Thư viện tri thức trực tuyến
Kho tài liệu với 50,000+ tài liệu học thuật
© 2023 Siêu thị PDF - Kho tài liệu học thuật hàng đầu Việt Nam
Building a Successful Board-Test Strategy
Nội dung xem thử
Mô tả chi tiết
Biolding a Successful Board-Test Strategy
This page intentionally left blank
Successful BoardTest Strategy
Second Edition
Stephen F. Scheiber
BUTTERWORT H
E I N E M A N N
Boston Oxford Johannesburg Melbourne New Delhi
Building a
Newnes is an imprint of Butterworth-Heinemann.
Copyright © 2001 by Butterworth-Heinemann
A member of the Reed Elsevier group
All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, or otherwise, without the prior written permission of the publisher.
Some material contained herein is derived from IEEE Std. 1014-1987, IEEE Standard for
Versatile Backplane Bus: VMEbus, IEEE Std. 1049.1-1990, IEEE Standard Test Methods
and Boundary-Scan Architecture, and IEEE Std. 1 1 55-1 992, IEEE Standard VMEbus Extensions for Instrumentation: VXIbus, copyrights by the Institute of Electrical and Electronics
Engineers, Inc. The IEEE takes no responsibility for and will assume no liability for damages
resulting from the reader's misinterpretation of said information resulting from the
placement and context in this publication. Information is reproduced with the permission
of the IEEE.
©Recognizing the importance of preserving what has been written, Butterworth-Heinemann
prints its books on acid-free paper whenever possible.
Library of Congress Cataloging-in-Publication Data
Scheiber, Stephen F.
Building a successful board-test strategy / Stephen F. Scheiber.
p. cm.
Includes bibliographical references and index.
ISBN 0-7506-7280-3 (pbk. : alk. paper)
1. Printed circuits-Testing. I. Title.
TK7868.P7S34 2001
621.3815'310287-dc21
2001032680
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
The publisher offers special discounts on bulk orders of this book.
For information, please contact:
Manager of Special Sales
Butterworth-Heinemann
225 Wildwood Avenue
Woburn, MA 01 801-2041
Tel: 781-904-2500
Fax: 781-904-2620
For information on all Newnes publications available, contact our World Wide Web home
page at: http://www.newnespress.com
1 0 98765432 1
Printed in the United States of America
Contents
Preface to the Second Edition x
Chapter 1 What Is a Test Strategy? I
1.1 Why Are You Here? 3
1.2 It Isn't Just Testing Anymore 3
1.3 Strategies and Tactics 4
1.3.1 The First Step 5
1.3.2 Life Cycles 6
1.4 The Design and Test Process 9
1.4.1 Breaking Down the Walls 10
1.4.2 Making the Product 15
1.4.3 New Challenges 16
1.5 Concurrent Engineering Is Not Going Away 17
1.6 The Newspaper Model 21
1.6.1 Error Functions 21
1.6.2 What Do You Test? 23
1.6.3 Board Characteristics 26
1.6.4 The Fault Spectrum 28
1.6.5 Other Considerations 34
1.6.6 The How of Testing 37
1.7 Test-Strategy Costs 39
1.7.1 Cost Components 40
1.7.2 Committed vs. Out-of-Pocket Costs 43
1.8 Project Scope 44
1.9 Statistical Process Control 46
1.10 Summary 50
Chapter 2 Test Methods 53
2.1 The Order-of-Magnitude Rule 53
2.2 A Brief (Somewhat Apocryphal) History of Test 55
2.3 Test Options 58
2.3.1 Analog Measurements 59
vi BUILDING A SUCCESSFUL BOARD-TEST STRATEGY
2.3.2 Shorts-and-Opens Testers 60
2.3.3 Manufacturing-Defects Analyzers 61
2.3.4 In-Circuit Testers 62
2.3.5 Bed-of-Nails Fixtures 68
2.3.6 Bed-of-Nails Probe Considerations 71
2.3.7 Opens Testing 76
2.3.8 Other Access Issues 79
2.3.9 Functional Testers 80
2.3.10 Functional Tester Architectures 83
2.3.11 Finding Faults with Functional Testers 88
2.3.12 Two Techniques, One Box 91
2.3.13 Hot-Mockup 92
2.3.14 Architectural Models 93
2.3.15 Other Options 96
2.4 Summary 96
Chapter 3 Inspection as Test 97
3.1 Striking a Balance 98
3.2 Post-Paste Inspection 101
3.3 Post-Placement/Post-Reflow 103
3.3.1 Manual Inspection 107
