IEC TS 62396-2:2008

IEC/TS 62396-2
Edition 1.0 2008-08
TECHNICAL
SPECIFICATION
Process management for avionics – Atmospheric radiation effects –
Part 2: Guidelines for single event effects testing for avionics systems

IEC/TS 62396-2:2008(E)
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by
any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from either IEC or
IEC's member National Committee in the country of the requester.
If you have any questions about IEC copyright or have an enquiry about obtaining additional rights to this publication,
please contact the address below or your local IEC member National Committee for further information.

Droits de reproduction réservés. Sauf indication contraire, aucune partie de cette publication ne peut être reproduite
ni utilisée sous quelque forme que ce soit et par aucun procédé, électronique ou mécanique, y compris la photocopie
et les microfilms, sans l'accord écrit de la CEI ou du Comité national de la CEI du pays du demandeur.
Si vous avez des questions sur le copyright de la CEI ou si vous désirez obtenir des droits supplémentaires sur cette
publication, utilisez les coordonnées ci-après ou contactez le Comité national de la CEI de votre pays de résidence.

IEC Central Office
3, rue de Varembé
CH-1211 Geneva 20
Switzerland
Email: inmail@iec.ch
Web: www.iec.ch
About IEC publications
The technical content of IEC publications is kept under constant review by the IEC. Please make sure that you have the
latest edition, a corrigenda or an amendment might have been published.
ƒ Catalogue of IEC publications: www.iec.ch/searchpub
The IEC on-line Catalogue enables you to search by a variety of criteria (reference number, text, technical committee,…).
It also gives information on projects, withdrawn and replaced publications.
ƒ IEC Just Published: www.iec.ch/online_news/justpub
Stay up to date on all new IEC publications. Just Published details twice a month all new publications released. Available
on-line and also by email.
ƒ Electropedia: www.electropedia.org
The world's leading online dictionary of electronic and electrical terms containing more than 20 000 terms and definitions
in English and French, with equivalent terms in additional languages. Also known as the International Electrotechnical
Vocabulary online.
ƒ Customer Service Centre: www.iec.ch/webstore/custserv
If you wish to give us your feedback on this publication or need further assistance, please visit the Customer Service
Centre FAQ or contact us:
Email: csc@iec.ch
Tel.: +41 22 919 02 11
Fax: +41 22 919 03 00
IEC/TS 62396-2
Edition 1.0 2008-08
TECHNICAL
SPECIFICATION
Process management for avionics – Atmospheric radiation effects –
Part 2: Guidelines for single event effects testing for avionics systems

INTERNATIONAL
ELECTROTECHNICAL
COMMISSION
PRICE CODE
U
ICS 03.100.50; 31.020; 49.060 ISBN 2-8318-9958-3
– 2 – TS 62396-2 © IEC:2008(E)
CONTENTS
FOREWORD.3
INTRODUCTION.5
1 Scope.6
2 Normative references .6
3 Terms and definitions .6
4 Abbreviations used in the document .6
5 Obtaining SEE data .7
5.1 Types of SEE data .7
5.2 Use of existing SEE data.7
5.3 Deciding to perform dedicated SEE tests.8
6 Availability of existing SEE data for avionics applications .8
6.1 Variability of SEE data .8
6.2 Types of existing SEE data that may be used.8
6.2.1 Sources of data, proprietary versus published data .9
6.2.2 Data based on the use of different sources.11
6.2.3 Ground level versus avionics applications .14
6.3 Sources of existing data .15
7 Considerations for SEE testing .16
7.1 General .16
7.2 Selection of hardware to be tested .17
7.3 Selection of test method.17
7.4 Selection of facility providing energetic particles .18
7.4.1 Radiation sources.18
7.4.2 Spallation neutron source .18
7.4.3 Monoenergetic and quasi-monoenergetic beam sources.19
7.4.4 Thermal neutron sources .20
8 Converting test results to avionics SEE rates .20
8.1 General .20
8.2 Use of spallation neutron source .20
8.3 Use of SEU cross section curve over energy .21

Bibliography.24

Figure 1 – Comparison of Los Alamos and TRIUMF neutron spectra with terrestrial
neutron spectrum.12
Figure 2 – Variation of high energy neutron SEU cross section per bit as a function of
device feature size.13
Figure 3 – Comparison of mono-energetic SEU cross sections with Weibull and Piece-
Wise Linear Fits.23

