Single Event Effects in Aerospace E-Book

Edward Petersen

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 594

Veröffentlichungsjahr: 2011

Beschreibung

This book introduces the basic concepts necessary to understand Single Event phenomena which could cause random performance errors and catastrophic failures to electronics devices. As miniaturization of electronics components advances, electronics components are more susceptible in the radiation environment. The book includes a discussion of the radiation environments in space and in the atmosphere, radiation rate prediction depending on the orbit to allow electronics engineers to design and select radiation tolerant components and systems, and single event prediction.

Leseprobe

Table of Contents

Series Page

Title Page

Copyright

Chapter 1: Introduction

1.1 Background

1.2 Analysis of Single Event Experiments

1.3 Modeling Space and Avionics See Rates

1.4 Overview of this Book

1.5 Scope of this Book

Chapter 2: Foundations of Single Event Analysis and Prediction

2.1 Overview of Single Particle Effects

2.2 Particle Energy Deposition

2.3 Single Event Environments

2.4 Charge Collection and Upset

2.5 Effective Let

2.6 Charge Collection Volume and the Rectangular Parallelepiped (RPP)

2.7 Upset Cross Section Curves

2.8 Critical Charge

2.9 Upset Sensitivity and Feature Size

2.10 Cross-Section Concepts

Chapter 3: Optimizing Heavy Ion Experiments for Analysis

3.1 Sample Heavy Ion Data

3.2 Test Requirements

3.3 Curve Parameters

3.4 Angular Steps

3.5 Stopping Data Accumulation When You Reach the Saturation Cross Section

3.6 Device Shadowing Effects

3.7 Choice of Ions

3.8 Determining the LET in the Device

3.9 Energy Loss Spread

3.10 Data Requirements

3.11 Experimental Statistics and Uncertainties

3.12 Effect of Dual Thresholds

3.13 Fitting Cross-Section Data

3.14 Other Sources of Error and Uncertainties

Chapter 4: Optimizing Proton Testing

4.1 Monitoring the Beam Intensity and Uniformity

4.2 Total Dose Limitations on Testing

4.3 Shape of the Cross-Section Curve

Chapter 5: Data Qualification and Interpretation

5.1 Data Characteristics

5.2 Approaches to Problem Data

5.3 Interpretation of Heavy Ion Experiments

5.4 Possible Problems with Least Square Fitting Using the Weibull Function

Chapter 6: Analysis of Various Types of SEU Data

6.1 Critical Charge

6.2 Depth and Critical Charge

6.3 Charge Collection Mechanisms

6.4 Charge Collection and the Cross-Section Curve

6.5 Efficacy (Variation of SEU Sensitivity within a Cell)

6.6 Mixed-Mode Simulations

6.7 Parametric Studies of Device Sensitivity

6.8 Influence of Ion Species and Energy

6.9 Device Geometry and the Limiting Cross Section

6.10 Track Size Effects

6.11 Cross-Section Curves and the Charge Collection Processes

6.12 Single Event Multiple-Bit Upset

6.13 SEU in Logic Systems

6.14 Transient Pulses

Chapter 7: Cosmic Ray Single Event Rate Calculations

7.1 Introduction to Rate Prediction Methods

7.2 The RPP Approach to Heavy Ion Upset Rates

7.3 The Integral RPP Approach

7.4 Shape of the Cross-Section Curve

7.5 Assumptions Behind the RPP and IRPP Methods

7.6 Effective Flux Approach

7.7 Upper Bound Approaches

7.8 Figure of Merit Upset Rate Equations

7.9 Generalized Figure of Merit

7.10 The FOM and the LOG Normal Distribution

7.11 Monte Carlo Approaches

7.12 PRIVIT

7.13 Integral Flux Method

Chapter 8: Proton Single Event Rate Calculations

8.1 Nuclear Reaction Analysis

8.2 Semiempirical Approaches and the Integral Cross-Section Calculation

8.3 Relationship of Proton and Heavy Ion Upsets

8.4 Correlation of the FOM with Proton Upset Cross Sections

8.5 Upsets Due to Rare High Energy Proton Reactions

8.6 Upset Due to Ionization by Stopping Protons, Helium Ions, and Iron Ions

Chapter 9: Neutron Induced Upset

9.1 Neutron Upsets in Avionics

9.2 Upsets at Ground Level

Chapter 10: Upsets Produced by Heavy Ion Nuclear Reactions

10.1 Heavy Ion Nuclear Reactions

10.2 Upset Rate Calculations for Combined Ionization and Reactions