3.3.2 Automated Optical Inspection (AOI) 108
3.3.3 Design for Inspection 111
3.3.4 Infrared Inspection—A New Look at an Old Alternative 111
3.3.4.1 A New Solution 113
3.3.4.2 Predicting Future Failures 114
3.3.4.3 The Infrared Test Process 115
3.3.4.4 No Good Deed ... 116
3.3.5 The New Jerusalem?—X-Ray Inspection 117
3.3.5.1 A Catalog of Techniques 121
3.3.5.2 X-Ray Imaging 122
3.3.5.3 Analyzing Ball-Grid Arrays 124
3.4 Summary 128
Chapter 4 Guidelines for a Cost-Effective "Test" Operation 129
4.1 Define Test Requirements 129
4.2 Is Automatic Test or Inspection Equipment Necessary? 133
4.3 Evaluate Test and Inspection Options 134
4.4 The Make-or-Buy Decision 137
4.5 Getting Ready 138
4.6 Programming—Another Make-or-Buy Decision 140
4.7 The Test Site 143
4.8 Training 144
4.9 Putting It All in Place 145
Contents vii
4.10 Managing Transition 147
4.11 Other Issues 149
4.12 Summary 149
Chapter 5 Reducing Test-Generation Pain with Boundary Scan 151
5.1 Latch-Scanning Arrangements 151
5.2 Enter Boundary Scan 153
5.3 Hardware Requirements 158
5.4 Modes and Instructions 161
5.5 Implementing Boundary Scan 163
5.6 Partial-Boundary-Scan Testing 166
5.6.1 Conventional Shorts Test 167
5.6.2 Boundary-Scan Integrity Test 167
5.6.3 Interactions Tests 168
5.6.4 Interconnect Test 169
5.7 Other Alternatives 170
5.8 Summary ' 172
Chapter 6 The VMEbus extension for Instrumentation 173
6.1 VME Background
6.2 VXI Extensions
6.3 Assembling VXI Systems
6.4 Configuration Techniques
6.5 Software Issues
6.6 Testing Boards
6.7 The VXIbus Project
6.8 Yin and Yang
6.9 Summary
Chapter 7 Environmental-Stress Screening 199
7.1 The "Bathtub Curve" 199
7.2 What Is Environmental-Stress Screening? 201
7.3 Screening Levels 202
7.4 Screening Methods 202
7.4.1 Burn-in 202
7.4.2 Temperature Cycling 204
7.4.3 Burn-in and Temperature-Cycling Equipment 206
7.4.4 Thermal Shock 207
7.4.5 Mechanical Shock and Vibration 208
7.4.6 Other Techniques 210
7.4.7 Combined Screens 210
7.5 Failure Analysis 212
7.6 ESS Costs 212
viii BUILDING A SUCCESSFUL BOARD-TEST STRATEGY
7.7 To Screen or Not to Screen 213
7.8 Implementation Realities 214
7.9 Long-Term Effects 215
7.10 Case Studies 217
7.10.1 Analogic 217
7.10.2 Bendix 217
7.10.3 Hewlett-Packard (now Agilent Technologies) 218
7.11 Summary 218
Chapter 8 Evaluating Real Tester Speeds 219
8.1 Resolution and Skew 220
8.2 Voltage vs. Time 222
8.3 Other Uncertainties 224
8.4 Impact of Test-Method Choices 225
8.5 Summary 228
Chapter 9 Test-Program Development and Simulation 230
9.1 The Program-Generation Process 230
9.2 Cutting Test-Programming Time and Costs 232
9.3 Simulation vs. Prototyping 236
9.4 Design for Testability 237
9.5 Summary 239
Chapter 10 Test-Strategy Economics 241
10.1 Manufacturing Costs 242
10.2 Test-Cost Breakdown 243
10.2.1 Startup costs 244
10.2.2 Operating costs 246
10.2.3 Maintenance and Repair 248
10.3 Workload Analysis 249
10.4 An Order-of-Magnitude Rule Counterexample 251
10.5 Comparing Test Strategies 253
10.6 Break-Even Analysis 256
10.6.1 Payback Period 257
10.6.2 Accounting Rate of Return 258
10.6.3 The Time Value of Money 259
10.6.4 Net Present Value 260
10.6.5 Internal Rate of Return 262
10.7 Estimating Cash Flows 263
10.8 Assessing the Costs 264
10.9 Summary 265
Contents tx
Chapter 11 Formulating a Board-Test Strategy 266
11.1 Modern Tester Classifications 267
11,2 Establishing and Monitoring Test Goals 268
11.3 Data Analysis and Management 270
11.4 Indicators of an Effective Strategy 273
11.5 Yin and Yang in Ease of Tester Operation 274
11.6 More "Make-or-Buy" Considerations 275
11.7 General-Purpose vs. Dedicated Testers 278
11.8 Used Equipment 279
11.9 Leasing 280
11.10 "Pay as You Go" 281
11.11 Other Considerations 282
11.12 The Ultimate "Buy" Decision—Contract Manufacturing 282
11.13 Summary 285
Chapter 12 Test-Strategy Decisions 286