Table 1 – Sources of existing data .16

TS 62396-2 © IEC:2008(E) – 3 –
INTERNATIONAL ELECTROTECHNICAL COMMISSION
____________
PROCESS MANAGEMENT FOR AVIONICS –
ATMOSPHERIC RADIATION EFFECTS –

Part 2: Guidelines for single event effects
testing for avionics systems
FOREWORD
1) The International Electrotechnical Commission (IEC) is a worldwide organization for standardization comprising
all national electrotechnical committees (IEC National Committees). The object of IEC is to promote
international co-operation on all questions concerning standardization in the electrical and electronic fields. To
this end and in addition to other activities, IEC publishes International Standards, Technical Specifications,
Technical Reports, Publicly Available Specifications (PAS) and Guides (hereafter referred to as “IEC
Publication(s)”). Their preparation is entrusted to technical committees; any IEC National Committee interested
in the subject dealt with may participate in this preparatory work. International, governmental and non-
governmental organizations liaising with the IEC also participate in this preparation. IEC collaborates closely
with the International Organization for Standardization (ISO) in accordance with conditions determined by
agreement between the two organizations.
2) The formal decisions or agreements of IEC on technical matters express, as nearly as possible, an international
consensus of opinion on the relevant subjects since each technical committee has representation from all
interested IEC National Committees.
3) IEC Publications have the form of recommendations for international use and are accepted by IEC National
Committees in that sense. While all reasonable efforts are made to ensure that the technical content of IEC
Publications is accurate, IEC cannot be held responsible for the way in which they are used or for any
misinterpretation by any end user.
4) In order to promote international uniformity, IEC National Committees undertake to apply IEC Publications
transparently to the maximum extent possible in their national and regional publications. Any divergence
between any IEC Publication and the corresponding national or regional publication shall be clearly indicated in
the latter.
5) IEC provides no marking procedure to indicate its approval and cannot be rendered responsible for any
equipment declared to be in conformity with an IEC Publication.
6) All users should ensure that they have the latest edition of this publication.
7) No liability shall attach to IEC or its directors, employees, servants or agents including individual experts and
members of its technical committees and IEC National Committees for any personal injury, property damage or
other damage of any nature whatsoever, whether direct or indirect, or for costs (including legal fees) and
expenses arising out of the publication, use of, or reliance upon, this IEC Publication or any other IEC
Publications.
8) Attention is drawn to the Normative references cited in this publication. Use of the referenced publications is
indispensable for the correct application of this publication.
9) Attention is drawn to the possibility that some of the elements of this IEC Publication may be the subject of
patent rights. IEC shall not be held responsible for identifying any or all such patent rights.
The main task of IEC technical committees is to prepare International Standards. In
exceptional circumstances, a technical committee may propose the publication of a technical
specification when
• the required support cannot be obtained for the publication of an International Standard,
despite repeated efforts, or
• the subject is still under technical development or where, for any other reason, there is the
future but no immediate possibility of an agreement on an International Standard.
Technical specifications are subject to review within three years of publication to decide
whether they can be transformed into International Standards.
IEC 62396-2, which is a technical specification, has been prepared by IEC technical
committee 107: Process management for avionics.

– 4 – TS 62396-2 © IEC:2008(E)
This standard cancels and replaces IEC/PAS 62396-2 published in 2007. This first edition
constitutes a technical revision.
The text of this standard is based on the following documents:
Enquiry draft Report on voting
107/80/DTS 107/86/RVC
Full information on the voting for the approval of this standard can be found in the report on
voting indicated in the above table.
This publication has been drafted in accordance with the ISO/IEC Directives, Part 2.
A list of all parts of the IEC 62396 series, under the general title Process management for
avionics – Atmospheric radiation effects, can be found on the IEC website.
The committee has decided that the contents of this publication will remain unchanged until
the maintenance result date indicated on the IEC web site under "http://webstore.iec.ch" in
the data related to the specific publication. At this date, the publication will be
• transformed into an International standard,
• reconfirmed;
• withdrawn;
• replaced by a revised edition, or
• amended
A bilingual edition of this document may be issued at a later date.