10.3 Heavy Nuclear Ion Reactions Summary

Chapter 11: Samples of Heavy Ion Rate Prediction

11.1 Low Threshold Studies

11.2 Comparison of Upset Rates for Weibull and Lognormal Functions

11.3 Low Threshold–Medium Lc data

11.4 See Sensitivity and LET Thresholds

11.5 Choosing Area and Depth for Rate Calculations

11.6 Running CREME96 Type Codes

11.7 CREME-MC and SPENVIS

11.8 Effect of Uncertainties in Cross Section on Upset Rates

Chapter 12: Samples of Proton Rate Predictions

12.1 Trapped Protons

12.2 Correlation of the FOM with Proton Upset Rates

Chapter 13: Combined Environments

13.1 Relative Proton and Cosmic Ray Upset Rates

13.2 Calculation of Combined Rates Using the Figure of Merit

13.3 Rate Coefficients for a Particular New Orbit

13.4 Rate Coefficients for Any Circular Orbit About the Earth

13.5 Ratio of Proton to Heavy Ion Upsets for Near Earth Circular Orbits

13.6 Single Events from Ground to Outer Space

Chapter 14: Samples of Solar Events and Extreme Situations

Chapter 15: Upset Rates in Neutral Particle Beam (NPB) Environments

15.1 Characteristics of NPB Weapons

15.2 Upsets in the NPB Beam

Chapter 16: Predictions and Observations of SEU Rates in Space

16.1 Results of Space Observations

16.2 Environmental Uncertainties

16.3 Examination of Outliers

16.4 Possible Reasons for Poor Upset Rate Predictions

16.5 Constituents of a Good Rate Comparison Paper

16.6 Summary and Conclusions

16.7 Recent Comparisons

16.8 Comparisons with Events During Solar Activity

Chapter 17: Limitations of the IRPP Approach

17.1 The IRPP and Deep Devices

17.2 The RPP When Two Hits are Required

17.3 The RPP Approaches Neglect Track Size

17.4 The IRPP Calculates Number of Events, not Total Number of Upsets

17.5 The RPP Approaches Neglect Effects that Arise Outside the Sensitive Volume

17.6 The IRPP Approaches Assume that the Effect of Different Particles with the Same LET is Equivalent

17.7 The IRPP Approaches Assume that the LET of the Particle is not Changing in the Sensitive Volume

17.8 The IRPP Approach Assumes that the Charge Collection Does Not Change with Device Orientation

17.9 The Status of Single Event Rate Analysis

Appendix A: Useful Numbers

Appendix B: Reference Equations

B.1 Weibull Function

B.2 Lognormal Distribution

B.3 Exponential Distribution

B.4 Least Squares Fitting

B.5 Weighted Least Squares Fitting

B.6 Calculation of Coefficient of Determination R2

B.7 The Bendel–Petersen Equation for the Proton Cross Section

B.8 The Figure of Merit and Upset Rate Calculations

B.9 RPP Differential and Integral Path Length Distributions

Appendix C: Quick Estimates of Upset Rates Using the Figure of Merit

Appendix D: Part Characteristics

Appendix E: Sources of Device Data

E.1 Nuclear and Space Radiation Effects Conference (NSREC)

E.2 Data Sets

References

Author Index

Subject Index

For further information visit: the book web page http://www.openmodelica.org, the Modelica Association web page http://www.modelica.org, the authors research page http://www.ida.liu.se/labs/pelab/modelica, or home page http://www.ida.liu.se/~petfr/, or email the author at [email protected]. Certain material from the Modelica Tutorial and the Modelica Language Specification available at http://www.modelica.org has been reproduced in this book with permission from the Modelica Association under the Modelica License 2 Copyright © 1998–2011, Modelica Association, see the license conditions (including the disclaimer of warranty) at http://www.modelica.org/modelica-legal-documents/ModelicaLicense2.html. Licensed by Modelica Association under the Modelica License 2.

Modelica© is a registered trademark of the Modelica Association. MathModelica© is a registered trademark of MathCore Engineering AB. Dymola© is a registered trademark of Dassault Syst`emes. MATLAB© and Simulink© are registered trademarks of MathWorks Inc. Java is a trademark of Sun MicroSystems AB. Mathematica© is a registered trademark of Wolfram Research Inc.

