that it requires more than simple data collection, an
examination of the following criteria may be helpful in choosing an
appropriate software approach:
Learning curve
Text-based vs. graphical syntax
Programming efficiency
Windows messaging and event management
Debugging
Application deployment
3.2.2.1 Learning Curve
Data acquisition software has benefited from the evolution of Windows
as a simplified, graphical environment. As a result, the user interfaces
of programming languages have become more visual in nature.
Further, instrumentation products frequently use ActiveX and COM
technology, which can simplify hardware control and other programming tasks. Therefore, language-based programming now approximates the look and feel of proprietary software packages more closely,
and the total amount of effort needed to learn programming and develop programs has been reduced. Language-based programs also allow
programs to be optimized and tailored more closely to applications
than might be possible with proprietary packages. User-developed
software components often can be recycled for new programming
tasks, reducing development time.
3.2.2.2 Text-Based vs. Graphical Syntax
The interface a software package presents to the user can be text-based
or graphical in nature. DOS and DOS-based programming languages
such as interpreter BASIC are classic examples of text-based software.
Windows and Windows-based programming languages provide varying degrees of graphically based software in which objects can be
dragged and dropped or processes can be connected by drawing sim-
SECTION 3
TEXT
GRAPHICAL
C=A+B
D=E+F
Figure 3-1.
A
B
E
F
Text-based programming vs. graphic-based programming.
ple lines (wires) from one icon to another. Although Visual Basic and
Visual C/C++ are graphics-intensive, they are still considered textbased languages because the code behind the Graphical User Interface
(GUI) is written in text format.
The issue of choosing between text-based or graphics-based software revolves around which method is the most intuitive and easy to
learn. The answer is mostly a matter of personal preference. For example, text-based programming has a top-to-bottom organization that is
similar to the English language. On the other hand, graphical syntax
can be considered more intuitive and easy to remember because it is
based on pictures rather than alphanumeric characters. Further, the
wires and icons approach is less susceptible to typing errors and other
syntactical errors that can occur in text-based languages.
The top-to-bottom organization of text-based programming is
invalid in a graphical language, so there must be another method to
establish the order in which instructions are to be implemented. In the
text-based example in Figure 3-1, C=A+B executes before D=E+F.
The graphical method requires an additional wire connecting both +
icons to accomplish the same thing. In LabVIEW, this wire is referred to
as artificial data dependency and its only purpose is to establish the
execution order.
Program documentation is another issue to consider when choosing between text- and graphic-based programming. Documentation is
more or less automatic in text-based programming because the program can be printed with non-executing comments to explain lines of
code. In the graphical method, wires and diagrams may be difficult to
consolidate on a printed document. This complicates documentation,
which may become quite tedious in a long and complicated program.
3.2.2.3 Programming Efficiency
Program execution speed and response can be important, even critical,
to the success of a test and measurement application. Typically, a proprietary programming package adds extra software to a basic data
acquisition engine to provide simplicity and automation for the user.
Every current proprietary development package is itself a Windows
application, typically written in a C/C++ environment. This architec-
SOFTWARE OVERVIEW
ture places additional software layers between the application software

and the hardware. The extra overhead that results can affect the ability
of the program to perform satisfactorily in critical applications. This
typically shows up as less-than-optimum execution latency, multithreading, and control over thread priority. Although there are
Windows tools that can help optimize a program, they may be hidden
by additional layers of abstraction in proprietary programming
methods.
Another point to consider is that proprietary programs are developed for a broad range of users, so a broad range of functions must be
built into the package. At the same time, code must be limited where
possible to keep the programming job manageable. As a result, some
data acquisition tools or Windows services might be excluded from the
software or the full range of functions available in a given hardware
product might not be available. That means the software may have significant limitations or may fall short in the future. In contrast, a customized test application developed in an open environment can be
limited to only the data acquisition tools and Windows services
required for the job, then expanded as needs change.
3.2.2.4 Windows Messaging and Event Management
Messages and events are the processes by which Windows manages its
multitasking system and shares keyboard, mouse, and other resources.
This is done by distributing information to applications, application
instances, and processes within an application. Typically, proprietary
packages dont make efficient use of Windows events and messages. By
contrast, events are the fundamental element of the Visual Basic programming environment.
For example, consider a data acquisition application running in
the background. With event-driven processing, the application can use
the CPU for additional tasks, such as database or network access rather
than constantly status-polling to determine if data are ready for collection. When data are available, a device driver notifies the application by
sending the proper event or message through Windows.
This tight coordination between data availability and CPU execution means that programs written for event-driven processing are
robust and portable across computers of vastly different CPU speeds. If
an application was developed with Visual C/C++ and ran successfully
on a 100MHz Intel Pentium processor, theres a high probability it will
run on a 433MHz Celeron processor. This may not be the case for an
application developed with a proprietary package, especially if it
employs simple, time-sensitive loops or other polling approaches.
SECTION 3
3.2.2.5 Debugging
Efficient debugging is critical to the success of a software development
project and limited debugging tools may increase development time
and stall progress. The component-based architecture of todays open
programming systems simplifies software development and debugging. Extensive and powerful debugging tools are available with Visual
C/C++ and Visual Basic.
3.2.2.6 Application Deployment
Application programs are sometimes created by software developers,
then distributed to end users. Deployment can be problematic for
applications developed with proprietary packages because the user
does not have the development package. The difficulty lies in the runtime program that is part of the proprietary layer. Deployment is much
less of a problem with COM-based applications because the compilers
are smaller and more efficient, and the runtime library is usually part
of Windows.
3.3
Software Development under Windows

To create a successful system, the interaction between the hardware
and software must be well understood, and application programs must
be developed for optimum results with them. This task is especially
challenging in the present environment due to rapid changes in PC
hardware, operating systems (OSs), communication buses, and software development tools.
Preventing problems requires application development planning
based on a knowledge of software structure. These principles will apply
to PC-controlled systems using both plug-in boards and stand-alone
instruments with GPIB (General Purpose Interface Bus, also known as
IEEE-488), USB (Universal Serial Bus), and other digital communication buses.
3.3.1
Software Structure Overview

Typical automated test application software can be represented as a
series of layers (Figure 3-2). In the Windows operating environment,
the test application communicates directly with the Application
Programming Interface (API) and device drivers to control the test and
measurement hardware devices.
3.3.2
Device Drivers
A device driver is the software interface between a computers operating system (such as Windows) and any data acquisition plug-in board,
GPIB instrument, or communication interface attached to the computer. The driver is the abstraction layer that shields the program
developer from the complexities of a hardware device. All the com-
SOFTWARE OVERVIEW
Test
Application
Developed in VB
API
Application Level
Driver
Package
Device Driver
Test & Measurement

Device
Buses: IEEE-488, RS-232, PCI,
ISA, PXI, USB, IEEE-1394, etc.
Figure 3-2.
Kernel Level
Device
Under Test
API software structure
mands to and from the hardware are ultimately executed by device

drivers. Therefore, they are key elements in both the development and
implementation of test application software.
Lets examine the role of a driver in more detail. Much like a printer or mouse device driver, a data acquisition driver is essential to the
hardwares operation. It permits communication with the data acquisition device. The PC subsystems that each hardware device needs to
use, such as RAM and I/O addresses, are protected by the operating
system (OS). A user cannot communicate directly with these subsystems, but requires a device driver that acts as mediator between the
test application and the OS kernel. (The kernel is a set of OS programs
that implement primary system functions).
This layered architecture provides robustness, portability, maintainability, flexibility, and other features and requirements that computer users demand. Imagine if the OS didnt prevent direct access to
its sensitive areas, and device users had to write their own protocols to
gain access to hardware functions. Such an OS would be intolerable.
Oddly, the pre-Windows DOS operating system behaved in much this
way. Most programs written for DOS included mouse, video, printer or
other specialized drivers designed to work only with that software.
3.3.2.1 Driver Development
A Windows device driver is written in Microsoft Visual C/C++ and is a
special extension of the core OS. A device driver (typically a .SYS file) is
executable code that loads in the PCs RAM. Device drivers receive
SECTION 3
more attention, flexibility, and full unprotected access to the system

memory and hardware. As such, device drivers must follow very specific rules and guidelines for interfacing with the rest of the OS. The
Windows Driver Model (WDM) is the latest specification by Microsoft
for coding common binary device drivers that are compatible with
Windows 98 and Windows 2000.
The multitasking, multithreading, and multiprocessing of todays
PCs further elevate the role of device drivers. A device driver should, at
a minimum, handle the most likely multitasking scenarios that could
affect operation of its hardware device. Since a typical PC has many drivers running (such as video, mouse, printer, keyboard, networking, and
disk controllers), each driver should ideally be designed to coexist with
all other drivers, to be robust (i.e., crash-resistant), and to consume a
minimum of memory, processing time, or other system resources.
Developing a robust device driver is challenging because of the
sensitive interrelationship between it and the OS, as well as the virtually unlimited variety of multitasking conditions where it might be used.
Developing a good device driver requires intimate knowledge of both
the hardware device and the OS software. Many Windows crashes are
due to unpredictable conflicts and malfunctioning device drivers.
3.3.2.2 The Evolution from DOS to Windows
The replacement of DOS by the Windows operating system, plus
todays increasingly complex test systems, have placed greater
demands on software and its development. Under DOS, programming
a device was relatively easy. A given program would usually have full
control of the computer and only had to do one task at a time. Direct
access to the I/O address space was available with relatively simple
commands in C language (inp and outp) or BASIC (peek and poke).
With the multitasking paradigm of Windows, Microsoft had to
establish a new set of rules for managing communication with peripheral devices. A Windows driver is no longer just a single .EXE file that is
loaded in memory or copied to a special directory.
Rather, it is a package of many files, such as .INF, .SYS, .VXD, .DLL,
.EXE, and .LIB files, for example. These require a specific installation
process, which might include modification of the Windows registry,
implementation of Plug-and-Play, power management setup, and
resource management, to name only a few. This process isnt just a
matter of copying files to a hard drive; it is an algorithm that handles
the different versions of Windows appropriately, including Windows 95
(versions a, b, c), Windows 98 (first and second edition), Windows Me,
Windows NT (SP1 to SP6), and Windows 2000. When Windows is
upgraded from one version to another, it can easily affect driver installations. Its expected that device driver developers will follow suit and
upgrade installations as necessary.
SOFTWARE OVERVIEW
A test and measurement application may simultaneously acquire

data from multiple devices, control a process, graph data, save it in a
database, and make it available on the Internet. The more demanding
the test requirements are, the greater the complexities of PC-based test
systems will be. This in turn imposes more conditions on the drivers
and other system software. In most cases, it will not be necessary for a
programmer to develop the necessary drivers because they will be
available from hardware and software suppliers, and can be interfaced
by the application program.
3.3.3
The Application Programming Interface

When creating an application program, a developer sees and interacts
with the Application Programming Interface. The API greatly simplifies
common tasks the developer performs with the hardware and makes
communications between the device driver and the test application
transparent. The API acts as liaison between the device driver and the
test application, so end users rarely have to communicate directly with
a device. Therefore, its important for a driver developer to know how
users will program or configure the hardware device before building
the driver foundation under the API.
The API typically is part of the driver package. It should be the first
piece designed when developing a device driver. The ActiveX control
interface is an example of an API, which plays a significant role in simplifying development of test programs for data acquisition and other
applications. Other types of APIs include COM objects, DLLs, libraries,
and support modules for Visual Basic, Visual C/C++, and other programming languages.
3.3.3.1 The Role of an API

From its definition, it follows that the API can be a major factor in
obtaining results from a test application. For example, streaming digital data to disk at 100kHz requires not only high-speed data acquisition
hardware, but also proper API tools to take advantage of this speed by
transferring data efficiently. Similarly, logging temperature data accurately requires not only good measurement techniques and coldjunction sensors, but also a cold-junction compensation algorithm in
the API. Ideally, the software automatically supports hardware features.
However, both developers and end-users need to pay close attention to
the features offered by the API and its possible limitations.
3.3.3.2 Hardware Independence
Hardware independence is an important feature of the ActiveX API
mentioned previously. This factor has the potential to increase the
longevity of test application software. Given fast-changing PC technology and the frequent obsolescence of data communication hardware,
its also a critical cost issue. For example, consider how quickly the
SECTION 3
transition from RS-232 to USB and FireWire is occurring. The degree of

portability of an API helps determine whether the software will have to
be rewritten with every hardware upgrade.
3.3.3.3 ActiveX Controls
Although a device driver must change with a significant hardware
change, the API exposed to the user doesnt have to change, especially
if it is ActiveX-based.
ActiveX is a set of rules governing how different applications or
software components should interact and share information in the
Windows environment. Developers have been shifting rapidly to
component-based software architectures, so ActiveX controls are among
the most important features in an API.
An ActiveX control (formerly known as an OLE or OCX control) is a
user interface element that takes advantage of the standards for information exchange and functional modularity among Windows-based
applications. ActiveX controls are based on Component Object Model
(COM) technology, a software architecture that allows building applications from software components.
COM technology provides many benefits, including easier integration, scalability, and reusability, as well as language independence,
cross-platform compatibility, and context-sensitive help.
3.3.3.4 The Benefit of ActiveX
A major benefit of ActiveX is its use of a single, simple object to replace
many lines of code for common functions. This lets programmers create reusable software components that can be interchanged without
the need to rewrite entire applications. Interchangeability reduces
development cost and extends the life of an application.
For example, an application originally written for a 16-channel
100kHz analog I/O board could be used with a 64-channel 300kHz analog I/O board simply by replacing an ActiveX control that describes the
characteristics and functionality of the data acquisition device.
The following snippet of code uses three different ActiveX controls.
ActiveX1 acquires the data from the measurement device, ActiveX2
performs the analysis, and ActiveX3 graphs the data.
Data() = ActiveX1.Acquire_Data( Channels() )
Analysis() = ActiveX2.Spectrum( Data() )
ActiveX3.Chart( Data(), Analysis() )
This code shows the simplicity of programming with ActiveX controls. In this component-based open architecture model, each of the
ActiveX controls is an independent entity, which could be supplied by
more than one vendor.
SOFTWARE OVERVIEW
Test and measurement manufacturers and third parties have

developed thousands of ActiveX controls, which strongly support the
concept and benefit of open architecture software systems. These
properties of ActiveX and its underlying model protect the users software investment by making applications more portable and simplifying hardware upgrades. For instance, DriverLINX is a standard data
acquisition software driver for Keithleys data acquisition hardware. Its
design is based on the common object model; therefore, it is truly hardware independent across all Keithleys boards.
While its natural for a given test and measurement vendor to
maintain the same driver interface and syntax across all its hardware
and languages, software reusability is particularly challenging when
using multiple vendors.
3.4
FIFO and Buffer Overrun Issues

A FIFO (first-in, first-out) is a temporary memory storage block found
on almost every data acquisition device. A FIFO memory operates on
the First In, First Out principle, and can serve as the on-board memory where data are stored before being retrieved by the device driver
(Figure 3-3). When a sample is read out of a FIFO, its space can be used
by incoming data. To make data-streaming more efficient, a FIFO typically asserts an interrupt when it is half-full, signaling the device driver
that its time to retrieve data. At that point, the driver launches an
Interrupt Service Routine (ISR) to read the data and wait for the next
interrupt.
FIFOs were designed to compensate for software and OS latency in
order to help prevent loss of data. For example, when a program
streams data to a hard drive, the FIFO helps keep sequential data contiguous while the application writes it to disk. However, during FIFO
use, there is a race going on between the data intake and data output
read by the device driver. Ideally, in order to avoid a buffer overflow
condition, the Data-out speed should be faster than the Data-in speed.
If the FIFO fills up faster than the driver can retrieve its output, a buffer
FIFO
Data In
Half Full Interrupt
Reading 3
Reading 2
Reading 1
Figure 3-3.
CPU
Data Out
FIFO operation
SECTION 3
overrun (or overflow) error occurs. Incoming data has no place to be

stored, so it is lost.
3.4.1
The Role of DMA

Direct Memory Access (DMA) is a means of performing I/O without
involving the computers CPU in the transaction. Bypassing the CPU
permits direct data transfer from the data acquisition device to memory. The advantage this offers is greater throughput.
During a DMA transfer, the CPU can be used for other tasks without affecting the data transfer speed. Typically, a complete set of data is
acquired by initiating multiple DMA transactions of 64 kilobytes. At the
end of each DMA transaction, an ISR is generated and is handled by the
device driver. Multiple DMA transactions can also be chained, making
the process even more efficient by reducing the number of ISRs. There
are two styles of DMA design. Bus Master DMA is used with the PCI
bus, while System DMA is used with the ISA bus.
With Bus Master DMA, all the control circuitry necessary to initiate
a DMA transfer exists on the PCI board. With System DMA, the support
circuitry is on the motherboard, shared among all ISA peripherals. Bus
Master DMA is much more efficient, and the PCI bus operates at a
much higher data transfer rate than the older ISA bus.
In either case, DMA requires critical coordination between hardware and software. The hardware must have the proper circuitry, and
the software must implement the proper procedures for successful
DMA operation.
3.4.2
Polled vs. Event-Driven Control

In a polled system, the computer checks many devices to see if theyre
ready to send or receive data. In the context of a data acquisition system, this typically involves reading or writing a single value from or to
a data I/O channel. In a Windows-based PC, the time between polled
readings is scheduled by Windows, so its nondeterministic. In other
words, the time at which Windows will initiate an operation cannot be
known precisely. Its operation can depend on any number of system
factors, such as computer speed, operating systems, programming languages, and code optimization.
By way of comparison, a deterministic data acquisition system is
one that collects data having discrete values, and direct cause-andeffect relationships with the physical phenomena they represent. This
implies a strong time relationship between the data and the underlying
phenomena.
Typically, data acquisition running in the background, using DMAand ISR-driven operations, is under the control of the data acquisition
hardware. The data acquisition hardware notifies the users application
when the task is complete. The time between background readings is
SOFTWARE OVERVIEW
clocked by the data acquisition hardware and is independent of

Windows timing.
During a background data transfer, the data acquisition hardware
interrupts the CPU. The driver handles these interrupts by transferring
data into the application memory space. When the requested background operation is finished, the driver posts a Windows message or
event to the application, which responds to the event and manipulates
the data. Events and messaging are the processes Windows uses to distribute information to applications and processes within an application, thereby managing its multitasking system. Event-driven programs result in a more deterministic system.
An application that uses polling is simple to program. Polling can
be appropriate with slower, non-time-sensitive operations, such as
programming discrete steps in a power supplys output or reading precise voltages from a nanovoltmeter.
Trying to squeeze high speeds out of a polled application will probably lead to less than acceptable results. In applications that use high
speed data acquisition boards for faster sampling rates, event-driven
programming is advised. Event-driven programming schemes are less
dependent on OS timing, reducing latency problems.
3.4.3
Tight Control
A programmer can take advantage of Windows events and messaging
by using Visual C/C++ and Visual Basic programming tools to create a
deterministic application that runs fast and provides tight control.
Rather than constantly status polling to determine if data are ready for
collection, such programs can use the CPU for additional tasks such as
database or network access until interrupted by the data acquisition
hardware.
This tight coordination between data availability and CPU execution also makes event-driven applications more robust and portable
across computers of vastly different CPU speeds.
The hardware deployed for real-time applications in Windows is
important. Event-driven programming may be necessary, but is not
necessarily sufficient for some real-time applications. Contrary to what
the words might indicate, real time tends to be a relative term, the
meaning of which varies with the application and user requirements.
In some applications, users consider a response time of 100 milliseconds to be a real-time response. In a different setting, a one microsecond response might be needed. The Windows operating system,
whether its Windows 95, 98, Me, NT, or 2000, isnt well suited for fast,
real-time applications.
Real-time applications tend to work better with dedicated operating systems running on embedded processors. This implies instrument
SECTION 3
or data acquisition hardware equipped with a digital signal processor

(DSP) or microprocessor. The DSP or on-board processor performs
many of the functions that would otherwise be performed by the computers CPU.
The dedicated DSP or CPU approach eliminates problems associated with Windows running tasks in the background that arent under a
users control. These background tasks use up CPU clocks and interrupt requests, so the test application must share these system
resources. Some applications will actually perform better when programmed and run on a DOS system, particularly if the hardware and
software were designed around that OS in the first place.
The design of every operating system (Windows, DOS, Linux, etc.)
attempts to balance several conflicting demands, but tends to be optimized for certain types of tasks. A test system developer must weigh
these tradeoffs. The ultimate selection will invariably involve compromises between flexibility, functionality, ease of use, robustness, and
reliability.
3.4.4
Managing Speed and Accuracy Tradeoffs

Throughput is an issue in most production test applications, so a
developer must look for programming techniques that speed up processing while maintaining accuracy. In particular, the programmer
should avoid an error that can occur when using a PC that is much
faster than the instruments attached to it. In this situation, the user
may gain a false sense that the system is faster than it actually is, which
can lead to inaccurate test results. This problem arises because the
application software outpaces the hardware. Sometimes, the hardware
driver package will take care of these timing issues; in other cases, the
user must make adjustments in the application program as outlined in
the following example.
In a typical production calibration system, the application program commands the calibrator to apply a certain voltage stimulus to
the device under test (DUT) before measuring its response. Both the
calibrator and DUT need a certain amount of time, often seconds, to
settle to the required state (stimulus or response) before the program
commands a measurement. If that amount of time isnt accounted for
in the program, the calibration process will be inaccurate.
Typically, programmers hard-code wait times, so this isnt an issue.
However, in production tests where throughput is critical, theres a
downside: the wait time may be longer than is actually needed.
Rather than hard-coding a specific wait time, the programmer can
create an algorithm that uses feedback from a calibrator (an instrument-ready bit) to acknowledge that the stimulus voltage level has
been reached. This means the program can proceed to the DUT
response measurement. The measuring instrument can provide a sim-
SOFTWARE OVERVIEW
ilar signal after it has acquired the data following a delay, based on its
own internal program. This kind of programming approach results in
higher production rates.
SECTION 3
SECTION 4
Basic Component
Theory
4.1
Introduction
This section of the handbook provides an overview of the types of electronic components used in data acquisition. Its intended to help users
understand and work with components used in the construction of test
setups, to give users insight into data acquisition hardware so they can
understand analog and digital measurements concepts, and to help
users optimize their test procedures.
Test and measurement systems are developed and operated in a
variety of ways, ranging from a single individual to a team of system
integrators, programmers, system maintenance staff, and operators.
These people often have widely varying degrees of technical expertise
and specialization. This section assumes that the reader is familiar with
the concepts of voltage, current, and resistance, and the mathematical
relationship between them.
4.2
Passive Components
Ideally, a sensor or signal source can be connected directly to an instrument or data acquisition board input without the need for specialized
signal conditioning. In practice, it is often necessary to add passive
components (resistors, capacitors, inductors, or diodes) to a circuit to
remove noise, alter signal levels, or achieve some other goal.
4.2.1
Resistors
The function of a resistor is to impede the flow of electric current.
Resistance is measured in ohms (). The most important characteristics
of resistors are resistance, power handling ability, and construction.
These factors are interrelated and most often mutually exclusive. For
instance, it may be difficult to find a resistor that combines very high
resistance and very high power handling ability; such a device might be
prohibitively large or expensive. Some types of resistors are more sensitive than others to signal frequency or ambient temperature.
Resistors are manufactured with several basic techniques. The
resistive element can be molded from a mixture of carbon and clay or
deposited as a resistive film on an insulating substrate. Alternately, a
resistive wire can be wound on a suitable form. Varying the thickness,
length, or composition of the resistive element allows the manufacturer to control resistance and other properties. The complete package
can be made in a variety of forms and sizes with different styles of wire
leads or no leads at all in the case of surface-mount devices.
The general differences in resistors manufactured with these techniques are summarized in Table 4-1. Note that the information is typical
and may vary considerably from one manufacturer to another. Also note
that some characteristics may be less relevant for certain applications.
SECTION 4
Table 4-1.
Resistor characteristics film vs. wirewound

Composition or
Film Resistors
Wirewound
Resistors
Value
Approx. 0.5 to
several million ohms.
Approx. 0.05 to
several thousand ohms.
Accuracy
0.1% to 5%. Technology

supports laser trimming
and other techniques to
achieve high accuracy.
0.1% to 5%.
Power Handling
Up to 5W.
1W to 100W.
Frequency Effects
Performance generally
Wound construction can
independent of frequency result in inductive
unless film has been
effects for AC signals.
deposited on the substrate
as a spiral.
Relative Physical
Size
Smallest. Includes chip

resistors used in
miniaturized products.
Characteristic
Medium to large.
4.2.1.1 Resistor Applications

Resistors are generally used to limit current, reduce voltages (as when
multiple resistors are configured as voltage dividers), or act as shunts
for current measurement. Example circuits are shown in Figures 4-1
through 4-3, along with pertinent formulas. These applications are typical of data acquisition.
Example: Current Limiting
The example circuit (Figure 4-1) demonstrates the use of a resistor to
limit current through a device that would otherwise be destroyed by
excess current. The light emitting diode (LED) is shown powered from
a 5V battery, but power could also be supplied by a source of voltage in
a data acquisition system. A typical LED will turn on at about 0.7VDC;
higher voltages will result in progressively greater current flow and
intensity. If not limited, current can exceed the safe operating limit of
the diode.
R
Current = 25mA
+
5V
Figure 4-1.
0.7V
Current limiting
BASIC COMPONENT THEORY
In this example, the voltage drop across the LED is 0.7V, leaving
4.3VDC (5.0 minus 0.7) dropped across resistor R. The desired current
is 25mA, so the value of the resistor can be calculated using Ohms Law:
R =V / I
R = 4.3V / 0.025 A = 172.
A 172 resistor may be difficult to find, and the value is not critical.
A slightly higher standard value of 180 will do nicely. To be safe, its
important to check the power to be dissipated by the resistor:
P =V I
P = 4.3V 0.025 A = 0.1075 watts.
Resistors are available in a variety of power ratings, including 0.125
and 0.25 watt types. Either would work in this application. The 0.25W
type would provide some margin of safety.
Example: Voltage Division
Figure 4-2 shows resistors used to scale down a voltage to a value compatible with existing measurement hardware. In this case, a signal with
a maximum amplitude of 36V must be measured with an A/D board
with an input limit of 10V.
For this example, assume that the signal has a low source impedance, which is represented by the resistor symbol in the Signal Source
box. Source impedance is an important concept that appears often in
test and measurement applications. For a complete discussion, see
Keithleys reference handbook, Low Level Measurements.
Briefly, any voltage source can be thought to include a resistance
through which the generated current must flow (Figure 4-2). This resistance, which is called the source resistance or source impedance, is
expressed in ohms. At a given voltage, current compliance varies
+36VDC
Max.
R1
Out HI
Signal
Source
R2
Out LO
Figure 4-2.
Voltage divider
SECTION 4
inversely with source resistance. When a voltmeter is connected to a

voltage source, a small current flows from the source into the meter to
facilitate the measurement. Ideally, any voltage measurement instrument will minimize this current to avoid loading down the source. Most
data acquisition A/D inputs offer an input resistance on the order of
several megohms to hundreds of megohms, which is sufficient for voltage measurements from ordinary sources and transducers, but may be
inadequate for sensitive measurements of high impedance signals.
In the case of the example circuit, assume a source impedance of
less than 100 for the signal source and an input resistance for the data
acquisition board of 100M. This means that a total resistance of a few
hundred kilo-ohms for the voltage divider R1 + R2 should give good
results. Therefore, lets assume a value of 100kW for R2. This could as
easily be 10k or 500k. In a real-world application, the goal of selecting a value would be to avoid loading down the source while providing
a signal that the A/D input will not load.
The voltage ratio that needs to be produced by the divider is 10V/
36V or 0.2778. The formula for calculating the resistors in the dividers is:
Voltage Ratio = (R2) / (R1 + R2)
0.2778 = 100,000 / (R1 + 100,000)
0.2778 (R1) = 72,220
R1 = 259,971 = 259.971k
Obviously, 259.971k is not a standard resistor value. A slightly
higher value 270k resistor would work well and result in slightly less
than 10V out when the sensor produces 36V. A full scale output of less
than 10V is preferable to avoid applying a voltage to the A/D input that
is outside its range.
The actual output when R1 = 270k and R2 = 100k would be:
Output = 36V (100,000)/ (100,000+270,000)
Output = 36V (0.2703)
Output = 9.73V
The actual voltage ratio is 0.2703, which can be used to scale the
reading back properly to 36V. For example, a reading of 6.8V would correspond to 6.8/0.2703 or 25.157V.
Example: Current Measurement
A third example of using resistors in data acquisition is as a sensing
resistor to facilitate current measurements. Few data acquisition
boards can measure current directly. However, it is a relatively simple
matter to send the current through a resistor, then measure the voltage
drop across the resistor.
020mA
Out HI
R
Current
Source
Figure 4-3.
Out LO
Current shunt
In Figure 4-3, the transducers output is a current ranging from 0 to