TS 62396-2 © IEC:2008(E) – 5 –
INTRODUCTION
This industry-wide technical specification provides additional guidance to avionics systems
designers, electronic equipment component manufacturers and their customers to determine
the susceptibility of microelectronic devices to single event effects. It expands on the
information and guidance provided in IEC/TS 62396-1.
Guidance is provided on the use of existing single event effects (SEE), SEE data, sources of
data and the types of accelerated radiation sources used. Where SEE data is not available
considerations for testing is introduced including the suitable radiation sources for providing
avionics SEE data. The conversion of data obtained from differing radiation sources into
avionics SEE rates is detailed.

– 6 – TS 62396-2 © IEC:2008(E)
PROCESS MANAGEMENT FOR AVIONICS –
ATMOSPHERIC RADIATION EFFECTS –

Part 2: Guidelines for single event effects
testing for avionics systems
1 Scope
The purpose of this technical specification is to provide guidance related to the testing of
microelectronic devices for purposes of measuring their susceptibility to single event effects
(SEE) induced by the atmospheric neutrons. Since the testing can be performed in a number
of different ways, using different kinds of radiation sources, it also shows how the test data
can be used to estimate the SEE rate of devices and boards due to the atmospheric neutrons
in the atmosphere at aircraft altitudes.
2 Normative references
The following referenced documents are indispensable for the application of this document,
only the edition cited applies. For undated references, the latest edition of the referenced
document (including any amendments) applies.
IEC/TS 62396-1, Process management for avionics – Atmospheric radiation effects – Part 1:
Accommodation of atmospheric radiation effects via single event effects within avionics
electronic equipment
3 Terms and definitions
For the purpose of this document, the terms and definitions of IEC/TS 62396-1 apply.
4 Abbreviations used in the document
BPSG Borophosphosilicate glass
CMOS Complimentary metal oxide semiconductor
COTS Commercial off-the-shelf
D-D Deuterium-deuterium
DRAM Dynamic random access memory
D-T Deuterium-tritium
DTS Draft technical specification
E
Energy
ESA European Space Agency
eV electron volt
FPGA Field programmable gate array
GeV Giga electron volt
GV Giga volt (rigidity unit)
IBM International Business Machines
ICE Irradiation of Chips and Electronics
IEEE Trans. Nucl. Sci. IEEE Transactions on Nuclear Science
JEDEC JEDEC Solid State Technology Association
JESD JEDEC standard
TS 62396-2 © IEC:2008(E) – 7 –
JPL Jet Propulsion Laboratory
LET Linear energy transfer
LETth Linear energy transfer threshold
MBU Multiple bit upset
MeV Mega electron volt
NASA National Aeronautical and Space Agency
RADECS Radiations, effets sur les composants et systèmes.
RAM Random access memory
RVC Result of voting (IEC)
SDRAM Synchronous dynamic random access memory
SEB Single event burn-out
SEE Single event effect
SEFI Single event functional interrupt
SEGR Single event gate rupture
SEL Single event latch
SEP Solar energetic particles
SER Soft error rate
SET Single event transient
SEU Single event upset
SHE Single event induced hard error
SRAM Static random access memory
SW Software
TID Total ionizing dose
TRIUMF Tri-University Meson Facility (Canada)
TSL Theodore Svedberg Laboratoriet (Sweden)
WNR Weapons Nuclear Research (Los Alamos USA)
5 Obtaining SEE data
5.1 Types of SEE data
The type of SEE data available can be viewed from many different perspectives. As indicated,
the SEE testing can be performed using a variety of radiation sources, all of which can induce
single event effects in ICs. In addition, many tests are performed on individual devices, but
some tests expose an entire single board computer to radiation fields that can induce SEE
effects. However, a key discriminator is deciding on whether existing SEE data that may be
used is available, or whether there really is no existing data and therefore a SEE test on the
device or board of interest has to be carried out.
5.2 Use of existing SEE data
The simplest solution is to find previous SEE data on a specific IC device. This is not nearly
as simple as it appears. First, the largest interest lies in SEE data that is directly usable for
purposes of estimating the SEE rate in avionics. Thus, SEE tests that have been carried out
on devices using heavy ions, data which is directly applicable for space missions, is data that
is not directly applicable for avionics purposes. This heavy ion SEE data can be used to
calculate SEE data from high energy neutrons and protons by utilizing a number of different
calculation methods, but this requires the active involvement of a radiation effects expert in
the process. Therefore, heavy ion SEE data should not be used for application to the
atmospheric neutron environment, except by scientists and engineers who have extensive
experience in using this kind of data. For that reason, unless otherwise stated explicitly, when
SEE data is discussed in the remainder of this technical specification, it refers only to single
event testing using a neutron or proton source, not to the results from testing with heavy ions.