Copyright © 2011 by the Institute of Electrical and Electronics Engineers, Inc.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey. All rights reserved.

Published simultaneously in Canada.

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, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, [phone], fax [phone]. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, [phone], fax [phone], or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at [phone], outside the United States at [phone] or fax [phone].

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress Cataloging-in-Publication Data:

Single Event Effects in Aerospace / Edward Petersen

ISBN: 978-0-470-76749-8

Chapter 1

Introduction

1.1 Background

In recent years the dominant radiation effect in space-borne electronic systems has become the family of single event effects (SEEs). SEEs arise through the action of a single ionizing particle as it penetrates sensitive nodes within electronic devices. Single events can lead to seemingly randomly appearing glitches in electronic systems— frustrating errors that may cause anything from annoying (at best) system responses to catastrophic (at worst) system failures. The problem is particularly insidious due to the combination of its random nature, the omnipresent spectrum of high energy particles in space, and the increasing sensitivity of devices to SEEs as miniaturization progresses.

A SEE is a phenomenon that follows from the continuing trend in electronic device design toward higher density devices with smaller feature sizes. This trend permits faster processing of information with smaller required quantities of electric charge. As the charge involved has decreased, it has entered the region where corresponding amounts of charge can be generated in the semiconductor by the passage of cosmic rays or alpha particles. This charge can look like a legitimate signal, temporarily changing memory contents or commands in an instruction stream.

The single event upset (SEU) phenomenon was first suggested in 1962 by Walkmark [1962] and first reported in an operating satellite system in 1975 by Binder, Smith, and Holman [1975]. Both of these reports were generally ignored as they suggested responses well out of the mainstream of radiation effect studies of the time. However, in 1978, May and Wood [1979] reported alpha particle upsets in dynamic RAMs and Pickel and Blandford [1978] analyzed upsets in RAM circuits in space due to heavy ion cosmic rays. It was in this time period that IBM started major on-again off-again programs with alpha emitters and terrestrial cosmic rays [1979, 1981, 1996, and 2004]. In 1979 Guenzer, Wolicki, and Allas [1979], and Wyatt, McNulty, and co-workers [1979] experimentally observed upsets due to high energy protons such as those present in the Earth's trapped proton belts. Gradually, as more and more upset related problems have been observed in spacecraft, SEU has come to be recognized as a very serious threat to system operations. Radiation hardening of devices and SEU tolerance approaches have alleviated the problem somewhat. It is still grave. Single event upset must be considered in all future space, missile, and avionics systems.

The early “common knowledge” of the effects was based on two papers. The paper by Binder, Smith, and Homan presented the basic information for cosmic ray induced upsets [1975]. They discussed the basic mechanisms and circuit effects, the cosmic ray environment, including the effects of shielding, and a basic approach to cosmic ray event rate. A paper by Petersen presented the proton environment with the variation in altitude and the effects of the South Atlantic anomaly and included the effects of shielding [1981]. He then discussed the possible contributions of the various proton reactions in silicon and presented calculations of proton induced upset rates.

Much of the interest has been driven by developments such as:

The critical errors caused by cosmic ions in the Voyager and Pioneer probes.The necessary retrofits, at great expense, of the Landsat D and Galileo systems due to heightened concern over single event upsets.The errors in the guidance system of the Hubble space telescope as its orbit carries it through the earth's radiation belts, requiring frequent scrub and reload of the guidance system.The loss of the Japanese satellite “Superbird” due to SEU followed by operator error [1992].

Table 1.1 lists a sampling of other space programs for which single event effects have had an impact. Some of these events were collected by Bedingfield and co-workers [1996]. Ritter has also discussed some of these events [1996].

Table 1.1 A Few of the Spacecraft for Which Single Event Effects Have Had an Impact

There was a parallel set of problems for ground based systems as described by Ziegler [2004]. Soft errors from radiation are the primary limit on digital electronic reliability. This phenomenon is now more important than all other causes of computing reliability put together. Since chip single event rates (SERs) are viewed by many as a legal liability (selling something that you know may fail), the public literature in this field is sparse and always makes management nervous [2004, Preface and Chapter 1].

