Handbook for Robustness Validation
Electronic Components
and Systems Handbook for Robustness Validation
of Semiconductor Devices
in Automotive Applications
2
Preface
Can you imagine hiking on a steep mountain trail in the black of night not knowing how close to the
edge of the cliff you are? Would you feel safe?
Electronic components, such as semiconductors, have technical limits that might be very close to
the edge of the customer’s speci? cation. When this occurs, the semiconductor can malfunction and
possibly cause an operational failure of a critical vehicle system.
As in the hiking analogy, wouldn’t it be better to have the information as to how close the
semiconductor actually performs with regard to the speci? cation limits, or better yet, to know
that there is a the safety zone, or guard band, between to semiconductor’s performance and the
speci? cation limits?
The basic philosophy behind the Robustness Validation methodology described in this Handbook is to
gain knowledge about the size of the guard band by testing the semiconductor to failure, or end-of-
life. The goal of Robustness Validation is to achieve lower ppm-failure rates by ensuring adequate
guard band between the “real-life” operating range of the semiconductor and the points at which the
semiconductor fails.
The current “test-to-pass” statistical method used to select and qualify semiconductor devices does
not provide information regarding the amount of guard band. This is very similar to hiking in the
dark without knowing where the edge of the cliff is.
The safer way is to use Robustness Validation approach. Please read on.
Sincerely
Yours
Helmut Keller
Chairman ZVEI Robustness Validation Committee
Jack Stein
Chairman SAE Automotive Electronics Reliability Committee
3
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form
or by any means (print, electronic, mechanical, photocopying) or otherwise without the prior permission
of ZVEI. Every effort is made to ensure that the information given herein is accurate, but no legal
responsibility is accepted for any errors, omissions or misleading statements in this information.
4
Foreword
The quality of the vehicles we buy and the competitiveness of the automotive industry depend on
being able to make quality and reliability predictions. Quali? cation measures must provide useful
and accurate data to provide added value. Increasingly, manufacturers of semiconductor components
must be able to show that they are producing meaningful results for the reliability of their products
under de? ned mission pro? les from the whole supply chain.
Reliability is the probability that a semiconductor component will perform in accordance with
expectations for a predetermined period of time in a given environment. To be ef? cient reliability
testing has to compress this time scale by accelerated stresses to generate knowledge on the time to
fail. To meet any reliability objective requires comprehensive knowledge of the interaction of failure
modes, failure mechanisms, the mission pro? le and the design of the product.
10 years ago you could read: “Quali? cation tests of prototypes must ensure that quality and
reliability targets have been reached”.
This approach is no longer suf? cient to guarantee robust electronic products for a failure free life
of the car, which is the intention of the Zero-Defect-Approach. The emphasis has now shifted from
merely the detection of failures to their prevention.
We started this way by introducing screening methods after the product had been produced after
product has successfully survived a standard quali? cation. Then the focus shifted to reliability
methodologies applied on technology level during development.
Now product quali? cation again changes from the detection of defects based on prede? ned sample
sizes towards the generation of knowledge by generating failure mechanisms speci? c data, combined
with the knowledge from the technology ? eld. Now we can generate real knowledge on the robustness
of products.
Quali? cation focuses on intrinsic topics of products and technologies, requiring only small sample
sizes. Defectivity issues now put a big load on monitoring measures, which are now needed to
demonstrate manufacturability and the control of extrinsic defects.
This handbook should give guidance to engineers how to apply robustness validation during
development and quali? cation of semiconductor components. It was made possible because many
companies, semiconductor manufacturers, component manufacturers (Tier1) and car manufacturers
(OEMs) worked together in a joint working group to bring in the knowledge of the complete supply
chain.
I would like to thank all teams, organizations and colleagues for actively supporting the robustness
validation approach.
Andreas Preussger
Core Team Leader RV Group
Editor in Chief
5
Ito Masuo, Nissan Motor Company Ltd
Jendro Brian, Siemens VDO and AEC Member
Kanekawa Nobuyasu, Hitachi Ltd.
Kanemaru Kenji, Tokai Rika Co Ltd.
Klauke Martin, Renesas Technology
Knoell Bob, Visteon Corporation and AEC Member
Koch Herbert, Robert Bosch GmbH
Liang Zhongning, NXP and AEC Member
Lycoudes Nick, Freescale Semiconductor and AEC Member
Maier Reinhold, BMW AG
Mori Satoshi, Tokai Rika Co Ltd.
Nakaguro Kunio, Nissan Motor Co., Ltd.
Narumi Kenji, TRAM Inc.
Petersen Frank, Elmos Semiconductor AG
Schilde Bernd, Brose Fahrzeugteile GmbH & Co
Schmidt Ernst, BMW AG
Senske Wilhelm, Daimler Chrysler Corporation
Takasu Yuji, Tokai Rika Co Ltd.