– 8 – TS 62396-2 © IEC:2008(E)
If SEE data on a device of interest is found from SEE tests using high energy neutrons or
protons, it will still require expertise regarding how the data is to be utilized in order to
calculate a SEE rate at aircraft altitudes. Data obtained by IC vendors for their standard
application to ground level systems are often expressed in totally different units, FIT units,
where one FIT is one error in 10 device hours, which is taken to apply at ground level.
IC devices are constantly changing. In some cases, devices which had been tested, become
obsolete and are replaced by new devices which have not been tested. The fact that a device
is made by the same IC vendor and is of the same type as the one it replaced does not mean
that the SEE data measured in the first device applies directly to the newer device. In some
cases, small changes in the IC design or manufacturing process can have a large effect in
altering the SEE response, but in other cases, the effect on the SEE response may be
minimal.
A continuing problem with the existing SEE data is that there is no single database that
contains all of the neutron or proton SEE data. Instead, portions of this kind of SEE data can
be found published in many diverse sources. The SEE data in the larger databases is mainly
on much older devices, dating from the 1990s and even 1980s, and is primarily from heavy
ion tests that were performed for space applications and not from testing with protons and
neutrons.
5.3 Deciding to perform dedicated SEE tests
If existing SEE data is not available, for any one of the many reasons discussed above and
which will be further expanded upon below, then there is no real alternative but to carry out
one’s own SEE testing. The advantage of such a test is that it pertains to the specific device
or board that is of interest, but the disadvantage is that it entails making a number of
important decisions on how the testing is to be carried out. These pertain to selecting the
most useful test article (single chip or entire board), nature of the test (static or dynamic
(mainly applicable to board testing)), assembling a test team, choosing the facility that
provides the best source of neutrons or protons for testing, scheduling and performing the test,
coping with uncertainties that appear during the test and, finally, using the test results to
calculate the desired SEE rate for avionics. Many of these issues will be discussed in the
following clauses.
6 Availability of existing SEE data for avionics applications
6.1 Variability of SEE data
Because of the diverse ways that SEE testing is carried out, and the multitude of venues for
how and where such data is published, the availability of SEE data for avionics applications is
not a simple matter.
6.2 Types of existing SEE data that may be used
6.2.1 General
SEE data can be derived from a number of different kinds of tests, and all of the differences
between these tests need to be understood in order to make comparisons meaningful.
Although there are many different types of single event effects, for the purposes of this
technical specification, the focus is on three of them: single event upset (SEU), single event
functional interrupt (SEFI) and single event latchup (SEL). SEU pertains to the energy
deposited by an energetic particle leading to a single bit being flipped in its logic state. The
main types of devices that are susceptible to SEU are random access memories (RAMs, both
SRAMs and DRAMs), field programmable gate arrays (FPGAs, especially those using SRAM-
based configuration) and microprocessors (the cache memory and register portions). A SEFI
refers to a bit flip in a complex device that results in the device itself or the board on which it
is operating not functioning properly. A typical example is an SEU in a control register, which
can affect the device itself, but can also be propagated to another device on the board,
leading to board malfunction. SEL refers to the energy deposited in a CMOS device that leads