20mA full scale. The transducer will act to adjust its output voltage to
force a specific current through the circuit, regardless of cable length
and the value of the current sensing resistor (R).
As with the previous voltage divider example, the circuit should be
set up so that the maximum voltage developed across the resistor is
compatible with the A/D input range. A resistor of 499 will cause a
voltage drop of 499 0.020 or 9.98V. This voltage is ideal for a 16-bit A/D
set for an input range of 0 to 10VDC. A 249 resistor will produce a
drop of 249 0.020 or 4.98V, which is suitable for a 05VDC A/D input.
Note that current-to-voltage conversions such as this are only valid
if the current source can generate the voltage needed to force the
desired current through the circuit. For example, if the dropping resistor (R) plus other resistances in the circuit total 600, and the current
source cannot reach a voltage greater than 10V, the source will not be
able to force 20mA through the circuit (20mA 600 = 12V). A lower
resistance would have to be chosen for R.
4.2.2
Capacitors
A capacitor stores electrical energy in the form of an electrostatic field.
The general mechanical principle of the capacitor is two conductive
surfaces (plates) separated by a dielectric (typically an insulator such
as Teflon, polystyrene, or Mylar polyester film). Capacitance is directly proportional to the total shared surface area of the plates, and
inversely proportional to the distance between them.
Capacitors are manufactured using a variety of construction techniques and materials. This can cause capacitors of identical capacitance value to differ in other electrical properties, making one type or
another more suitable for certain applications. Some of these characteristics are listed in Table 4-2.
SECTION 4
Table 4-2.
Capacitor characteristics
Characteristic
Variation
Polarity
Some types of capacitors, notably electrolytics and

tantalum capacitors, are polar. One terminal is marked
+. Polarity must be observed for proper operation and
to avoid possible damage.
Working Voltage
Working voltage should at least equal the highest

voltage occurring in a circuit. However, it is of no benefit
to use a capacitor with a working voltage much higher
than that which will be applied in the application.
Frequency
The construction of some capacitors results in their

conductors exhibiting inductive effects that can become
significant at higher frequencies. Such capacitors are less
suitable for bypassing and noise filtering.
Leakage
An ideal capacitor with perfectly insulating dielectric

would retain a charge indefinitely, and also not dissipate
any power as heat energy. In practice, a charged
capacitor will gradually lose its charge, although it
may take quite some time for this to occur. Leakage is
important in sample-and-hold circuits where a capacitor
may have to retain charge for a long time.
ESR
Equivalent Series Resistance A real-world capacitor can

be thought of as a pure capacitor in series with a
resistance that results from leads, materials, and other
physical attributes of the capacitor. This resistance is
called the ESR.
The engineering unit used to define capacitance is the farad (F). A

farad is an impractically large capacitance for typical electronic circuits; therefore, it is more common to find capacitors with values of
microfarads (F), nanofarads (nF), and picofarads (pF)a range of
106F to 1012F.
Working voltage is a second parameter associated with capacitors.
A capacitor is designed to provide the rated capacitance at a specific
voltage. Exceeding this voltage can destroy the capacitor.
A third characteristic of capacitors is Equivalent Series Resistance
(ESR). A perfect capacitor would exhibit only capacitive properties, but
a practical capacitor behaves more like a pure capacitance in series
with a pure resistance, and in some cases, a pure inductance (Figure
4-4). Together, the resistance and inductance can be responsible for
energy being lost in the capacitor or they can interfere with the capacitors ability to store or empty a charge. For some frequencies or applications, these factors can overwhelm the capacitive properties.
When a capacitor is connected to a DC source, a momentary current surge occurs as positive and negative charges build up on the
plates and stabilize. Otherwise, the capacitor does not conduct DC
C
Equivalent
Capacitor
Figure 4-4.
Equivalent capacitor
DC Current
+
5V
Figure 4-5.
Capacitor connected to a source of DC
AC Current
AC
Source
Figure 4-6.
Capacitor connected to a source of AC, showing current flow

during positive and negative AC cycles
(Figure 4-5). When connected to AC, charges move in and out of the
capacitor as the polarity of the voltage alternates (Figure 4-6). The
apparent result is that the capacitor blocks DC, but conducts AC.
Note that a perfect capacitor presents a resistance to the flow of AC
that decreases with frequency. This resistance is called capacitive reactance or XC and can be calculated as:
SECTION 4
XC = 1 / (2 F C)
where: F = the frequency
C = the capacitance in farads
As an example, the reactance of a 0.047F capacitor at 10kHz
would be calculated as:
XC = 1 / (2 F C)
XC = 1 / (2 3.14159 10 103 0.047 106)
XC = 1 / 2.953 103
XC = 338.63
Qualitatively, the equation reveals that at DC, capacitive reactance
(XC)is infinite, while at very high frequencies, XC approaches zero. This
property makes capacitors useful for selectively passing high frequencies while blocking DC; a common application is bypassing an A/D
input in order to filter noise from a signal. Capacitors can also be used
in rudimentary anti-aliasing filters, although the roll-off of a capacitor
is only 6dB per octave.
Another capacitor application that may be useful in data acquisition is based on the fact that a resistor in series with a capacitor produces a circuit that charges at a fixed time constant. The circuit in
Figure 4-7 exhibits an RC Time Constant, which is the time needed
for the voltage across the capacitor to rise to 63% of the applied voltage.
The RC time constant for the circuit in Figure 4-7 can be calculated as:
T = RC
T = 470000 10 106
T = 4700000 106
T = 4.7 seconds
After five RC time constants (5RC) have elapsed, the voltage on the
capacitor will have risen to more than 99% of the power supply voltage.
Conversely, a fully charged capacitor will take 1RC to discharge to 37%
4.7s
9.4s
100%
80%
63%
R = 470k
5V
0
1
2
RC Time Constants
Figure 4-7.
C = 10F
RC time constant
of its initial voltage, and 5RC to discharge to less than 1% of its fully
charged voltage.
The characteristics of capacitors make them useful in a number of
areas in test and measurement. Their ability to pass higher frequency
signals permits capacitors to be used as simple high frequency noise
filters, particularly where the signal of interest is a DC voltage.
Similarly, a knowledge of time constants and capacitor settling behavior is important in designing test and measurement systems where circuit capacitance is relatively high and signals change rapidly.
4.2.3
Inductors
The inductor is the third passive component. The principle behind the
inductor is that if a magnetic field moves past a conductor, a current
will be induced in the conductor. Conversely, when an electric current
flows through a conductor, a magnetic field is generated around the
conductor. If the conductor (wire) is wound in the form of a coil, both
effects become more pronounced and form the basis of the electromagnets used in motors, generators, transformers, and related devices.
However, even a single loop can increase the inductive effect.
Understanding this principle can be of value in minimizing some types
of noise pickup in data acquisition setups.
A closer analysis of inductors shows that current flow sets up a
magnetic field in the coil, which causes the inductor to generate a back
voltage (or back EMF) equal to the forward voltage. The faster the current changes, the greater the back EMF will be. An important aspect of
back EMF is that when the source current is instantaneously disconnected from an inductor, the resulting magnetic collapse can generate
a substantial voltage spike, which induces noise in surrounding circuits. This is why a diode is usually connected across a relay coil; the
diode suppresses the back EMF that results when the coil is deenergized.
The unit of inductance is the henry (H), with typical inductor values in circuits on the order of millihenries or microhenries.
Like capacitors, inductors are constructed using a variety of materials and techniques that affect inductance. An ideal inductor would be
a lossless coil exhibiting only inductive effects. However, a real-world
inductor behaves more as a pure inductance in series with a resistance
(Figure 4-8). This resistance is a result of the conductive properties of
the wire winding and can vary with gauge, length, and composition of
the wire. This resistance is important because it is a potential source of
energy loss in the inductor.
The core of an inductor is also extremely important in establishing
inductive properties. Inductor cores are fashioned from a variety of
metallic materials in different shapes and dimensions, but torroids and
cylindrical bars are the most common. Common core materials
SECTION 4
L
Equivalent
Inductor
R
Figure 4-8.
Equivalent Inductor
include iron and ferrite. However, the core can be omitted, resulting in
an air core inductor. Small-value inductors can also be etched flat on
printed circuit substrates.
The resistance of an inductor is frequency dependent. However,
inductive reactance mirrors capacitive reactance in that it is 0 at DC,
then increases with frequency. Inductive reactance (XL) can be calculated as:
XL = 2 F L
where: F = the frequency
L = the inductance in henries
As an example, the reactance of a 0.5 millihenry inductor at 3kHz
would be calculated as:
XL = 2 F C
XL = 2 3.14159 3 103 0.5 103
XL = 9.42
Their coiled construction means that inductors tend to be more
expensive than resistors and capacitors and are available in fewer values. Therefore, circuits are usually designed to achieve the desired performance using resistors, capacitors, and other components where
possible.
4.3
Op Amp Theory
An active electronic device used extensively in solid-state electronic
circuits is the operational amplifier or op amp. Op amps get their
name from their original application, in which they were used to perform mathematical operations in analog computers. Today, op amps
are used as general analog building blocks in a wide variety of circuits.
Unlike digital electronics, which have two valid output states (e.g.,
high or low, on or off), op amps are linear circuits where the output
represents a mathematical function applied to the input signal. The

most common application of op amps is simple, linear amplification,
where the op amps output voltage is a multiple of the input voltage.
However, integrators, differentiators, logarithmic amplifiers, and other
functions can be constructed using op amps. Many uses exist for op
amp circuits in data acquisition; it is often faster and less expensive to
build these circuits rather than buy them.
The goals of this section are to summarize op amp theory and to
highlight information that relates directly to these data acquisition
applications. Many texts have been written on the topic of op amps
and semiconductor manufacturers offer a variety of data books and
application guides. These resources can supply detailed specifications
and design information for a wide range of circuits.
4.3.1
Types
Op amps are transistor circuits fabricated using a variety of semiconductor processes, including bipolar, JFET, CMOS, and mixed processes.
Some operating parameters associated with op amps relate directly to
the fabrication process, such as input impedance, power consumption,
noise, drive capability, and bandwidth. These factors need to be considered when selecting components to ensure the desired result. For
example, it is sometimes necessary to work with sensors with a low
output level or high output impedance. In both cases, an op amp can
be used to build a simple buffer amplifier to condition the signal. This
application would best be served with an op amp with a very high
input impedance, so one with FET inputs would be a good choice.
4.3.2
Power Supply
The power requirement for most op amps, especially older components, is a symmetrical, positive and negative supply in the range of 3
to 30VDC. Power supplies in AC-powered equipment containing op
amps typically provide 12 to 18VDC for op amp circuits. More recent
trends in op amp design have been toward low power, applicationspecific devices that operate at lower voltages and currents, frequently
5VDC or less.
As a class of devices, op amps require a relatively low operating current. However, current requirements vary sufficiently to make some
families suitable for battery-powered circuits, while others are best
used in AC-powered equipment. Again, consult specific op amp manufacturers data sheets when contemplating an op amp project. Low
power consumption is usually accomplished at the expense of other
performance criteria.
4.3.3
Input and Output Impedance

The input impedance of an ideal op amp is infinite, which would result
in the op amps inputs (V+ and V) absorbing zero current. In reality,
SECTION 4
(+) Supply
V+
+
Gain
(A)
VOUT
V
() Supply
Figure 4-9. Op amp
while the input impedance of an op amp is not infinite, it is still very

high. This characteristic makes op amp circuits useful for buffer and
amplifier applications involving signals with a high source impedance.
Actual input impedance depends on both the op amps characteristics
and the components used to construct the circuit. The highest input
impedance is provided by op amps fabricated with MOSFET input
transistors.
4.3.4
Gain
The schematic symbol and simplified model of an op amp are shown
in Figure 4-9. An op amp is a differential amplifier and will amplify
the difference between the voltages applied to the inverting (V) and a
non-inverting (V+) inputs. The equation describing an op amps function is:
VOUT = ((V+) (V)) A
where V+ and V are input voltages, VOUT is the output voltage, and A is
the gain.
Note that the maximum output voltage from an op amp circuit
depends on the supply voltage. A calculated output is valid only if the
power supply voltages exceed the voltage by one to two volts. This is
why most data acquisition boards can tolerate input voltages of only
up to approximately 10VDC. The boards are powered through an
expansion slot from the computers 12VDC supply.
4.3.5
Feedback
Op amps possess extremely high open loop gain (up to 100,000 or
more), which is usually reduced with other components to a level suitable to the application. These components comprise the negative
feedback loop of an op amp circuit, and are illustrated by R1 and R2 in
Figures 4-10 and 4-11. Negative feedback is the mechanism by which
op amp closed loop gain can be adjusted and a circuit made stable.
R2
R1
Gain
+
Figure 4-10. Inverting amplifier
+
Gain
VIN
R1
VOUT
R2
Figure 4-11. Non-inverting amplifier
A central operating principle of negative feedback is that an op

amp will attempt to keep the voltage equal at both its inputs. This principle applies to inverting as well as non-inverting circuit configurations. In its simplest form, negative feedback uses a resistive divider
between the output and inverting input of the op amp to send a portion of the output voltage back to the input. The op amp adjusts its output voltage as needed to maintain V equal to V+. However, diodes and
capacitors are frequently used in the feedback loop to configure integrators, differentiators, logarithmic amplifiers, and other types of nonlinear circuits, or to adjust the response of linear circuits.
4.3.6
Inverting vs. Non-Inverting Operation

An op amp can be configured to operate as an inverting or non-inverting voltage amplifier. In the inverting mode (Figure 4-10), a signal is
SECTION 4
connected to the inverting () input, while the non-inverting input (+)

is held at ground potential or some other reference voltage. The feedback loop acts to maintain both inputs at the same voltage (0V or a
virtual ground in Figure 4-10). The input impedance of the inverting
amplifier therefore equals R1, and circuits gain (A) can be calculated as
R2/R1. The output voltage VOUT for any given input VIN equals VIN
R2/R1. The phase of the input signal will be shifted 180 at the output.
In non-inverting mode (Figure 4-11), no phase shifting occurs.
Gain (A) can be calculated as 1+R1/R2, and the output voltage VOUT calculated as VIN (1 + R1/R2). In this circuit, the input impedance at the
(+) input equals that of the op amp, although a resistor or other component may be connected between the (+) input and other circuit
points in some circuits, affecting input impedance. The circuit in
Figure 4-11 can be further simplified by removing R2 and making R1
equal to 0. In this case, the circuit becomes a high input impedance
voltage follower where the output voltage equals the input voltage (a
gain of 1).
4.3.7
Normal Mode and Common Mode Voltages

Two terms that appear frequently in discussions of analog inputs are
normal mode voltage and common mode voltage. An understanding of
these terms is important in selecting and making the best possible use
of data acquisition equipment, particularly in minimizing the effects
of noise.
One definition of normal mode voltage (VNM in Figure 4-12) is an
error voltage that appears across the inputs of an amplifier, thereby
adding to the input. Normal mode specifications are usually given at
frequencies or frequency ranges where sources of noise are most common, such as 50Hz or 60Hz. The measure of an amplifiers ability to
reject such noise while passing DC and low frequency signals is Normal
Mode Rejection Ratio (NMRR), which is expressed by the equation:
NMRR (in dB) = 20 log [peak NM noise/peak measurement deviation]
Qualitatively, the ideal amplifier would be completely unaffected
by normal mode noise, making the peak measurement deviation 0, and
the resulting ratio infinitely large. A more typical value for NMRR is
80dB. Each 20dB of NMRR reduces normal mode voltage by a factor of
10. Thus, an 80dB ratio will reduce a normal mode voltage by a factor
of 10,000.
A practical example of normal mode voltage is noise pickup resulting from insufficient shielding, improper cable routing, etc. The effects
of normal mode voltage are sometimes easy to discern, and can be
eliminated through filtering, averaging, or post-processing of the signal. However, it is usually best to treat the problem through careful
hardware layout. In particular, if a signal has components of interest in
the same frequency range as the noise, its critical to eliminate noise at
VNM
V+
+
Gain
V
VOUT
VCM
Figure 4-12. Normal Mode vs. Common Mode Voltages
its source, because it cannot be filtered out of the final data without
affecting the data as well.
In contrast to normal mode voltage, common mode voltage (VCM in
Figure 4-12) appears between an amplifiers inputs and ground. Both
inputs see the same common mode voltage in addition to the voltage
differences attributable to the signal. As mentioned previously, an op
amp amplifies the difference in voltage appearing at its (+) and ()
inputs, and will naturally reject a signal appearing at both inputs. The
degree of this rejection is called Common Mode Rejection Ratio
(CMRR). A typical value for CMRR is 120dB, meaning that a voltage
appearing on both sides of a differential input will be reduced by a factor of one million.
In data acquisition, the effects of common mode voltage are sometimes noted as noisy measurements or inexplicable measurement
errors. One situation where common mode voltage can present a problem occurs when individual circuit common points in a test setup are
tied to ground at different locations. If each ground point is not a true,
low impedance path to ground, a voltage gradient can exist across the
grounds and current will flow (see Figure 4-13). This phenomenon is
known as a ground loop. A ground loop may be produced when each
end of the shield of a long cable is connected to separate chassis
grounds. The voltage above true ground at one end of the shield can be
higher than the other, floating one ground above the other, which
results in current flow through the ground system. As a result, the
acquired data will have an offset because it was not referenced to
ground. The solution is to ground the shield at only one end.
A second example of a problem related to common mode voltage
is where a sensor is allowed to float, without a bias current return to
ground. For instance, if a thermocouple is connected to an A/D input,
the circuit may stop working after a few minutes. What has happened
SECTION 4
Signal
HI
LO
GND
Ground Loop
Current Flow
Voltage
Gradient
Figure 4-13. Ground loop
Thermocouple
V+
+
Gain
V (A)
VOUT
Figure 4-14. Thermocouple A/D input with ground return resistor (R)
is that the inputs gradually charge from the applied signal, increasing
the common mode voltage on the inputs until they are too close to the
power supply rails for the op amp to function. The solution is to install
a resistor (R) between input high or input low and ground (Figure
4-14). A typical value for this resistor is one mega-ohm or more. This
situation highlights the fact that the possibility of common mode voltage must be considered when a signal is connected to a data acquisition A/D input, and the total voltage at an input kept within 10V
relative to power supply ground.
4.3.8
Single-Ended vs. Differential, Bipolar vs. Unipolar

The terms single-ended and differential are usually used to describe an
amplifier (or op amp) input. A single-ended signal is one referenced to
0V. The signal can be carried by a pair of conductors, but one of the
conductors is, by definition, tied to ground (0V). A differential signal
requires a minimum of two conductors for transmission, but neither is
Unipolar
Bipolar
+10 V
12-Bit
4096 Counts
2.44 mV/Bit
+10 V
4.88 mV/Bit
0V
0V
10 V
Resolution
10 V
Common Input Ranges
8-Bits = 256 Divisions

12-Bits = 4096 Divisions
16-Bits = 64 K Divisions
10 V
5
2.5
1.25
10 V
1
.1
.02
Figure 4-15. Bipolar and unipolar input ranges
necessarily tied to ground. In the case of a differential input, the measurement instrument responds to the difference in voltage between
signal high and signal low. The differential method is often used to
carry low-level signals, so the signal-carrying cable can actually consist
of two signal leads plus a ground shield.
The terms bipolar and unipolar are normally used to describe
whether a signal remains positive with respect to ground or whether it
can also assume a negative level. These terms can apply to inputs as
well as outputs. A bipolar signal can be positive or negative with
respect to ground, while a unipolar signal generally goes no lower than
0V. The main benefit of using a unipolar range for digitized analog
input or output is to enhance resolution. The available number of A/D
or D/A bits remains the same, but is divided between only half the total
voltage range of a bipolar configuration. Therefore, each bit provides
twice the resolution. Figure 4-15 shows the resolution on a 12-bit A/D
in unipolar and bipolar configurations.
SECTION 4
Channel N HI
Channel N LO
Signal
Source
RS
Rb
Rb
Data Acquisition
Board
LL GND
where RS > 100

Rb = 2000 RS
Channel N HI
Channel N LO
Signal
Source
RS
Rb
Data Acquisition
Board
LL GND
where RS < 100

Rb = 2000 RS
Figure 4-16. Differential input configurations
4.3.9
Single-Ended vs. Differential Inputs for Signal Measurement

A differential input will usually offer better noise immunity than a
single-ended input. This is especially important in systems that take
data from a number of different sensors or are located a long distance
from the sensors. Figure 4-16 shows two connection schemes for
wiring a signal source to a channel of a data acquisition board configured for differential input mode. The two circuits of the diagram
require the addition of resistors to provide a bias current return. The
value of the bias return resistors (Rb) can be determined from the value
of the source resistance (Rs) by using the following relationships:
When Rs is greater than 100, use the connections in the upper
circuit. The resistance of each of the two bias return resistors
must equal 2000 Rs.
When Rs is less than 100, use the connections in the lower circuit. The resistance of the bias return resistor must be greater
than 2000 Rs.
4.4
Filters
Some data acquisition applications can benefit from the use of filters to
condition a signal. There are many types of filter designs, which can
broadly be divided into passive and active types. Passive filters are relatively simple circuits constructed with resistors, capacitors, and
Low-Pass
High-Pass
Band-Pass
Band-Reject
Figure 4-17. Passive L-C filter designs
inductors, but are generally less effective than active types. Active filters can provide more aggressive filtering action than passive filters.
However, active filters are more complex in design, and also require a
source of operating power. Bessel, Butterworth, Chebyshev, and Cauer
(elliptical) filters are common active filter designs that use a combination of op amps and passive components.
The response curve or cut-off of a filter describes how well the filter
blocks or passes a band of frequencies. Response is usually expressed
as decibels (dB) of amplitude change per frequency span. Typically, the
frequency span is specified per octave (amplitude change between fre-
SECTION 4
quency f and 2 f ) or per decade (amplitude change between frequency f and 10 f ).

Passive filters are useful for a maximum of about 18dB per octave;
beyond this point, a passive filter can affect the signal of interest
adversely. Passive filters always impose some insertion loss, which
becomes more severe as elements are added to tighten the response
curve. Figure 4-17 illustrates basic circuits for 12dB per octave highpass, low-pass, band-pass, and band-reject (or notch) filters.
Active filter designs are suitable for response curves up to 80dB per
octave. However, active filters can exhibit ringing, phase delay, ripple,
distortion, loss, or less-than-perfect cut-off characteristics, depending
on design and sharpness of response. Therefore, its important to
choose the type and design parameters for active filters carefully to
optimize performance and avoid problems. The design of active filters
is most easily accomplished with software programs, which are available on the Internet or from commercial sources. Active filters can also
be purchased to pass or block specific frequencies.
4.5
Digital I/O
Many data acquisition processes involve digital signals that are either
on or off, high or low, etc. This is in contrast to analog signals where the
signal voltage can range anywhere between an upper and lower voltage
limit. One gauge of the quality of analog I/O is bits of resolution, with
12- and 16-bit A/D and D/A being common in data acquisition.
Conversely, a digital signal is a 1-bit phenomenon; the signal is ultimately represented as a single 1 or a 0. While this sounds simple in
concept, there are many factors that can complicate digital measurement and control.
Digital signals are usually generated or read by digital gates, which
can have a single output and single or multiple inputs. Normally, an
integrated circuit package will contain two to six gates, depending on
the number of pins the package can support. There are a variety of
Boolean logic functions for gates with multiple inputsAND, NAND,
OR, and NOR being the most common (see Figure 4-18). For more
information on this topic, consult a text such as the TTL Cookbook or a
manufacturers data book.
Note in Figure 4-18 that the various types of digital gates can be
subdivided into non-inverting and inverting types. A non-inverting
gate produces a logic 1 output when the logical input function of the
gate is satisfied. For example, an AND gate requires all inputs to be at
logic 1 for an output of logic 1. The inverted form, the NAND gate, produces a logic 0 output when all inputs are at logic 1. Also note that versions of the inverting and non-inverting gates exist where there is only
one input. The inverting version is simply called an inverter, while the
non-inverting version is called a buffer.
AND GATE
NAND GATE
OR GATE
All Inputs = 1
For Output = 1
All Inputs = 1
For Output = 0
Any Input = 1 Any Input = 1

For Output = 1 For Output = 0
Any Input = 0
For Output = 0
Any Input = 0
For Output = 1
All Inputs = 0 All Inputs = 0

For Output = 0 For Output = 1
A
V
AND
B
V
NAND
C
V
NOR GATE
OR
D
V
NOR
Figure 4-18. Digital gates
The remainder of this discussion on gates concentrates on how to

control individual gate inputs and use outputs.
4.5.1
Digital Logic Types and Logic Levels

The type of semiconductor process used to fabricate a digital gate
determines the operating parameters for the device. The most important parameters include:
Voltage corresponding to logic 1 and logic 0.
Compatibility with other families of analog or digital circuitry.
SECTION 4
Voltage and power needed to operate the device.

Speed capability (generally important only when dealing with
rapidly changing signals).
The usual method for activating a digital input is to connect the
input to ground or 0V, which causes current to flow from the input to
ground. This process is commonly referred to as sinking current, and
can be accomplished using a mechanical switch, a transistor wired as a
switch, the output of another digital gate, or a sensor that includes one
of these devices (Figure 4-18 shows mechanical switches). Note that a
typical digital output can usually sink enough current to control up to
10 or more inputs of the same logic family. The term for this capability
is fan-out. Exceeding the recommended fan-out can result in unreliable operation of a digital circuit. Similarly, mixing different logic families can result in improper operation.
4.5.2
TTL Logic
One of the first and most common semiconductor processes used to
fabricate digital gates is Transistor-Transistor Logic (TTL). TTL gates
are constructed with bipolar transistors, which provide relatively high
current source and sink capability and high speed, but also consume
more operating power than some newer types of devices. The standard
power supply voltage for TTL logic is +5VDC.
The voltages and current levels corresponding to logic 1 and logic 0
for TTL are:
Logic 0 = 0.00.8 volts. For conventional TTL, control signal
must be capable of sinking at least 1.6mA from a digital input.
Newer implementations of TTL-type logic have substantially
reduced this current requirement to as little as 10A.
Logic 1 = 2.05.0 volts. The actual output voltage of TTL
devices is usually between 3.5V and 5V. Typical output source
current is 0.4mA (400A). New types of TTL logic may have lower
source current capability.
For TTL, note that the input range is neither symmetrical nor continuous. There is a gap in the values between 0.8 and 2.0 volts in which
the signal is ambiguous, defined neither as a 1 nor a 0. In some situations, such as when trying to read an analog or slow-moving signal,
this can cause problems. However, several options are available for
dealing with ambiguous or slow-moving digital input signals. These are
discussed elsewhere in this book.
Because of its design, a TTL output is effective at sinking current to
ground to create a logic 0. However, sufficient variation exists in the
design of TTL output stages that a logic 1 output voltage must be
defined more broadly. Some TTL outputs swing to 3.5V or higher and
Table 4-3.
Typical TTL input and output specifications
Input High Voltage:
2.0V minimum, 5.0V maximum
Input Low Voltage:
Input High Current:
0.02mA
Input Low Current:
0.4mA
Output High Voltage:
2.7V minimum
Output Low Voltage:
0.5V maximum
Output High Current:
0.4mA
Output Low Current:
8.0mA
can source a small amount of current, while others (open collector

types) float when set to logic 1, and have no real drive capability.
Although TTL logic was an early development in digital integrated
circuitry, it remains the de facto standard for digital I/O for several reasons. TTL can easily drive high load currents and is much less susceptible to damage from static electricity. Devices implemented with other
types of circuitry often use TTLs logic levels for backward compatibility with TTL. Further, TTL has spun off related device families where
various operating parameters such as power consumption and speed
are optimized for certain types of tasks. These optimizations permit
engineers to design in terms of TTL conventions. Low Power Shottky
TTL (LS-TTL) is one such family, which is often used in manufacturing
digital I/O boards. As the name implies, LS-TTL requires less power to
operate than standard TTL.
4.5.3
CMOS Logic
CMOS (Complementary Metal Oxide Semiconductor) technology was
developed after TTL as a low power alternative to existing bipolar transistor technology. CMOS technology is used in linear as well as digital
devices, and offers such advantages as operation over a wide range of
supply voltages and extremely low current consumption. Therefore,
many products that use CMOS devices can be operated from batteries
or from much smaller power supplies than those required for TTL.
The disadvantages of CMOS logic relative to digital I/O applications include sensitivity to static damage and incompatibility with TTL
signal levels. The threshold voltage that differentiates logic 1 from logic
0 with a CMOS input is roughly half the power supply voltage. The transition from log 1 to logic 0 does not contain any appreciable dead zone,
as does TTL. The current sink capability of a CMOS gate is also relatively low. However, buffer devices are available that can drive TTL
from CMOS.
SECTION 4
SECTION 5
Basic Analog
and Digital I/O
5.1
A/D Conversion
Voltage measurement during data acquisition relies on a process
known as analog-to-digital conversion (often abbreviated as A/D or
A-to-D). An analog input board contains an A/D converter and support
circuitry (Figure 5-1), which conditions and digitizes the incoming
voltage. The following list summarizes the individual circuit stages and
operation of a typical complete A/D circuit. Specialized analog input
boards may depart from this description, with multiple A/D converters, large FIFO buffers, circular buffers, triggering, or other features.
Signal conditioning (optional)
- Sensor excitation
- Filtering
- Input protection
Multiplexer (selects a channel on multi-input A/D boards)
Programmable instrumentation amplifier (applies gain)
A/D converter (digitizes the signal)
FIFO buffer (temporarily stores measurement data)
Control circuitry (retrieves data from FIFO buffer)
Input
Protection
Multiplexer
Instrumentation
Amplifier
.
.
.
A/D
Conversion
A
Gain
Select
Data
FIFO
Channel
Select
Bus
Data
CTRL
Control
Figure 5-1.
5.1.1
Typical A/D converter and associated circuitry
A/D Resolution and Speed

Three of the most important specifications involved in choosing an
analog input board are A/D converter resolution, accuracy, and speed.
These specifications and other A/D characteristics are interrelated,
because higher performance in one area may come at the expense of
performance in other areas. For example, high speed and high resolution are usually mutually exclusive to some degree, and achieving both
SECTION 5
in a single product usually results in an elevated price. However, the

premium for higher performance is not as high as it once was.
5.1.2
Resolution
The function of an A/D converter is to generate a series of unique digital output states corresponding to a specific range of analog input
voltages. The ideal A/D converter would accept an infinite range of
input voltages and digitize the range into an infinite number of output
states. This is, of course, technically impossible. Fortunately, the factors
that limit real-world A/D resolution are easy to identify and understand.
As a general rule, the input voltage range of any A/D input is limited by the voltage used to power the circuit. A plug-in A/D board for a
computer receives operating power (a nominal 12VDC) from the
computer expansion slot in which it resides. The analog input board
requires 12V of headroom, limiting the actual input range to about
10VDC. This is, in fact, a very common input voltage limit for analog
input boards. Gain stages in front of the A/D converter can increase
sensitivity and reduce the permissible input to 5VDC, 2.5VDC,
1.25VDC, or some other fraction of 10VDC, but the maximum signal
voltage farther into the A/D circuitry will not exceed 10V. Stand-alone
instruments can be powered by internal supplies that arent affected by
this limitation, so they may offer a broader input dynamic range.
Standard resolutions of plug-in A/D boards are 8, 10, 12, and 16
bits, while stand-alone instruments can offer 1824 bits of resolution or
more. This represents the number of output bits the A/D converter has
available to digitize an analog input voltage. The voltage resolution per
bit (VRES) can be calculated as:
VINPUT
VRES = _________
2(No. of bits)
For example, an 8-bit converter can output 0000 0000 to 1111 1111
(binary). This corresponds to 28 or 256 discrete steps. An input range of
10V divided by 256 steps equals 20V/256 or 78.125mV per step. A 16bit converter would have a resolution of 20V/216 or 1.22mV per step for
the same 10V input.
Most A/D boards can also be configured for unipolar operation,
where the input range extends from 0 to +10VDC, 0 to +5VDC, etc. The
method for calculating the resolution of bipolar measurements is also
applicable to unipolar ones. Table 5-1 compares resolution and other
A/D characteristics as a function of A/D converter bits.
This table highlights several facts concerning A/D converters. First,
an A/D can not read a voltage equal to its maximum range (i.e., 10V on
a 10V range), even in unipolar mode. In the case of each converter
range shown in Table 5-1, the maximum readable voltage is one bit less
BASIC ANALOG AND DIGITAL I/O
Table 5-1.
Measurement resolution and maximum ranges for different A/D

resolutions
Converter Bits (n)
Output States (2n)
10
12
16
256
1024
4096
65,536
Resolution, 010V input
39.06 mV
9.765 mV
2.441 mV
152.59 V
19.53 mV
4.883 mV
1.221 mV
76.29 V
78.12 mV
19.53 mV
4.883 mV
305.2 mV
39.06 mV
9.765 mV
2.441 mV
152.59 V
Resolution, 2.5V input
19.53 mV
4.883 mV
1.221 mV
76.29 V
Resolution, 1.25V input
9.76 mV
2.441 mV 610.35 V
38.15 V
Max. input, 010V (res 2n1)
9.960
9.990
V 9.9976
V 9.99985
than the range. Second, an averaged reading at zero volts will equal
zero only for a measurement made in bipolar mode; in unipolar
modes, there are no negative readings to bring the average to zero.
Typical reading jitter with a digital input is at least one A/D count, placing the average for a series of readings made in unipolar range somewhere between zero and a few A/D counts. These observations apply
more to plug-in A/D data acquisition boards than to instruments.
Third, and perhaps less obvious, is that A/D offset errors can
swamp gain errors, especially for lower resolution boards. For the sake
of comparison, consider the maximum reading on a 010V scale for 8and 16-bit A/D converters. For the 8-bit converter, one bit of uncertainty represents 39mV, which calculates as 100 (0.039V/9.96V), or
0.39%. At 16 bits, the error is 152V, which corresponds to 0.0015%. In
comparison, gain errors of 0.01% to 0.05% are common for analog
input boards. This observation can apply to stand-alone instruments
as well.
Total error figures for a specific measurement situation depend on
the input voltage as well as the equipment and the environment. The
key to specifying the correct resolution for an application is to match
the resolution of the board to the resolution required of the measurement. In general, higher resolution A/D converters are more expensive
than their lower resolution counterparts. Buying capability somewhat
beyond current needs can provide insurance against obsolescence. On
the other hand, purchasing more capability than will ever be required
has no benefit. For example, if an application uses a pressure sensor
accurate to 1% (1 part in 100), a 16-bit A/D board will likely add needless expense to the system. Also note that, even if the application justifies the cost, a very high resolution A/D can still be a waste of money if
the less significant A/D bits are swamped by noise in the measurement
environment or from the sensor. Signal averaging can help the situa-
SECTION 5
tion in this case. Furthermore, higher resolution A/D converters are

usually slower than lower resolution versions.
5.1.3
Input Accuracy
Input accuracy is related to, but not equal to, input resolution. The
accuracy of a data acquisition board depends on its whole analog front
end, including the input multiplexer, the programmable gain amplifier, and the A/D converter.
Accuracy can be specified as absolute accuracy or relative accuracy. Absolute accuracy at a given A/D output code is the difference
between the actual and the theoretical voltage required to produce that
code. Relative accuracy is the deviation from the theoretical value after
the full-scale range has been calibrated.
Input accuracy can be specified in a number of ways. Three of the
more common specification methods, along with the formulas
required to convert the specifications into voltage accuracies, are
shown in Table 5-2. All calculations assume a 12-bit A/D converter and
a 10V full-scale input.
Table 5-2.
Comparison of analog accuracy specification formats
SPEC: 0.024% of reading 1 bit (Gain + Offset method)

Measurement Accuracy = 10V
0.024 ____
1
______
+
= 4.8mV
100
212
SPEC: 2 bits (Total Bits of Deviation)

10V
Measurement Accuracy = 2 ____ = 4.8mV
212
SPEC: 0.048% of FSR (Percentage of Full Scale Range)
0.048
Measurement Accuracy = 10V ______ = 4.8mV
100
Note that in the examples, the A/D accuracies are specified differently, but the results are identical.
5.1.4
Maximum A/D Speed

The maximum speed of an A/D converter relates to the number of digitizations that can be performed in a unit of time. Typically, the maximum sample rate is specified in samples/second (S/s), kilosamples/
second (kS/s) or megasamples/second (MS/s), not in hertz (Hz).
Most multi-channel A/D products contain a single A/D converter
and input multiplexer (Figure 5-1). The multiplexer acts as a switch
that allows each of the input channels to be sampled independently.
The maximum sample rate per channel is no faster than the maximum
sample rate of the A/D converter divided by the number of channels
sampled. For example, an 8-channel analog input board might be

capable of 100,000 samples per second. A single channel could be sampled at the full 100,000 S/s, two channels at 50,000 S/s each, and so on.
In many cases, the maximum sample rate is specified with all channels set to the same gain. If different gains are set for channels in a scan
list, the overall sampling speed can suffer unless there is an onboard
gain or gain/channel queue. A gain queue permits specifying a different gain for each channel. As the channel multiplexer is incremented,
the associated channel gain is set. However, high gain settings may
impose long settling times and slower A/D conversions, even where a
gain queue is used.
5.1.5
A/D Techniques
The mainstream A/D conversion methods for data acquisition and
measurement instruments include successive approximation, integrating, and flash converters. Each conversion method offers a different
combination of performance and price that makes it suitable for a specific set of data acquisition applications. There are also several other
varieties of A/D conversion, which may be used in more specialized
data acquisition applications.
Normally, instrument and board manufacturers select an A/D
technology appropriate for the primary goals of a product. The three
most important characteristics of A/D product design are speed, resolution, and cost. Frequently, the user has to accept compromises in one
area to obtain essential performance in another area. It is important to
be aware of the limitations associated with different conversion methods. Table 5-3 briefly describes some of the main characteristics and
tradeoffs of mainstream A/D converter technologies.
Table 5-3.
Converter
Type
A/D converter types popular for data acquisition

Maximum
Speed
Typical
Resolution
Noise
Immunity
Relative
Cost
Successive
Medium
Approximation (10kHz to 1MHz)
616 bits
Little
Low
Integrating
Slow
(10Hz to 30Hz)
1224 bits
Very Good
Low
Flash
Very Fast
(1MHz to 500MHz)
48 bits
None
High
Sigma-Delta
Slow to Medium
(Up to 1MHz
or higher)
16 bits
or more
High
Low
SECTION 5
5.1.5.1 Successive Approximation A/D