SEU in space originates from two sources in the natural environment. Satellites at geosynchronous orbit and corresponding regions outside Earth's radiation belts experience upsets due to heavy ions from either cosmic rays or solar flares. The natural cosmic ray heavy ion flux has approximately 100 particles/cm2 per day. In very sensitive devices this flux can lead to daily upsets. Many devices can be upset in these environments at a rate of about 10−6 upsets/bit-day.

Also, upsets can occur within the proton radiation belts. Even though energy loss rates by direct ionization from protons are too low to upset most devices, proton induced nuclear reactions in silicon can result in heavy recoil nuclei capable of upsetting most memory cells. About one proton in 105 will undergo a nuclear reaction capable of SEU. Considering the large population of high energy protons capable of causing these reactions, proton induced upsets become a significant SEU mechanism. In the heart of the proton belts there are about 107 to 109 protons/cm2 per day with energies above 30 MeV (approximately the minimum energy that will penetrate a spacecraft and then cause upsets). Thus, for example, a 1K memory with a proton upset cross section of 10−11 cm2 per bit would have 10 upsets per day in the most intense part of the belt.

As the cosmic rays penetrate the atmosphere, there is a chain of nuclear reactions that produce high energy neutrons and protons. A nominal figure is 6000 neutrons per square centimeter per hour at 40,000 feet altitude and 45 degrees latitude. In the 1990s these were shown to produce single event upsets in complex integrated circuits in avionics equipment.

There are a variety of possible single event effects (SEEs). These are important as they can cause malfunctions in microelectronics devices operating in the space ionizing radiation environment. The principal effects are upset, transients, and latchup, but the others need also to be kept in mind. The basic effects are as follows:

SEUUPSETTemporary change of memory or control bitSETTRANSIENTTransient introduced by single eventSELLATCHUPDevice latches in high current stateSESSNAPBACKRegenerative current mode in NMOSSEBBURNOUTDevice draws high current and burns outSEGRGATE RUPTUREGate destroyed in power MOSFETsSEFIFUNCTIONAL INTERRUPTControl path corrupted by an upsetMBUMULTIBIT UPSETSeveral bits upset by the same event

Single events acquire that name because they depend on the interaction of a single particle. Most other radiation effects depend on the dose or damage deposited by large numbers of particles. SEEs can be caused by the passage of a single heavy ion—a cosmic ray in space, for example. As the cosmic ray passes through the silicon of the device, it deposits a track of ions. In space the cosmic rays are ordinarily energetic enough that they pass through the device. If these resulting ions are in the presence of the natural or applied field in an electronic device, they are collected at the device electrodes. See Figure 1.1. This produces an electric pulse or signal that may appear to the device as a signal to which it should respond. If the electrical characteristics of the device are such that the signal appears valid, then there may be a bit upset or the production of a signal in a logic device that triggers a latch later in the device.

Figure 1.1 Ionization path due to direct passage of a heavy ion.

High energy protons can also initiate single event effects. It is not the proton passage that produces the effect. The proton itself produces only a very small amount of ionization. Very few devices are sensitive enough to respond to the proton ionization. However, 1 proton in 105 will have a nuclear reaction in the silicon device. These reactions can produce heavy ions that in turn can deposit enough energy to cause upset. See Figure 1.2. Although this seems like a very small number of cases, in space the protons in the proton radiation belts are intense enough so that they can cause many more upsets than the heavy ion cosmic rays in the same environment.

Figure 1.2 Ionization paths due to proton reaction in a device.

The basic concepts are similar for both heavy ion and proton induced upsets. The prime emphasis in the present work is the heavy ion induced upsets. We will also discuss proton and neutron upsets for comparison later. The prediction of single event effect rates depends on a number of independent models of various aspects of the phenomena involved.

Single event effects can be thought of as one of nature's ways of enforcing Murphy's Law. They can occur at any time, at any place in an electronic system. They do not depend on the cumulative exposure to the space environment and are as likely to occur during or shortly after launch as after a long time in orbit. As for location, the sorrowful words of one space system designer express it: “I know that you said I was going to have upsets, but I didn't expect an upset in that bit.”

Because of these and other real world problems in space systems due to cosmic ions, an understanding of single particle errors in integrated circuit (IC) electronics has become an important part of the design and qualification of IC parts for space-based use.