TS 62396-2 © IEC:2008(E) – 9 –
to the turning on of a parasitic p-n-p-n structure, which usually results in a high current in the
device and a non-functioning state. The high energy neutrons in the atmosphere can induce
all of these effects: SEU, SEFI and SEL. Where semiconductor devices are operated at high
voltage stress (200 V and above) they may be subject to single event burn-out, SEB or single
event gate rupture, SEGR; these effects are covered in detail in IEC/TS 62396-4
One of the important simplifying assumptions to be used in this technical specification is that, for
single event effects, including SEU, SEFI and SEL, the response from high energy protons, i.e.,
those with E > 100 MeV, is the same as that from high energy neutrons of the same energy. The
SEE response is generally measured in terms of a cross section (cm²/dev), which is the number of
errors of a given type divided by the fluence of particles to which the device was exposed.
Therefore, for the SEU, SEFI and SEL cross sections, measurements made with high energy
protons can be used as the same cross section from the atmospheric neutrons. This is far more
than an assumption, since it has been demonstrated by direct measurement in many different
1)
devices see [1] to [5] and IEC/TS 62396-1. In these references, SEU was measured in the
same devices using monoenergetic proton beams and using the neutron beam from the Weapons
Neutron Research (WNR) facility at the Los Alamos National Laboratory. The energy spectrum of
the neutrons in the WNR is almost identical to the spectrum of neutrons in the atmosphere. An
estimate of the SEE rate at aircraft altitudes in a device can be obtained by the simplified equation:
2 2
SEE rate per device = 6 000 [n/cm h] × avionics SEE cross section [cm per device] (1)
Here, the integral neutron flux in the atmosphere, E >10 MeV, is taken to be 6 000 n/cm h,
the approximate flux at 40 000 ft (12,2 km) and 45° latitude as in IEC/TS 62396-1. This shows
the importance of the SEE cross section. As indicated above, the avionics SEE cross section
is taken to be the SEE cross section obtained from SEE tests with a spallation neutron source
such as the WNR, and also with a proton or neutron beam at energies > 100 MeV. The
simplified approach of Equation (1) is used in IEC/TS 62396-1and is the nominal flux under
the above conditions.
A more elaborate approach for calculating the SEE rate is to utilize a number of
measurements of the SEE cross section as a function of neutron or proton energy, and
integrate the curve of the SEE cross section over energy with the differential neutron flux. The
details for this approach are given in the standard JESD-89A [6], although the neutron flux
given in this standard is at ground level and would have to be multiplied by approximately a
factor of 300 to make it relevant to avionics applications (see 6.2.3).
Thus the data that is most valuable for estimating the SEE rate in avionics is from SEE cross
section measurements made with: a) a spallation neutron source such as the WNR, b) a
monoenergetic proton beam and c) a quasi-monoenergetic neutron beam. Other SEE data
that are also valuable are SEU cross sections made with a monoenergetic 14 MeV neutron
beam. Based on comparisons of SEU cross section measurements with a 14 MeV neutron
beam and the WNR, the WNR SEU cross section is approximately a factor of 1,5 to 2 higher
than the 14 MeV SEU cross section for relatively recent devices ([3], feature size < 0,5 μm),
and a factor of 4 times higher for older devices [4]. For some of the very latest devices, the
factor is close to 1.
6.2.2 Sources of data, proprietary versus published data
As indicated above, SEE cross section measurements that are relevant to avionics SEE rates
are being made by a variety of different groups. These include:
a) Space organizations that use only monoenergetic proton beams for their SEE testing,
b) IC vendors who use neutron sources to measure the upset rate at ground level [which they
refer to as the soft error rate (SER), rather than the SEU rate, although the terms have the
same meaning],
___________
1)
Numbers in square brackets refer to the bibliography.