Most general-purpose data acquisition boards use successive approximation converters. Successive approximation converters offer an optimal compromise between high speed, high resolution, and cost.
The operating principle of the successive approximation A/D converter is that the unknown voltage and the output of a digital-to-analog
(D/A) converter are both fed to a comparator. The output of the D/A is
adjusted until the inputs to the comparator are equal and the comparator balances. The binary output code of the D/A then represents
the voltage of the unknown signal. The A/D simply tracks the input, so
any noise on the signal of interest will also appear in the digital output.
5.1.5.2 Integrating A/D
The general operating principle of the integrating A/D converter is
based on the charging and discharging of a capacitor by an unknown
signal and a reference voltage. The capacitor is charged first by the
unknown signal for a set time interval. Next, the capacitor is discharged
back to zero at a fixed rate and the time needed to discharge the capacitor is measured. This time is a measure of the integrated input voltage
and can be used to deduce the unknown voltage.
A benefit of the integrating A/D conversion process is that it can
average out noise over time, which results in good noise rejection. The
integration time can be selected to match the frequency of a known
noise source, which allows the integrating A/D to be particularly effective against certain types of noise. Commonly, integration times that
are multiples of 50Hz and 60Hz are used (sometimes referred to as line
cycle integration) to minimize the effects of AC noise on the
measurement.
Integrating A/D converters are more accurate and linear than successive approximation converters, so they are a good choice for lowlevel measurements. While integrating A/D converters are often used
in DMMs and other stand-alone instruments, they are not as common
in data acquisition boards.
5.1.5.3 Flash Conversion
Very high-speed analog input boards use flash converters to achieve
acquisition rates of megasamples per second or even gigasamples per
second. The heart of the flash converter is an array of voltage comparators that simultaneously sample the unknown voltage in parallel.
Collectively, the comparator outputs represent a progressive series of
bits, which are then encoded into a standard binary output.
The chief drawback of flash A/D converters is that the number of
comparators increases exponentially with output resolution. For
example, an 8-bit flash A/D requires 28 1 or 255 comparators, while a
16-bit flash A/D requires 216 1 or 16,383 comparators. Therefore, the
cost of a flash converter escalates quickly for higher resolutions. For

this reason, very high speed analog input boards most often provide
low to medium A/D resolution.
Like the successive approximation converter, the flash A/D simply
tracks the input, so any signal noise will also appear in the digital output. Further, because flash converters are intended for high speed signal capture, it can be difficult to filter high frequency noise from a
signal without affecting the signal itself.
5.1.5.4 Sigma-Delta Conversion
The Sigma-Delta (also called Delta-Sigma or one-bit) conversion
method is based on theoretical technology that has existed for many
years. It is only with the realization of high speed digital and analog circuitry that the hardware necessary to implement sigma-delta converters has been feasible.
Sigma-delta conversion uses an oversampling modulator (a voltage-to-frequency converter) and a digital filter to digitize an analog
voltage. The modulator loop oversamples and processes the analog
input at a rate much higher than the bandwidth of interest. The modulators output provides information to the filter one bit at a time, at a
very high rate, and in a format that the digital filter can process to
extract higher resolution (such as 16 bits) at a lower rate.
With sigma-delta conversion, there is a tradeoff between speed and
resolution. The hardware has to operate at a much higher (oversampled) rate than the signal bandwidth, which places greater
demands on the digital circuitry. Thus, sigma-delta converters typically are used for high resolution, relatively low frequency applications. In
return, this technique provides many advantages, such as good temperature stability, low cost, highly linear operation, and minimal
requirements for post-A/D anti-aliasing filters.
5.1.6
Aliasing and Anti-Aliasing Filters

An alias is a false signal component that appears in sampled data
acquired at too low a sampling rate. Aliasing errors occur when a signal
contains frequency components that are faster than one half the sampling frequency (the Nyquist rate). For example, if acquiring data from
one channel at 100kS/s, the signal cannot contain any frequency components greater than 50kHz, or false, low frequency artifacts (aliasing)
will appear when the data is used to reconstruct the waveform. If eight
channels are sampled by the same A/D converter, the maximum sample rate for each channel is 12.5kS/s and the aliasing errors will be
caused by any frequency component greater than 6.25kHz.
Aliasing errors are hard to detect and almost impossible to remove
using software. The solution to aliasing is to use a high enough sampling rate or, if this is not possible, to use an anti-aliasing filter in front
SECTION 5
Cutoff Frequency
Attenuation Slope
Stop-Band
Amplitude
Pass Band
Stop-Band
Ripple
Frequency
Figure 5-2.
Response of an anti-aliasing filter
of the analog-to-digital converter to eliminate the high frequency components before they get into the data acquisition system. Among the
more common types of anti-aliasing filters are Butterworth, Bessel,
and Cauer, each of which has specific filter characteristics. Figure 5-2
shows a typical anti-aliasing response curve. The pass band indicates
the frequencies that pass through the filter unchanged. The stop band
includes the frequencies that are attenuated by the filter. The filter type
determines the slope of the attenuation curve and the amount of ripple in the stop band. Cauer filters have the sharpest cutoff; however,
their transient response is not as good as that of the others. The Bessel
filter has the slowest cutoff of the three types and also has the best transient response.
5.2
D/A Conversion
Digital-to-Analog (D/A or D-to-A) conversion is the process used to
generate an analog output voltage, usually in response to digital data
supplied by a computer or other control circuitry. To some degree, D/A
conversion is the inverse of A/D.
A common D/A converter design couples a reference voltage with
a resistor network and control logic to generate specific voltages in
response to binary input data. As with A/D converters, multiplexers
can be included in D/A designs to provide multiple output channels
using one converter.
The output characteristics of D/A converters are also similar to

those of A/Ds. Twelve- and 16-bit resolutions are common, as are typical full-scale ranges based on 010V, 10V, 05V, 5V, 02.5V, 2.5V, etc.
One important characteristic of D/A converters is that they normally offer only a few milliamps of drive current. Therefore, applications that require higher currents need to incorporate an external
circuit or programmable power supply to boost available current.
5.2.1
Four-Wire Remote Sensing

The ability to perform four-wire measurement techniques in conjunction with D/A conversion is a concept borrowed directly from benchtop instrumentation and is a relatively rare capability with PC-based
analog output boards. Keithleys Series KPCI-3130 analog output
boards provide this capability.
This technique uses a pair of sense leads that extend from the
device under test, back to a high input impedance voltmeter stage on
the D/A board. Theoretically, the current flowing in the sensing loop
will be nil and will be unaffected by any voltage drop associated with
the leads connecting the analog output to the DUT. The sensed voltage
can be used to adjust the boards analog output level quickly and transparently, until precisely the desired voltage is present at the DUT,
regardless of cable lengths and interconnection losses.
The ability to perform four-wire remote sense operations provides
a significant advantage over standard analog output PC hardware. This
feature becomes of critical importance when generating control signals over long distances. Significant voltage drops can occur as the
result of cable resistance, device interconnections, and terminations;
therefore, the programmed output value may not be the actual voltage
delivered to the device under test (DUT) that is located at the end of
the cable.
5.3
Interfacing Digital I/O to Applications

Communication between digital circuits depends on the ability of outputs and inputs to generate and interpret valid ON and OFF states.
Several different digital logic standards exist, the most common being
Transistor-Transistor Logic (TTL) and Complementary Metal Oxide
Semiconductor (CMOS). TTL circuits appear to be more common, perhaps because they have been around longer than CMOS; their design
rules are well understood and they are relatively immune to static damage. Each logic family has different specifications for voltage and current levels that define ON and OFF, as well as any ambiguous signal
states that might exist between ON and OFF. The I/O specifications for
a typical Low-Power Shottky (LS-TTL) compatible digital I/O board are
shown in Table 5-4. LS-TTL is a more recent version of TTL logic, offering lower power consumption.
SECTION 5
Table 5-4.
Typical TTL input and output specifications
Supply Voltage:
5VDC
Input High Voltage:
Input Low Voltage:
Input High Current:
0.02mA
Input Low Current:
0.4mA
Output High Voltage:
2.7V minimum
Output Low Voltage:
0.5V maximum
Output High Current:
0.4mA
Output Low Current:
8.0mA
The specifications for commercially available digital I/O boards

and sensors normally describe I/O capabilities sufficiently to allow
matching inputs and outputs without concern for what sort of logic has
been used in either.
Most Keithley digital I/O boards are intended to read or control
TTL-level signals. Most feature protection against occasional transients, but are not intended to dissipate significant amounts of energy
where the board is connected to voltages far outside the range of 05V.
However, some digital I/O boards can read a much higher maximum
input voltage than 5V. Always check equipment specifications for compatibility with the intended signals.
5.3.1
Interfacing with Mechanical Switches

Switch contacts can be interfaced to a digital input by adding a pull-up
resistor, as shown in Figure 5-3. Some digital I/O boards and instruments provide a socket and header to facilitate user installation of pullups. The use of a pull-up resistor ensures that the digital input will
receive a reliable TTL high input level when the switch is open.
Generally, all unused digital inputs should be tied to the positive supply or ground. Unterminated digital inputs can be affected by noise
and cause unreliable operation.
+V
To digital input
Figure 5-3.
Pull-up resistor
5.3.2
Contact Debouncing
When a mechanical switch snaps shut, there is a short periodtypically
15mswhen the contact surfaces bounce against each other. Some
logic inputs, notably digital counters, read this contact bounce as a
burst of pulses or ON/OFF signals (Figure 5-4). Solutions to contact
bounce can be implemented in software or hardware, although
debouncing in software is often preferred because it is less expensive.
A common software technique involves performing multiple readings
until the signal becomes stable. This may impose a slight delay in the
control program. Hardware debouncing is relatively instantaneous and
also more suitable where software is not used to read switch contacts.
Hardware debouncing requires a flip-flop circuit for each digital input
channel to provide a clean signal. Figure 5-5 shows a common
example.
Contact Bounce
Period
1
0
Switch Activated
Figure 5-4.
Contact bounce
+V
OUT
OUT
Figure 5-5.
Hardware debounce circuit
SECTION 5
5.3.3
Dry Switching
Ordinary relay contacts designed for relatively high currents (>100mA)
have large surface areas made of arc-resistant materials. These materials can be unreliable for switching currents of a few milliamps or less,
because the limited signal energy cannot break through the film that
tends to build up on contact surfaces. Switches designed for dry circuits
have softer crosspoint contacts that will work well with small voltages
and currents. This is one case in which specifying a higher rating for the
contact may actually be less reliable.
5.3.4
Slow Moving Signals

A TTL ordinary digital input is designed to read signal levels that satisfy the specifications describing ON and OFF (Table 5-4), but can provide unpredictable operation when connected to a slowly changing
voltage. A slow moving signal might remain in the ambiguous
0.8V2.0V range long enough to cause multiple triggerings, rather than
a smooth transition from one logic level to the other (Figure 5-6).
Sensed as 1
2.0
Logic level is ambiguous
0.8
Sensed as 0
1
0
Figure 5-6.
Slow moving digital signal can result in unreliable switching
There are several ways to deal with this situation:

Correct the input signal so it changes more quickly.
If it is necessary to detect a specific voltage level, read the signal
using an analog comparator. If necessary, condition the comparators output with a digital gate.
If the exact switching level is not important, but clean switching
from slow or noisy signals is required, use a digital input with
hysteresis (a Schmidt trigger). Hysteresis establishes different
voltages for logic 1 and for logic 0, depending on whether the
signal is rising or falling. The result is a snap action when the
input detects a logic 1, because the voltage needed to trigger a
logic 0 is considerably lower.
Somewhat the same effect can be achieved in software by ignoring
state changes that occur too soon after the previous change. The effectiveness of this kind of pulse-width discrimination depends on the
characteristics of the input signal and on how costly the additional programming required will be when compared with a hardware solution.
5.3.5
Dealing with TTL-Incompatible Signal Levels

TTL switching levels have become the de facto standard for digital I/O
boards, although the upper voltage limit that can be applied to digital
inputs varies. The input circuits on some digital I/O boards are intended to be used with 05V signals, so they are protected with internal
diodes that clip signals greater than 5.5 V or less than 0.5 V. Although
these diodes provide adequate protection against occasional transients, they are not intended to dissipate significant amounts of energy.
Signals far outside TTL limits can damage the board and/or the signal
source. Therefore, its important to use proper interfacing when dealing
with non-TTL logic levels.
Signals in the range of 012VDC, such as those found in CMOS digital circuits, cannot be connected directly to TTL inputs. Possible solutions include the use of the CD4049 hex inverting buffer or CD4050
non-inverting buffer, which are designed to convert CMOS to TTL logic
levels (Figure 5-7).
5.3.6
Digital High Current or High Voltage

If an output load requires more current or a higher voltage than a digital board can provide, the drive current and/or voltage can be boosted
using the circuit shown in Figure 5-8. For drive current requirements
from 15mA to 100mA, select an NPN transistor rated for the required
supply voltage with a collector current rating no higher than 0.5A. If
higher current is needed, substitute a Darlington NPN transistor.
Solid-state relay modules allow TTL signals to control outputs up to
250VDC or 280Vrms and to read inputs up to 280VDC/VAC as a TTL sig-
SECTION 5
+5V
012V
Logic
Level
TTL
Logic Level
to Board
CD4049
(1 of 6 gates)
Figure 5-7.
CMOS-to-TTL buffer/inverter
Supply
Voltage
470
min.
NAND
Figure 5-8.
High voltage or high current drive for TTL gate
nal. These modules have two additional advantages: they are usually
mounted on external circuit boards, which locates these elevated voltages at a distance from the computer, and they provide up to 4000V of
electrical isolation. Isolation ensures that the test and measurement
equipment can handle signals referenced to different grounds safely.
Solid-state relay modules also offer an advantage in controlling
inductive AC loads such as solenoids. Unlike mechanical relays, where
contacts can be subjected to large voltage arcs when they are opened,
the solid-state relay will continue to carry current until it drops to zero,
then turn off.
5.4
Isolation
Electrical isolation refers to a condition where no electrical connection
or any common reference exists between two signals. An isolated
measurement is a measurement performed without an electrical connection between the signal source and the data acquisition input chan-
nel. Similarly, isolated control involves no electrical connection

between the control circuitry and the equipment being controlled. This
isolation includes the signal and ground connections.
Isolation is possible for both analog and digital measurements and
can be beneficial or necessary in a number of situations:
A voltage differential exists between the grounds of the data
acquisition system and external equipment. If this difference is
large enough, it could lead to measurement problems or equipment failure.
The external equipment contains internal voltages that could
damage the data acquisition system in the event of equipment
failure.
Electrical isolation specifications range up to thousands of volts.
The specified voltage gradient can exist between the isolation circuits,
with little or no leakage current between them.
5.4.1
Digital Isolation
Digital isolation is simpler and less expensive to implement than analog isolation. Figures 5-9 and 5-10 show methods suitable for isolating
digital inputs and outputs. Both use an opto-isolator that provides an
electrical barrier between the digital input or output and the external
equipment. Note that both circuits maintain separate grounds and
power supplies on each side of the isolation barrier.
In the case of the isolated digital input in Figure 5-9, the diode side
of the opto-isolator is powered and controlled by external equipment,
Equipment
Digital I/O
+5VDC
10k
External
Ground
OptoIsolator
Figure 5-9.
Isolated digital input
SECTION 5
Digital I/O
Equipment
+5VDC
Digital
Ground
External
Power
OptoIsolator
Load
External
Ground
Figure 5-10. Isolated digital output
while the transistor switch in the isolator controls a digital input. For
the isolated output shown in Figure 5-10, a digital output channel controls the opto-isolator diode, while the opto transistor switches power
to external equipment. Generally, opto-isolators are low-power
devices, but they can be connected to a relay, power transistor, or other
device capable of controlling higher currents or voltages.
These circuits can be added externally to digital I/O channels,
although some test and measurement boards and instruments offer
isolated channels. External, solid-state, or mechanical relays can also
provide digital isolation.
5.4.2
Analog Isolation
Although analog isolation can be desirable for the same reasons as digital isolation, the nature of analog signals makes isolation more complex. Typically, isolation of analog signals involves converting the signal
into a form that can be coupled from one circuit to another without
physical connection. Further, the complexity and space requirements
of analog isolation make it difficult to implement a multichannel isolated analog input board that plugs into a PC. It is also questionable
practice to plug a board that carries high voltages or other potentially
dangerous signals into a PC. Therefore, it is more common to find isolated analog I/O channels on external or industrial data acquisition
systems where adequate room is available for circuit boards.
5.5
Ground Loops
Improper grounding is a common source of problems affecting analog
measurements. A perfect circuit ground sits at 0V with reference to
earth ground and represents an impedance of 0 at any frequency. In
reality, a good DC ground requires some effort to achieve and still may
not constitute a perfect ground at all frequencies.
When a piece of equipment is plugged into a three-wire grounded
AC outlet, in theory, the chassis is at ground potential. However, there
can be a considerable electrical path and a slight but significant resistance between the chassis and true ground. No problems should result
as long as the chassis is used as a central grounding point for all other
sensors and equipment. When instruments, sensors, and other components of a data acquisition system are plugged in AC mains at different physical locations in a facility, the chassis of each can be
established at a slightly different ground potential. The resulting voltage gradient between these different grounds can cause current to flow
through cable shields or other parts of the ground system, resulting in
a ground loop (Figure 5-11).
HI
Signal
A/D In
LO
GND
Ground Loop
Current Flow
Voltage
Gradient
Figure 5-11. Ground loop
Ground loops can occur in any type of measurement setup, but are
most troublesome for analog measurements. For single-ended measurements, ground loops result in current flow through circuit common
or cable shields. In the case of differential measurements, the voltage
gradient can add to the common mode voltage. As a result, measurements can be degraded through noise or suffer from reduced signal
input range.
SECTION 5
Follow these practices to minimize the possibility of ground loops:

Where possible, use a single AC distribution strip and AC line
cords of similar length to power equipment. Make sure all line
cords are in good condition and that ground pins are present on
three-conductor AC line cords.
Use a single location as the central ground point for the entire
test setup.
When using shielded cables, connect the shield at only one end
of the cable.
Keep signal wiring and cabling as short as possible and avoid
looping excess cable.
While still a valid concern, ground loops are less critical for digital
signals because the large voltage difference between logic levels
reduces the possibility of misinterpreting a signal. Be sure, however,
that all digital grounds (except for isolated inputs) and the grounds of
any auxiliary power supplies for accessories are connected to the digital I/O instruments ground reference.
SECTION 6
Temperature
Measurement
6.1
Temperature
Temperature is one of the most frequently measured physical phenomena. Several sensor technologies are available for this purpose,
including thermocouples, resistive temperature detectors (RTDs),
thermistors, and various semiconductor devices. The measurement
range, accuracy, and ease-of-use of these sensors differ considerably,
so the best choice of a sensor depends on the application and temperature range.
6.2
Thermocouples
Thermocouples are probably the most widely used sensor for temperature measurement; they are used in labs, industry, and even consumer devices. As such, thermocouples are used in a large base of data
acquisition applications, including those employing many channels,
automated data logging, and process control, to name a few.
Thermocouple cards are also available for mainframe scanners, making it possible to add temperature monitoring to bench-type or rackbased systems. Even some low cost, hand-held DMMs have
thermocouple inputs to support temperature monitoring.
6.2.1
Features and Operating Principle of the Thermocouple

Despite their widespread use, thermocouples may be the least understood of temperature sensors. When compared to some temperature
sensors (notably, semiconductor types), thermocouples are easy to
work with and are based on a simple operating principle. However,
there are many different types of thermocouples, and special attention
to metallurgy, operating principles, limitations, and treatment of measurement data is required to achieve valid results.
6.2.1.1 Advantages and Disadvantages of Thermocouples

Thermocouples offer several advantages over other types of temperature sensors:
The basic thermocouple is relatively inexpensive, although protective sheaths, cabling, and connectors can contribute to overall expense.
Thermocouples are mechanically simple, durable, and reliable.
Properties of typical metals used in thermocouples provide predictable output voltages. This allows users to adapt thermocouples to a variety of applications, including those in reactive
or caustic environments.
The physical construction of a basic thermocouple is simple
all thats necessary is twisting together wires of the appropriate
alloys. Commercial thermocouples are assembled through welding, crimping, or soldering. All methods produce similar results.
SECTION 6
Thermocouples lend themselves to a variety of packaging techniques that can be adapted to many types of applications.
Thermocouples offer a very wide overall temperature measurement range, spanning about 100C to higher than 2500C.
The typical accuracy of thermocouples is on the order of 12C,
which is more than adequate for the accuracy requirements of
most industrial applications.
While thermocouples have relatively few disadvantages, these disadvantages affect their usage and the hardware needed to read them
significantly. Thermocouple output is on the order of microvolts per
degree, and thermocouples are sometimes located far from the data
acquisition equipment. In order to compensate for these factors, differential measurement mode, high gain, filtering, and other signal conditioning techniques are used to maximize the signal and minimize
noise. These practices result in relatively slow measurement rates for
thermocouples, typically, a maximum of only a few hundred readings
per second. Furthermore, thermocouple output is non-linear, so linearization routines must be built into the hardware and/or software
used to convert thermocouple voltages to a temperature reading. This
is more of a concern for custom-written software, because commercial
software normally provides linearization capabilities.
Finally, thermocouple measurements require the use of a reference
junction. Rather than resorting to a separate thermocouple junction
and reference temperature for each channel, A/D boards designed for
thermocouples usually incorporate an isothermal block with an
embedded reference temperature sensor. The isothermal block provides a large thermal mass, which ensures that all thermocouple terminals on the A/D board are at the same temperature. Although these
features result in higher complexity and higher hardware costs, they
simplify the use of thermocouples.
6.2.1.2 Operating Principle of the Thermocouple
In the early 1800s, Thomas Seebeck discovered that the junction
between two metals generates a voltage that is a function of temperature. A thermocouple is simply a practical application of the Seebeck
Effect. It is a temperature sensor that consists of two wires of dissimilar metals joined at one end. These metals are shown as Alloy 1 and
Alloy 2 in Figure 6-1 and form junctions J1 and J4.
Historically, temperature measurement with thermocouples relied
on a second thermocouple element to sense a known temperature as a
reference. The easiest and most precise way of producing a reference
temperature was to immerse this reference junction (J4) in an ice bath,
which gave it the name cold junction. The magnitude of the voltage
generated in this scenario now depends on the temperature difference
TEMPERATURE MEASUREMENT
Alloy 3 (Cu)
J2
J3
Alloy 2
Alloy 2
Voltage proportional
to temperature difference
Alloy 1
J1
J4
Unknown
Temperature
Reference
Junction
(Ice water, 32F)
Figure 6-1.
Thermocouple (Seebeck) principle.
between J1 and J4, and the types of metals used for Alloy 1 and Alloy 2.
The result can be expressed by the following equation:
V = (TUNKNOWN TREF)
where is the Seebeck Coefficient. Different types of thermocouples
have different coefficients, which are listed in most thermocouple references. With this configuration, it was only necessary to read the voltage, then look up the corresponding temperature in a table for the Alloy
1/Alloy 2 thermocouple type referenced to 0C.
Note that the connection of the thermocouple to the voltmeter
forms additional, potentially unwanted junctions J2 and J3. In effect,
these junctions are also thermocouples, but they are of like composition and opposing polarities. If the temperatures of J2 and J3 are equal
(a condition that can be reached relatively easily through proper hardware design), these junctions will have no effect on the measurement.
We now have a basic model that can be used to develop a more sophisticated instrumentation system for reading thermocouples.
6.2.1.3 Simplifying the Measurement System
Elimination of the ice bath and corresponding reference junction for
each thermocouple would be desirable for most applications, but particularly for industrial applications. It would not only simplify the use
SECTION 6
of thermocouples, but eliminate the need for a potentially large number of additional data acquisition input channels and sensor readings
for reference junctions.
As a first step in this process, we can assume that the thermocouple
measurement instrument or data acquisition board has been designed
and constructed so that its internal circuitry provides proper compensation against thermoelectric EMFs. This is, in fact, the case with good
circuit designs, and even truer for those designed to read low-level
voltages.
Next, we can concentrate on what happens between the input terminals and the thermocouple. The chief reason for immersing the reference junction in an ice bath was to force the junction to a known
temperature (0C). However, any temperature will suffice, as long as it
is known. Recall from Figure 6-1 that connecting the thermocouple to
a voltmeter input introduces extra junctions into the circuit at the
points of connection, each of which can also generate thermoelectric
EMFs. Ideally, these terminals (J2 and J3) will be at the same temperature. This can be ensured by mounting them on an isothermal block
that offers sufficient mass to withstand fluctuations in ambient temperature while maintaining the terminals at the same temperature
(Figure 6-2a).
Next, we can move the reference junction out of the ice bath, onto
the isothermal block to produce the circuit in Figure 6-2b. This ensures
that the instrument terminals and the reference junction (J2 , J3 , and
J4) are all at the same temperature. This temperature can be read with
a sensor that does not require a reference junction, such as a thermistor or semiconductor temperature sensor in contact with the isothermal block. Thus, we have succeeded in eliminating the need for a
separate reference temperature source and can measure the temperature of the terminals and the reference junction.
The final step in simplifying this thermocouple input circuit is to
eliminate the length of Alloy 2 wire that extends from the reference
junction (J4) to the instrumentation input (J3). The Law of Intermediate Metals states that a third metal inserted between two dissimilar
metals of a thermocouple junction will have no effect on the output
voltage, as long as the two junctions formed by the additional metal are
at the same temperature. As these junctions are all mounted to the
isothermal block, their temperatures are equal. The thermocouple
junction in this case must be visualized as being formed by copper and
Alloy 1, with Alloy 2 as the intermediate metal.
By removing Alloy 2, we have achieved the input circuit commonly
used for modern thermocouple instrument inputs (Figure 6-2d). A
fixed reference temperature and thermocouple are no longer required
because a non-thermocouple reference sensor is now used to read the
temperature of the isothermal block and input terminals. Multiple
a.
HI
Cu
LO
Cu
Data
Acquisition
System
Alloy 2
J2
J1
Alloy 1
Alloy 2
J3
Isothermal Block
J4 Ice Bath
TREF = 0C
TREF = ?
T
b.
Data
Acquisition HI
System
Cu
LO
Cu
Alloy 2
J2
J1
Alloy 1
Alloy 2
J3
J4
Isothermal Block
TREF = ?
T
c.
Data
Acquisition HI
System
Cu
LO
Cu
Alloy 2
J2
J1
Alloy 1
J3
Isothermal Block
Cold
Amplifier
d.
A/D
Converter
Cu
Fe
Cu
Fe
C
+5V
(Constantan)
Alternate
Temperature
Sensor
Typical input for a J-type thermocouple, using a semiconductor
temperature sensor as the reference junction.
Figure 6-2.
Development of a typical thermocouple instrument input
SECTION 6
thermocouple inputs can be populated on the isothermal block, and

the reference temperature sensor used to compensate all of them.
6.2.1.4 Linearization
Within the usable temperature range of any thermocouple, there is a
proportional relationship between thermocouple voltage and temperature. However, this relationship is by no means a linear one. In fact,
most thermocouples are extremely non-linear over their operating
ranges. In order to obtain temperature data from a thermocouple, it is
necessary to convert the non-linear thermocouple voltage to temperature units. This process is called linearization.
Several methods are commonly used to linearize thermocouples.
At the low cost end of the solution spectrum, one can restrict the thermocouple operating range to where the thermocouple is nearly linear.
At the opposite end of the spectrum, special thermocouple interface
components (integrated circuits or modules) are available to perform
both linearization and reference junction compensation in the analog
domain. In general, neither of these methods is well-suited for costeffective, multi-point data acquisition systems.
In addition to linearizing thermocouples in the analog domain, it is
possible to perform such linearizations in the digital domain. This is
accomplished by either piecewise linear approximations (using lookup tables) or arithmetic approximations or, in some cases, a combination of these two methods. For example, to demonstrate how
thermocouple linearization is typically performed in the digital
domain, lets look at a simple example. Lets assume that we have the
hot junction of a Type J thermocouple in boiling water at 100C and the
cold junction, near the measuring device, at room temperature of 25C.
Obviously, the system should provide a final answer of 100C. The following describes the steps leading to that answer.
First, the following equation determines the measured output of
the thermocouple:
Vmeasured = Vhot Vcold
where: Vmeasured is the voltage measured by the data acquisition
instrument
Vhot is the voltage of the hot junction based on the look-up
tables, and
Vcold is the voltage of the cold junction read from the look-up
tables as well.
According to Type J look-up tables, which always assume a 0C reference:
Vhot (100C) = 5.278mV
Vcold (25C) = 1.019mV
Therefore, the voltage read by the data acquisition instrument would be:
Vmeasured = 5.268 1.019 = 4.249mV
In reality, given that the hot junction temperature is unknown
(100C), the reverse process is applied. The data acquisition device
takes two voltage measurements: one from the thermocouple
(4.249mV) and another one from the cold reference junction Vnew (note
that Vnew is different from Vcold). Then, the software would take over; it
would convert the Vnew to an actual cold junction temperature, Tcold
(25C), based on the properties of the cold junction sensor. Having
Tcold, the software would figure out the equivalent voltage (1.019mV)
according to the Type J look-up tables. It would then add this voltage to
the measured voltage (1.019 + 4.249 = 5.268mV) to obtain the final hot
junction voltage. The final step would be to convert the hot junction
voltage (5.268mV) to an actual temperature (100C) by using the same
Type J look-up tables.
6.2.1.5 Thermocouple Alloys, Extensions, Terminal Pins,
and Other Interconnects
When thermocouples are connected to data acquisition board terminals or other readout devices, the connections form additional junctions that can generate unwanted thermoelectric voltages. A copper
terminal pin plugged into a copper socket will not generate a thermoelectric EMF. However, a constantan pin or socket crimped to a copper
wire results in a J-type thermocouple junction that will generate a
thermoelectric EMF. Extension wire and connector pins made from
thermocouple metals are available to permit connection of like metals.
Attention must be paid to every conductor and termination throughout a thermocouple circuit to ensure that unwanted junctions are not
introduced into the circuit.
Also note that the purity of alloys directly affects the accuracy of
temperatures calculated from a junction voltage. The wire used for
making sensors has a higher accuracy than the extension grade
thermocouple cabling used for long cable runs. Its important to keep
this fact in mind when constructing thermocouple circuits.
6.2.2
Physical Construction of Commercial Thermocouples

In addition to the wire alloy selection, packaging may affect the suitability of a thermocouple to a given application. A working thermocouple can be made by twisting together the stripped ends of a pair of
thermocouple wires. However, the most reliable and consistent operation is provided by thermocouples that have been welded. Real-world
applications often require that thermocouples be enclosed and protected from the environment or fitted with mounts, probe tips, or other
features that best suit a specific application. The sheath is extremely
important because it protects the thermocouple element from con-
SECTION 6
Thermocouple
Wire
Sheath
Extension
Figure 6-3.
Connector
Typical industrial thermocouple
tamination and physical damage due to caustic materials, liquids, and

other environment elements. Common sheath materials include iron,
steel, stainless steel, iconel, ceramics, and porcelain. Figure 6-3 shows
one of several typical industrial thermocouple designs.
Thermocouple Element: Two wires of dissimilar alloys that produce a voltage when exposed to a temperature gradient.
Sheath: A tube of metal or other material, usually closed at the
end, which protects the thermocouple element from the environment.
Terminal Block: The connector assembly (optional) that facilitates connection of the thermocouple to the measurement
instrument or a thermocouple extension. The physical design of
the terminal pins prevents backward connection.
Thermocouple Extension: An extension wire manufactured
from metal alloys compatible with the thermocouple element.
Temperature range must also be considered when choosing sheath
materials; some are better suited to high temperatures or offer greater
service life in hostile environments. Likewise, the insulation on thermocouple and extension wire must be evaluated for its ability to withstand
temperatures and physical abuse encountered in an application.
Another important point to consider with thermocouples is the
mechanical and electrical interface between the thermocouple element and the outside world:
Exposed Junction: Thermocouple wires are unprotected.
Overall, the sensor possesses a lower thermal mass and is directly exposed to outside media, so it responds the fastest to temperature changes.
Grounded Junction: Thermocouple wires are completely
enclosed within and joined to the thermocouple sheath. The
grounded junction provides medium response time and an electrical connection to the sheath.
Ungrounded Junction: Thermocouple wires are completely

enclosed within the thermocouple sheath, but are electrically
isolated from the sheath. The ungrounded junction is the slowest to respond to temperature changes.
The overall response time of a thermocouple depends not only on
the tip design, but also on the sheath material and diameter, and the
surrounding medium. Response times can vary from a tenth of a second to several seconds.
6.2.3
Thermocouple Types and Applications

Several different metal alloys are used to construct thermocouples.
Each alloy offers characteristics that are advantageous for specific
applications. As shown in Table 6-1, these alloys have been assigned a
series of standardized letter codes. Each type of thermocouple wire can
be identified by a color code for the individual conductors. There are
several color coding systems used around the world, but most indicate
the negative thermocouple lead with red. However, the colors of the
positive conductor, thermocouple wire jacket, and extension wire jacket can vary. The color code system used in the United States is shown in
Table 6-2.
6.2.3.1 Base Metal Thermocouples

Thermocouple types J, K, N, E, and T are economical, reliable, and reasonably accurate. These types represent more than 90% of all thermocouples. They are well suited for temperatures ranging from 200 to
1700C.
Type E: Suitable for 200 to 871C. Applicable to atmospheres
ranging from vacuum to mildly oxidizing, and for very low temperatures. Type E provides the highest output of any of the base
metal thermocouples.
Type J: Suitable for lower temperatures (0 to 600C). Should not
be used above 760C. Economical and reliable. Popular in the
plastics industry, but useful as a general-purpose thermocouple
within the prescribed temperature range.
Type K: Industry standard for temperatures up to 1250C. Can
corrode in chemically reducing environments.
Type N: Similar to Type K but more resistant to oxidation.
Type T: Suitable for -200 to 350C. Commonly used in food processing industry.
6.2.3.2 Noble Metal Thermocouples
Thermocouple types R, S, and B are constructed of platinum and
rhodium, so they are referred to as noble metal thermocouples. As a
class, these thermocouples are more accurate and stable than base
metal types, but also more expensive. They are used for applications up
SECTION 6
Table 6-1.
Thermocouple types
Type
Gauge
F Range
C Range
J (Iron vs. Constantan)
8
14
20
24
70 to 1400
70 to 1100
70 to 900
70 to 700
57 to 760
57 to 593
57 to 482
57 to 371
K (Chromel vs. Alumel)
8
14
20
24
70 to 2300
70 to 2000
70 to 1800
70 to 1600
57 to 1260
57 to 1093
57 to 982
57 to 870
N (Nicrosil vs. Nisil)
8
14
20
24
70 to 2300
70 to 2000
70 to 1800
70 to 1600
57 to 1260
57 to 1093
57 to 982
57 to 870
T (Copper vs. Constantan)
14
20
24
70 to 700
70 to 500
70 to 400
57 to 371
57 to 260
57 to 200
E (Chromel vs. Constantan)
8
14
20
70 to 1600
70 to 1200
70 to 1000
57 to 871
57 to 649
57 to 538
R, S Platinum vs.
Platinum/13% Rhodium
24
50 to 2650
46 to 1454
B (Platinum/6% Rhodium vs.