The issue becomes even more important as device dimensions scale, and denser, more powerful integrated systems are placed in space or satellite applications. Electronics is reaching integration levels where a single bit of information is represented by an extremely small value of charge, and noise margins are very tight. For example, if a typical dynamic random access memory (DRAM) cell can tolerate approximately 100 mV of noise on the bit storage node with 100 fF (10−15 farads) of storage capacitance, then this value of noise corresponds to a charge of only 62,500 electrons. Any perturbation of this delicate balance by an impinging cosmic ion is intolerable. So, a recognition of, and familiarity with, the effects of space radiation on the electronics to be placed in that hostile environment is essential. Single event modeling plays a key role in the understanding of the observed-error mechanisms in existing systems, as well as the prediction of errors in newly designed systems.

There are two different aspects of interest. First is the analysis of various types of single event experiments to help understand the phenomena. Second is the modeling of the various aspects of the phenomena that allow prediction of SEE rates in space.

1.2 Analysis of Single Event Experiments

1.2.1 Analysis of Data Integrity and Initial Data Corrections

Single event experiments involve a mixture of electronic measurements and nuclear physics experimental techniques. This leads to a number of possible errors in the measurements, some of which are not common in standard electrical engineering experiments. Therefore, the first step in the analysis of SEEs data is the examination of the integrity of the data itself.

There are some aspects of the data that are an artifact of experimental approach. The data needs to be corrected for these before it is used for analysis and prediction.

1.2.2 Analysis of Charge Collection Experiments

The initial studies of SEEs assumed that it took a unique amount of charge to cause the effect. This corresponded to a step function as one examined device sensitivity as a function of energy deposition. The energy deposition was measured by the linear energy transfer (LET) of the ionizing particle (see Section 2.2). However, all of the experiments indicted an “S” curve of the variation of SEE rate as a function of LET. It was initially assumed that this was caused by a variation from memory cell to memory cell. Then a number of charge collection experimental and theoretical studies indicated that the variation in sensitivity corresponded to a variation across the transistor and across the memory cell.

1.2.3 Analysis of Device Characteristics from Cross-Section Data

The SEE cross-section curve is a reflection of the various charge collection processes and circuit characteristics that are involved in the SEE. Study of the characteristics of this curve gives some insight into the processes involved in a particular device.

1.2.4 Analysis of Parametric Studies of Device Sensitivity

There are a large number of device parameters that can be adjusted when testing the device. These may be parameters that affect device design, or factors corresponding to different modes of operation in space. The variation in the experimental cross sections with parametric changes gives a direct measure of the space SEE rates dependence on these factors.

1.3 Modeling Space and Avionics See Rates

1.3.1 Modeling the Radiation Environment at the Device

This involves modeling the proton and heavy ion radiation environment in space and the neutron environment in the atmosphere. (We will often refer to modeling in space, with the understanding that the discussion refers to both space and avionics environments.) The sources are cosmic rays, solar particles, and geomagnetically trapped radiation. Basic factors are intensities, elemental compositions, and energy distributions. The environment at the device is influenced by the component material shielding surrounding the device, geomagnetic shielding effects (which depend on orbit parameters), and time variations associated with the solar cycle. For single event effects by heavy ions, it is really the energy loss characteristics of the environment that are important. Therefore, the cosmic ray composition and energy variation are translated to the number of particles as a function of their energy loss per unit path length (dE/dX), normalized to the material and called linear energy transfer (LET). The LET spectrum can be expressed in either a differential form (number of particles with given energy loss) or in an integral form (number with energy loss greater than a given LET).

1.3.2 Modeling the Charge Collection at the Device

The upset depends on the amount of charge that the electronic circuit detects. This in turn depends on the amount of charge deposited by the ion. It is not sufficient for an ion to hit a device or an individual nucleus in the device, but instead it must have an appreciable track length. The amount of charge deposited depends on the product of the path length and the LET of the ion. Therefore, one must model the sensitive volume of the device. That is, you must know the area, depth, and shape of the sensitive region. With this information, the distribution of possible path lengths in the device can be modeled. These path length distributions can be expressed in either a differential or an integral form, that is, number of paths of a given length, or number of paths that equal or exceed a given length.

The charge collection may be complicated by the fact that there are several possible charge collection processes that can take place. The charge may be collected by the intrinsic and applied field along the track in the device itself. There may be additional charge collected as charge diffuses into the device. It is also possible that the fields are distorted along the ion track, leading to additional charge in the track being brought into the device.