– 10 – TS 62396-2 © IEC:2008(E)
c) Avionics vendors who use neutron sources to measure the upset rate at aircraft levels.
Generally, SEE data taken and reported by government agencies contains most if not all of
the relevant information, including identifying the specific IC devices tested and providing the
measured SEU cross sections in unambiguous units. This applies to most of the proton data
taken and reported by NASA in the open literature by the NASA centres at GSFC and JPL.
GSFC and JPL invariably publish almost all of the proton SEE data that they take. However,
even though they disseminate essentially all of the results from the proton SEE testing that
they carry out, this is data that is usually reported in the open literature in an inclusive
compilation that contains results from SEE testing with both heavy ions and protons, thus the
proton SEE data has to be carefully sought out. Examples of the most recent NASA-GSFC
compilations of SEE testing containing proton SEE test results are given in [7-10], and
examples of JPL reports of SEE testing containing proton SEE test results are given in [11-13].
Other governmental agencies do not necessarily publish the results from all of the proton SEE
tests that they perform.
Data from the other sources, primarily private companies, is not nearly as accessible. IC
vendors perform a large number of tests, but only a small fraction of that data is reported
upon in the open literature. Furthermore, when the SEE data from IC vendors is published,
the results are often disguised, so that the identity of the devices tested or the part number
are usually hidden by using an arbitrary designation and the results are expressed in units
that are ambiguous at best and often of little use quantitatively. Sometimes, the data is
expressed in FIT units, which means errors per 10 device hours; however, this does not
incorporate information on how many bits are included in the device. If only the FIT value is
given, this can be converted into a SEE cross section by using the FIT definition and dividing
by 14 (14 n/cm ·h is the flux of high energy neutrons (E > 10 MeV) at ground level in New
York City, which is the value recommended by the JESD-89A standard and so most often
–9 2
used.) Thus, FIT×10 /14 gives the SEE cross section in cm /device.
Some reports give the SER rate in units of FIT/Mbit, which allows the SEE cross section per
–15
bit to be calculated by multiplying as follows (FIT/Mbit) × 10 /14 to obtain the SEE cross
section in cm /bit. Other papers report the FIT value in arbitrary units (a.u.) which allows the
authors to show how the FIT rate varies with a particular parameter (e.g., applied voltage),
but it allows no quantitative assessment to be made of the SEE cross section. Examples of
such reports using FIT rates are given in [2], [14] to [17].
Most of the SEE data that we have been discussing comes from the SEE testing of individual
components, placing those devices in a beam of neutrons or protons and monitoring changes
in the status of the device for errors. A typical procedure is to fill a portion of memory in a
RAM with a specified bit pattern and monitor that memory for bit flips in one or more
addresses. However, some tests are done using an entire board to monitor when an error has
occurred. In this case, the malfunction of the board is an indication that an error has occurred,
and such an error is referred to as a SEFI, but the functional interruption is in the board rather
than the actual device being irradiated. If the beam is collimated such that only one or two
devices are exposed to the particles in the beam during each test, the likely source of error is
a SEE error in those devices. However, this is a dynamic type of test and it may be that the
device in the beam experienced the initial error which was propagated to another device on
the board, and faulty performance of the latter device is what led to the board malfunctioning.
There are some reports of such board level tests in the open literature, but they are less
common. NASA-JSC has a requirement to perform such testing on all electronic boards that
will be going on the Space Shuttle and related programs. This testing is carried out with a
beam of protons, and while it is recorded in a NASA-JSC report, these reports are not widely
available, examples are given in [18] to [20]. Furthermore, the main purpose of the test is to
screen all of the devices for the potential of a hard error induced by the protons, such as a
single event latchup, so recoverable errors are not analyzed in great detail in these reports.
Other government agency groups also perform such board level SEE testing, and the results
of these tests are often reported in the literature, but are not included in any organized
database. In addition, private companies carry out such board level testing, often for the
benefit of specific programs for avionics applications (neutron tests for avionics vendors) or

TS 62396-2 © IEC:2008(E) – 11 –
space applications (proton tests for low earth orbit spacecraft contractors), and this data is
rarely reported in the open literature.
6.2.3 Data based on the use of different sources
6.2.3.1 Obtaining SEE data using radiation sources
In general, all SEE testing is carried out using an accelerated source of neutrons or protons,
meaning that the device or board to be tested will receive a larger fluence of particles over a
given period of time in the test environment compared to the fluence it would receive during
that same time period in the intended vehicle in the atmosphere or space. In the past, testing
was usually carried out with only one type of source, but in recent times, some engineering
groups have been exposing devices to more than one type of particle environment and
comparing the SEE responses. Two main types of sources have been used for this SEE
testing for avionics applications, neutrons and protons, although there are a variety of
different kinds of neutron sources that have been used, as will be discussed below.
6.2.3.2 Data obtained using neutron sources
Single event effects, in particular, single event upset, can be induced by neutrons in two
distinct energy ranges, at high energies and at very low energies, called thermal neutron
energy. The high energy neutrons cause the SEU by the nuclear reaction with the silicon in
the IC that creates a recoil, and it is the energy from this recoil that is locally deposited in
other silicon atoms that directly causes the upset. For the purposes of simplification, neutrons
with energies > 10 MeV are of greatest concern, but it is true that neutrons with lower
energies, e.g. (2 to 3) MeV, can also cause SEUs. However, since the SEU cross section for E
< 10 MeV is considerably lower than the cross section for E > 10 MeV, 10 MeV is used as an
effective cut-off. Estimates of the SEU contribution for electronics technology with geometry
greater than 0,2 μm by neutrons with E < 10 MeV to the total SEU rate from the entire WNR
neutron spectrum is < 10 %, but for lower feature sizes, this fraction is expected to increase.
This is roughly consistent with SEU measurements made with monoenergetic neutrons (3 and
14 MeV) on devices of the mid 1990’s (feature size greater than 0,5 μm), showing that the
SEU cross section at 3 MeV for these older devices was about a factor of 100 lower than that
at 14 MeV for most of the SRAMs tested [21]. However, for more recent devices, especially
μm and even down to 45 nm, the contribution of
those with feature sizes less than 0,2
neutrons with energies below 10 MeV, is expected to be in the (8 to 10) % range.
For high energy neutrons, there are three different types of sources:
a) a spallation neutron source which has neutrons with energies over a wide energy
spectrum similar to that of the atmospheric neutrons,
b) a quasi-monoenergetic neutron source that has a peculiar energy spectrum, roughly half
of the neutrons are at a peak energy and the other half are evenly distributed between
close to the peak and ~1 MeV, and
c) a 14 MeV neutron generator, the only source that is close to being truly monoenergetic.
The WNR at Los Alamos which was mentioned previously is the best example of a spallation
neutron source, although the neutron irradiation facility at TRIUMF (Tri University Meson
Facility, in Vancouver, Canada) is another such source. Since the WNR facility was upgraded
around the year 2000, it is sometimes referred to by its new name, the ICE (Irradiation of
Chips and Electronics) House [22]. Figure 1 compares the neutron spectra from Los Alamos
(the ICE House), the neutron facility at TRIUMF and the atmospheric neutron spectrum at
ground level [23], [24].
– 12 – TS 62396-2 © IEC:2008(E)