12.1 A Sample Test Philosophy 286
12.2 Big vs. Small 288
12.3 Do You Need a High-End Tester? 290
12.4 Assembling the Strategy 291
12.5 The Benefits of Sampling 294
12.6 Tester Trends 295
12.7 Sample Strategies 297
12.8 A Real-Life Example 301
12.9 Changing Horses 304
12.10 Summary 305
Chapter 13 Conclusions 307
Appendix A Time-Value-of-Money Tables 309
Appendix B Acronym Glossary 318
Works Cited and Additional Readings 321
Index 329
Preface to the
Second Edition
When I wrote the first edition of Building a Successful Board-Test Strategy, my
intent was to avoid (as much as possible) the malady that plagues many books in
our industry: like the products they deal with, they become obsolete before release
to the public. To accomplish this goal, the book discussed tools, alternatives, and
ways to evaluate and select test strategies, rather than dictating what those strategies should be.
In many respects, I succeeded. Most of the comments in the original edition
are as true today as when they were written. Nevertheless, the industry refuses
to stand still. Test has undergone something of a transformation in the past few
years. The migration of production capacity away from traditional manufacturers
toward contractors continues to accelerate. Today's army of contractors ranges
from "garage shops" catering to complex very-low-volume products to multibillion-dollar megaliths handling board volumes in the millions. This continuing
evolution brings with it new challenges, the most significant of which is how to
select a contract manufacturer. Such vendors are not like commodity products. As
with pieces of test equipment, contractors offer a wide range of strengths and areas
of expertise. Choosing one requires finding a combination of skills and capabilities that best matches your needs. Discussions throughout the new edition take this
trend into account.
One development that I missed completely in the first edition was the plague
that open circuits bring to our surface-mount world. Hidden nodes, board coplanarity (flatness), and other characteristics of today's boards require another look
at test methods. A new section of Chapter 2 explores these issues.
The concept of what constitutes a "test" strategy is evolving as well. Various
forms of inspection, once mere adjuncts to the quality process, have become intimately linked with more traditional forms of conventional test. Then, too, inspection is not a single technique, but in fact a menu of approaches, each of which has
advantages and drawbacks. The new Chapter 3, "Inspection as Test," examines this
solution in considerable detail.
This new edition also updates information in many places, adding examples
and figures to prior discussions. Much of the additional material comes from
seminars that I have given in the past few years—both my own work and material
Preface to the Second Edition xi
from attendees. Some of those contributions are attributed to their sources. Other
examples must remain by their nature anonymous. Nevertheless, I appreciate all
of the assistance I have received.
At the risk of leaving out some important names, I would like to thank certain
people explicitly for their help. Bob Stasonis at GenRad, Jim Hutchinson at Agilent
Technologies, Charla Gabert at Teradyne, and Robin Reid at CyberOptics provided
considerable assistance for the chapter on inspection. Jon Titus, Editorial Director
at Test & Measurement World, has provided constant encouragement along with a
stream of contact suggestions and source recommendations over many years. And,
of course, my family has once again had to endure my particular brand of craziness as I rushed to complete this project.
Stephen F. Scheiber
December 18, 2000
This page intentionally left blank
CHAPTER
What Is a Test Strategy?