Platinum/30% Rhodium)
24
32 to 2650
0 to 1454
Table 6-2.
Type
Thermocouple color codes, United States

(+)
Conductor
()
Conductor
Thermocouple
Jacket
Extension
Jacket
White
Red
Brown
Black
Yellow
Red
Brown
Yellow
Orange
Red
Brown
Orange
Blue
Red
Brown
Blue
Purple
Red
Brown
Purple
Black
Red
Green
Black
Red
Green
Gray
Red
Gray
to 1700C, and as references for testing other types. To avoid the possibility of contamination at high temperatures from metal vapors, they
should be used inside a non-metallic sheath.
Type R: Industrial standard for high temperature (to 1450C).
Prone to contamination when contacting other metals. Stable in
oxidizing atmospheres, but degrade rapidly in vacuum or reducing atmospheres.
Type S: Similar to Type R. Not used extensively as an industrial
sensor.
Type B: Similar to Types R and S, but useful to 1700C. Best used
at temperatures higher than 250C. A weak, non-linear output at
low temperatures and a dip in output voltage from 0C to 50C
make the B type thermocouple unusable at temperatures less
than 50C.
6.2.3.3 Other Types of Thermocouples
Type C, D, and E thermocouples are more difficult to use because of
their brittle nature and susceptibility to oxidation and breakage. These
types must be used in inert atmospheres, and are useful to 2315C.
Type C: Tungsten/5% Rhenium vs. Tungsten/26% Rhenium
Type D: Tungsten/5% Rhenium vs. Tungsten/26% Rhenium
Type G: Tungsten vs. Tungsten/26% Rhenium
6.3
Resistive Temperature Detectors

Resistive Temperature Detectors (RTDs) are among the most stable and
accurate temperature sensors available. RTDs offer a narrower measurement range than thermocouples, covering approximately 200C to
+800C. The actual range for a particular RTD depends on its composition and construction, but wont vary appreciably from this range.
RTDs are used where high accuracy and repeatability are required,
such as in food, laboratory, and pharmaceutical applications. Accuracy
Figure 6-4.
Simple RTD
SECTION 6
is often expressed as a percentage of resistance at a specified temperature. For instance, a Class B RTD is specified by the International
Electrotechnical Commission (IEC) as 100 0.12% at 0C. IEC Class A
accuracy provides 0.15C at 0C. However, the calibration is performed at two or more temperatures suitably spaced over the stated
RTD working range. Frequently, an interchangeability tolerance is also
specified for RTDs. This specification indicates the maximum tolerance error an RTD of a particular design will have. A typical value is
0.1C.
6.3.1
Construction
Several techniques are used to manufacture RTDs. These techniques
provide various tradeoffs between cost, durability, ease-of-use, and
performance.
The classic RTD configuration is a length of platinum wire wound
on a glass or ceramic bobbin (Figure 6-4). The RTD is then encapsulated in glass or other protective material.
A second variety of RTD is constructed by depositing a conductive
film on a non-conductive substrate, which is then encapsulated or
coated to protect the film. RTD assemblies often include connectors,
metallic sheaths, and handles that make them resemble thermocouple
probes.
6.3.2
Principle of Operation
RTDs are based on the principle that the resistance of most metals
increases with an increase with temperature. An ideal metal for RTDs
would exhibit the following characteristics:
High resistivity (resistance per unit of length), which minimizes
the amount of wire required to provide a high resistance.
The change in resistivity vs. temperature is adequate to provide
the desired measurement resolution.
The change in resistivity is linear, simplifying the conversion to
a corresponding temperature.
Mechanical properties and durability of the metal facilitate construction and ensure the reliability of the measurement device.
Tungsten, nickel, and platinum all possess relatively high resistivities. However, tungsten is fragile and nickel has a non-linear response.
Platinum wire is fragile, but it is only slightly non-linear in its response.
Platinum is highly resistant to contamination and exhibits a predictable change in resistance with temperature. Therefore, most
general-purpose RTDs are made of platinum wire.
The resistance of platinum RTDs ranges from tens of ohms to several thousand ohms, but most platinum RTDs have been standardized
to a value of 100 at 0C. Depending on the purity of the platinum
used, the temperature coefficient () of a platinum RTD is 0.00385/

/C (the European curve) to 0.00392//C (American curve).
6.3.3
Application of RTDs
Unlike a thermocouple, an RTD needs no reference junction. It might
seem a simple matter to connect a standard DMM to the RTD, measure
the resistance of the RTD, then convert to a corresponding temperature. In practice, the resistive properties of the RTD and associated
wiring usually require sensitive instrumentation optimized for low
resistance measurements. For example, a 100 RTD having =
0.00385//C produces a resistance change of only 100 0.00385/
/C or 0.385/C. The wire leads connecting the RTD to the ohmmeter might have a value of several ohms. With a 100 RTD, 1 amounts
to an equivalent temperature error of about 2.5C.
Traditionally, RTDs have been implemented either as part of a
Wheatstone bridge (ratiometric) circuit or in a four-wire (direct voltage
or resistance measurement) configuration. Both methods are capable
of minimizing the effects of lead resistance to provide accurate readings, but each imposes certain requirements. Wheatstone bridges
require additional resistive components to complete the bridge and
the resistance of wire leads from the bridge to the RTD must also be
taken into consideration (Figure 6-5). The bridge design shown in
Figure 6-6 is a refinement of the basic bridge and uses a separate voltage sensing lead from the voltmeter to the RTD to minimize the effects
of lead resistance (a three-wire arrangement). Further developing and
simplifying the bridge produces the four-wire measurement circuit in
Figure 6-7. Here, a current source is used to excite the bridge.
VS
Rref
VS
Voltage
Source
VS
2
Rref
Voltmeter
VM = VR
VS
2
RL
RTD
RL
Lead Resistance
Figure 6-5.
Bridge with two-wire RTD
SECTION 6

VS
Rref
VS
VS
2
Voltage
Source
Voltmeter
V
VM = VR S
2
Sense Current = 0
VR
Rref
RL
RL
RTD
Lead Resistance
Figure 6-6.
Bridge with three-wire RTD
RL
Current = I
RM
0
Current
Source
VM
VR
Current = 0
RM
RTD
RL
Lead Resistance
Figure 6-7.
Four-wire RTD
Today, RTDs are available in two-, three-, or four-wire models to

accommodate the corresponding measurement configuration,
although advances in instrumentation and data acquisition hardware
have largely eliminated the need to build bridge circuits. Typically,
modern RTD analog input modules or plug-in boards provide a regulated excitation source, plus terminals that can be set up for two-,
three-, or four-wire RTD configurations. The software and drivers for
these boards usually contain algorithms to convert raw data to temperature, greatly simplifying the manipulation and conversion of test
data. When a precision ohmmeter is used to read RTD resistance values, the equivalent circuit approximates Figure 6-7 closely. Such
meters usually offer a four-wire measurement mode, and use an onboard excitation source optimized for precise, high resolution measurements of low resistances.
The following subsections describe the theory and calculations for
three-wire and four-wire RTDs. These are offered for reference purposes. The two-wire RTD is not discussed in detail here, although a diagram of it is shown in Figure 6-5. The two-wire configuration is more
suitable for applications where the distance from the bridge to the RTD
is short and some measurement error can be tolerated. Higher gauge
lead wires can be used to minimize these errors. However, the four-wire
RTD configuration is recommended in most applications.
6.3.4
Three-Wire Bridge Configuration

Figure 6-6 illustrates a Wheatstone bridge containing a three-wire RTD.
The standard method of using a Wheatstone bridge for other applications involves balancing the bridge so that the voltage (VM) measured
across the bridge is 0. However, an RTDs value shifts with temperature,
so a bridge containing an RTD will not necessarily be balanced.
Calculations of RTD resistance can be simplified if we first apply a
few constraints to the circuit. First, selecting reference resistors (Rref)
with the same value and composition will result in a reference voltage
(VS/2) of half the supply voltage. Even if the temperature coefficient of
these resistors is not 0, any temperature-induced drift should be in the
same direction, resulting in a stable reference voltage. The other side of
the bridge contains the RTD and a reference resistor (R). Where possible, the value of R should be chosen to be close to the resistance of the
RTD at the temperature of greatest interest.
The circuit in Figure 6-6 shows an RTD with two lead wires connected to one end and a single lead connected to the other end. The
additional lead allows the RTD bias circuit to be partially separated
from the sensing circuit. If the current-carrying leads are of the same
length and gauge, their resistances will tend to cancel when the RTD
value equals resistor R.
The formula to determine the RTDs resistance in the three-wire
configuration (with the RTD on the ground side of the bridge) is:
(VS + 2VM)
RL
RTD = (R + RL) ___________
(VS 2VM)
The current through the RTD side of the bridge is:
VS + 2VM
IRTD = ___________
2R
If the RTD is on the supply side of the bridge, the formula is:
SECTION 6
(VS 2VM)
RL
RTD = (R + RL) ___________
(VS + 2VM)
and the current through the RTD is:
VS 2VM
IRTD = ___________
2R
If RL is known, it can be included in the formula. If RL is not
entered, the worst-case error for the value of the RTD will be about half
that obtained in the two-wire configuration. As the value of the RTD
approaches R, the error will approach zeroan advantage of choosing
R equal to the RTDs value at the temperature of greatest interest.
6.3.5
Four-Wire RTD Configuration

Four-wire RTDs have two wire leads terminated at each end of the RTD
element. The complete four-wire circuit shown in Figure 6-7 uses a
current source, rather than a voltage source. Current flows to the RTD
through one pair of leads, while the voltage drop across the RTD is
measured using the other pair of leads. Lead resistance (RL) has no
effect on the current passing through the RTD. Furthermore, virtually
no current flows in the measurement side of the circuit, so any error
contributed by resistance of the measurement leads (RM) is insignificant as long as a high impedance analog input voltmeter is used.
The voltage (VM) measured across the RTD can be used to calculate
the resistance of the RTD:
VM
RTD = _____
IRTD
To minimize self-heating errors, the current should be limited to
1mA or less, meaning that VM will be a maximum of 0.1V for a 100
RTD. Therefore, a high gain, low noise analog input channel will provide best results. Filtering or other forms of signal conditioning may be
helpful.
6.3.6
Converting RTD Resistance to Temperature

There are two options for converting resistance to temperature. One is
simply to consult a look-up table and find the temperature corresponding to a specific resistance. This method is workable in software
programs where an event will be triggered at a certain temperature (the
corresponding resistance or voltage can be used as a trigger level), but
is not suitable for real-time readout of temperature based on RTD
resistance values.
A second method of converting resistance to temperature is by
means of an equation. The most commonly cited equation for this purpose is a polynomial that uses a set of constants called the CallendarVan Dusen coefficients.
The general equation for the relationship between RTD resistance

and temperature is:
RTD = R0 [1 + At + Bt2 + C(t 100)3]
where: RTD is the resistance of the RTD at temperature t,
R0 is the resistance of the RTD at 0C, and
A, B, and C are the Callendar-Van Dusen coefficients shown in
Table 6-3.
For temperatures higher than 0C, the C coefficient is 0, and the
equation becomes:
RTD = R0 [1 + At + Bt2]
If a current IRTD is passed through the RTD, and the voltage VRTD
measured, this equation can be solved for t:
t=
2 (VRTD IRTDR0)
IRTDR0 [
A2
+ A2 + 4B (VRTD IRTDR0) / IRTDR0 ]
Table 6-3.
Callendar-Van Dusen coefficients for common RTD alphas
Standard
RTD
Temperature
Coefficient ()
C*
DIN 43760
0.003850
3.9080 103
5.8019 107
4.2735 1012
American
0.003911
3.9692 103
5.8495 107
4.2325 1012
ITS-90
0.003926
3.9848 103
5.870 107
4.0000 1012
* Used for temperatures less than 0C only. For temperatures higher than 0C, C = 0.
6.3.7
Excitation Current and Joule Heating

One aspect of using RTDs and most other resistive sensors is resistive
(joule) heating that results from excitation current passing through
the sensor (power = excitation current2 RTD resistance). Although the
amount of heat energy may be slight, it can affect measurement accuracy nonetheless. Self-heating is typically specified as the amount of
power that will raise the RTD temperature by 1C. Its typical value is
about 1mW/C.
Inaccuracy caused by joule heating is aggravated by higher excitation currents and stagnant surrounding media of low specific heat.
These effects can be minimized if the surrounding medium is in
motion or is agitated to carry heat away from the RTD.
Typical RTD data acquisition boards supply an excitation current
of 100A to 1mA to minimize joule heating. A 100 RTD will pass a current of 1.0mA when biased with only 0.1V. While lower current is better,
currents lower than 100A result in lower signal levels, which are more
difficult to measure.
SECTION 6
6.4
Thermistors
The thermistor (thermally sensitive resistor) is a second variety of resistive temperature detector commonly used in data acquisition applications. Although RTDs and thermistors are both resistive devices, they
differ substantially in operation and usage.
Thermistors are passive semiconductor devices. Both negative
temperature coefficient (NTC) and positive temperature coefficient
(PTC) thermistors are available. The resistance of an NTC thermistor
decreases as its temperature increases, while the resistance of a PTC
thermistor increases as its temperature increases. For temperature
measurement applications, NTC types are used more commonly than
PTC thermistors.
Very small thermistors can be manufactured and this small size
allows them to respond quickly to slight temperature changes.
However, they can be prone to self-heating errors. Thermistors are also
relatively fragile, so they must be handled and mounted carefully to
avoid damage.
Thermistors offer a significantly broader range of base resistance
values than RTDs do, with base resistance values of kilo-ohms to megaohms readily available. Compared to RTDs, the temperature coefficient
of a typical thermistor is relatively largeon the order of several percent or more per degree Celsius. This high temperature coefficient
results in a resistance change of up to several thousand ohms per
degree Celsius. Therefore, the resistance of the wires connecting the
instrumentation to the thermistor is insignificant, so special techniques such as high-gain instrument inputs and three- or four-wire
measurement configurations are unnecessary to achieve high accuracy.
While thermistors have relatively few drawbacks associated with
them, its important to be aware of these limitations in order to achieve
accurate, reliable measurements. For example, thermistors are relatively low temperature devices, with a typical measurement range of
50C to 150C, although some thermistors can be used at temperatures up to 300C. This range is significantly narrower than that of
thermocouples and RTDs. Exposure to higher temperatures can decalibrate a thermistor permanently, producing measurement inaccuracies. Thermistors are highly non-linear in their response, and are not
as standardized as thermocouples and RTDs. They tend to be more
appropriate for applications that require sensitive measurements over
a relatively restricted temperature range, rather than for generalpurpose temperature measurements.
6.4.1
Thermistor Circuit Configuration

As discussed previously, thermistors have a higher base resistance
value and a higher temperature coefficient of resistance than RTDs do.
Therefore, techniques such as four-wire configurations and sensitive

measurement capability are required only in more critical thermistor
applications, because any resistance in the test leads is relatively
insignificant when compared to the resistance of the thermistor itself.
Figure 6-8 shows a standard two-wire measurement setup. The calculation of the thermistors resistance is a straightforward exercise in
Ohms Law. In cases where the series resistance of the lead configuration is significant, the four-wire configuration can be used. The current
applied to the thermistor should always be limited to the minimum
needed to produce a readable voltage.
6.4.2
Converting Thermistor Resistance to Temperature

The output of most thermistors is highly non-linear, and their response
has been standardized much less than for thermocouples or RTDs.
Therefore, manufacturers frequently supply resistance-temperature
curves, tables, or constants for their specific products. Typical thermistor alphas () range from 2% to 8% per C, and are generally larger at
the lower end of the temperature range. Linearized thermistors also
Current
Source
Lead Resistance
Current
Source
Lead Resistance
Figure 6-8.
Two- and four-wire thermistor circuits
SECTION 6
exist, although the use of computerized data acquisition systems and

software make them unnecessary unless the readout hardware must be
used with a linearized type.
For computerized applications, relatively accurate thermistor
curves can be approximated with the Steinhart-Hart equation:
1
T = ________________________________
A + B ln(RT) + C[ ln(RT) ]3
T is the temperature in degrees Kelvin, which is equal to the Celsius
temperature plus 273.15. RT is the resistance of the thermistor. The
constants A, B, and C for a given thermistor should be provided by the
thermistor manufacturer.
6.5
Semiconductor Linear Temperature Sensors

Monolithic linear temperature sensors constitute yet another type of
temperature transducer. Typically, these sensors are two- to three-pin
active electronic devices that operate from a nominal 5VDC to 30VDC
supply voltage, and pass an output current or voltage proportional to
temperature.
The first of these sensors were designed and calibrated such that
current output increased linearly by 1V per Kelvin degree, with an
output of 0V corresponding to 0K. These sensors are available in a variety of packages and models that provide voltage or current output.
Later-generation devices are available scaled to read directly in Celsius
or Fahrenheit values, eliminating the need to convert sensor output to
the desired temperature scale.
The temperature range of monolithic linear temperature sensors is
approximately 50C to 150C, making them suitable for a relatively
narrow range of temperatures as compared to thermocouples and
RTDs. Within this range, output is extremely linear with respect to temperature, and no reference junctions or complex calculations are
required to use them. Although these sensors require a power source,
self-heating is a minor concern. Supply currents are generally in the
range of 75100A, so power consumption is very low, and typical maximum self-heating in still air is just 0.10.2C.
Using solid-state temperature sensors is fairly straightforward. The
current sensor in Figure 6-9 is inserted in series with a resistor that
provides a voltage drop that can be read by a digital voltmeter. The voltage sensor outputs a voltage directly.
6.6
Thermal Shunting
All temperature detectors possess some mass in the form of a sensor
element, protective sheath or encapsulation, leads, and other physical
components. When the sensor is placed in contact with a medium to
i = 1A/K
10mV/K
10k
To DVM
Current
Sensor
Figure 6-9.
To
DVM
Voltage
Sensor
Semiconductor temperature sensors
measure its temperature, the sensor will absorb some heat energy from
the medium, thereby altering its heat content and temperature. This
process is called thermal shunting.
Thermal shunting can be minimized by using temperature sensors
of the smallest possible mass. However, the choice sometimes imposes
tradeoffs. For example, although thermocouples generally have lower
mass than RTDs, they are less accurate. Powered resistive sensors of
lower mass are more prone to joule heating than more massive sensors.
Temperature sensors with lower mass can also be more susceptible to
damage or other problems.
SECTION 6
SECTION 7
Strain
Measurement
7.1
Strain
Studying the physical behavior of mechanical structures frequently
includes measuring a phenomenon known as strain. Strain is defined
as a physical distortion of an object in response to one or more external stimuli applied to the object. Typical stimuli include linear forces,
pressures, torsion, and expansion or contraction due to temperature
differentials. Strain can result in elongation or contraction; these phenomena are noted by the use of a (+) sign for expansion or a () sign for
compression.
Strain information is important in structural engineering because
it can be used to determine the stress present in an object. Stress data
can then be used to assess factors such as the structural reliability and
service life of the object. The basic principle of strain measurement is
also used in other types of force-related measurements, such as pressure, torque, and weight.
Strain is calculated as the change in length of an object divided by
the unit length of the object. Normally, this change is extremely small
in relation to the objects length. For example, the length of a onemeter rod might change by only 0.01mm under compression. This
change can be calculated as 0.01mm/1000mm, and is expressed as
0.00001 strain or ten microstrain.
L
Strain () = ____
L
7.2
Poissons Strain
When a bar is strained with a force along the length, causing the bar to
elongate, a phenomenon known as Poissons Strain causes the girth of
the bar to contract. This contraction is a material property denoted by
Poissons Ratio () for the specific material (Table 7-1). Poissons Ratio
is defined as the negative ratio of the strain in the transverse direction
(perpendicular to the force) to the strain in the axial direction (parallel
to the force) or .
t
= ____
In a half bridge or full bridge, one or two strain gauges can be

mounted in the transverse direction, so that their resistance changes in
response to Poissons Strain.
7.3
Strain Gauges
Strain is measured using strain gauge sensors. While there are piezoelectric and semiconductor strain gauges, the majority are passive
resistive devices constructed by depositing or etching a wire or foil
sensing grid on a substrate known as the carrier matrix. Figure 7-1
SECTION 7
Table 7-1.
Poissons Ratio for some common materials
Material
Poissons Ratio ()
Aluminum
0.32
Red Brass
0.33
Cast Iron
Polycarbonate
Steel 1018
0.285
0.285
Steel 4130/4340
0.280.29
Stainless Steel - 304
0.25
Stainless Steel 410
0.270.29
Titanium
0.34
illustrates one of the many varieties of strain gauges available. Gauges

may be constructed with a single grid or with multiple grids, and in a
variety of form factors. Furthermore, multiple grids can be arranged on
the carrier matrix in a variety of patterns to allow measuring strain in
different directions.
Gauge Length
Figure 7-1.
Solder
Tabs
A typical strain gauge
Usually, strain gauges are attached to a test object using an adhesive designed to transmit strain from the object to the gauge, as well as
to insulate the gauge and dissipate heat from the gauge to the object. A
protective coating may also be applied to the outside of the mounted
strain gauge to seal it from the environment.
Resistive strain gauges are used in a Wheatstone bridge configuration (Figure 7-2). The Wheatstone bridge is effective in detecting an
imbalance between resistances comprising the bridge. However, a
four-wire measurement setup similar to that used for RTDs can also be
used for measuring the resistance of one strain gauge (see Section 6,
STRAIN MEASUREMENT
R1
R2
G
R3
Figure 7-2.
R4
Wheatstone bridge configuration
Temperature Measurement). In this case, the strain gauge is excited

using a current source and a precision microvoltmeter is used to measure the voltage drop across the gauge. This measurement can be used
to calculate the change in resistance. A micro-ohmmeter can also be
used to measure resistance directly.
The Wheatstone bridge strain gauge can contain one, two, or four
active strain gauge elements (Figure 7-3, R1 through R4). The resistance
value of a bridge constructed with four equal resistances is equal to the
value of one of the bridge legs. Common bridge values include 120, 350,
600, 700, and 1000 ohms, but gauge values can vary from 2530 ohms
to several thousand ohms. At a given supply voltage, a higher resistance
bridge will dissipate less heat than a lower resistance bridge, and the
resistance of measurement leads will have less effect on accuracy.
Four active elements allow a bridge to produce the maximum
imbalance and VOUT for a given strain level, provided the elements are
properly positioned on the structure. For bridges containing two active
elements, VOUT at a particular level of strain will be about one-half that
of a bridge with four active elements. For a one-element bridge, VOUT
will be approximately one-fourth as great.
In the full bridge configuration, two gauges must be mounted on
the tensile surface of the structure, with two more on the opposite
compressive surface, so that two gauges increase in resistance while
the opposite gauges decrease in resistance (Figure 7-4). Note that the
gauge elements must be aligned with respect to the test object so that
the sensing grid can detect flexing of the test object. Also note that different types of strain measurement, such as torsional and shear strain,
can require different gauge orientations.
If only two active strain gauge elements are used in a bridge, one
active element should be mounted on each side of the object (e.g.,
positions R1 and R2 or positions R3 and R4). The bridge should be
completed with resistors with values and temperature characteristics
similar to the active element(s).
SECTION 7
VS
VS
Regulated
Voltage
Source
R4
R1
Voltmeter
(VOUT)
R3
R2
+
Strain
VOUT
Figure 7-3.
Strain gauge bridge with four strain gauge elements
Force
R1
R3
R2
R4
Gauge Placement (1, 2, 3, 4)
Figure 7-4.
Strain gauge placement for a four-element bridge
Accurate measurement of strain gauges requires a low noise, regulated voltage source to eliminate unwanted variation in VOUT. The current requirement can be calculated according to Ohms Law as I = VS/R,
where R is the total resistance of the bridge. These current requirements are typically quite low; a 120 bridge operating from 5V would
require 5/120 or about 42mA. Recommended maximums for the strain
gauge supply voltage (VS) range from 12V up to 25V or more. The
applied voltage should not exceed the strain gauge manufacturers recommendations.
STRAIN MEASUREMENT
7.4
Gauge Factor
The term gauge factor (GF) describes the ratio of the change in resistance with respect to strain. The complete expression for GF is:
R/R
R/R
GF = ______ = ______
L/L
A typical gauge factor is two; however, the strain gauge manufacturer should supply the actual gauge factor and tolerance for individual
product samples. Ten microstrain applied to a gauge with a gauge factor of two results in a resistance change of 20 micro-ohms (20 106).
The maximum strain likely to be encountered with most metallic
objects is about 5,000 microstrain, which, with GF = 2, corresponds to
a change in resistance of 10,000 micro-ohms (0.01). Therefore, measuring strain accurately requires instrumentation that can resolve at
least 1 or 1V.
7.5
Sources of Error
Like other types of resistive sensors, the accuracy of strain gauges can
be affected by electrical noise, thermoelectric EMFs, and joule heating.
Therefore, strain gauges should be installed using shielding where necessary and taking connector metals, supply voltage, and ambient temperature into consideration.
Note that quarter and half bridge configurations are more susceptible to non-linearity errors than a full bridge as the bridge moves out
of balance, as well as to errors caused by the resistance of measurement leads. Using a full bridge will minimize or eliminate these error
sources.
To minimize self-heating, the voltage used to power a strain gauge
bridge should be only high enough to achieve the desired result, and
should not exceed the manufacturers recommendations. Note that a
rated maximum supply voltage might only be permitted for applications where the test object can adequately sink heat from the gauge.
Excitation can be applied only during measurements as a means of
reducing self-heating.
Strain gauges are often accompanied by a reference temperature or
temperature range for which the gauges specifications are valid.
A full bridge strain gauge can tolerate temperature change with
minimal effect on the bridges basic accuracy, because all four legs of
the bridge presumably have the same temperature characteristics and
are affected equally. For quarter or half bridge configurations, temperature-dependent errors can be compounded by the possibility that
bridge completion resistors are located some distance from the strain
gauge, and that they have thermal characteristics that differ from the
rest of the circuit. Therefore, it may be necessary to compensate for
SECTION 7
temperature effects. This process requires measurement of gauge temperature, and application of temperature coefficient information supplied with the strain gauge.
In actual applications, temperature changes can result in dissimilar rates of expansion between the strain gauge and the test object.
These dissimilar expansion rates appear as mechanical strain in the
test object, even though the object has not been subjected to any distorting force. This apparent strain can be minimized by maintaining
the test object and strain gauges at a consistent temperature, and by
matching strain gauges to test objects according to thermal expansion
characteristics. Strain gauge vendors frequently offer gauges matched
to the thermal expansion properties of plastics, aluminum, steel alloys,
and other common structural materials.
7.6
Operation
Unlike bridge-based measurements with resistive temperature sensors, strain gauge bridges are used in applications where the incremental change in resistance is of greater interest than the absolute
output level of the bridge. Normally, bridge output vs. strain closely
approximates a linear relationship. The output of a strain gauge is relatively small, up to a few thousand microstrain, because the bridge does
not move very far out of balance, even with the test object loaded to
maximum.
A variety of techniques exist for converting bridge behavior into
strain. These range from bridge rebalancing to a simple reading of the
voltage across the bridge. This direct reading method assumes a linear
relationship between bridge output voltage (VOUT) and strain (). A
bridge that is balanced under no load conditions is helpful in minimizing possible linearity errors.
The following paragraphs describe formulas for calculating basic
strain measurements for a Wheatstone bridge, based on the supply
voltage (VS), the bridge type, gauge factor (GF), and the bridge output
voltage (VOUT). The formulas make use of an additional variable, voltage ratio (VR), which is calculated as follows:
VOUT (Strained) VOUT (Unstrained)
VR = __________________________________
VS
7.6.1
Full Bridge Configuration

The full bridge provides several advantages over half and quarter
bridges, including maximum sensitivity, fewest error components,
greatest immunity to noise, and easier management of temperature
effects. Figure 7-5 shows the four strain gauge elements in a full bridge
configuration. The following formulas can be used to calculate strain
for the full bridge:
STRAIN MEASUREMENT
Where R1, R2, R3 and R4

measure normal strain:
VR
GF
Where R1 and R2 measure

Poissons Strain, and R3 and R4
2VR
GF( + 1)
Where R1 and R3 measure

Poissons Strain, and R2 and R4
2VR
GF[( + 1) VR( 1)]
VS
VS
Regulated
Voltage
Source
R2
R4
R1
+
Voltmeter
(VOUT)
R3
+
Strain
VOUT
Figure 7-5.
7.6.2
Strain gauge bridge with four strain gauge elements
Half Bridge Configuration

A half bridge uses two strain gauges mounted on a test member and
two bridge completion resistors (Figure 7-6). Where the active strain
gauge elements are some distance from the instrumentation, the
resistance of lead wires (RL) must also be considered. Assuming that all
wires leading to the remote bridge are the same length and gauge, then
RL should be the same value for all leads. Thus, the lead resistances in
the bridge legs will cancel each other out, leaving the resistance in the
(+) voltmeter lead as the only one of importance. Use of a high impedance voltmeter will minimize the effect of this resistance. The following
equation applies to the half bridge and provides compensation for lead
resistance:
SECTION 7
Where R3 and R4
measure normal
strain:
2VR
R
1+ L
GF
RG
Where R3 measures
Poissons Strain and
R4 measures normal
strain:
2VR
4VR
R
1+ L
GF
GF[( + 1) 2VR( 1)]
RG
VS
R1
VS
Regulated
Voltage
Source
R2
RL
Voltmeter
+
0
R4
RL
(VOUT)
RL
R3
Lead Resistance
Figure 7-6.
Half bridge strain gauge
RG is the resistance of one of the strain gauge elements and RL is

the lead resistance.
Note that wire gauge strongly influences the effect of lead resistance. Ten feet of 18 gauge wire (0.0066/foot) and a 120 bridge results
in the multiplier term (1 + RL/RG) equaling 1.0005, which shows an
error of 0.05%. Therefore, the best method of reducing lead resistance
effects is to use the shortest, heaviest gauge wire practical in the application. Refer to Appendix C for information on the diameter and resistance of various wire gauges.
7.6.3
Quarter Bridge Configuration

A quarter bridge (Figure 7-7) uses one strain gauge and three bridge
completion resistors. Completion resistor R3 can be located with the
instrumentation and other completion resistors or remotely with the
strain gauge element. Remote location of R3 can minimize errors due
to a temperature differential between R3 and R4, but can present a
problem with mounting the resistor.
As with the half bridge, the lead resistance (RL) must be considered.
Figure 7-7 uses a three-wire measurement scheme that cancels lead
resistance in the leads carrying excitation and ground to the remote
gauge. The following equation applies to the quarter bridge, and pro-
STRAIN MEASUREMENT
VS
VS
R1
+
Regulated
Voltage
Source
R2
RL
Voltmeter
+
0
R4
RL
(VOUT)
RL
R3
Lead Resistance
Figure 7-7.
Quarter bridge strain gauge
vides compensation for resistance of the voltmeter lead. The same

observations concerning wire gauge apply to the half bridge:
Use of one active strain
gauge R4 precludes measurement of Poissons Strain.
7.7
2VR
4VR
R
1+ L
GF
GF(2VR + 1)
RG
Strain Gauge Signal Conditioning