KEY
10 FACILITY Multiplication
Factor
Ground
Line plot is
Spectrum
9 ground level
[latitude
multiplied by
45º North] 8
3 × 10
ICE House 1
(WNR)
Measured
Spectrum
TRIUMF at 1
100 μA
0 1 2 3
10 10 10 10
Neutron energy  (MeV)
IEC  1366/08
Figure 1 – Comparison of Los Alamos and TRIUMF neutron spectra
with terrestrial neutron spectrum
SEU data on devices that were exposed to the WNR neutron beam have been published in a
number of papers [2] to [4], [25] to [26], however, many more devices have been tested at Los
Alamos and those results are considered to be proprietary. These results have not been
published, nor are they expected to be published. Reference [27] indicates that in the year
2001, at least eight different groups carried out SEE testing, and of these, we estimate that
maybe two of the testing groups may publish some of their results, an American national
laboratory and a university. The six private companies, both IC manufacturers and avionics
vendors, will keep their test results proprietary.
The TRIUMF facility in Canada, called the TNF (TRIUMF Neutron Facility) also provides a
spallation neutron source. Until 2004, it had received limited use, but since that time, a
number of papers on SEU results from the testing of IC devices at the TNF have been
published [28].
There are a number of quasi-monoenergetic neutron sources around the world, including
some in the United States of America, but until recently they had not been used for testing
microelectronics for SEE. The site with the most experience with such tests is the Theodor
Svedberg Laboratory (TSL) at Uppsala University, Uppsala, Sweden [29]. A few papers have
been published reporting on the results of microelectronics devices being exposed to the TSL
neutron beam [5], [30], [31]. Methodologies have been developed for extracting SEU cross
section data at the pseudo-peak energy [30], [31]. In addition, a similar facility has been
operating in Japan at Tohoku University [32] which also has been used to make some SEU
measurements. A different methodology from that of the Swedish researchers has been
developed for extracting SEU cross section data at the pseudo-peak energy [33], [34].
In Figure 2, we combine SEU measurements made by several different groups at these
various facilities to illustrate how the high energy SEU cross section per bit for SRAMs has
varied with feature size over the last 5 or more years. The trend that is illustrated in Figure 2
shows a consistency within an approximate plateau region of 10 to 30 times between
maximum and minimum values, however we cannot predict how this might change in the
future, as feature sizes continue to decline below 0,1 μm.
Differential neutron flux  (n/cm MeVh)