Unger Walter, Daimler Chrysler Corporation
Vanzeveren Vincent, Melexis
Wilson Peter, On Semiconductor
Wulfert Friedrich-W., Freescale Semiconductor
7
Contents
1Introduction – Objectives of the Handbook 10
2Scope 11
3Terms, De? nitions, and Abbreviations 12
3.1 Terms and De? nitions
3.2 Abbreviations
4De? nition and Description of Robustness Validation 15
4.1 Robustness Validation Flow
4.2 De? nition of Robustness Validation
4.3 Robustness Diagrams
4.4 Difference between Robustness Validation Approach
and Stress Test Driven Quali? cation Standards
5Mission Pro? le / Vehicle Requirements 21
6Technology Development 26
7Product Development 28
8Potential Risks and Failure Mechanisms 29
8.1 The Knowledge Matrix
8.2 How to Use the Knowledge Matrix
9Creation of the Quali? cation Plan 32
9.1 Reliability Test Plan
9.2 De? nition of a Quali? cation Family
9.2.1 Wafer Fab
9.2.2 Assembly Processes
9.3 Quali? cation Envelope
9.4 Characterization Plan
9.4.1 Process Characterization
9.4.2 Device (Semiconductor Component) Characterization
9.4.3 Production Part Lot Variation Characterization
8
10Stress and Characterization 38
11Robustness Assessment 39
11.1 Lifetime t P as a Function of Stress Value S i
11.2 Determine Boundary of the Safe Operating Area
11.3 Determine Robustness Target and Area
11.4 Determine Robustness Margin
12Improvement 44
13Monitoring 47
13.1 Planning
13.2 High Volume Production
14Reporting 49
14.1 Content, Structure
14.2 Documents for Communication, Handouts and General Remarks
15Examples 50
15.1 Examples of the Lack of or Poor Quali? cation
15.1.1 Delamination between Mould Compound and Die/lead Frame
15.1.2 Quali? cation of a New Lead Frame Finish
15.1.3 Via-problems in Semiconductor Component Metallization
15.2 Integrated Capacitor Design
15.3 Requirement Temperature Cycles
16Annex Knowledge Matrix 55
17Annex Reporting Template 55
18References and Additional Reading 56
9
1 Introduction – Objectives of the Handbook
Members of SAE International Automotive Electronic Systems Reliability Standards Committee, ZVEI
(German Electrical and Electronic Manufacturers` Association), AEC (Automotive Electronics Council)
and JSAE (Japanese Society of Automotive Engineers) formed a joint task force and met to update
SAE Recommended Practice J1879-October-1988 (General Quali? cation and Production Acceptance
Criteria for Integrated Circuits in Automotive Applications). This version did not describe methods
to demonstrate that a device under test would meet the customer demand for failure levels in the
single-digit parts per million (ppm) range. Additionally, with the old quali? cation “test-to-pass”
approach, there is very little knowledge generated about the relevant component failure mechanisms
that may occur at the boundaries of the speci? cation limits. Extending the old approach to single-
digit ppm levels is unfeasible with respect to both economics and time. A new knowledge-based
approach to understanding and preventing the occurrence of the relevant component failure
mechanisms was required.
The joint task force concluded that the J1879 Recommended Practice should be revised to encompass
a Robustness Validation approach and that an Automotive Electronics Robustness Validation
Handbook should be published. This handbook is based on information from a wide number of
sources including international Automotive OEMs and their full supply chain, engineering societies,
and other related organizations.
This Robustness Validation Handbook provides the automotive electronics community with a common
quali? cation methodology to demonstrate acceptable reliability. The Robustness Validation approach
requires testing the component to failure, or end-of-life (EOL), without introducing invalid failure
mechanisms, and evaluation of the Robustness Margin between the outer limits of the customer
speci? cation and the actual performance of the component.
The principles de? ned in this handbook are also applicable to automotive electronic modules and
systems. Publications addressing these topics are currently under development by the SAE/ZVEI Joint
Task Force.
10
2 Scope
This document will primarily address intrinsic reliability of electronic components for use in
automotive electronics. Where practical, methods of extrinsic reliability detection and prevention
will also be addressed. This document primarily deals with integrated circuit issues, but can easily be
adapted for use in discrete or passive component quali? cation with the generation of a list of failure
mechanisms relevant to those devices. Component quali? cation is the main scope of this document.
Other procedures addressing extrinsic defects are speci? cally addressed in the monitoring chapter.
This document is to be used within the context of achieving Zero Defect in component manufacturing
and product use. If the handbook is adopted as a standard, the term “shall” indicates a binding
requirement.
11