This book examines various board-test techniques, relating how they fit into an
overall product design/manufacturing/test strategy. It discusses economic, management, and technical issues, and attempts to weave them into a coherent fabric.
Looking at that fabric as a whole is much more rewarding than paying too close
attention to any individual thread. Although some of the specific issues have
changed in the past few years, the basic principles remain relatively constant,
Printed-circuit boards do not exist in a vacuum. They consist of components
and electrical connections and represent the heart of electronic systems. Components, boards, and systems, in turn, do not spring to life full-blown. Designers
conceive them, manufacturing engineers construct them, and test engineers make
sure that they work. Each group has a set of tools, criteria, and goals. To be
successful, any test strategy must take all of these steps into account.
Test managers coined the briefly popular buzzword "concurrent engineering"
to describe this shared relationship. More recently, enthusiasm for concurrent engineering has waned. Yet the ideas behind it are the same ones that the test industry
has been touting for as long as anyone can remember. The term represents merely
a compendium of techniques for "design-for-marketability," "design-for-manufacturability," "design-for-testability," "design-for-repairability," and so on. The
fact that the term "concurrent engineering" caught on for awhile was great. A
company's overall performance depends heavily on everyone working together.
Regardless of what you call it, many manufacturers continue to follow "design-forwhatever" principles. For those who do not understand this "we are all in it
together" philosophy, a new term for it will not help.
Concurrent engineering boils down to simple common sense. Unfortunately,
as one basic law of human nature so succinctly puts it, "Common sense isn't." In
many organizations, for example, each department is responsible only for its own
costs. Yet, minimizing each department's costs does not necessarily minimize costs
across an entire project. Reducing the costs in one department may simply push
them off to someone else. Achieving highest efficiency at the lowest cost requires
that all of a project's participants consider their activities' impact on other departments as well as their own.
2 BUILDING A SUCCESSFUL BOARD-TEST STRATEGY
The test-engineering industry is already feeling the effects of this more global
approach to test problems. Trade shows geared exclusively toward testing electronics—aside from the annual International Test Conference sponsored by the
IEEE—have largely passed into the pages of history. Instead, test has become an
integral part of trade shows geared to printed-circuit-board manufacturing. There
are two basic reasons for this phenomenon. When test shows first appeared, test
operations enjoyed little visibility within most organizations. The shows helped
focus attention on testing and disseminated information on how to make it work,
In addition, most companies regarded testing as an isolated activity, adopting the
"over-the-wall" approach to product design. That is, "I designed it, now you figure
out how to test it."
Today, neither of those situations exists. Everyone is aware of the challenges
of product test, even as they strive to eliminate its huge costs and its impact on
time to market. Managers in particular dislike its constant reminders that the
manufacturing process is not perfect. They feel that if engineering and manufacturing personnel had done their jobs right the first time, testing would not be
necessary.
Also, in the past few years, product-manufacturing philosophy has migrated
away from the vertically integrated approach that served the industry for so long.
Companies still design and market their creations, but someone else often produces
them and makes sure that they work. Even within large companies that technically
perform this task themselves, production flows through one or a few dedicated
facilities. These facilities may differ legally from contract manufacturers, but from
a practical standpoint they serve the same purpose, possessing both the same
advantages and the same drawbacks.
Because of the popularity of at least the concept of concurrent engineering,
considering test activities as distinct from the rest of a manufacturing process is
no longer fashionable. Design engineers must deliver a clean product to either inhouse or contract manufacturing to facilitate assembly, testing, and prompt
shipment to customers. Depot repair and field-service engineers may need to cope
with that product's failure years later. With the constant rapid evolution of
electronic products, by the time a product returns for repair, the factory may no
longer make it at all.
Therefore, although this book is specifically about building board-test strategies, its principles and recommendations stray far afield from that relatively narrow
venue. The most successful board-test strategy must include all steps necessary
to ship a quality product, whether or not those steps relate directly to the test
process itself.
The aim of this book is not to provide the ultimate test strategy for any specific situation. No general discussion can do that. Nobody understands a particular manufacturing situation better than the individuals involved. This book will
describe technical and management tools and fit them into the sociology and
politics of an organization. You must decide for yourself how to adapt these tools
to your needs.