TS 62396-2 © IEC:2008(E) – 13 –
–12
KEY
Symbol Data
Intel
–13
microprocessors, L1
Data Cache
SRAMs – Granlund
(>2000) Various
Vendors
SRAMs – Slayman
–14
(>2000) Various
Vendors
–15
0 0,1 0,2 0,3 0,4 0,5
Feature size of process  (μm)
IEC  1367/08
Figure 2 – Variation of high energy neutron SEU cross section per bit
as a function of device feature size
The third kind of high neutron facility is one that provides essentially monoenergetic neutrons,
and 14 MeV, from the D-T reaction, is the highest energy of such a monoenergetic neutron
beam. A number of facilities in the United States and abroad have such neutron generators.
Tests on SRAM devices fabricated in the mid-1990s indicated that the SEU response per bit
from a spallation neutron source was 3 to 5 times higher than from a 14 MeV neutron source
[4]. Tests on more recent devices have shown a closer agreement in the SEU response
between a spallation neutron source and 14 MeV neutron sources [3], [5]. This indicates that
for current, low voltage devices, 14 MeV neutrons provide a fairly good simulation of the
atmospheric neutrons with respect to inducing SEUs. However, 14 MeV neutrons do not
provide a good simulation with respect to inducing single event latchup (SEL) [35].
In 2006 and 2007, it has been shown [36], [37] that for devices with feature sizes smaller than
0,25 µm [37], neutrons with lower energies, between (3 to 10) MeV, are much more
susceptible to SEU than was the case in older technology devices. Previously, the
contribution of such lower energy neutrons had been largely ignored, since it was very small.
For future devices with even smaller feature sizes (< 90 nm), the contribution to the SEU rate
from these lower energy neutrons is likely to grow, and so SEU testing of such devices using
neutron sources covering this energy range [35], [36] may be needed to accurately assess the
SEU rate.
Furthermore, the extrapolation of data points in curves that display trends in SEE
susceptibility, such as Figure 2, to future reduced feature sizes is not warranted without
newer data to back it up. The situation of the higher SEU susceptibility to neutrons in the
(3 to 10) MeV range is one such example showing that extrapolations are not justified
because of the potential for new SEE susceptibilities that have not been observed in older
devices. This is also true for other SEE effects, such as SEL and SEFI in a number of
different types of devices. Only through a continuing commitment to updating trend curves like
Figure 2 with data on newer devices can the user be assured of bounding SEE susceptibilities
of future devices.
There is a fourth type of neutron facility that should be considered for testing devices for
inducing SEUs: that of thermal neutrons. Thermal neutrons cause SEUs through the neutron
reactions with the isotope Boron, which can be present in high enough concentrations to be
of concern mainly as a constituent of the glassivation layer above an IC, i.e., in BPSG
(borophosphosilicate glass). Many devices use a different type of glassivation (e.g., PSG) and
11 10
in some cases, the boron in the BPSG is Boron, so there are no B reactions leading to
7 10
SEU from the reaction products (alpha particle and Li) of the B interaction. A limited
Atm neutron SEU cross section  (cm /bit)

– 14 – TS 62396-2 © IEC:2008(E)
amount of data has been published on the SEU cross section induced by thermal neutrons [5,
16, 38].
6.2.3.3 Data obtained using proton sources
It was demonstrated nearly 25 years ago [39] that high energy protons cause SEUs in
microelectronics. It was also recognized that at high energies, the protons, even though they
are charged particles, cause the upsets by the same mechanism as the high energy neutrons,
by nuclear reactions with the silicon, rather than by direct ionization in the silicon. Proton SEU
cross sections have therefore been published over the years, but the effectiveness of the low
energy protons in causing upsets has increased over time, as the applied voltage to the ICs
has decreased below 5 V. Thus, for DRAMs made during the 1980s and tested with protons,
the SEU cross section decreased by more than an order of magnitude for proton energies
< 50 MeV [40], [41]. For more recent devices, the SEU cross section has generally not
decreased very much with energy, the cross section due to 50 MeV protons being only about
a factor of 2 higher than the cross section due to 14 MeV neutrons [3]. A very useful
compendium of SEU cross sections in more than 120 different SRAMs and DRAMs was
compiled by ESA in 1997 [42], mostly on 5 V devices, but a few at 3,3 V. However, few if any,
of these devices are used today. In contrast, most other papers in the open literature contain
measured proton SEU response data for fewer devices, roughly 4 to 8 devices.
6.2.4 Ground level versus avionics applications
There are a number of important differences between the SEU considerations for devices in
avionics applications and those on the ground. First and foremost, the neutron flux in the
atmosphere is much higher than it is on the ground, so the SEU rate is going to be
proportionally higher. The nominal difference is taken to be a factor of 300 between the
neutron flux at 40 000 ft (12,2 km) and on the ground. As explained in JESD-89A and in
various technical papers [43], [44], there are two main sources of upsets in devices on the
ground, the atmospheric neutrons and alpha particles from trace amounts of radioactive
materials within the IC package. As the nature of IC packaging has evolved over the years,
the specific components responsible for most of the alpha
...