1.3.3 Modeling the Electrical Characteristic and Circuit Sensitivity for Upset

The circuit that the device is embedded in determines what the applied pulse shape must be if it is to appear as a legitimate electrical signal. In most cases, these effects can be modeled by standard SPICE circuit modeling of the circuit. In some case, there are more complicated interactions between the charge collection and the circuit, so that the effects are better modeled if considered together.

It is not always possible to obtain enough information about the detailed device characteristics to perform electrical modeling. It has now become common to base the estimates of device sensitivity on experimental upset measurements in the laboratory.

The prediction of rate effects that occur in space then depends on models that include three basic factors: the environment, the device dimensions, and the device sensitivity. There are possible complications that involve the details of the shape of the device sensitive volumes and of the charge collection. We will analyze the various approaches to upset rate predictions, and examine how they consider both the basic factors and the complications.

One of the modeling complications that we will discuss is the treatment of charge collection in the device. The basic aspects of charge collection are included in all models. However, there are a number of possible complications that can have a significant impact on the interpretation of experimental data. These factors can also sometimes impact the rate predictions. These factors will be introduced where appropriate.

1.4 Overview of this Book

We present an overview of the various factors and approaches that are important for single event effect measurements, rate modeling, and predictions. We then develop the concepts that are important for rate prediction, develop the history of rate predictions, outline the approaches that have developed, and present standards that are being developed. We have an extensive discussion of the interpretation of experimental results. We will also discuss some of the issues that remain.

One of the possible complications is the charge collection in the device. The basic aspects of charge collection are included in all models. We will consider the charge collection as it shows up in the cross-section measurements, but not consider the various detailed micromodeling approaches to the problem.

The book presents an overview of the methods and procedures involved in computer modeling of single event phenomena that have been proposed and utilized in recent years. The goal of these procedures is to model the interaction of a radiation environment with microelectronic circuits, and to predict the resulting influences on proper IC operation.

This book is based on published materials but does include some new material and some new examples. We hope that some of the examples will be helpful for newcomers in the field. Other examples are aimed at those experienced in the field who we believe have misunderstood some of the concepts of the integral rectangular parallelepiped (IRPP) approach to upset rate calculations.

1.5 Scope of this Book

This book is intended to be a tutorial, covering the basic terminology and concepts of single event effects and rate prediction. It is based on the notes for 2008 NSREC Short Course with additions, corrections, and indexes [2008]. It is hoped that it presents enough information for the reader to find the relevant literature of a specific topic of interest so that he/she can study it in depth. The book will go into some depth on the issues that lay at the foundation of the subject or that are important for SEU measurements or for the interpretation of SEU measurements. Attention is paid to a number of experimental aspects from nuclear physics that may not be covered in electrical engineering courses. In some sections we may go into too much detail for the reader. Skip those sections until you need them.

There is little discussion of the actual effects in devices or circuits. It is possible to write a book or major review article on modeling and simulation of the electronic aspects of single event effects that has virtually no overlap with this book [2004, 2008, 2008, 2008]. The electronic modeling has been well presented in an IEEE NSREC (Institute of Electrical and Electronics Engineers, Nuclear and Space Radiation Effects Conference) short course by Lloyd Massengill [1993]. The short course presentation of Sexton is also highly recommended [1992]. A very good update of Massengill is given by Black and Holman [2006]. Reed discusses some of the basic physics underlying the electrical effects [2008].

There is a wealth of information in important IEEE NSREC short course presentations and other reviews. [Benedetto 2008, Bourdarie 2008, Hafer 2008, Munteanu 2008, Petersen 2008b, Reed 2008, Wilkinson 2008, Kastensmidt 2007, Ladbury 2007a, Black 2006, Law 2006, Santin 2006, Xapsos 2006b, Bauman 2005, Buchner 2005, Cressler 2003, Oldham 2003, Weatherford 2002, Buchner 2001, Hoffmann 2000, Dodd 1999, Dressendorfer 1998, Barth 1997, Petersen 1997b, Alexander 1996, Galloway 1996, Ritter 1996, Stapor 1995, Normand 1994a, Massengill 1993c, Sexton 1992, McNulty 1990, Petersen 1983c, Pickel 1983]. The reader who wants a comprehensive background in single event effects should study these. At the same time, be aware that our knowledge has been changing continuously. The material in this book has been changed significantly from the corresponding material in 1997 [1997].