The output of a strain gauge is on the order of a few millivolts per volt
of excitation, so the instrumentation used must be able to resolve
microvolt signal levels (or micro-ohm levels if an ohmmeter is used for
direct resistance readings). Therefore, all the standard techniques for
low noise, sensitive measurements apply to strain gauges, including
shielding, filtering, differential voltage measurement, and signal averaging. This signal conditioning should be available as part of the A/D
input used to read the strain gauges. For remote installations, note that
many manufacturers offer signal conditioning pre-amps that can be
installed near the strain gauges to boost the signal level to 010V or
420mA output ranges. (The term transmitter is used in process industries for devices that convert input signals to isolated two-wire current
loop outputs.)
7.8
Shunt Calibration
The normal procedure to verify the output of a strain gauge measurement system relative to some predetermined mechanical input or
strain is called shunt calibration. Shunt calibration involves simulating
strain gauge loading by connecting a known high resistance (R1) across
one leg of the bridge (R2). The resulting resistance (RT) for that bridge
leg can be calculated by:
SECTION 7
R1 R2
RT = ________
R1 + R2
The output of the bridge can then be measured and compared to
the expected voltage value. It can also be used to correct span errors in
the entire measurement path or to verify general operation of the test
setup.
Caution: The shunt calibration is a convenient way to simulate
strain electrically. However, it does not check the mechanical function
of the sensor, so it may not detect errors caused by cracks, deformation, or other mechanical problems. To check for these kinds of problems, apply a known weight or other suitable physical stimulus
periodically to verify that the sensor responds appropriately.
7.9
Load Cells, Pressure Sensors, and Flow Sensors

A variety of physical phenomena can be measured with strain gaugebased transducers coupled with specialized mechanical elements.
These phenomena include load, gas or liquid pressure, and flow rate,
among others.
A load cell is simply a packaged strain gauge designed to measure
force under different load conditions. A typical load cell body is
machined from metal to provide a rigid but compressible structure.
The load cell body is usually designed with mounting holes or other fittings to allow the unit to be mounted permanently on a supporting
structure.
Typical transducers for pressure sensing are electromechanical
devices that combine a sensor with mechanical elements such as a
diaphragm, piston, or bellows. The mechanism is designed to respond
to a pressure differential and expand against a spring, which determines the range of the sensor. The pressure sensing element also actuates a strain or positional sensor that produces an electrical output that
can be read by an instrument.
Flow can be measured in a variety of ways, most of which are indirect measurements that depend on the materials viscosity, conductance, or other properties. One very common way to infer flow is to
measure the materials pressure drop when passing through an orifice,
using two pressure sensors. The equations for converting this delta-P
to flow are beyond the scope of this handbook, but are discussed extensively in other literature.
Several general rules apply to the transducers described in this section:
1. These transducers are ready-to-use, self-contained devices fitted
with wire leads for excitation input and signal output. Packaging
can include electrical quick disconnects, mounting holes, threads,
pressure fittings, or other features to simplify attachment and use.
STRAIN MEASUREMENT
2. These transducers are designed for process control or other commercial applications, as opposed to traditional strain gauges that
are used for experimentation or research.
3. Transducers may be based on traditional resistive strain gauges or
on other sensor technologies. They must be powered from an
external excitation voltage specified by the manufacturer. The output is frequently a low-level signal, so it often requires amplification and signal conditioning for best results. External drift
compensation, calibration, and zero adjustment may be required
to provide accurate, long-term results.
4. Sensing elements are fully integrated into the transducer. They are
frequently sealed from the environment and inaccessible to the
user.
7.10 Acceleration, Shock, and Vibration

Acceleration, shock, and vibration are important parameters in
mechanical applications, including automotive, aerospace, product
packaging, seismology, navigation and guidance systems, motion
detection, and machine maintenance. These parameters can all be
measured using a sensor known as an accelerometer.
7.10.1 Acceleration
Acceleration is the rate of change of velocity that a mass undergoes
when it is subjected to a force. It is defined by the formula:
Force = mass acceleration
Assuming that mass and force remain the same, then acceleration
is constant, meaning that an objects velocity will continue to change at
a uniform rate. Acceleration is expressed as distance per time squared,
with meters/sec2 and feet/sec2 being common units. Earths gravity
exerts a force of 9.8m/sec2, or 32ft/sec2; this amount of acceleration is
often referred to as 1g (g for gravity). Sensors designed to measure
acceleration can measure values ranging from millionths of g (g) to
hundreds of g or more. They are frequently designed to withstand overloads of thousands of g without damage.
Note that velocity is a vector quantity and consists of components
for speed and direction. Therefore, an object undergoes acceleration
when its speed, direction, or both change. Furthermore, the direction
of the acceleration can be multi-dimensional, requiring a multi-axis
accelerometer for complete measurement. Accelerometer mounting
can be critical for achieving satisfactory results; a misalignment of only
a few degrees can have profound effects on the final result.
SECTION 7
7.10.2 Shock
Shock is a special case of linear acceleration in which the acceleration
(or deceleration) time approaches zero. In the real world, the time cannot equal zero, but it can equal a very small fraction of a second. The
result of a large change in velocity over a short time can produce an
acceleration of hundreds of g or more.
For shock studies, the measurement range of the accelerometer
must be able to handle the acceleration expected in the specific shock
situation. This value can be calculated by dividing the change in velocity by the time interval:
V
a = ____
t
For example, an object traveling at 100m/sec that strikes a surface
and stops in 0.5 seconds (a relatively mild shock, by some standards)
would experience a deceleration of 200m/sec2. The force of gravity is
9.8m/sec2, so this deceleration corresponds to approximately 20g. For
purposes of comparison, this shock might be fatal to a human, but
some computer hard drives are specified to withstand a 30g shock
while operating, and up to 300g while unpowered.
7.10.3 Vibration
Vibration is a continuing change in the position of a body, typically
occurring in a cyclic pattern with a constant or near-constant period.
Vibration is a common characteristic of machines; it is of interest in
manufacturing because its measurement can provide an indication of
the health of machinery. Periodic monitoring and analysis of vibration
levels can reveal specific problems, such as worn bearings, loose fasteners, or out-of-balance rotating components, before they become
severe enough to be noticed or to cause complete failure. Vibration is a
cyclic changing of position, so it can be detected by using an
accelerometer.
7.10.4 Resonance and Q

Accelerometers are based on a mechanical operating principle, so they
are governed by the physics of vibrating bodies, notably, the characteristics of resonance and Q. All bodies possess a resonant frequency at
which the body will vibrate when excited mechanically. Generally,
more massive bodies have a lower resonant frequency. The Q parameter describes the degree to which the effect of excitation energy is
amplified by a body as the frequency of excitation nears the bodys resonant frequency. This amplification is a process of the body storing the
energy and re-emitting it in phase with the driving energy. The higher
the Q factor, the more pronounced this amplification effect will be.
STRAIN MEASUREMENT
When the frequency of a vibrating body nears the sensors resonant

frequency, the sensor begins to absorb energy. This drives the sensor
into mechanical oscillation, and the measurement becomes progressively less accurate. In extreme cases, the sensor itself can go into
uncontrolled oscillation and be damaged.
Every physical sensors response consists of two components: a
forced response in reaction to an input signal (the useful output for
measuring that signal), and a natural response, which are attributable
to the sensors own resonant frequency and Q. Resonance and Q are
properties of the sensor itself, so they do not help characterize the
input signal. As far as the measurement is concerned, they are noise.
Thus, its important to consider resonance and Q to choose an
accelerometer that will provide accurate results in a given application.
The frequency of vibration to be studied should be well below the resonant frequency of the sensor; some sensors will be suitable only for
steady-state acceleration or low frequency vibration measurements.
Generally, sensors with a low mechanical Q (e.g., Q=1) can be used to
measure vibration at frequencies up to approximately one-half the
sensors resonant frequency before error becomes objectionable. HighQ sensors (e.g., Q=10,000 or more) may be used for vibration frequencies up to about one-third the sensors resonant frequency. These
figures are approximations; consult the manufacturers data for the
specific accelerometer to be used.
7.10.5 Accelerometer Types

An accelerometer measures acceleration by detecting a change in the
speed and/or direction of a body. The operating principle of the
accelerometer couples a mass to a sensor that can generate an electrical signal. The sensor produces an output signal that is proportional to
the acceleration.
Accelerometers exist in a variety of forms, some of which are listed
in Table 7-2. There are also accelerometers that use inductive, capacitive, Hall effect, or other varieties of basic sensing technology to detect
acceleration. Each offers specific advantages in the form of output, frequency response, etc.
Another important consideration with accelerometers is that
attaching the sensor to the test structure alters the mass of the structure, so it can affect its frequency of vibration significantly. For an accurate assessment of vibration, this loading effect should be minimized
by using a low mass sensor, which is placed to provide maximum signal while minimizing loading effects. A rule of thumb for the accelerometers mass is that it should not exceed one-tenth the mass of the
structure under test.
SECTION 7
Table 7-2.
Selected accelerometer types
TYPE
FEATURES
Strain Gauge/Mass
(or strain)
Accelerometer
Combines a traditional strain gauge type of sensor

with an inertial mass. Suitable for steady-state (noncyclic) acceleration measurements and low frequency
vibration studies.
Piezoelectric
Accelerometer
Uses a piezoelectric (quartz) element coupled to an

inertial mass to generate a voltage or charge in
response to acceleration. Suitable for vibration studies
at frequencies of a few Hertz and higher. Suitable for
high shock environments. Not recommended for
steady-state measurements. Very high resonant
frequency permits measuring vibration of 3040kHz
or more.
Spring-resistive
Accelerometer
Uses a spring/mass mechanism that operates the

wiper of a potentiometer in response to acceleration.
Simple operating principle. Suitable for steady-state
and low frequency vibration measurements (~7 to
10Hz or less).
7.10.5.1 Strain Gauge Accelerometer

The strain accelerometer uses a strain gauge in conjunction with a cantilevered mass. When the sensor is subjected to an accelerating force,
the masss inertia causes a distortion of the cantilever, and the strain
gauge measures the resulting strain. Obviously, the positioning of the
sensor relative to the accelerating force is important in detecting the
acceleration and generating the highest possible output signal.
Circuits for interfacing to strain gauge accelerometers are similar
to those used for standard strain gauges. A Wheatstone bridge configuration may be used, along with appropriate signal conditioning.
However, the possibility that the gauge will be used to measure vibration makes it important to consider the frequency of the signal. Where
vibration is measured, a sampling rate and signal conditioning chosen
must capture the desired information while avoiding aliasing. If the
goal is to reconstruct a waveform in great detail, then a sampling rate
many times the highest frequency component of the vibration will be
required.
The resonant frequency of an undamped strain gauge accelerometer is relatively low (typically, a few kilohertz), while its Q is relatively
high, making this type of sensor better suited to applications with frequencies of DC to ~1000Hz. Some are suitable for frequencies of only a
few hertz. A damping material, such as oil, is sometimes incorporated
into the design of an accelerometer to lower the mechanical Q of the
sensor and increase the frequency range over which it can be used.
STRAIN MEASUREMENT
7.10.5.2 Piezoelectric Accelerometer

The piezoelectric effect is well known; it is used in many types of
devices that produce an output voltage in response to a mechanical
force. A compressive or shear force applied to a piezoelectric crystal
(typically, quartz) produces a charge within the crystal that is proportional to the force. The crystal acts like a capacitor to store the charge.
The output impedance of a piezo crystal is very high, so specialized
voltage or charge amplifiers are needed to measure the unconditioned
output of the crystal. Special attention must be given to the length and
properties of cabling between the sensor and amplifier.
A more common variation of piezo-based accelerometers uses
built-in amplification and signal-conditioning circuitry to provide a
low impedance output signal. The sensitivity of this type of sensor is
specified by the manufacturer; its output can be read with conventional high speed voltage input on a data acquisition board. These conditioned sensors offer high noise immunity and simpler requirements for
cabling between the sensor and measurement instrument.
The piezoelectric accelerometer is characteristically a low Q device
with high resonant frequency. Piezoelectric accelerometers with resonant frequencies as high as 120kHz are available, providing a usable
frequency range from a few hertz to 3540kHz. Piezoelectric accelerometers are not recommended for steady-state acceleration measurements.
7.10.5.3 Spring-Resistive Accelerometer
The spring-resistive accelerometer is a simple design that is similar in
concept to a spring scale with a suspended mass. Acceleration causes
the mass to exert a force on the spring, which causes the spring to
expand or contract. If this linear motion is coupled to a potentiometer
wiper, the changing resistance value will provide an electrical signal
indicating acceleration.
7.10.6 Instrumentation Requirements

Acceleration, vibration, and shock present a very wide dynamic range
of physical phenomena to be measured in terms of magnitude, duration, sensor characteristics, and final information the measurement
must provide. The goal of any of these measurements might be simply
to determine maximum acceleration or it might be as complex as
reconstructing a profile of velocity with respect to time.
Assuming that a suitable accelerometer has been chosen for the
application, the primary hardware concerns for reading the accelerometer become sampling speed and signal conditioning. The sampling
rate of the analog input board must be sufficient to capture the desired
phenomenon, and also satisfy the Nyquist rule: the sampling rate must
be at least twice the highest frequency component of the signal to be
SECTION 7
captured. For a vibration study, in particular, the test subject may

vibrate at a fundamental frequency plus selected harmonics of the frequency. These harmonics are multiples of the fundamental, and can be
useful for identifying particular problems with a machine.
Conversely, it is also important to use anti-aliasing filters to remove
high frequency components that are more than one-half the sampling
rate in order to eliminate false characteristics in the reconstructed
waveform. Obviously, where greater detail is required in a vibration
measurement, it may be necessary to sample the signal at many times
the highest frequency component to reconstruct the waveform accurately. Depending on the type of signal and distance between the sensor and A/D input, additional signal conditioning in the form of
amplification and impedance conversion can be required.
Given that the maximum frequency of vibration to be measured
with an accelerometer might be on the order of 3040kHz and probably 510kHz in more typical applications, an A/D board with a minimum 100kHz sampling capability would be a good choice.
Signal levels can vary from a few millivolts for strain gauge
accelerometers to several volts for sensors using integrated signal conditioning. Some integrated sensors may output a current (420mA)
rather than a voltage. Therefore, a suitable A/D board for use with
accelerometers should have sufficient gain and noise performance to
provide readings with the necessary resolution.
STRAIN MEASUREMENT
SECTION 8
Related Topics
of Interest
8.1
Current Measurements
Electrical currents can be measured with data acquisition systems, but
the method selected will depend on the current level and number of
required channels. An important characteristic of a current loop is that
any lead resistance in the circuit does not affect current flow as long as
the source can supply the required voltage. That makes current loops
ideal where theres an appreciable distance (i.e., >2m) between the signal source and the instrumentation.
Measuring current is necessary under two sets of circumstances.
The first case is where the sensor or signal source outputs a current,
rather than a voltage. In these circumstances, the current is an indicator of some other phenomenon to be measured, so it must be converted into appropriate engineering units. A pressure transducer with
420mA output is one example of this situation. A reading of 4mA corresponds to no pressure (zero pounds per square inch), while 20mA is
the transducers full-scale output in pounds per square inch.
The second application of current measurement is where the actual current reading is the parameter of interest, such as tracking the load
current drawn by a device over time. Here, its important to remember
that the current may be relatively high, which requires extra care in
selecting a dropping resistor.
8.1.1
Voltage Burden
The concept of voltage burden is important to achieving the best
results in a current measurement setup. Voltage burden is defined as
the voltage drop across the input of an ammeter when it is inserted into
a circuit. In Figure 8-1, the dropping resistor (R) and A/D voltage input
constitute an ammeter, and the current flow can be calculated from the
voltage drop across a resistor.
The resistor value will normally be selected to provide a voltage
drop corresponding to the A/D boards input range when the maximum anticipated current flows through the resistor. For example, a
20mA current produces a 10V drop across a 500 resistor. A 490
resistor will provide some safety margin in the measurement without a
significant loss of resolution. Note that the sensor or current source
must be capable of a minimum 1011V output to achieve the full voltage drop across the resistor. If the circuit is powered only by 6V, the circuit can not drive more than 12mA through the resistor.
The power dissipated by the resistor must also be considered. For a
20mA current through a 500 resistor, dissipation (P) can be calculated as:
P = I 2 R = (0.02)2 500 = 0.2W
This application could be handled with a 0.25 or 0.5 watt resistor.
SECTION 8
Cable
Resistance
Sensor
or
Signal
Source
Current
I
Dropping
Resistor
R
A/D Input
+
V
I=
V
R
Figure 8-1. Current measurement using a dropping resistor
In a case where a higher current2A, for exampleneeds to be

measured, the voltage drop across a 500 resistor would be 1000V, with
a power dissipation of 2000W. Clearly, a lower resistance value is
required. A 5 resistor would produce a voltage drop of 10V, but the
resistor would still dissipate 20W. A resistance value or 1, or even
0.5, would produce an easily measurable signal and result in much
lower power dissipation and voltage burden.
8.1.2
DMM vs. A/D Board for Current Measurements

A programmable DMM or external data acquisition system based on
DMM architecture (such as Keithleys Model 2700 or 2750) can measure
current directly. Making connections to the signal source is straightforward, and such meters are designed to impose a minimal voltage burden on the signal. They are also suitable for measuring currents up to
several amperes. The advantage of this method is that such meters
have built-in signal conditioning circuitry and can provide high channel counts (up to 200) at a low cost per channel.
A typical plug-in A/D card offers many voltage input channels, and
some include on-board sockets that allow dropping resistors to be
mounted on the board. The advantage of using an A/D board for current measurements is that the A/D board will typically run significantly faster than a DMM. However, currents generally have fewer high
frequency components, so speed may not be a primary concern. The
chief limitation of current measurement with an A/D card is that there
is no simple way to change the dropping resistor under program
control in order to change ranges.
RELATED TOPICS OF INTEREST
8.2
Connection Theory
Most electronic measurements, including those typical of data acquisition, require an electrical connection between the data acquisition
hardware and signal source. The validity of these measurements
depends largely on how well cables and connectors are matched to the
signals, and how carefully the measurement setup is constructed and
routed.
Data acquisition applications usually deal with DC signals in the
range of 1mV to 10V, but stand-alone instruments can be used to measure signals far above and below this range. While many concerns common to low level or high level measurements, such as triboelectric
effects, thermoelectric effects, and insulation properties, are less of an
issue in data acquisition applications, they are still critical concerns in
other measurement scenarios. Electromagnetic noise, cable length, and
cable capacitance can affect the quality of any type of measurement.
8.2.1
Cable Type
A particular type of wire or cables suitability for data acquisition
depends largely on the nature of the signal and the test environment.
When evaluating a cable for this application, consider these issues:
How much electrical noise exists in the environment? Noise
can be defined as any undesirable signal that is impressed upon
a signal of interest. Sources of electromagnetic noise include AC
power lines, motors and generators, transformers, fluorescent
lights, CRT displays, computers, radio transmitters, etc.
Depending on the nature of the signal and the noise, it may not
be possible to separate them once the signal has been acquired
What is the distance between the signal and data acquisition
input? Wire exhibits some electrical resistance, which depends
on the composition, length, and gauge of the wire. Resistance
increases with increasing wire length and with decreasing wire
diameter. This resistance becomes a part of the analog input circuit as shown in Figure 8-2. For most voltage-based analog
measurements, the source resistance of the signal will be low
(i.e., less than 100), while the input resistance of the analog
channel will be on the order of 10100M or higher. Therefore, a
few ohms resistance in interconnect cabling will not appreciably
affect measurements. However, a high cable resistance in conjunction with low A/D input resistance can result in a significant
voltage drop through the interconnect wiring. For data on the
resistances of common wire gauges, refer to Appendix C.
Is the data acquisition channel a single-ended or differential
input? Single-ended signals, i.e., those referenced to ground,
can be transmitted with two wires or with a shielded cable where
SECTION 8
Cable Effects
A/D Input
Sensor
Source
Resistance
Input Bias
Current
Input
Resistance
+
V
Figure 8-2. Circuit for voltage measurement
the shield is tied to ground. For differential signals, at least two

wires are needed to transmit the signal, which consists of a signal high and a signal low, neither of which is referenced to
ground. Two individual conductors will work, but a twisted pair
or shielded twisted pair provides greater noise immunity.
8.2.2
Conductors
The conductors used in shielded or unshielded cable can be solid or
stranded wire. Solid wire results in minimum signal attenuation, but
stranded conductors provide more flexibility. Often, conductors are
plated with silver or tinned with solder.
8.2.3
Shielding
Shielding is employed to reduce the amount of noise picked up by signal leads, but shielding can also be helpful in reducing signal radiation
from conductors carrying high frequency signals. Shielding can be
constructed with different types of wire braid or a combination of wire
braid and foil. Multi-layer or multi-braid shields are more effective
than single layer shields in attenuating signal pickup or radiation, but
can also make cables stiffer and more difficult to position.
Consider these tips when selecting shielded cable:

Depending on frequency, shielding may not attenuate noise
completely. Higher frequency signals are more difficult to attenuate than lower frequency signals.
Simple spiral wire wrap is the least effective type of shielding.
Tight braiding, double braiding, or braiding plus foil offer more
effective shielding.
Caustic atmospheres, moisture, etc. can change the effectiveness of shielding. In some cases, these contaminants can leach
into a cable and degrade the shielding far beneath the outer
insulating jacket.
Table 8-1 lists typical wire and cable types that may be encountered in data acquisition. These range from single, unshielded wires to
various types of shielded single or multi-conductor cables. Cable vendors normally supply specifications concerning the shielding properties of their cable products.
Table 8-1. Characteristics of hookup wire and cables
TYPE
ADVANTAGES
DISADVANTAGES
Single Conductor
Inexpensive.
Easy to connect.
Susceptible to noise pickup.
Twisted pair
Suitable for differential

Marginally more expensive
signals.
than single conductor.
Provides some protection
against noise pickup.
Single Conductor
with Shield (Coax)
Better protection against

noise pickup.
Higher cost.
Connection to equipment
more difficult.
Shielded Twisted
Pair or Shielded
Multi-conductor
Cabling
Best for differential

signals in noisy
environments.
Higher cost.
Connection to equipment
more difficult.
Ribbon Cable
Convenient for multiconductor hookup to

data acquisition cards.
Available as flat ribbon
cable or a series of
twisted pairs.
Can be more susceptible to

noise pickup and crosstalk.
May not be suitable for
higher speed signals.
More difficult to attach
connectors.
Insulation displacement
connectors may have lower
insulation ratings.
SECTION 8
8.2.4
Cable Capacitance
A typical signal behaves like a voltage source in series with a resistance.
Similarly, an analog instrument input resembles a meter with infinite
input resistance in parallel with the instruments actual input resistance.
During a measurement, the instrument input absorbs a small bias current that the source must be able to generate. The interconnect cabling
is an essential part of this circuit, and can introduce resistance, capacitance, and inductive effects that depend on length, gauge, composition,
routing, and environment. Figure 8-2 shows a dashed box around a portion of the signal path that represents the sum of these effects.
For high speed, rapidly changing signals, circuit inductance and
capacitance can be serious obstacles to measurement speed, even if
signal and instrument impedances are properly matched. Generally,
high impedance signals take longer to stabilize at the instrument
because the signals limited current requires more time to charge cable
capacitance.
SECTION 9
Application
Examples
9.1
Introduction
This section provides examples of real-world applications that employ
data acquisition products to monitor or control industrial processes.
These examples are intended simply to demonstrate what is possible,
rather than providing details such as programming examples and
wiring diagrams. For in-depth information, contact Keithley Instruments Applications Engineering department.
9.2
OEM/Factory Automation and Data Acquisition

The primary purpose of implementing data acquisition hardware in
OEM and factory automation systems is to capture process-specific
parameters, which are evaluated in a control processor. The control
processor generates signals to maintain or adjust the process based on
these inputs. Additionally, the acquired data can be used to provide
operator indications and feedback.
The types of transducers typically interfaced to data acquisition
systems include thermocouples, resistive temperature devices (RTDs),
thermistors, pressure sensors, strain gauges, and flow sensors. The outputs from these transducers are generally unipolar or bipolar voltages
from 5V to 10V or a current in the range of 420mA. These outputs can
be a linear representation of the input to the transducer, such as flow
or pressure, or they can be a non-linear representation, such as the
response of a thermocouple to temperature.
Typically, the systems control outputs are used to change the state
of some external device:
Actuating a relay
Increasing or decreasing flow or pressure by driving a solenoid
valve
Repositioning an object through the use of stepper motors
Regulating the temperature of an environment
Primarily, these output signals are a combination of digital TTL levels (0 to 5VDC) and analog levels (-10V to +10V, 420mA current, or
120220VAC). High voltage outputs will most likely be controlled using a
digital output that drives a relay or optically isolated accessory. However,
boards are available that can supply these signal levels directly.
The devices used to measure and generate these voltages and currents vary according to computer platform, and can include ISA, EISA,
and PCI plug-in hardware, IEEE-488 bus-based hardware, VXI/VME,
PLC, or compact PCI. These devices seem to be evolving into distributed I/O components connected via Ethernet or high speed serial and
parallel links. The fundamental purpose behind each of these platform
options is to convert real-world analog signals into digital signals via an
analog-to-digital converter or A/D. These digital signals can then be
SECTION 9
transferred to, and manipulated by, a computer. The control outputs

are then converted from digital signals to analog (D/A) and applied to
the external devices.
9.2.1
Design Considerations
It would be nearly impossible to characterize a typical OEM or factory automation application. Systems that can measure and control
almost every parameter in a process have been implemented throughout nearly every industry. The nature and physical environment of an
application will help determine its basic system design requirements.
For example, consider what would be involved in an application
designed to control a 2000-ton hydraulic press and its associated
robotic parts-handling equipment. In this situation, a high degree of
determinism (i.e., precisely known and controlled response times to
stimuli) is required, along with control signals and transducers that can
function in electrically noisy environments. If the system builder were
to implement a solution based on a PC platform, plug-in data acquisition boards, TTL inputs and outputs (I/O), and the Windows 98 operating system, its likelihood of success would be questionable. The
system requirements listed above would make a system based on
Programmable Logic Controllers (PLCs) with 115VAC control I/O much
more appropriate for this application. PLC processors are designed to
operate with very deterministic control loops, and the 115VAC control
signals are inherently immune to noisy environments.
Next, consider an application that involves measuring a number of
thermocouple inputs from a process every 10 seconds, generating outputs to maintain a specific temperature, routing unit-under-test (UUT)
signals, and measuring DC test parameters. In this case, a PC-based
system with a GPIB interface, a Keithley Model 2700 Multimeter/Data
Acquisition System would function well. Given that temperature is a
slow-moving parameter by nature, temperature control can easily be
maintained through PC-based I/O and control software. One measurement per second is typically sufficient. The time-critical UUT switching and measurements can be supported through the deterministic
control provided by the Model 2700.
To build such systems successfully, the designer is responsible for
determining the following requirements:
Critical process parameters
Measurement of process parameters
Generation of control algorithms
Generation of control signals
APPLICATION EXAMPLES
9.2.2
Measurement Integrity
It not unusual to obtain significantly different test results from an automated test system than from a manual system that uses the same
instrumentation. In other words, during the system design phase, an
engineer may obtain perfectly acceptable test results from a manual
test sequence on a bench-top system. However, the test results are significantly different once this system is transferred to the test rack,
where signal switching and routing are under computer control and
additional fixturing and longer signal paths are involved. Whats going
wrong?
In data acquisition and control systems, signal integrity depends
on many factors unrelated to the actual measurement device. The
measurement device might be an 8-digit multimeter capable of
nanovolt resolution. If the system switching and interconnects introduce many microvolts of thermal offset, the result will contain these
errors as well. Furthermore, if the signal from a D/A board is transmitted through many feet of cable and switching, the resulting signal at
the UUT may not be the value that was programmed.
So what are the primary areas of concern when trying to maintain
signal integrity? A variety of factors can contribute to measurement
error, including:
Thermal offsets
EMI/RFI
Grounding
Cable lengths/cable routing/switch speed
A/D input configuration
Thermal offsets are generated where two dissimilar metals are
joined. This junction can be at a screw terminal panel, quick disconnect panel, or even at relay contacts. When a system contains multiple
connections of this type, each at a different temperature, significant
measurement errors can be introduced. Errors of this type can often be
ignored if the signal of interest is a relatively high level voltage, such as
10V. However, if the source signal level is in the millivolt or microvolt
range, these offsets may introduce unacceptable levels of error.
System switching speeds can also introduce errors, depending on
the length and type of cables and interconnects being used. Even if a
switch can actuate in five milliseconds, it doesnt necessarily follow
that the signal is ready to be measured in that time. The actual time
constant of the signal path can be computed and compared to the
manufacturers specifications. This information must be considered
when programming measurement times. The actuation time of the
switching components may only be a few milliseconds. However, if the
time constant of the entire path is 100 milliseconds, a delay (typically
SECTION 9
equal to several time constants) must be programmed to occur

between the switch closure and the measurement.
The effects generated from ground loops can also contribute to
system measurement errors. The voltage potential between widely separated ground terminals within a system can be significant, and can
generate a current that flows through the measurement ground system. One proven approach for reducing the effects of ground loops is
to ground all equipment at a single point, and use isolated instruments
and sources when possible.
The input configuration of programmable A/D hardware can have
a significant effect on the acquired data. With some equipment, the
user can choose single-ended or differential inputs. Generally speaking, single-ended inputs perform well in a low noise environment and
when the signal levels are high (>1.0V). Differential inputs have greater
immunity to noise than single-ended ones, and should be selected
when measuring low-level signals.
Interference from EMI/RFI sources can also contribute to system
measurement error. Shielded cabling can be utilized to reduce the
effect of EMI/RFI radiation both internal and external to the system. A
wide variety of commercially available cabinets and enclosures incorporate extensive shielding.
9.3
Semiconductor CVD Application

The continuing demand for semiconductor devices that combine
increased density and higher speed has generated the need for increasingly complex manufacturing equipment to produce these devices.
Chemical Vapor Deposition (CVD), the process of using a controlled
chemical reaction to deposit layers on wafers, is widely used throughout the industry today.
A typical CVD system contains a reaction chamber, gas inlets, gas
outlets, a heater, an exhaust system, and a substrate loading and
unloading assembly. Systems can operate at or below atmospheric
pressure and employ temperatures greater than 300C. Clearly, if the
system is to function properly, many parameters must be measured
and evaluated, and control signals must generated. These parameters
include temperature, pressure, gas flow rate, and deposition time.
An industrial PC with a passive backplane configuration was
selected for this application. It houses the processor card, as well as
plug-in boards that perform the digital I/O, analog I/O, and
counter/timer functions. The Model KPCI-3108 and KPCI-3130 boards
are ideal for this application. Typically, the processor runs a real-time
operating system, permitting deterministic response to process data.
Multiple thermocouples mounted beneath the wafer monitor the
temperature within the chamber. The thermocouple cabling is routed
through a hermetically sealed interface and connected to a terminal

block assembly with a cold junction reference. The millivolt-level thermocouple signals are measured by the KPCI-3108 board and then converted to temperature, evaluated, and the heater outputs adjusted
based on the control algorithm. The heater control signals are digital
outputs that control relays that isolate the TTL signals from the high
voltage heater supply. Figure 9-1 illustrates this application.
Equipment Control Cabinet
Meaurement Signals For:

Temperature, Vacuum, Flow, Position, Speed
Process Vacuum Chamber
PC
I/O Chassis
Water Handling
Equipment
Servo Motor
KPCI-3108
Embedded Control Software

KPCI-3130
Figure 9-1.
Control Signals For:

Pumps, Valves, Relays, Motors,
Heaters, RF Power
Semiconductor fabrication equipment control
System pressure and vacuum levels are measured by the KPCI-3108

via pressure transducers, which may provide either amplified or millivolt signal levels. Unlike thermocouples, pressure transducers typically operate linearly, based on the full-scale rating of the device. The gas
flow through the inlet is monitored with a flow transducer that also
exhibits linear characteristics based on the full-scale ratings. These signals are routed to the A/D hardware for conversion to flow rate and
pressure readings. Consider the environment in the chamber when
choosing transducers to ensure reliable operation
The operation of vacuum pumps and valves is controlled with digital output hardware. The digital input and output signals, which may
be exposed to hazardous voltages and spikes, are isolated through
relays or optical isolation. Signal conditioning and isolation are accomplished with industry-standard solid-state I/O modules. The KPCI3130 analog output board generates discrete and variable control
voltages for devices such as motors, and also for device excitation.
9.4
Process Monitoring in a Nuclear Power Plant

Process monitoring is a crucial application for many manufacturers
because poorly controlled processes can produce defective products
SECTION 9
and waste resources. A PC-based process/machine monitoring system

provides an ideal solution for this application. It can collect data for
analysis by statistical process control applications to ensure the manufacturing process is operating within specifications and in compliance
with applicable governmental regulations.
PC-based systems are used widely by discrete, process, and batch
manufacturers because they offer a low cost solution. For highly regulated industries, such as nuclear power plants, PC-based systems offer
the added benefit of easy integration with existing processes and equipment without disturbing or modifying established control systems.
The challenge one nuclear power utility faced was the need to follow
equipment manufacturers preventive maintenance specifications strictly,
as required by the Nuclear Regulatory Commission (NRC). The company
also wanted the monitoring system to provide information in sufficient
detail to allow engineers to detect and predict diesel generator abnormalities before the equipment experienced any damage. The parameters
to be monitored included temperature, pressure, flow, and stress.
The power company wanted a system that was easy to use and versatile, so operators could acquire and analyze information gathered
from diesel generators located throughout the facility. Finally, the company needed a low cost, turnkey solution.
The solution to this application was a networked system using a
Keithley KPCI-1802HC plug-in data acquisition board in a PC (Figure
9-2). The KPCI-1802HC is a multifunction PCI board with 32 differential or 64 single-ended inputs, and two analog outputs. Analog inputs
feature gains of 1, 2, 4, and 8. Any single channel can be sampled at any
gain at up to 333ksamples/second, while multiple channels can be
sampled at aggregate rates up to 312.5ksamples/second.
The system used an industrial PC, a dual redundant power supply,
a network interface board, and Keithleys KPCI-1802HC data acquisition boards, all of which were integrated in a purged and cooled industrial enclosure. Each signal input to the monitoring system was
continuously logged by the remotely located data acquisition PC, then
transferred automatically over the plant LAN to a system manager PC.
The system tracked equipment performance automatically and
generated the summary reports needed to satisfy the OEMs and the
NRCs requirements. While the past practice had been to generate
reports from operator readings taken every 30 minutes, the new reports
could be based on up-to-the-minute information compiled from millions of samples. The reports were much more accurate, and allowed
plant personnel to capture and analyze any unexpected transients.
Today, the system monitors the diesel generators continuously
year-round and captures all the data from every run. An added advantage of this system architecture is that since the system uses standard,
Generator Signals
Coax Cable
Future Expansion
Communications Cable
System Layout A typical diesel

generator monitoring layout.
It illustrates the main components
that comprise the monitoring
system for an individual generator.
This system utilizes Keithleys
KPCI-1800 Series Data
Acquisition Boards.
KPCI-1802HC
Local Area Network

Network Interface Card
Diesel Manager PC
Rack-mounted Data Acquisition PC
Remote Equipment Cabinet
Figure 9-2.
Diesel monitoring in a nuclear power plant
off-the-shelf products, such as Keithley data acquisition boards, it can

be readily expanded incorporate new capabilities.
9.5
Tensile Test Stand Application

The increased use of rubber-based products across many industries
has created the need for detailed characterization of these products
prior to fabrication. This characterization process can include tensile,
adhesion, tear, and compression testing.
In this application, a sample is placed in a test chamber, where the
required environmental conditions are applied. A KCPI-3104 12-bit
A/D board is used to measure process parameters from thermocouples
(temperature), pressure transducers, and strain gauges. The application program converts the millivolt signals from the thermocouples
into temperature readings. Strain and pressure inputs are processed in
a similar manner, based on the functional specifications of the transducer. The application program evaluates these inputs, and uses a control algorithm to adjust the system temperature and force parameters.
System control outputs are generated using the KPCI-3104 boards
digital outputs. Some applications also implement variable motor con-
SECTION 9
trol with the boards analog output features. The input and output signals are optically isolated using industry-standard solid-state I/O modules. These modules can also be used to condition input and output
signals to usable levels.
9.6
Burn-In and Stress Testing of Electronic Devices

The rapid growth of the telecommunications, desktop computing, and
network server markets has created a burgeoning demand for switching power supplies and DC-to-DC converters. While these power supplies are relatively inexpensive, careful production testing is required
to maintain product quality.
Highly Accelerated Life Testing (HALT) and Highly Accelerated
Stress Screening (HASS) are common production burn-in procedures
for switching power supplies designed for computers and file servers.
Extended environmental testing is performed to ensure the product will
continue to function properly over its entire service life. It is not uncommon to age and monitor thousands of power supplies at one time.
The fundamental proposition underlying HALT/HASS testing is
that if a product and its manufacturing processes are properly
designed and verified with HASS, then production operations should
turn out reliable units. Consequently, if HASS detects units failing prematurely in production, this usually indicates improper manufacturing practices and/or random part failures.
Typically, the reliability of a population of products can be characterized by the bathtub-shaped curve shown in Figure 9-3. This curve
has three distinct failure rate regions. The first region is the infant
mortality section of the curve, which has a decreasing failure rate and
is associated with built-in, rather than designed-in, defects. These are
the types of defects often identified by HALT. The amount of time
required for a subsequent HASS test is determined by the width of the
Instantaneous
Failure Rate
Infant
Mortality
Useful
Life
Wearout
Log Time
Figure 9-3.
Product reliability curve
infant mortality region of the reliability curve. In general, the higher the
stress applied, the sooner the failures will occur, which narrows the
infant mortality region, and shortens the required HASS test period.
When designing this type of test system (Figure 9-4), the greatest
challenges are dealing with the high number of channels the system
must monitor and the test system surroundings. Large numbers of
switching power supplies can produce tremendous amounts of electrical noise, which can affect the test systems measurement performance
significantly.
Environmental Burn-in Chamber
+
Model 2700
w/ 7708 Modules
Thermocouples
Model 7002
w/ 7011 cards
KPCI-488
PIO-32/IO
To All
Instruments
GPIB
System Triggering &

Alarm Functions
Digital I/O
Figure 9-4.
Trigger Link
TestPoint
Model 2000
HI
LO
Model 7002
w/ 7011 cards
System
Cabling
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
PS
Power Supplies Under Test
Configuration of an 800-channel switching power supply burn-in

test system
The basic requirement for burn-in testing of power supplies is to

measure the voltage drop across a load resistor placed across the output of each switching power supply during the entire test cycle. Test
cycle duration can range from less than an hour to many days, depending on the manufacturers quality requirements.
A typical specification for a power supply is to output 5V with 10%
accuracy, which can be verified easily with a 6-digit DMM. The output is cycled every 15 seconds, and the DMM must make measurements on 800 channels during the 15 second on time. Setting the
integration rate or measurement time to be as fast as possible (NPLC =
0.01), disabling all filters, and using the Trigger Link feature found on
many Keithley instruments greatly simplifies measurement execution.
The Model 2700 Multimeter/Data Acquisition System is used to
verify the temperature profile of the environmental temperature
chamber independently. Plugging two Model 7700 20-channel differential multiplexer modules into the Model 2700 allows the system to
accommodate up to 40 thermocouples. Two Model 7708 40-channel
differential multiplexer modules can support up to 80 thermocouples.
SECTION 9
The PIO-32 I/O relay board within the PC is used for various triggering and alarm functions. The KPCI-488 is the GPIB controller board
that communicates with the instruments. This system has to be optimized by using synchronization and triggering techniques, and by taking advantage of the environmental noise rejection built into the
instruments.
9.7
Performance Characterization of Shock Absorbers

During R&D and production test processes, manufacturers perform
multiple tests to ensure devices meet the necessary specifications. In
this example, the device under test is a shock absorber. The objective is
to characterize shock absorber performance by plotting load vs. velocity, acceleration, linear position, or other parameters. This is a typical
low cost data acquisition test system that involves a low channel count
(<16 channels), high sampling rates, tight synchronization, triggering,
and a high level of integration. The test system must be cost-effective
because it will be duplicated on multiple stations. It must also be a fully
automated, turnkey system.
The KPCI-3102 board and MB signal conditioning modules are well
suited to this application. Figure 9-5 describes the data acquisition
system. The MB-05 is the signal conditioning accessory and the STA300 is the screw terminal accessory.
16-Bit Encoder
MB-05
Pressures,
RTD,
Load Cell,
Vibration
Figure 9-5.
STA-300
KPCI-3102
Trigger
Motion
Control
Shock absorber test system
As shown in Figure 9-5, a third-party motion control assembly triggers the data collection using a load cell, an RTD, multiple pressure sensors, accelerometers, and the binary output from a 16-bit encoder that
senses the linear position of the shock absorber. To synchronize the
analog inputs with the encoder, the plug-in board must include a digital channel in the scan list. The KPCI-3102 board provides this capability. The test also requires burst mode operation to minimize time skew
(~10s) between channels, so that the measured parameters are synchronized in the time domain and represent the true behavior of the
DUT. The A/D board should be able to achieve a sampling rate of up to
100kHz to capture vibration parameters, pressure, and valve flutter.
9.8
Instrument-Grade, Low Cost Analog Output Control

Today, the electronics community is moving away from traditional test
and measurement philosophies and toward a more hybrid approach.
Until relatively recently, if the application required generating or
measuring very precise data points, few solutions were available other
than bench-top instruments. However, the continuing evolution of
electronic components has led to the fabrication of precise voltage
sources and high resolution A/D and D/A hardware, creating a variety
of new solutions.
Series KPCI-3130 analog output boards offer the ability to generate
instrument quality analog voltages. Designed to the PCI-bus standard,
the KPCI-3130 Series provides a wide feature set that is typically not
seen in PC-based plug-in analog output hardware. Two unique characteristics of the board are the four-wire remote sense feature and the
four-quadrant source/sink operation. The following test examples
illustrate the functionality of these features.
9.8.1
Four-Wire Remote Sense Test Description

The ability of Series KPCI-3130 boards to perform four-wire remote
sense operations is a significant advantage over standard analog output hardware. This feature becomes critically important when transmitting control signals over long distances because significant voltage
drops can occur as the result of cable resistance, device interconnections, and terminations. Therefore, the programmed output value may
not be the voltage delivered to the device under test (DUT) when conventional sourcing methods are used.
Incorporating four-wire remote sense functionality makes it possible to connect sense leads directly to the input of the DUT. The sense
high and low connections allow the board to measure (sense) the actual voltage present at the DUT. The board can then adjust its output automatically, without user intervention, to ensure that the required voltage
is supplied, regardless of cable length and interconnection losses.
9.8.2
Four-Wire Remote Sense Test Procedure

In this test, a known voltage level is programmed from the KPCI-3130
and connected to the DUT using several different cable lengths. The
output voltage is programmed with ExceLINX software (Keithleys
Excel add-in), and the resistance of the cable measured using a
Keithley Model 2700 Multimeter/ Data Acquisition System that is set to
SECTION 9
four-wire ohms mode. For comparison purposes, the voltage at the

DUT is also measured without remote sense (two-wire mode).
The load is connected to the analog output in two-wire mode as
shown in Figure 9-6, and then connected in four-wire mode using the
sense line inputs as shown in Figure 9-7. The OUT0 and GND signals
are connected to pins 1 and 19 respectively; the S0H and S0L signals are
connected to pins 2 and 20.
Analog
OUT0
Analog
Output
DUT
Return
GND
Figure 9-6.
Conventional two-wire analog output connections
Analog
OUT0
Sense HI
S0H
Analog
Output
Adjust Vsource
(Feedback)
DUT
Sense LO
S0L
Return
GND
Figure 9-7.
Analog output with four-wire remote sense
The data in Table 9-1 was obtained using the KPCI-3130 boards
source feature and a Model 2700 Multimeter/Data Acquisition System.
The data clearly indicates that cabling and device interconnections can
be significant sources of error.
Table 9-1.
Lead resistance and programmed voltage in two-wire and

four-wire modes
Signal Path
(Per lead)
9.8.3
Resistance Programmed
( per lead)
Value (V)
Two-Wire
Mode
Measured
Value (V)
Four-Wire
Mode
Measured
Value (V)
10ft. cable
0.5266
8.000
7.999
8.000
100ft. cable
2.5
8.000
7.920
8.000
6 relays,
4 interconnections,
20ft. cable
4.06
8.000
7.360
8.000
Constant Current Source Test Description

Another unique feature of the KPCI-3130 is its ability to operate in all
four voltage/current quadrants (Figure 9-8). Four-quadrant operation
is the ability of a device to sink or source any combination of current or
voltage. The KPCI-3130 can perform these operations on each of its
eight channels simultaneously. Additionally, the robust nature of the
analog outputs permits sinking or sourcing up to 10V @ 20mA, without the need for external excitation. This feature is usually available
only on bench-top systems.
+20mA
IV
Sink
I
Source
10V
+10V
III
Source
II
Sink
20mA
Figure 9-8.
Four-quadrant operation
The examples outlined in Sections 9.8.4 and 9.8.5 illustrate the

process of configuring a constant current supply that will function in
Quadrants I and III. The key to generating the desired output current is
SECTION 9
selecting the proper shunt resistor. This shunt resistor is placed in

series with the load, and the analog output across the shunt is maintained at a level that results in the desired current passing through the
load/shunt pair.
9.8.4
Quadrant I, Resistive Load Test Procedure

This example explores an application with a test sample (resistive load)
that requires a constant current to be applied and controlled, independent of the load. Typically, the load resistance and current requirements are known. From this information, we can derive the proper
shunt resistor value using Ohms Law and the sum of the series resistance of the load/shunt combination.
This example assumes the load is 490, the desired current is
20mA, and the applied voltage is 10V. From Ohms Law, we can calculate Rshunt:
Vao = (Rtotal I)
= (Rshunt + Rload) 20mA
10V = (Rshunt + 490) 20mA
Therefore:
Rshunt + 490 = 10V/20mA
Rshunt + 490 = 500
Rshunt = 500 490 = 10
The analog output voltage will now be a function of the voltage
programmed, and sensed, across the shunt resistor. The acceptable
shunt voltage levels can once again be determined using Ohms Law
and the specified current:
Vmax shunt = 20mA Rshunt
= 20mA 10
= 0.2V
Therefore, if 0.2V is programmed, referenced to the shunt resistor,
the output across the series resistor network results in 10V @ 20mA.
Other current levels can be obtained in the same manner. For example,
if the required drive current were 10mA, the control voltage would be
0.1V.
For this test, connect the load and shunt resistors to the analog output and sense lines as shown in Figure 9-9. Connect the OUT0 and
GND signals to pins 1 and 19 respectively, and then connect the S0H
and S0L signals to pins 2 and 20.
Note that the shunt resistor must be connected on the ground side
of the load if the analog output is to function properly (Figure 9-9).
Failure to do so will result in unpredictable behavior from the analog
output circuit, and may damage the circuit under test.
Analog
OUT0
Sense HI
RLOAD
S0H
Analog
Output
Adjust Vsource
(Feedback)
RSHUNT
Sense LO
S0L
Return
GND
Figure 9-9.
9.8.5
Constant current configuration
Quadrant I and III, Battery Charge/Discharge Test Procedure

Lets consider another application that requires two-quadrant operation. This application involves battery charge/discharge cycle testing
using a constant current source. The first half of the test requires charging the battery to a specific voltage level (9.6V), then discharging the
battery to another predetermined level (1.0V).
The charge cycle occurs in Quadrant I, where both the voltage and
current are positive. The discharge cycle occurs in Quadrant III, where
the voltage and current are negative. A shunt resistor will be selected in
order to limit the charge and discharge currents to a predetermined
safe level (in this case, 10mA). The load resistance of the battery source
is quite low and is assumed to be negligible (less than 0.1). Therefore,
the following calculations assume the load resistance is less than 0.1
and therefore negligible. A 10 shunt resistor is used in the following
calculation.
The analog output voltage will now be a function of the voltage
programmed and sensed across the shunt resistor. The acceptable programmed shunt voltage levels can once again be determined using
Ohms Law and the required current:
Vmax shunt = 10mA Rshunt
= 10mA 10
= 0.1V
The shunt circuit analog output can now be set to 0.1V for the
charge cycle and 0.1V for the discharge cycle. The actual battery volt-
SECTION 9
Analog
OUT0
DUT
Sense HI
S0H
Analog
Output
Adjust Vsource
(Feedback)
RSHUNT
Sense LO
S0L
Return
GND
Figure 9-10. Constant current battery test configuration
age levels can be monitored using an analog input board, such as the
KPCI-3101. A user-written program would command and control the
specific voltage levels on the KPCI-3130.
For this test, connect the load and shunt resistor to the analog output and sense lines as shown in Figure 9-10. Connect the OUT0 and
GND signals to pins 1 and 19 respectively, and connect the S0H and S0L
signals to pins 2 and 20.
9.8.6
Source/Sink Test Description

Finally, we discuss the steps necessary to perform a battery charge/discharge cycle test in constant voltage operating mode. Unlike the previous example, the KPCI-3130 both sources voltage and sinks current.
9.8.7
Quadrant I and II, Battery Charge/Discharge Test Procedure

This test involves charging the battery to a specific voltage level, then
discharging it to another predetermined level. The charge cycle occurs
in Quadrant I, where both the voltage and current are positive. The discharge cycle occurs in Quadrant II, where the voltage is positive and
the current is negative. A series resistor must be selected to limit current. The resistor value is based on the voltage and current specifications of the DUT. The load resistance of the battery source is quite low.
Therefore, the following calculations assume load resistance is less
than 0.1 and therefore negligible.
The analog output voltage will be a function of the voltage programmed and sensed across the series combination of the battery and
resistor. This example assumes the maximum charge current is 10mA,
Analog
OUT0
Sense HI
S0H
Analog
Output
DUT
Adjust Vsource
(Feedback)
Sense LO
RSERIES
S0L
Return
GND
Figure 9-11. Constant voltage battery test configuration
the maximum voltage is 9V, and the minimum voltage is 3V. A suitable
series resistance can now be selected based on the test requirements:
Rseries = Vchange/Imax
= (9.0 3.0)/10mA
= 600
The analog output can now be set to 9.0V for the charge cycle and
3.0V for the discharge cycle. The actual voltage levels can be monitored
using an analog input board, such as the KPCI-3101, and the test cycle
can be controlled via the user program.
Note that, unlike the constant current example outlined previously, the exact current level sourced and sunk by the KPCI-3130 will vary
with the level of charge of the DUT (a battery in this case). Also, the current flow will be reduced as the potential of the DUT approaches that
of the programmed analog output. Therefore, it is critical to select the
correct series resistance to limit the maximum current from the analog
outputs. Failure to do so can result in damage to the DUT.
For this test, connect the load and shunt resistor to the analog output and sense lines as shown in Figure 9-11. Connect the OUT0 and
GND signals to pins 1 and 19 respectively, and connect the S0H and S0L
signals to pins 2 and 20.
Series KPCI-3130 universal analog output boards offer a versatile
feature set, which makes them appropriate for a wide range of control
applications. Furthermore, the four-wire sense function ensures that
the DUT is actually being controlled at the voltage level that was pro-
SECTION 9
grammed; this was evident in the first test example. This four-wire
sense feature is only available from Keithleys analog output boards.
Additionally, the four-quadrant operation and the 20mA sink/source
current capability can simplify test setup and reduce equipment
requirements.
APPENDIX A
Selection Guide for

Plug-In Boards vs.
External Data Acquisition
Instruments
APPENDIX A
Sampling (successive approximation).

High speed. >1kHz.
Moderate. Approx. 100V.
Moderate (8-16 bit).
NIST traceable.
Solid-state. High speed.
Data throughput, measurement to user
Measurement accuracy
Resolution/Sensitivity
Calibration
Multiplexing/Switching
Low to moderate.
10V.
Additional modules required.
Required. Included in software.
Required. Included in software.
Noise immunity
Common mode voltage
Signal conditioning (res., TC, current, freq.)
Scaling/Engineering units/Limits
Digital filtering
Tightly coupled.
Highly dependent. Device drivers required. Independent. Standard bus (GPIB, RS-232).
High. <1ms.
High. Fast response-time <1ms.
System software dependence
Application program response time
Control capabilities (analog out, digital out)
Moderate. Response time >1ms.
Moderate, >1ms.
Loosely coupled.
PCI/ISA/PCMCIA.
Computer dependence/coupling
GPIB, RS-232.
Built in.
Built in.
Built in conditioning hardware.
Up to 1000V.
High.
Built in, up to 1000V.
Mechanical. High isolation and accuracy.
NIST traceable.
High (22-28 bit).
High. Approx. 1V.
Low. <1kHz.
Integrating (power line cycle integration).
Low. <1kHz.
External Instruments
Communication bus
SOFTWARE
None. Additional modules required.
Electrical isolation to earth ground
SIGNAL CONDITIONING
High. >1kHz.
A/D technology
Plug-in Boards
Sampling rate
MEASUREMENT
FEATURES/CAPABILITIES
SELECTION GUIDE FOR PLUG-IN BOARDS VS.

EXTERNAL DATA ACQUISITION INSTRUMENTS
High.
High. A la carte.
None. Plugs into the computer chassis.
High. Ideal for OEM integration.
Modular/Customizable
Rack space
Compactness, embedding capabilities
Moderate. Requires software.

None. Requires GUI software.
Component level. Integration required.
Low. <$1000
Ease of use
Front panel display
Usage
Cost
OTHER
Limited. PC slot dependent.
Multifunctionality (AI, AO, DIO, CT)
Plug-in Boards
Channel expandability
FORM FACTOR
Moderate. >$1000.
System level.
Pushbutton display - troubleshooting.
High. Pushbutton and display interface.
Moderate, stand-alone.
Rack/stack, benchtop.
Moderate. All functions built in.
Moderate.
Almost unlimited. Daisy-chained.
External Instruments
FEATURES/CAPABILITIES
APPENDIX B
Glossary
2S COMPLEMENT. A 2s complement is derived by reversing

the digits in a binary number
(changing 1s to 0s and 0s to
1s) and adding 1 to the result.
When 2s complements are
used to represent negative
numbers, the most significant
(leftmost) digit is always 1.
A/D CONVERTER. See ANALOG-TODIGITAL CONVERTER.
A/D. Abbreviation for analog-todigital. Also See ANALOG-TODIGITAL CONVERTER.
ABOUT-TRIGGER ACQUISITION MODE.
A data acquisition triggering
mode in which the data
acquisition is started by an
internal or external trigger,
then continues after a second
trigger event, thus retaining
data both before and after the
second trigger. See also
TRIGGER (DATA ACQUISITION)
and TRIGGER MODES.
ABSOLUTE ACCURACY. A measure of
the uncertainty of an instrument reading compared to
that of a primary standard
having absolute traceability to
the National Institute of
Standards and Technology,
expressed in ppm. Accuracy is
often separated into gain and
offset terms. See also RATED
ACCURACY.
ACQUISITION RATE. The rate at
which the board acquires
analog or digital data from an
external signal input to the
board. In the case of a scanning A/D, the aggregate conversion rate for all channels.
ACQUISITION TIME. In general, the

minimum amount of time
that an analog signal must be
present at the input of an A/D
converter for a conversion to
take place. For a sampling
A/D converter, the acquisition
time specifies the time that
the analog signal must be present at the SH front end
before the A/D conversion
starts. Also referred to as
Aperture Time and Sample
Window. See also ANALOG-TODIGITAL CONVERTER, SAMPLING
ANALOG-TO-DIGITAL CONVERTER,
and SAMPLE-AND-HOLD.
ACTIVE EDGE. The definition of
which edge of a trigger signal
(positive, rising edge, or the
negative, falling edge) will be
used to initiate an action.
Data acquisition systems typically may be configured to
specify either edge as a trigger
event. See also TRIGGER (DATA
ACQUISITION) and TRIGGER
POLARITY.
ADC. See ANALOG-TO-DIGITAL
CONVERTER.
ADDRESS. A number specifying a
location in memory where
data is stored.
ADMITTANCE. The reciprocal of the
impedance. The admittance is
the complex ratio of the current flowing through divided
by the voltage across a device,
circuit element, or network.
ALIASING. False artifacts produced
in data when sampling rate is
less than twice the input signals highest frequency content.
APPENDIX B
ANALOG OUTPUT. An output that

provides an analog signal
derived from the input signal
or digital information within
an instrument.
APERTURE JITTER. The short-term

variation of aperture delay.
Also called Aperture
Uncertainty. See also
APERTURE DELAY.
ANALOG RAMP. A voltage output of

constant slope, dV/dt
(volts/second).
APERTURE TIME. The time interval

during which an amplifier or
measuring instrument
acquires a sample of the signal. Also known as the sample
window. See also ACQUISITION
TIME.
ANALOG TRIGGER. An event that

occurs at a user-selected
point on an analog input signal. The polarity, sensitivity,
and hysteresis of the analog
trigger can often be programmed. See also TRIGGER,
TRIGGER CONDITIONS, TRIGGER
HYSTERESIS, TRIGGER MODE,
TRIGGER POLARITY, and TRIGGER
SENSITIVITY.
ANALOG-TO-DIGITAL CONVERTER. An
electronic device that converts an analog voltage to a
digital value. All digital instruments use analog-to-digital
converters to convert the
input signals into digital
information. Also called an
A/D converter or an ADC.
ANGSTROM. A unit of length equal
to 1010 meters. Thus, there
are ten angstroms to one
nanometer (nm).
APERTURE DELAY. The time delay
between when an analog-todigital converter receives a
conversion command and
when it starts the conversion
process. See also ANALOG-TODIGITAL CONVERTER, APERTURE
JITTER.
GLOSSARY
APERTURE UNCERTAINTY. See

APERTURE JITTER.
API. See APPLICATION PROGRAMMING
INTERFACE.
APPLICATION PROGRAM. A computer
program used to perform a
particular kind of work, such as
data acquisition. Examples of
application programs include
high-level packages such as
TestPoint or user programs
written using function call
drivers or low-level calls.
APPLICATION PROGRAMMING
INTERFACE. A set of routines
used by an application program to direct the performance of a procedure by the
computers operating system.
ASYNCHRONOUS. In hardware, an
unsynchronized event that
occurs independent of other
events. In software, a function
that begins an operation and
returns to the calling program
prior to the completion of the
operation.
AUTOPOLARITY. The ability of an
instrument to measure and
display an input of either
polarity without switching the
input leads.
AUTORANGING TIME. For instruments with autoranging capability, the time interval
between application of a step
input signal and its display,
including the time for determining and changing to the
correct range.
AUTORANGING. The ability of an
instrument to switch among
ranges automatically. The
ranges are usually in decade
steps.
AVERAGE RESPONDING. A measurement where the displayed
value is proportional to the
average of the absolute values
of all input waveforms within
a specified frequency range. It
is calibrated in the rms value
of a sine wave.
BACKGROUND TASK. An operation
that can take place while
another program or processing routine is running without
apparent interruption to that
program or routine; for example, an interrupt or DMA
operation.
BANDWIDTH (DATA ACQUISITION). The
range of frequencies that can
be switched, conducted, or
amplified within certain limits. Under given load conditions, bandwidth is defined by
the 3dB (half-power) points.
Also, the highest frequency
signal component that can
pass through an amplifier or
filter without being
attenuated.
BASE ADDRESS. An I/O address that

is the starting address for programmable registers. All subsequent registers are accessed
by adding to the base address.
BETA. The ratio of the collector
current to the base current of
a bipolar transistor, commonly referred to as either the
common-emitter current gain
or the current amplification
factor.
BIAS CURRENT (IN DIFFERENTIAL
AMPLIFIER). A small current
drawn through an input
terminal of a differential
amplifier.
BIAS VOLTAGE. A voltage applied to
a circuit or device to establish
a reference level or operating
point of the device during
testing.
BIPOLAR. An analog signal range
that includes both positive
and negative values.
BREAK-BEFORE-MAKE.
Disconnecting the present
circuit before connecting a
new circuit.
BUFFER MEMORY. Temporary storage area for acquired or generated data. See also LOCAL
BUFFER.
BURN-IN. The operation of items
prior to their ultimate application, intended to stabilize
their characteristics and identify early failures.
BURST CLOCK FREQUENCY. See BURST
CLOCK RATE.
BANK. A group of relays with a

common connection for scanning or multiplexing applications.
APPENDIX B
BURST CLOCK RATE. The rate at

which timing pulses are emitted from a pacer clock. See
also BURST CLOCK, BURST
CONVERSION MODE.
BURST CLOCK. For a data acquisition board operating in the
burst mode, a pulse-emitting
circuit that determines the
analog data conversion rate.
See also BURST CONVERSION
MODE and CONVERSION RATE.
BURST CONVERSION MODE. A data
acquisition mode in which a
group of analog input channels are scanned at a rate
determined by the pacer clock
and each channel within the
group is converted at a higher
rate determined by the burst
clock. This mode minimizes
the skew between channels.
See also BURST CLOCK and
PACER CLOCK.
BUS MASTERING. On a microcomputer bus such as the PCI bus,
the ability of an expansion
board to take control of the
bus and transfer data to
memory at high speed, independently of the CPU.
Replaces direct memory
access (DMA).
BUS. An interconnection system
that allows each part of a
computer to communicate
with the other parts.
CAPACITANCE. In a capacitor or
system of conductors and
dielectrics, the property that
permits the storage of electrically separated charges when
potential differences exist
between the conductors.
Capacitance is related to
charge and voltage as follows:
C = Q/V, where C is the capacitance in farads, Q is the
charge in coulombs, and V is
the voltage in volts.
CARRY CURRENT. The maximum
continuous current of closed
relay contacts. Most relays are
rated higher for carry current
than switched current. (Heat
is generated by I2R losses for
carry current and I2R losses
plus arcing for switched current.) See also SWITCHED
CURRENT.
CHANNEL (SWITCHING). One of several signal paths on a switching card. For scanner or multiplex cards, the channel is
used as a switched input in
measuring circuits or as a
switched output in sourcing
circuits. For switch cards,
each channels signal paths
are independent of other
channels. For matrix cards, a
channel is established by the
actuation of a relay at a row
and column crosspoint. See
also PATH.
BYTE. A group of eight bits.
GLOSSARY
CHANNEL ISOLATION. On a switching card, the isolation from

signal high and low of one
channel to signal high and
low of any other channel (or
the output on switch or scanner cards). Specified as resistance and capacitance, except
for RF cards (decibels and frequency range). See also PATH
ISOLATION.
CHANNEL. On a data acquisition
board, one of several input or
output paths on the board.
Multiple analog input channels are commonly connected
to one analog-to-digital converter, one at a time, using a
multiplexer. See also
MULTIPLEXER and ANALOG-TODIGITAL CONVERTER.
CHANNEL-GAIN QUEUE. A userdefined scan sequence in a
data acquisition device. It
specifies both the position in
the sequence and the gain at
which an analog input channel is scanned, or in some
cases, the output range at
which an analog output channel is updated. It can also
specify whether the input or
output mode is bipolar or
unipolar and whether the
input mode is single-ended or
differential. See also CHANNEL
and SCAN (DATA ACQUISITION).
CHARGE COUPLED DEVICE (CCD). A
semiconductor device, often
used for sensing light, which
operates by storing charge on
capacitors and selectively
moving that charge through
the device by manipulating
voltages on its electrodes.
CMRR. See COMMON MODE

REJECTION RATIO (CMRR).
COIL RESISTANCE. A nominal value
of the resistance of a relay coil
winding at a specified ambient temperature.
COLD JUNCTION. The junction in a
thermocouple circuit that is
held at a stable, known temperature. Also known as a
reference junction.
COLD JUNCTION COMPENSATION
(CJC). A method of compensating for ambient temperature variations in thermocouple circuits by using either a
physical cold junction or by
measuring ambient temperature to adjust collected values.
COLD SWITCHING. Closing the relay
contacts before applying voltage and current and removing
voltage and current before
opening the contacts.
(Contacts do not make or
break current.) See also DRY
CIRCUIT SWITCHING.
COLUMN. As viewed on the
schematic of a matrix relay
card, the vertical signal lines
or a vertical group of relays.
COMMON MODE INPUT ISOLATION.
On a switching card, the isolation from signal high and low
to guard (or shield) for a
three-pole circuit, or from signal high and low to chassis
ground for a two-pole circuit.
Specified as resistance and
capacitance.
APPENDIX B
COMMON MODE REJECTION RATIO

(CMRR). The ability of an
instrument to reject interference from a common voltage
at its input terminals with
respect to ground. Usually
expressed in decibels at a frequency.
COMMON MODE VOLTAGE. A voltage
between input low and chassis ground of an instrument. A
differential input sees the
common mode voltage as a
common component of the
voltages at both the inputhigh and input-low terminals
and rejects all but a small
fraction. See also COMMON
MODE REJECTION RATIO (CMRR)
and DIFFERENTIAL INPUT.
COMPLIANCE CURRENT. The maximum output current of a constant voltage source. Also
known as current limit.
COMPLIANCE VOLTAGE. The maximum output voltage of a constant current source. Also
known as voltage limit.
CONDUCTANCE (G). The ability to
conduct electricity. Defined
by G = 1/R = I/V, where G is
the conductance in Siemens, I
is the current in Amps, and V
is the voltage in Volts.
CONNECTION PATH. The cables,
connectors, switch cards, etc.
between a device under test
(DUT) and test instrumentation. Its major parts are the
conductors making the connection and the insulators
isolating the conductors from
the rest of the world.
GLOSSARY
CONTACT BOUNCE. The intermittent and undesired opening of

relay contacts during closure,
or closing of relay contacts
during opening.
CONTACT LIFE. The maximum
number of expected closures
before failure. Life is dependent on the switched voltage,
current, and power. Failure is
usually when the contact
resistance exceeds an end of
life value.
CONTACT OFFSET VOLTAGE. See
CONTACT POTENTIAL.
CONTACT POTENTIAL. A voltage produced between contact terminals due to the temperature
gradient across the relay contacts (typically caused by
power dissipated by the energized coil.), and the reed-toterminal junctions of dissimilar metals. Also known as
contact offset voltage. See
THERMOELECTRIC VOLTAGE.
Contact Rating. The voltage, current, and power capacities of
relay contacts under specified
environmental conditions. See
CARRY CURRENT and SWITCHED
CURRENT.
CONTACT RESISTANCE. For a relay,
the resistance in ohms across
closed contacts. For a Keithley
switching card, also includes
the tape resistance and connector terminal resistance.
See also PATH RESISTANCE.
CONVERSION RATE. The rate at
which sampled analog data is
converted to digital data or
digital data is converted to
analog data.
CONVERSION TIME. The time

required to complete an analog-to-digital or digital-toanalog conversion.
D/A. Abbreviation for digital-toanalog.
CONVERSION. A process where a

signal is changed from an
analog-to-digital (A-D) representation, or digital-to-analog
(D-A).
DARLINGTON. A high-gain current

amplifier composed of two
cascaded bipolar transistors,
typically integrated in a single
package.
COULOMB. The unit of electric

charge, defined as the amount
of charge accumulated in one
second by a current of one
ampere. One coulomb represents the charge on approximately 6.24 1018 electrons.
Named for French physicist
Charles-Augustin de Coulomb
(1736-1806).
DAS. Data Acquisition System.
COUNTER/TIMER. A circuit that

counts pulses or measures
pulse duration.
CREST FACTOR. The ratio of the
peak value to the root-meansquare (rms) value of a waveform.
CROSSPOINT. The intersecting
point of a column and row in
a relay matrix. Specified as
(column,row) or (row,column). A crosspoint generally
consists of one or more poles
of Form A (normally open)
relay switching.
CROSSTALK. The coupling of a signal from one input to another
(or from one channel to
another or to the output) by
conduction or radiation.
Crosstalk is expressed in decibels at a specified load and up
to a specific frequency.
D/A CONVERTER. See DIGITAL-TOANALOG CONVERTER.
DAC. See DIGITAL-TO-ANALOG

CONVERTER.
DATA TRANSFER. Refers to the way

data is transferred to and
from memory, such as programmed I/O or DMA mode.
DDE. See DYNAMIC DATA EXCHANGE.
DEVICE HANDLE. A name that
uniquely identifies a hardware
device in an application program.
DIELECTRIC. An insulating layer. A
material that has high resistance. This term is usually
used when the insulating
layer separates the plates of a
capacitor.
DIFFERENTIAL AMPLIFIER. An analog
input circuit that measures
the difference between the
voltages at two input terminals input high and input
low, each of which is referenced to a common ground. A
differential input rejects the
common mode voltagethe
common voltage relative to
ground, as measured at the
input low terminalto an
extent limited by the common-mode rejection ratio of
the circuit. See also COMMON
MODE VOLTAGE, COMMON-MODE
TEJECTION RATIO (CMRR), and
SINGLE-ENDED INPUT.
APPENDIX B
DIFFERENTIAL INPUT. The condition

where the low terminal of a
two-terminal instrument is
connected to a reference
point which is not necessarily
a power line common, earth
ground, or circuit common.
Typically, a data acquisition
board with multiple differential input channels may have
each low terminal at a different potential. See also SINGLEENDED INPUT.
DIFFERENTIAL INPUT ISOLATION. On
a switching card, the isolation
from signal high to low.
capacitance.
DIFFERENTIAL LINEARITY. See
DIFFERENTIAL NONLINEARITY
(DNL).
DIFFERENTIAL NONLINEARITY (DNL).
The maximum deviation of a
real digitized step width or
height from the ideal digitized
step width or height. The
input range of a data acquisition board is divided into a
series of discrete steps, each
step ideally having a height of
one least significant bit (LSB).
See also LEAST SIGNIFICANT BIT
(LSB), ANALOG-TO-DIGITAL
CONVERTER, and QUANTIZATION.
DIGITAL I/O. Abbreviation for digital input/output.
DIGITAL LINES/PORTS/BITS/
CHANNELS. In hardware, a digital line is physical hardware
connection to a pin with a
digital signal. A digital port is
a physical grouping of digital
lines. In software, a digital bit
(1 or 0) is a logical representation of a digital line. A digital
channel is a logical grouping
of digital bits.
DIGITAL TRIGGER. An event that
occurs at a user-selected
point on a digital input signal.
The polarity and sensitivity of
the digital trigger can often be
programmed. See also
TRIGGER, TRIGGER CONDITIONS,
TRIGGER POLARITY, and TRIGGER
SENSITIVITY.
DIGITAL-TO-ANALOG CONVERTER. A
device that translates digital
data to an analog signal. A
digital-to-analog converter
takes a succession of discrete
digital values as input and
creates an analog signal
whose amplitude, moment by
moment, corresponds to each
digital value. Compare to
ANALOG-TO-DIGITAL CONVERTER.
DIRECT DIGITAL SYNTHESIS. A technique for signal generation
where the signal is directly
synthesized using only digital
techniques. This technique
generates very precise waveforms, even at low frequencies. Waveforms with correct
phase and frequency are
obtained immediately after a
shift to a new frequency.
DIRECT MEMORY ACCESS (DMA). See
DMA (DIRECT MEMORY ACCESS)
MODE.
GLOSSARY
DISCRETE DEVICE. A class of electronic components that contain one active element, such
as a transistor or diode.
However, hybrids, optoelectronic devices, and intelligent
discretes may contain more
than one active element.
DLL. See DYNAMIC LINK LIBRARY.
CHANNELS. ISA bus PCs offer
eight parallel channels for
DMA mode data transfers. A
number of these are reserved
for exclusive use by the computer. The remaining channels are available for use by
user-supplied I/O options,
such as plug-in data acquisition boards. Also called DMA
levels.
LEVELS. See DMA (DIRECT
MEMORY ACCESS) CHANNELS.
MODE. A mode in which data
transfers directly between an
I/O device and computer
memory, bypassing the CPU.
Most commonly, DMA mode
refers to data transfers across
the ISA bus, using special circuitry on the computer motherboard. In the most general
sense, PCI bus mastering is a
DMA mode. See also BUS
MASTERING and OPERATION
MODES.
DMM. An electronic instrument

that measures voltage, current, resistance, or other electrical parameters by converting the analog signal to digital
information and display. The
typical five-function DMM
measures DC volts, DC amps,
AC volts, AC amps, and resistance.
DNL. See DIFFERENTIAL
NONLINEARITY.
DRAM. Dynamic Random Access
Memory. A semiconductor
read/write memory chip, in
which the presence or
absence of a capacitive charge
represents the state of a binary storage element (zero or
one). The charge must be
periodically refreshed.
DRAIN TERMINAL. Along with the
gate and source, one of the
three terminals of a fieldeffect transistor (FET). The
low impedance path through
a turned-on FET is through
the Source and Drain terminals. The Gate is a high
impedance, voltagecontrolled input.
DRIFT. A gradual change of a
reading or an amplifier output
over time with no changes in
the input signal or operating
conditions.
DRIVER. Software that controls a
specific hardware device, such
as a data acquisition board.
APPENDIX B
DRY CIRCUIT SWITCHING. Switching

below specified levels of voltage (e.g., 20mV) and current
to minimize any physical and
electrical changes in the contact junction. See also COLD
SWITCHING.
DRY REED RELAY. A glass-enclosed,
hermetically sealed, magnetically actuated contact. No
mercury or other wetting
material is used.
DSP. Abbreviation for Digital
Signal Processing.
DUTY RATIO. The ratio of pulse
width to repetition period.
Also known as Duty Cycle.
DYNAMIC DATA EXCHANGE (DDE). A
Microsoft Windows standard
mechanism for communication between programs. It
allows your application to
send and share data with
other applications such as
spreadsheets.
DYNAMIC LINK LIBRARY (DLL). A
software module in Microsoft
Windows containing executable code and data that
can be called or used by
Windows applications or
other DLLs. DLL functions
and data are loaded and
linked at run time when they
are referenced by a Windows
application or other DLLs.
GLOSSARY
EEPROM. Electrically Erasable

Programmable Read-Only
Memory. Similar to PROM,
but with the capability of
selective erasure of information through special electrical
stimulus. Information stored
in EEPROM chips is retained
when the power is turned off.
EISA BUS. Extended Industry
Standard Architecture. A 32bit wide, upwards-compatible
extension of the 16-bit wide
ISA bus.
ELECTROMETER. A highly refined
DC multimeter. When compared with a digital multimeter, an electrometer is characterized by higher input resistance and greater sensitivity.
It can also have functions not
generally available on DMMs
(e.g., measuring electrical
charge, sourcing voltage).
ENOB. Effective Number Of Bits.
EPROM. Electrically Erasable
Programmable Read-Only
Memory. See EEPROM.
EPUT. Events Per Unit Time.
ERROR. The deviation (difference
or ratio) of a measurement
from its true value.
EXPANSION SLOT. A socket in a
computer designed to hold
expansion boards and connect them to the system bus.
EXTERNAL PACER CLOCK SOURCE. A

source of pulses that is connected externally to a data
acquisition board and is used
to pace or time events such as
analog-to-digital conversions,
digital-to-analog conversions,
data sampling, interrupt generation, digital I/O transfers,
etc.
EXTERNAL TRIGGER. An analog or
digital hardware event from
an external source that starts
an operation. See also
INTERNAL TRIGGER.
FALL TIME. The time required for a
signal to change from a large
percentage (usually 90%) to a
small percentage (usually
10%) of its peak-to-peak
amplitude. See also RISE TIME.
FIFO. First-In/First-Out memory
buffer. The first data into the
buffer is the first data out of
the buffer. On a data acquisition board, a FIFO allows data
collection to continue while
the board waits for data
transfer access to the host
computer.
FIREWIRE. A communication standard and external bus, also
designated IEEE-1394, that
supports Plug-and-Play, hot
plugging, and data transfer
rates of up to 400Mbps.
FLASH MEMORY. It is a non-volatile
memory technique with fast
access times; rewriteable
many times, and uses a block
erase technique (as opposed
to EEPROM, which erases one
bit at a time).
FLOATING. The condition where a

common mode voltage exists
between an earth ground and
the instrument or circuit of
interest. (Low of circuit is not
at earth potential.)
FOREGROUND TASK. An operation,
such as those that occur in
the single or synchronous
mode, that cannot take place
while another program or
routine is running.
FOUR-TERMINAL RESISTANCE
MEASUREMENT. A measurement
in which two leads are used to
supply current to the
unknown and two different
leads are used to sense the
voltage drop across the resistance.
FPGA. See FIELD-PROGRAMMABLE
GATE ARRAY (FPGA).
FRAME. A data structure that consists of one or more elements
corresponding to an operations defining attributes.
FUNCTION CALL DRIVER. A type of
Keithley board driver that
provides a high-level alternative to register-level programming.
GAIN. The factor by which an
incoming signal is multiplied
by an amplifier.
GATE (SIGNAL). A signal that in the
active state enables an operation and in the inactive state
inhibits the operation.
APPENDIX B
GATE TERMINAL. Along with the

source and drain, one of the
the Source and Drain terminals. The Gate is a high
GUARDING. A technique that

reduces leakage errors and
decreases response time.
Consists of a guard conductor
driven by a low impedance
source surrounding the lead
of a high impedance signal.
The guard voltage is kept at or
near the potential of the
signal.
GLITCH ENERGY. A measure of the

energy of an unwanted transient superimposed on the
output of a digital-to-analog
converter. A simple figure of
merit is an integral of the
transient voltage with time.
Also called glitch charge or
glitch impulse.
HARDWARE. The physical parts of a

computer-controlled system,
such as circuit boards, chassis, peripheral devices, cables,
etc.
GPIB. Abbreviation for General

Purpose Interface Bus, also
referred to as the IEEE-488
bus. It is a standard for parallel interfaces.
GROUND LOOP. A current loop created when a signal source and
a signal measurement device
are grounded at two separate
points on a ground bus
through which noise currents
and/or currents from other
devices flow. These currents
generate voltage drops
between the two ground connection points, which can
cause errors and noise in the
signal measurement.
GROUND. A common reference
point for an electrical system.
GLOSSARY
HARMONICS. Signal components

in a waveform that occur at
integer multiples of the
fundamental frequency. In
electronic circuits, harmonics
are usually caused by
nonlinearity.
HEAT SINK. A part used to absorb
heat.
HOT JUNCTION. The junction of
two dissimilar metals in a
thermocouple circuit that is
used to measure an unknown
temperature. Also known as
measurement junction.
IEEE. Abbreviation for Institute of
Electrical and Electronics
Engineers.
IEEE-488. See GPIB.
IMPEDANCE. The reciprocal of
admittance. Admittance is the
complex ratio of the voltage
across divided by the current
flowing through a device, circuit element, or network.
INPUT BIAS CURRENT. The current

that flows at the input of an
analog measurement circuit
due to internal circuitry and
bias voltage. Also, at conditions of zero input signal and
offset voltage, the current that
must be supplied to the
input-high measuring terminal to reduce the output indication to zero. The input bias
current is drawn through the
source resistance of a signal
source. Therefore, in critical
and/or low-level measurements, bias current compensation or attention to source
resistance may be required to
minimize errors.
INPUT IMPEDANCE. The shunt
resistance and capacitance
(or inductance) as measured
at the input terminals, not
including effects of input bias
or offset currents.
INPUT ISOLATION. On a switching
card, the isolation between
signal high to low (or guard)
for a two-pole circuit.
capacitance.
INPUT OFFSET CURRENT. The difference between the two currents that must be supplied to
the input measuring terminals of a differential instrument to reduce the output
indication to zero (with zero
input voltage and zero offset
voltage).
INPUT OFFSET VOLTAGE. The voltage

that must be applied directly
between the input measuring
terminals, with bias current
supplied by a resistance path,
to reduce the output indication to zero.
INPUT/OUTPUT (I/O). The process
of transferring data to and
from a computer-controlled
system using its communication channels, operator interface devices, data acquisition
devices, or control interfaces.
Also refers to the electrical
inputs and outputs for data
signals.
INPUT/OUTPUT PORT. A channel
through which data is transferred between an input or
output device and the processor.
INSERTION LOSS. The attenuation
of signals caused by routing
them through a switching
card or other intermediate
circuit element. Specified as a
decibel value over a frequency
range.
INSTRUMENTATION AMPLIFIER. A
high performance differential
amplifier having high input
impedance at both the input
high and input low terminals
and typically characterized by
high common mode rejection
ratio (CMRR) and low drift.
See also DIFFERENTIAL
AMPLIFIER, DIFFERENTIAL INPUT,
DRIFT, and COMMON MODE
REJECTION RATIO (CMRR).
APPENDIX B
INSULATION RESISTANCE. The ohmic

resistance of insulation.
Insulation resistance can
degrade quickly as humidity
increases.
INSULATOR. A material that does
not significantly conduct electrical current. Insulators have
wider bandgaps than semiconductor materials.
INTEGRAL LINEARITY. See LINEARITY.
INTEGRATION CONVERSION. An analog to digital conversion
process where the output
results in a digital representation of the integral of the
input signal over a specified
time interval.
INTERNAL PACER CLOCK. See PACER
CLOCK.
INTERNAL TRIGGER. A softwaregenerated event that starts an
operation. See also EXTERNAL
TRIGGER.
INTERRUPT LEVEL. A method of
assigning priority to hardware
interrupts so that high priority tasks interrupting a system
(or CPU) can be serviced
before lower priority tasks.
INTERRUPT SERVICE ROUTINE (ISR). A
software program that handles interrupts.
INTERRUPT. For a data acquisition

board, a signal to the CPU
indicating that the board
detected a condition or event
calling for special processing.
An interrupt causes the CPU
to temporarily stop the current processing task, complete the special processing
task, and then return to the
original processing task. See
also INTERRUPT LEVEL,
INTERRUPT-MODE OPERATION,
and INTERRUPT SERVICE ROUTINE
(ISR).
INTERRUPT-MODE OPERATION. Mode
in which a data acquisition
board acquires or generates
samples using an Interrupt
Service Routine (ISR). See also
OPERATION MODES.
ISA BUS. Industry Standard
Architecture. A PC bus architecture (8 or 16 bits wide)
used in most MS-DOS and
Windows computers.
Sometimes called the AT bus.
ISA bus PCs offer eight parallel channels for DMA mode
data transfers. Replaced by
the PCI bus. Also called DMA
levels. See also DMA MODE.
ISOLATED OUTPUTS. Output signals
where a common reference is
not connected to either input
terminal.
JITTER. The short-term variation
of timed events. See APERTURE
JITTER, TIMING JITTER, and
TRIGGER JITTER.
GLOSSARY
KELVIN CONTACTS. A means for

testing or making measurements in electronic devices
and circuits, particularly
when low values are being
measured. Two sets of leads
are used at each test point,
similar with respect to thickness, material and length; one
set carries the test signal and
the other connects with the
measuring instrument. The
effect of resistance in the
leads is thus eliminated.
LEAST SIGNIFICANT BIT (LSB). The

lowest order bit, usually the
rightmost bit, in the binary
representation of a digital
quantity. Measurement precision or accuracy is sometimes
specified in terms of multiple
Least Significant Bits (LSBs).
In that case, the precision or
accuracy is represented by the
binary number that results
from counting the specified
number of least significant
bits.
LATENCY. The time required by a

system to begin to acknowledge an event.
LINEARITY. For a curve relating

instrument readings to known
inputs, the maximum deviation of readings from a
straight line drawn between
readings at zero and full
range.
LCZ METER. Inductance (L),

capacitance (C), impedance
(Z) meter. A general-purpose
instrument for measuring
component L, C. and Z.
Sometimes called LCR meter.
This instrument may be
applied to C-V testing, but
typically lacks features optimized for C-V. See C-V METER.
LEAKAGE CURRENT. Any unwanted
current that flows when test
voltage is applied. The ideal
leakage current is zero.
Leakage currents can originate in instruments, cables, or
the device being tested. Even
high resistance paths between
low current conductors and
nearby voltage sources can
generate significant leakage
currents.
LOCAL BUFFER. Temporary memory location within an application programs memory area.
It is always available to the
application program. See also
BUFFER MEMORY.
LONG-TERM ACCURACY. The limit
that errors will not exceed
during a 90-day or longer
time period. It is expressed as
a percentage of reading (or
sourced value) plus a number
of counts over a specified
temperature range.
LSB. See LEAST SIGNIFICANT BIT.
LSTTL. Schottky-clamp TTL logic
typically using one-third the
power of TTL, but maintaining TTL speeds. See also TTL.
APPENDIX B
MAINFRAME. A self-contained
instrument in a cabinet,
which provides a measurement or connection capability
without requiring other
instruments in the circuit.
Some mainframes may be
designated as a master or
slave. See also SWITCHING
MAINFRAME, MASTER, and SLAVE.
MAKE-BEFORE-BREAK. Connecting
a new circuit before disconnecting the present circuit.
MAP. Any representation of the
structure of an object. For
example, a memory map
describes the layout of objects
in an area of memory, and a
symbol map lists the association between symbol names
and memory addresses in a
pro-gram.
MASTER. A mainframe that has
control of other mainframes
(slaves) through an external
connection. A slave unit adds
capacity or functions to the
master. The master/slave
combination has one IEEE488 bus address. See also SLAVE
and MAINFRAME.
MATRIX CARD. A type of card with
a switching configuration that
has columns and rows of relay
crosspoints. (Also called a
coordinate switch.) A matrix
card supports simultaneously
connection of one input to
multiple outputs, multiple
inputs to one output, or
multiple inputs to multiple
outputs.
GLOSSARY
MAXIMUM ALLOWABLE INPUT. The

maximum DC plus peak AC
value (voltage or current) that
can be applied between the
high and low input measuring
terminals without damaging
the instrument.
MERCURY WETTED RELAY. A reed
relay in which the contacts
are wetted by a film of mercury. Usually has a required
operating position to avoid
liquid mercury shorting the
contacts; other types are position insensitive.
MICRON. A unit of length equal to
106 meters. Also called a
micrometer. There are 1000
microns per millimeter.
MULTIPLEX. Connecting one
instrument to multiple
devices under test or multiple
instruments to one device
under test. See also SCAN.
MULTIPLEXER (MUX). A circuit that
switches multiple signals into
one input, one signal at a time
in a specified sequence.
MULTIPLEXING. A technique
whereby multiple signals are
sent to one input, one signal
at a time in a specified
sequence.
MUX. See MULTIPLEXER.
NANOVOLTMETER. A sensitive DC
voltmeter (typically one
decade more sensitive than a
digital multimeter) with a low
thermal input connection.
NEGATIVE-EDGE TRIGGERING. Digital

trigger mode in which the
triggering action starts on the
falling edge of the signal. See
also TRIGGER POLARITY and
DIGITAL TRIGGER.
NIBBLE. A group of four bits or
half a byte.
NOISE. An undesirable electrical
signal from an external source
such as an AC power line,
motors, generators, transformers, fluorescent lights,
CRT displays, computers,
radio transmitters, and others.
NORMAL MODE REJECTION RATIO
(NMRR). The ability of an
instrument to reject interference (usually of line frequency) across its input terminals.
Usually expressed in decibels
at a frequency.
NORMAL MODE VOLTAGE. A voltage
applied between the input
high and input low terminals
of an instrument.
OBJECT LINKING AND EMBEDDING
(OLE). A Microsoft Windows
standard mechanism for
embedding one program
within another. For example,
an Excel spreadsheet can be
pasted into a Visual Basic program. If a file is linked to an
OLE control, the data stored
in that file is displayed in the
OLE control.
OCX. Abbreviation for OLE
Custom Control. Also referred
to as ActiveX control. Software
modules packaged as OCX are
usable from a variety of programming languages.
OFFSET CURRENT. A current that

comes from a switching card
even though no user signals
are applied. It comes mostly
from the finite coil to contact
impedance. It is also generated by triboelectric, piezoelectric, and electrochemical
effects present on the card.
OFFSET VOLTAGE. An error voltage
that appears in series with an
analog input terminal of a
data acquisition board and is
generated by the input circuits of the board.
OLE. Abbreviation for Object
Linking and Embedding.
OPERATING SYSTEM. Base-level software that organizes the computers resources and capabilities, runs application programs, interacts with users,
and communicates with
installed and peripheral
devices. Popular operating
systems include DOS,
Windows, OS/2, and UNIX.
OPERATION MODES. Refers to single, synchronous, interrupt,
DMA, and on-board memory
(OBM) operations. See also
DMA MODE and INTERRUPTMODE OPERATION.
OPERATION MODES. Refers to single, synchronous, interrupt,
DMA, and on-board memory
(OBM) operations. See also
DMA MODE and INTERRUPTMODE OPERATION.
OVERLOAD PROTECTION. A circuit
that protects an instrument
against excessive current at
the input terminals.
APPENDIX B
PACED MODE. A data-acquisition

analog-to-digital conversion
mode in which one sample is
converted following each
pulse of a pacer clock. That is,
the conversion rate equals the
pacer clock rate. See also
PACER CLOCK, CONVERSION RATE,
SAMPLE RATE, and ANALOG-TODIGITAL CONVERTER.
PACER CLOCK RATE. The rate at
which timing pulses are emitted from a pacer clock. See
also PACER CLOCK and PACED
MODE.
PACER CLOCK. An on-board or
external clock that paces or
times events such as analogto-digital conversions, digitalto-analog conversions, data
sampling, interrupt generation, digital I/O transfers, etc.
PARAMETRIC TESTS. Tests that measure DC conditions of a chip,
such as maximum current,
leakage, and output drive.
PASS-THROUGH MODE. See TARGET
MODE.
PATH ISOLATION. On a matrix
switching card, the isolation
from signal high and low of
one path to signal high and
low of any other path.
capacitance. See also CHANNEL
ISOLATION.
PATH RESISTANCE. On a matrix
switching card, the resistance
per conductor of a closed
path, including the tape resistance and connector terminal
resistance. See also CONTACT
RESISTANCE.
GLOSSARY
PATH. One of many signal paths

on a matrix switching card. A
path is established by the
actuation of a relay at a row
and column crosspoint. See
also CHANNEL.
PC. Abbreviation for Personal
Computer.
PCI. Abbreviation for Peripheral
Component Interconnect. It is
a standard for a personal
computer local bus.
Compared to the older ISA
bus, PCI offers higher speed
and improved bus arbitration.
PCMCIA. Abbreviation for
Personal Computer Memory
Card International
Association. This organization
establishes standards for
removable function boards for
computers. PCMCIA was originally conceived for adding
memory, but now includes a
broad range of functions,
including modems, communication interfaces, storage
and data acquisition boards.
PEAK RESPONDING. A measurement
where the displayed value is
equal to the peak value of the
input signal.
PEAK. The highest magnitude,
either positive or negative. See
also PEAK-TO-PEAK.
PEAK-TO-PEAK. The difference
between the minimum value
and maximum value of an
alternating signal.
PER CHANNEL RATE. The sample
rate for each channel of a
scanning A/D system.
PGIA. See INSTRUMENTATION
AMPLIFIER.
PICOAMMETER. A measuring
instrument that is similar in
function to the ammeter of an
electrometer. However, a
picoammeter generally provides voltage burden that is as
low or lower than the ammeter of an electrometer, faster
readings, and less sensitivity.
PORT GROUP. For digital I/O emulating the I/O of an 8255 programmable peripheral interface chip, a group of three 8bit ports, commonly labeled
PA, PB, and PC. Digital I/O
that emulates multiple 8255
chips is typically divided into
multiple port groups.
PLUG AND PLAY. A set of specifications developed by Intel that

allows a PC to configure itself
automatically to work with
peripherals such as monitors,
modems, and printers. A user
can plug in a peripheral and
play it without manually
configuring the system. The
PC requires both a BIOS that
supports Plug and Play and a
Plug and Play expansion card.
PORT I/O CALL. A software program statement that assigns

bit values to an I/O port or
retrieves bit values from an
I/O port. Examples include a
C/C++ statement containing
an inp or outp function or a
BASIC statement containing a
PEEK or POKE function.
PNP. See PLUG AND PLAY.

P-N-P. A type of bipolar transistor
junction. Opposite of N-P-N.
POLARITY MODE. The mode that
specifies whether a data
acquisition channel inputs or
outputs both positive and
negative signals (bipolar
mode) or only positive signals
(unipolar mode) relative to
analog ground. See also
BIPOLAR and UNIPOLAR.
POLE. A combination of mating
relay contacts: normally open,
normally closed, or both.
Port. See INPUT/OUTPUT PORT.

POSITIVE-EDGE TRIGGERING. A digital trigger mode in which the
triggering action starts on the
rising edge of the signal. See
also TRIGGER POLARITY and
DIGITAL TRIGGER.
POST-TRIGGER ACQUISITION MODE.
A data acquisition triggering
acquisition starts after an
internal or external trigger
event and continues until a
specified number of samples
has been acquired or until the
operation is stopped by software. See also TRIGGER (DATA
ACQUISITION) and TRIGGER
MODES.
P-P. See PEAK-TO-PEAK.
APPENDIX B
PRE-TRIGGER ACQUISITION MODE. A

data acquisition triggering
acquisition is started before
an internal or external trigger
occurs, and stops when the
trigger occurs. See also
TRIGGER (DATA ACQUISITION)
and TRIGGER MODES.
PROGRAMMABLE GAIN
INSTRUMENTATION AMPLIFIER
(PGIA). See INSTRUMENTATION
AMPLIFIER.
PROGRAMMED I/O. A standard
method of accessing an I/O
devicethe CPU reads each
byte of data from or writes
each byte of data to the
device, with no external timing interrupts.
PSEUDO-SIMULTANEOUS SAMPLE AND
HOLD. Emulating
Simultaneous Sample and
Hold (SSH) by scanning a
group of data acquisition
channels at the highest practical rate while repeating scans
at a much slower rate. This is
commonly done in the burst
data-conversion mode, by
running the burst clock at a
rate close to maximum
throughput while running the
pacer clock at a much slower
rate. Typically used when
multiple parameters must be
compared at essentially the
same instant in time but
slight timing variations are
acceptable. See also BURST
CLOCK, BURST CONVERSION
MODE, PACER CLOCK,
SIMULTANEOUS SAMPLE AND
HOLD (SSH), SCAN (DATA
ACQUISITION), and THROUGHPUT.
GLOSSARY
PSEUDO-SSH. See PSEUDOSIMULTANEOUS SAMPLE AND

HOLD.
PULSE DURATION. See PULSE WIDTH.
PULSE WIDTH. The time interval
between the rising and falling
edges of a pulse, specified at a
certain percentage of the peak
amplitudecommonly 50%
for a rectangular pulse. Also
referred to as pulse duration.
QRAM. Queue RAM. Onboard
memory on a data acquisition
board that holds information
about the channel number
and gain, and sometimes
other settings, for each position in the channel-gain
queue. See also CHANNEL-GAIN
QUEUE.
QUANTIZATION. A process where
the continuous range of values of an input signal is divided into non-overlapping subranges and, to each subrange, a discrete value of the
output is uniquely assigned.
RANGE. A continuous band of signal values that can be measured or sourced. In bipolar
instruments, range includes
positive and negative values.
RATED ACCURACY. The limit that
errors will not exceed when
the instrument is used under
specified operating conditions. It is expressed as a percentage (of input or output)
plus a number of counts. See
also ABSOLUTE ACCURACY.
RATIO MEASUREMENT. The measurement of a signal input with
relation to an external reference input.
READING RATE. The rate at which

the displayed number is
updated.
READING. The displayed number
that is proportional to the
measured magnitude of the
input signal.
REAL TIME. A property of an event
or system in which data is
processed as it is aquired
instead of being accumulated
and processed at a later time.
REAL-TIME PROCESSING. A technique where events occur
within the required time
interval.
REGISTER. A set of bits of high
speed memory within a
microprocessor or other electronic device used to hold
data for a particular purpose.
On data acquisition boards,
any of the data or control registers that are generally accessible from the computer
through I/O or memory reads
and writes.
REGISTER LEVEL PROGRAMMING. A
type of low-level programming performed through
direct manipulation of the
hardwares data and control
registers, as opposed to programming through high-level
program calls.
RELAY MUST-OPERATE VALUE. A
specified functioning value
(typically, the coil voltage) at
which all relays meeting the
specifications must operate.
Also known as relay pick-up
or pull-in value.
RELAY MUST-RELEASE VALUE. A

specified functioning value
(typically, the coil voltage) at
which all relays meeting the
specifications must release.
Also known as relay drop-out
value.
RELEASE TIME. The time between
the removal of the relay coil
voltage and the stabilized
opening of the relay contacts.
RELIABILITY. The ability of a device
to perform within the desired
range over a measured period
of time.
REPEATABILITY. The ability of an
instrument to measure the
same input to the same value,
defined over a short period of
time and over a narrow temperature range.
RESISTANCE INSERTION. A current
measuring technique where a
known resistor is connected
in series with the circuit to be
measured. The voltage drop
across the resistor is proportional to the unknown
current.
RESOLUTION. The smallest increment of a signal that can be
measured, sourced, or displayed. Also called sensitivity
or minimum resolvable quantity. For a digitized signal, resolution is typically expressed
in bits or digits. By contrast,
sensitivity is expressed in
engineering units.
APPENDIX B
RESPONSE TIME. For a measuring

instrument, the time between
application of a step input
signal and the indication of its
magnitude within a rated
accuracy. For a sourcing
a programmed change and
the availability of the value at
its output terminals. Also
known as Settling Time.
ROOT-MEAN-SQUARE (RMS). For an

alternating signal, RMS equals
the square root of the time
average of the square of that
signal. A sinusoidal alternating current having a particular rms value and a DC current having that same value
produce the same joule heating when connected to a
given resistor.
RINGING (IN DIGITAL-TO-ANALOG

CONVERTER). A transient oscillation in the output of a
Digital-to-Analog Converter
(DAC) that follows an abrupt
change in input, analogous to
the decaying vibrations of a
clapped bell. Susceptibility to
ringing in a DAC is caused by
excessive capacitance in the
driven load.
ROW. As viewed on the schematic

of a matrix relay card, the horizontal signal lines or a horizontal group of relays.
RISE TIME. The time required for a

signal to change from a small
percentage (usually 10%) to a
large percentage (usually
90%) of its peak-to-peak
amplitude. See also FALL TIME.
RMS. See ROOT-MEAN-SQUARE
(RMS).
RMS RESPONDING. A measurement
where the displayed value is
equal to the root-meansquare (rms) of the input signal, for all input waveforms
having components within
the specified frequency range
and crest factor limit. See also
CREST FACTOR.
GLOSSARY
S. Abbreviation for the Sample or

Samples unit. See SAMPLE
(DATA ACQUISITION).
SAMPLE (DATA ACQUISITION). A single value that is read from or
written to one channel. See
also CHANNEL.
SAMPLE AND HOLD (SH). An operation, or electronic circuit, in
which an analog input signal
is stored briefly as a voltage
on a capacitor, typically until
it can be digitized by an analog-to-digital converter.
SAMPLE RATE. The rate at which a
continuous-time signal is
sampled. It is frequently
expressed as samples/second
(S/s), kilosamples/second
(kS/s), or megasamples/
second (MS/s).
SAMPLING ANALOG-TO-DIGITAL
CONVERTER. An analog-to-digital converter containing a
sample-and-hold circuit at
the front end, which captures
the incoming analog signal
and holds it for the duration
of the analog-to-digital conversion process. See also
ANALOG-TO-DIGITAL CONVERTER
and SAMPLE-AND-HOLD (SH).
SATURATION (AMPLIFIER). Amplifier
condition in which an
increase of the input signal
produces no further increase
in the output signal.
SCADA. Supervisory Control and
Data Acquisition. A type of
architecture in control systems.
SCAN (DATA ACQUISITION). Sample a
group of channels once at the
acquisition or burst-mode
rate; also can refer to a group
of channels. These channels
may be sequential (start to
stop channel specified) or
nonsequential (channel-gain
queue used). See also
MULTIPLEX.
SCAN RATE. The rate at which a
group of channels is sampled,
as measured from the start of
one scan to the start of the
next scan.
SCATTER-GATHER. A very high

speed, direct memory access
data transfer method under
PCI bus mastering. Data written to memory may be scattered into noncontiguous
memory blocks. When reading data, the memory block
locations are first supplied to
the bus master, and then data
is rapidly gathered from the
noncontiguous memory
blocks.
SCSI. Small Computer System
Interface. This is a standard
interface that supports up to
seven devices.
SENSITIVITY. See RESOLUTION.
SERIES RC MODEL. Interpretation
of the real and imaginary
components of admittance
measured by a capacitance
meter as resistance in series
with capacitance.
SETTLING TIME. For a measuring
application of a step input
signal and the indication of its
magnitude within a rated
accuracy. For a sourcing
a programmed change and
the availability of the value at
its output terminals. For a
switching card, the time
required for establishing relay
connections and stabilizing
user circuits. Also known as
Response Time.
SH. See SAMPLE AND HOLD.
APPENDIX B
SHIELDING. A metal enclosure for

the circuit being measured or
a metal sleeve surrounding
wire conductors (coax or triax
cable) to lessen interference,
interaction, or current leakage. The shield is usually
grounded.
SHORT-TERM ACCURACY. The limit
that errors will not exceed
during a 24-hour period of
continuous operation. Unless
specified, no zeroing or
adjustments of any kind are
permitted. It is expressed as a
percentage of reading (or
sourced value) plus a number
of counts over a specified
temperature range.
SIGNAL/NOISE RATIO. The ratio of
the maximum signal that can
be measured to the level
detected with no signal present (noise level). It is
expressed in decibels. A figure
of merit for characterizing
how noise-free a system is.
SIMULTANEOUS SAMPLE AND HOLD.
Operation in which the analog input channels are sampled at the same time and the
values held until sequentially
read by a scanning A/D
system.
SINGLE-ENDED INPUT. The condition where the low terminal of
a two-terminal instrument is
connected to a specific reference point, such as power line
common, earth ground, or
circuit common. All signals
must therefore be measured
against the same reference.
See also DIFFERENTIAL INPUT.
GLOSSARY
SLAVE. A mainframe that is externally connected to a controlling mainframe (master). A

slave unit adds capacity or
functions to the master. See
also MASTER and MAINFRAME.
SLEW RATE. The maximum charge
rate of the signal sampling
capacitor in the sample and
hold circuit of an A/D converter. It is expressed in
volts/microsecond.
SMALL SIGNAL LEVEL. The amplitude of AC test signal voltage
producing a linear response
of the device under test. In
other words, within the small
signal range, the shape of the
device response curve can be
assumed to be linear without
distorting the results.
SOFTWARE TRIGGER. A programmed
event that starts an operation
such as data acquisition.
SOURCE IMPEDANCE. The combination of resistance and reactance that a source presents
to the input terminals of a
measuring instrument.
SOURCE TERMINAL. Along with the
gate and drain, one of the
the source and drain terminals. The gate is a high
SOURCE-MEASURE UNIT (SMU). An

electronic instrument that
sources and measures DC
voltage and current. Standard
modes of operation include
source voltage and measure
current or source current and
measure voltage. Also known
as a source/monitor unit or a
stimulus/measurement unit.
SPECTRAL PURITY. A description of
the distortion components of
an oscillators output signal at
a specified amplitude and
load. Includes total harmonic
distortion, harmonics, and
spurious components.
SPURIOUS COMPONENTS. Undesired
signals in the output of a signal generator that have a
defined amplitude and frequency. They are not harmonically related to the fundamental frequency. Specified as
decibels below the carrier frequency.
SSH. See SIMULTANEOUS SAMPLE
AND HOLD (SSH).
STAIRCASE WAVEFORM. A waveform
in which the voltage is incremented in uniform steps from
the start voltage to the stop
voltage.
STANDOFF POTENTIAL. The breakdown voltage across relay
contacts.
STROBE. A timing signal that initiates and coordinates the passage of data, typically through
an input or output device
interface.
SUBROUTINE. A set of software
instructions invoked by a single calling line of code.
SWITCH CARD. A type of card with
independent and isolated
relays for switching inputs
and outputs on each channel.
SWITCHED CURRENT. The maximum
current level that can be reliably handled while opening
and closing contacts. See also
CARRY CURRENT.
SWITCHING MAINFRAME. A switching instrument that operates
according to user commands
to connect signals among
sourcing and measuring
instruments and devices
under test. A mainframe is
also referred to as a scanner,
multiplexer, matrix, or
programmable switch.
SYNCHRONOUS. In hardware, an
event that occurs in a fixed
time relationship to another
event. In software, a function
that begins an operation and
returns to the calling program
only when the operation is
complete.
STRESS GRADIENT. Stress gradients

are produced between layers
due to different thermal coefficients of expansion between
the layers.
APPENDIX B
TARGET MODE. A PCI bus mode in

which data from a data acquisition board is transferred
indirectly to the computer
memory in the foreground,
via the host computer CPU,
instead of directly, via bus
mastering. Sometimes
referred to as pass-through
operation. See also BUS
MASTERING and FOREGROUND
TASK.
TEMPERATURE COEFFICIENT OF
RESISTANCE. The amount of
resistance change of a material per degree of temperature
change.
TEMPERATURE COEFFICIENT. A
change in reading (or sourced
value) with a change in temperature. It is expressed as a
percentage of reading (or
sourced value), plus a number
of counts per degree change
in temperature.
TERMINATION MODE. Refers to the
mode in which a user connects signals to the multiplexerand, internally, to the differential amplifierof a data
acquisition board. The choices are single-ended mode or
differential mode. See also
SINGLE-ENDED INPUT (DATA
ACQUISITION) and DIFFERENTIAL
INPUT.
THERMOELECTRIC EMFS.
Temperature-dependent voltages that develop across junctions of dissimilar metals.
THERMOELECTRIC VOLTAGE. Voltages

resulting from temperature
differences within a measuring circuit or when conductors of dissimilar materials are
joined together. Also known
as thermoelectric EMF or
thermal offset. See CONTACT
POTENTIAL.
THRESHOLD VOLTAGE (VT). The voltage needed to turn on a
device or initiate a process.
THROUGHPUT. The maximum rate
at which a data conversion
system can perform repetitive
conversions within a specified
accuracy. It is determined by
summing the various times
required for each part of the
conversion system and then
taking the inverse of this time.
The throughput rate takes
into account the total time
required to process a signal
and store the value in either
on-board or system memory.
TIMEBASE ACCURACY. A measure of
how closely the internal timebase of an instrument tracks a
known time standard.
TIMING JITTER. The short-term
variation of the time period
between sample points.
TOTAL HARMONIC DISTORTION. The
percentage of harmonic distortion present in an output
signal over a specified frequency range.
THERMOCOUPLE. A temperature
sensor created by joining two
dissimilar metals. This junction creates a small voltage as
a function of the temperature.
GLOSSARY
TRANSCONDUCTANCE. The ratio of

the incremental change in the
output current of any amplifying circuit or device to the
incremental change of input
voltage causing it, when the
output voltage is held constant.
TRANSFER RATE. The rate at which
data is transferred to or from
memory.
TRANSISTOR. A semiconductor
device in which a small control signal is used to control a
larger current flow.
TRAP (v.). To intercept an action
or event before it occurs, usually in order to do something
else. Trapping is commonly
used by debuggers to allow
interruption of program execution at a given spot.
TRIGGER. An event that starts or
stops an operation. A trigger
can be a specific analog, digital, or software condition. See
also ANALOG TRIGGER and
DIGITAL TRIGGER.
TRIGGER CONDITIONS. Refers to
trigger sensitivity, polarity,
etc.
TRIGGER HYSTERESIS. Applies only
to analog triggers. A specified
voltage change, opposite in
polarity to the trigger polarity,
through which an analog trigger signal must move before
triggering can occur. For positive-edge triggering to occur,
the signal must first fall below
the specified trigger voltage
by at least the amount of the
hysteresis value. For negativeedge triggering to occur, the
signal must rise above the

specified trigger voltage level
by at least the amount of the
hysteresis value. Trigger hysteresis helps prevent false
triggering due to noise. See
also ANALOG TRIGGER and
TRIGGER POLARITY.
TRIGGER JITTER. The short-term
variation of the time period
between a trigger event and
the first sample point. See also
TRIGGER LATENCY.
TRIGGER LATENCY. The fixed time
offset between the trigger
event and the first sample
point.
TRIGGER MODE. Refers to when
data acquisition begins and
ends in relation-ship to the
trigger. Trigger modes include
normal-trigger, pre-trigger,
about-trigger, post-trigger,
trigger-to-trigger, and triggerto-about-trigger. See also PRETRIGGER ACQUISITION MODE,
ABOUT-TRIGGER ACQUISITION
MODE, and POST-TRIGGER
ACQUISITION MODE
TRIGGER POLARITY. For edge-sensitive triggers: trigger polarity
defines whether the trigger
occurs when the signal is rising (positive direction) or
falling (negative direction).
For level-sensitive triggers:
trigger polarity defines
whether the trigger occurs
when the signal is above a
level (positive) or below a
level (negative).
APPENDIX B
TRIGGER SENSITIVITY. Refers to edge

and/or level of a trigger. For
analog triggers, trigger sensitivity defines whether the trigger occurs on a transition
across a specified value (edge)
or whether the trigger occurs
when it is above or below a
specified value (level). For
digital triggers, trigger sensitivity defines whether the trigger occurs on a transition
from one state to another
state (edge) or whether the
trigger occurs when it is at a
specified value (level).
TTL. Abbreviation for TransistorTransistor-Logic. A popular
logic circuit family that uses
multiple-emitter transistors. A
low signal state is defined as a
signal 0.8V and below. A high
signal state is defined as a signal +2.0V and above.
TWO-TERMINAL RESISTANCE
MEASUREMENT. A measurement
where the same current flows
through the unknown and the
test leads.
UNIPOLAR. An analog signal range
that extends from 0, in one
direction. Typically, an instrument or data acquisition
board operating in unipolar
mode measures a positive
voltage or current starting at 0.
VOLTAGE BURDEN. The voltage drop

across the resistor for the
resistance insertion technique
of current measurement.
VOLTAGE STANDING WAVE RATIO
(VSWR). For a switching card,
the loss due to the mismatch
introduced into the signal by
the card contacts and conductors. Expressed as a ratio
of the highest voltage to the
lowest voltage found in the
signal.
WARM-UP TIME. The time required
after power is applied to an
instrument to achieve rated
accuracy at referenced conditions.
WORD. The standard number of
bits that a processor manipulates at one time. Microprocessors typically use 16-,
or 32-bit words (2 bytes and 4
bytes, respectively).
YIELD. The ratio of good units
obtained divided by the total
units produced; the percent of
product conforming to specifications.
ZERO OFFSET. The reading (desired
or undesired) that occurs
when the input terminals of a
measuring instrument are
shorted.
USB. Universal Serial Bus.

Version 1.1 of this bus supports up to 127 peripherals
and data transfer rates of up
to 12MB/second; USB 2.0
offers a faster rate and other
new features.
GLOSSARY
The glossary definitions listed

here were drawn from a variety
of sources, including:
ANSI C39.6-1983. American
National Standard for Electrical
Instrumentation. Digital
Measuring Instruments.
American National Standards
Institute, 1983.
ANSI/EIA RS-473-1981.
Definitions and Terminology for
Relays for Electronic Equipment.
American National Standards
Institute, 1981.
ANSI/IEEE Std. 100-1984. IEEE
Standard Dictionary of Electrical
and Electronics Terms. 3rd edition. Institute of Electrical and
Electronics Engineers, 1984.
Engineers Relay Handbook. 3rd
edition. National Association of
Relay Manufacturers, 1980.
Grove, A.S. Physics and
Technology of Semiconductor
Devices.
Horowitz, Paul and Winfield Hill.
The Art of Electronics. 2nd edition, Cambridge University Press,
1989.
Microsoft Press, Computer

Dictionary, Third Edition (ISBN:
1-57231-446-X) by Microsoft
Press. Reproduced by permission
of Microsoft Press.
Nicollian, E.H. and Brews, J.R.,
MOS Physics and Technology.
Parker, Sylvia P., editor. McGrawHill Dictionary of Electronics and
Computer Technology. McGrawHill Book Company, 1984.
Schroder, Dieter K.
Semiconductor Material and
Device Characterization.
Sze, S.M. Physics of
Semiconductor Devices.
Sze, S.M., ed. VLSI Technology.
Semiconductor Technology
Handbook. Technology
Associates, 1982.
Van Zant, Peter. Microchip
Fabrication, A Practical Guide to
Semiconductor Processing.
Wolf, H.F. Semiconductors.
Young , E.C. The New Penguin
Dictionary of Electronics.
All rights reserved.
Howe, Denis, editor, The Free Online Dictionary of Computing,

http://wombat.doc .ic.ac.uk/.
Keithley Instruments, Inc.
Keithley Catalog and Reference
Guide (full line catalog), Glossary,
1998.
McGuire, Gary E., ed.
Semiconductor Materials and
Process Technology Handbook.
APPENDIX B
APPENDIX C
Diameter and
Resistance of Various
Wire Gauges
The following table provides resistance data for a range of pure copper
wire sizes that may be encountered in test and measurement applications. These values may vary slightly from the resistance of specific
cable samples.
AWG Wire Sizes
AWG
Diameter
(in.)
Resistance
(/1000 ft.)
Resistance
(/foot)
46
0.0016
4207
4.2070
44
0.0020
2592
2.5920
42
0.0024
1660
1.6600
41
0.0028
1323
1.3230
40
0.0031
1080
1.0800
39
0.0035
847
0.8470
38
0.0039
648
0.6480
37
0.0043
512
0.5120
36
0.0051
415
0.4150
35
0.0055
331
0.3310
34
0.0063
261
0.2610
33
0.0071
206
0.2060
32
0.0079
162
0.1620
30
0.0098
104
28
0.0130
65.4
0.0654
27
0.0142
51.5
0.0515
26
0.0161
41.0
0.0410
25
0.0177
32.4
0.0324
24
0.0201
25.7
0.0257
22
0.0252
16.2
0.0162
20
0.0319
10.2
0.0102
18
0.0402
6.40
0.0064
16
0.0508
4.00
0.0040
14
0.0642
2.50
0.0025
12
0.0808
1.60
0.0016
10
0.1019
1.00
0.0010
0.1285
0.63
0.0006
0.162
0.40
0.0004
0.1040
APPENDIX C
Metric Wire Sizes

Diameter
(mm)
Resistance
(/km)
Resistance
(/m)
0.05
8740
8.740
0.08
3414
3.414
0.14
1115
1.115
0.25
350.0
0.350
0.34
189.0
0.189
0.38
151.0
0.151
0.50
87.40
0.087
0.75
38.40
0.038
1.00
21.90
0.0219
1.50
9.710
0.0097
2.50
3.500
0.0035
4.00
1.370
0.0014
Approximate Correspondence of Metric Wire Sizes to AWG

Metric
(mm)
AWG
0.05
44
0.08
40
0.14
35
0.25
30
0.34
2728*
0.38
2627*
0.5
24
0.75
2022*
1.0
18
2.5
10
4.0
*Metric size is approximately equidistant from nearest AWG sizes.
DIAMETER AND RESISTANCE OF VARIOUS WIRE GAUGES
APPENDIX D
Safety
Considerations
Test System Safety

Many electrical test systems or instruments are capable of measuring
or sourcing hazardous voltage and power levels. It is also possible,
under single fault conditions (e.g., a programming error or an instrument failure), to output hazardous levels even when the system indicates no hazard is present.
These high voltage and power levels make it essential to protect operators from any of these hazards at all times.
Protection methods include:
Design test fixtures to prevent operator contact with any hazardous
circuit.
Make sure the device under test is fully enclosed to protect the
operator from any flying debris.
Double insulate all electrical connections that an operator could
touch. Double insulation ensures the operator is still protected, even
if one insulation layer fails.
Use high-reliability, fail-safe interlock switches to disconnect power
sources when a test fixture cover is opened.
Where possible, use automated handlers so operators do not require
access to the inside of the test fixture or have a need to open guards.
Provide proper training to all users of the system so they understand
all potential hazards and know how to protect themselves from
injury.
CAUTION: During power-up, the states of board outputs are uncontrolled until hardware and software initialization has been completed.
Users must make sure their designs can tolerate this or provide suitable
interlocks to prevent dangerous voltages or actions from reaching
users.
General Safety Considerations

It is the responsibility of the test system designers, integrators, and
installers to make sure operator and maintenance personnel protection is in place and effective.
The following safety precautions should be observed before using any
Keithley product and any associated instrumentation. Although some
instruments and accessories would normally be used with nonhazardous voltages, there are situations where hazardous conditions
may be present. Keithley products are intended for use by qualified
personnel who recognize shock hazards and are familiar with the safety precautions required to avoid possible injury. Read the operating
information provided in each product's manual carefully before using
any Keithley product.
APPENDIX D
The types of product users are:

Responsible body is the individual or group responsible for the use
and maintenance of equipment, for ensuring that the equipment is
operated within its specifications and operating limits, and for ensuring that operators are adequately trained.
Operators use the product for its intended function. They must be
trained in electrical safety procedures and proper use of the instrument. They must be protected from electric shock and contact with
hazardous live circuits.
Maintenance personnel perform routine procedures on the product to
keep it operating, for example, setting the line voltage or replacing consumable materials. Maintenance procedures are described in the manual. The procedures explicitly state if the operator may perform them.
Otherwise, they should be performed only by service personnel.
Service personnel are trained to work on live circuits, and perform safe
installations and repairs of products. Only properly trained service personnel may perform installation and service procedures.
Exercise extreme caution when a shock hazard is present. Lethal voltage may be present on cable connector jacks or test fixtures. The
American National Standards Institute (ANSI) states that a shock hazard exists when voltage levels greater than 30V RMS, 42.4V peak, or
60VDC are present. A good safety practice is to expect that hazardous
voltage is present in any unknown circuit before measuring.
Users of these products must be protected from electric shock at all
times. The responsible body must ensure that users are prevented
access and/or insulated from every connection point. In some cases,
connections must be exposed to potential human contact. Product
users in these circumstances must be trained to protect themselves
from the risk of electric shock. If the circuit is capable of operating at or
above 1000 volts, no conductive part of the circuit may be exposed.
As described in the International Electrotechnical Commission (IEC)
Standard IEC 664, these instruments are Installation Category I, and
signal lines must not be directly connected to AC mains.
For rack mounted equipment in which the power cord is not accessible, in the event of fire or other catastrophic failure, the user must provide a separate power disconnect switch.
Do not connect switching cards directly to unlimited power circuits.
They are intended to be used with impedance limited sources. NEVER
connect switching cards directly to AC mains. When connecting
sources to switching cards, install protective devices to limit fault current and voltage to the card.
Before operating an instrument, make sure the line cord is connected
to a properly grounded power receptacle. Inspect the connecting
SAFETY CONSIDERATIONS
cables, test leads, and jumpers for possible wear, cracks, or breaks
before each use.
For maximum safety, do not touch the product, test cables, or any
other instruments while power is applied to the circuit under test.
ALWAYS remove power from the entire test system and discharge any
capacitors before: connecting or disconnecting cables or jumpers,
installing or removing switching cards, or making internal changes,
such as installing or removing jumpers.
Do not touch any object that could provide a current path to the common side of the circuit under test or power line (earth) ground. Always
make measurements with dry hands while standing on a dry, insulated
surface capable of withstanding the voltage being measured.
Instruments and accessories must be used in accordance with specifications and operating instructions or the safety of the equipment may
be impaired.
Do not exceed the maximum signal levels of the instruments and
accessories, as defined in the specifications and operating information, and as shown on the instrument or test fixture panels, or switching card.
When fuses are used in a product, replace with same type and rating for
continued protection against fire hazard.
Chassis connections must only be used as shield connections for
measuring circuits, NOT as safety earth ground connections.
If you are using a test fixture, keep the lid closed while power is applied
to the device under test. Safe operation requires the use of a lid
interlock.
If a
screw is present, connect it to safety earth ground using the
wire recommended in the user documentation.
The
symbol on an instrument indicates that the user should refer
to the operating instructions located in the manual.
The
symbol on an instrument shows that it can source or measure
1000 volts or more, including the combined effect of normal and
common mode voltages. Use standard safety precautions to avoid
personal contact with these voltages.
The WARNING heading in a manual explains dangers that might result
in personal injury or death. Always read the associated information
very carefully before performing the indicated procedure.
The CAUTION heading in a manual explains hazards that could damage the instrument. Such damage may invalidate the warranty.
Instrumentation and accessories shall not be connected to humans.
Before performing any maintenance, disconnect the line cord and all
test cables.
APPENDIX D
To maintain protection from electric shock and fire, replacement components in mains circuits, including the power transformer, test leads,
and input jacks, must be purchased from Keithley Instruments.
Standard fuses, with applicable national safety approvals, may be used
if the rating and type are the same. Other components that are not safety related may be purchased from other suppliers as long as they are
equivalent to the original component. (Note that selected parts should
be purchased only through Keithley Instruments to maintain accuracy
and functionality of the product.) If you are unsure about the applicability of a replacement component, call a Keithley Instruments office
for information.
To clean an instrument, use a damp cloth or mild, water based cleaner.
Clean the exterior of the instrument only. Do not apply cleaner directly to the instrument or allow liquids to enter or spill on the instrument.
Products that consist of a circuit board with no case or chassis (e.g.,
data acquisition board for installation into a computer) should never
require cleaning if handled according to instructions. If the board
becomes contaminated and operation is affected, the board should be
returned to the factory for proper cleaning/servicing.
SAFETY CONSIDERATIONS
Index
A
Acceleration, 7-12 to 7-17
Accelerometers, 7-12 to 7-17
ActiveX(r) controls, 3-3, 3-4, 3-10,
3-11 to 3-12
A/D (analog-to-digital) conversion, 5-2 to 5-9
A/D converters
flash, 5-6 to 5-8
integrating, 5-6 to 5-7
resolution, 5-2 to 5-5
Sigma-Delta, 5-6 to 5-8
speed, 5-5 to 5-6
successive approximation, 5-6
to 5-7
Aliasing and anti-aliasing filters,
4-9, 5-8 to 5-9, 7-17
Analog isolation, 5-17
Analog output control, 9-12 to
9-19
API (Application Program
Interface), 3-7 to 3-8, 3-10 to
3-11
Cauer filter, 4-20, 5-9

Capactive reactance, 4-8 to 4-9
Capacitors, 4-6 to 4-10, 4-14,
4-19
CMOS (Complementary Metal
Oxide Semiconductor) logic,
4-24, 5-10, 5-14
Cold junction, 6-3
Color codes (thermocouples),
6-10 to 6-11
Common mode voltage, 4-16 to
4-17
COM (Common Object Model)
objects, 3-3, 3-4, 3-7, 3-10,
3-11
Contact bounce, 5-12
Converters (protocol or interface), 2-12 to 2-13
Current measurements, 8-2 to
8-4
Application deployment, 3-7
CVD (Chemical Vapor

Deposition) system control,
9-5 to 9-6
Base metal thermocouples, 6-10
D/A (digital-to-analog) conversion, 5-9 to 5-10
Battery charge/discharge testing,

9-16 to 9-18
Bessel filter, 5-9
Bipolar signals, 4-18
Buffer (non-inverting digital
gate), 4-21
Burn-in testing, 9-9 to 9-11
Bus architecture, 2-3 to 2-5
Debugging, 3-7
Deterministic data acquisition
system, 3-13 to 3-14, 9-3
Device drivers, 2-9, 2-10, 3-2, 3-3,
3-6, 3-7 to 3-15
Differential input, 4-17 to 4-19,
8-4 to 8-5
Bus mastering, 2-3, 2-4, 3-13
Digital I/O (Input/Output), 4-21

to 4-24
Butterworth filter, 4-20, 5-9
Digital isolation, 5-16 to 5-17
Discrete instruments, 1-5
Cables and connectors, 8-4 to 8-7
I-2
Callendar-Van Dusen coefficients, 6-17 to 6-18
DMA (Direct Memory Access),

2-3, 2-4, 3-13
INDEX
DOS (Disk Operating System),

2-4, 3-2, 3-3, 3-4, 3-8, 3-9 to
3-10
Drive current requirements, 5-14
to 5-15
Dry switching, 5-13
E
Electrical isolation, 5-15 to 5-17,
6-10
EMI/RFI (Electromagnetic
Interference/Radio Frequency
Interference) radiation, 9-5
ESR (Equivalent Series
Resistance), 4-7
Ethernet, 1-4, 2-11 to 2-13
G
Gauge factor, 7-6
GPIB (General Purpose Interface
Bus), 2-8 to 2-9, 2-10, 2-11, 212, 2-13, 3-7 (See also IEEE488)
Graphical programming, 3-4 to
3-5
Ground loops, 4-16 to 4-17, 5-18
to 5-19, 9-5
H
HALT/HASS (Highly Accelerated
Life Testing/Highly
Accelerated Stress Screening)
procedures, 9-9 to 9-11
Event-driven control, 3-6, 3-13 to

3-14
Harmonics, 7-17
Excitation current, 6-18

Expansion slots, 2-3, 2-4, 5-3
Hybrid data acquisition systems,

1-5 to 1-6
Extenders, 2-12 to 2-13
Hysteresis, 5-14
External data acquisition systems, 1-3 to 1-4
F
FIFO (First In, First Out) memory, 3-12 to 3-13, 5-2
Filters, 4-7, 4-10, 4-15, 4-19 to
4-21, 5-2, 6-3, 6-17, 7-10
aliasing and anti-aliasing, 4-9,
5-8 to 5-9, 7-17
Bessel, 5-9
Butterworth, 4-20, 5-9
Cauer, 4-20, 5-9
Chebyshev, 4-20
FireWire, 2-10 to 2-11, 3-11
Flash A/D converters, 5-6 to 5-8
Flip-flop circuit, 5-12
Flow sensors, 7-11 to 7-12
Hot swapping, 2-5
IEEE-488 interface, 1-5, 2-8 to

2-9, 2-11, 2-13, 2-5, 2-7, 3-7
(See also GPIB)
IEEE-1394 bus, 2-10 to 2-11 (See
also FireWire)
Inductors, 4-10 to 4-11
Input accuracy, 5-5
Integra Series data acquisition
systems, 1-5 to 1-6
Integrating A/D converter, 5-6 to
5-7
Intermediate Metals, Law of, 6-5
Inverter (inverting digital gate),
4-21
ISA (Industry Standard
Architecture) bus, 1-3, 2-3 to
2-4
Four-wire remote sensing, 5-10,

9-12
INDEX
I-3
Personal computer
configurations for data acquisition, 2-2 to 2-3
Joule heating, 6-18, 7-6
L
Laptop computers, 2-5
Law of Intermediate Metals, 6-5
Linearization of thermocouple
outputs, 6-3, 6-7 to 6-8
Plug-in data acquisition boards,

1-3 to 1-5, 3-7, 5-3, 5-4, 8-3
Load cells, 7-11 to 7-12
Poissons ratio, 7-2 to 7-3
Look-up tables, 6-4, 6-7 to 6-8,

6-17
Polling, 3-6, 3-13 to 3-14
Process monitoring, 9-6 to 9-8
Multiplexers, 5-2, 5-5, 5-6, 9-10
Program documentation, 3-9
Noble metal thermocouples, 6-10

to 6-12
Q (parameter), 7-13 to 7-16
Normal mode voltage, 4-15 to

4-16
Nyquist rate, 5-8, 7-16
O
OEM (Original Equipment
Manufacturer)/factory
automation applications, 9-2
to 9-5
Pressure sensors, 7-11 to 7-12
R
RAM (Random Access Memory)
requirements, 2-2
RC time constant, 4-9 to 4-10
Real-time data acquisition and
control applications, 1-4 to
1-5, 3-14 to 3-15
Reference temperature sensors,
6-3
Op amp theory, 4-11 to 4-19
Repeaters, 2-9, 2-12 to 2-13
Open vs. closed programming

environment, 3-2 to 3-7
Resistors, 4-2 to 4-6, 4-17, 4-19

dropping, 8-2 to 8-3
Opto-isolators, 5-16 to 5-17
load, 9-10
pull-up, 5-11
Parallel port, 2-5, 2-8
reference, 6-16
Pass band (filter), 5-9
series, 9-17
Passive components, 4-2 to 4-11,

4-19 to 4-20, 6-19, 7-2
PCI (Peripheral Component
Interconnect) bus, 1-3, 2-3,
2-4 to 2-5, 2-8, 2-9, 3-13
PCMCIA (Personal Computer
Memory Card International
Association) bus, 2-5, 2-9
I-4
definition, 1-2
Plug-and-Play (PnP) operation,
2-3 to 2-5, 2-10, 3-9
shunt, 9-14, 9-16, 9-17, 9-18

Resonance, 7-13 to 7-14
Response curve or cut-off (filter),
4-20 to 4-21
RS-232, 2-5 to 2-8, 2-11, 2-13,
3-11
RS-422, 2-5 to 2-8, 2-12
INDEX
RS-485, 2-5 to 2-8, 2-12

RTDs (Resistive Temperature
Detectors), 6-12 to 6-19
Successive approximation A/D

converter, 5-6 to 5-7
Runtime license, 3-3
System DMA (Direct Memory

Access), 3-13
Schmidt trigger, 5-14
TCP/IP (Transmission Control

Protocol/Internet Protocol),
2-12
Seebeck effect, 6-3

Semiconductor linear temperature sensors, 6-21
Tensile testing, 9-8 to 9-9
Serial ports, 2-5 to 2-8, 2-12, 2-13
Text-based programming, 3-4 to

3-5
Shielding, 8-5 to 8-6, 9-5
Thermal shunting, 6-21 to 6-22
Shock, 7-13
Thermistors, 6-19 to 6-21
Shock absorber testing, 9-11 to

9-12
Thermocouples, 6-2 to 6-12
Shunt calibration, 7-10 to 7-11

Sigma-Delta A/D converter, 5-6
to 5-8
Single-ended input, 4-17 to 4-19,
8-4 to 8-5
Thermoelectric EMFs (electromotive forces), 6-8, 7-6, 8-4

TTL (Transistor-Transistor
Logic), 4-23 to 4-24, 5-10 to
5-11, 5-13, 5-14 to 5-15
Slow moving signals, 5-13 to 5-14
Software development under

Windows(r), 3-7 to 3-12
USB (Universal Serial Bus), 1-4,

1-5, 2-6, 2-9 to 2-11, 3-7, 3-11
Software structure, 3-7 to 3-8
Unipolar signals, 4-18
Solenoids, 5-15
Solid-state relays, 5-14 to 5-15
Vibration, 7-13
Source resistance, 4-4 to 4-5,

4-19, 8-4
Voltage burden, 8-2 to 8-3
SourceMeter instruments, 1-5
Speed vs. accuracy tradeoffs,

3-15 to 3-16, 5-8
Wheatstone bridges
full bridge, 7-7 to 7-8
Steinhart-Hart equation, 6-21
half bridge, 7-8 to 7-9
Stop band (filter), 5-9
quarter bridge, 7-9 to 7-10
Strain
use as strain gauges, 7-3 to

7-11, 7-15
apparent, 7-7
definition, 7-2
gauges, 7-2 to 7-6, 7-7 to 7-12
Poisson's, 7-2
with RTDs, 6-14, 6-16

Windows Messaging and Event
Management, 3-6
Stress testing, 9-9 to 9-11
INDEX
I-5
Specifications are subject to change without notice.

All Keithley trademarks and trade names are the property of Keithley Instruments, Inc. All other
trademarks and trade names are the property of their respective companies.
Keithley Instruments, Inc.

Copyright 2001 Keithley Instruments, Inc.
Printed in the U.S.A.
28775 Aurora Road Cleveland, Ohio 44139 440-248-0400 Fax: 440-248-6168

1-888-KEITHLEY (534-8453) www.keithley.com
No. 1716
70130KSI

Data Acquisition KEYTHLEY

Uploaded by

Copyright:

Available Formats

Data Acquisition KEYTHLEY

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Acquisition KEYTHLEY

Uploaded by

Copyright:

Available Formats

A

A Guide to Hardware and Software for

Section 1: Data Acquisition and Control Overview

Section 2: Communication Buses and Protocols

2.4 Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2-5

Section 3: Software Overview

3.2.2.3 Programming Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . .3-5

3.3 Software Development under Windows . . . . . . . . . . . . . . . . . . . . . .3-7

3.4 FIFO and Buffer Overrun Issues . . . . . . . . . . . . . . . . . . . . . . . . . . .3-12

Section 4: Basic Component Theory

4.3 Op Amp Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-11

4.3.8 Single-Ended vs. Differential, Bipolar vs. Unipolar . . . . . . . . . .4-17

4.4 Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4-19

Section 5: Basic Analog and Digital I/O

5.2 D/A Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-9

5.3 Interfacing Digital I/O to Applications . . . . . . . . . . . . . . . . . . . . . .5-10

5.4 Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-15

5.5 Ground Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5-18

Section 6: Temperature Measurement

6.3 Resistive Temperature Detectors . . . . . . . . . . . . . . . . . . . . . . . . . . .6-12

6.4 Thermistors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6-19

6.5 Semiconductor Linear Temperature Sensors . . . . . . . . . . . . . . . .6-21

Section 7: Strain Measurement

7.4 Gauge Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7-6

7.7 Strain Gauge Signal Conditioning . . . . . . . . . . . . . . . . . . . . . . . . . .7-10

Section 8: Related Topics of Interest

8.2 Connection Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8-4

Section 9: Application Examples

9.2.2 Measurement Integrity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-4

9.3 Semiconductor CVD Application . . . . . . . . . . . . . . . . . . . . . . . . . . . .9-5

Appendix A: Selection Guide for Plug-in Boards vs.

Data Acquisition and

Data Acquisition and Control Hardware

tures and benefits generally associated with the various categories,

Plug-in Data Acquisition Boards

Features of plug-in data acquisition boards

Least expensive method of computerized measurement and control.

External Data Acquisition Systems

DATA ACQUISITION AND CONTROL OVERVIEW

High sensitivity to low-level voltage signals, i.e., approximately

Features of external data acquisition chassis

Multiple board slots permit mixing-and-matching boards to support

1.2.2.1 Real-Time Data Acquisition and Control

measurement and control applications. Therefore, the solution is to

Features of discrete instruments for data acquisition

Support measurement ranges and sensitivities generally beyond the limits of

1.2.2.3 Hybrid Data Acquisition Systems

DATA ACQUISITION AND CONTROL OVERVIEW

Keithley Model 2700 hybrid data acquisition system

Features of a hybrid data acquisition system

Delivers accuracy, measurement range, and sensitivity typical of bench

Computer Hardware Overview

COMMUNICATION BUSES AND PROTOCOLS

boards to coincide with the introduction of Windows 95. PnP enabled

Serial Ports (RS-232, RS-422, RS-485)

COMMUNICATION BUSES AND PROTOCOLS

Standard serial interface.

One transmitter may drive

Popular for printers,

A standard fixture on many