IEC 62396-2:2012
(Main)Process management for avionics - Atmospheric radiation effects - Part 2: Guidelines for single event effects testing for avionics systems
Process management for avionics - Atmospheric radiation effects - Part 2: Guidelines for single event effects testing for avionics systems
IEC 62396-2:2012 aims to provide guidance related to the testing of microelectronic devices for purposes of measuring their susceptibility to single event effects (SEE) induced by 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 atmospheric neutrons at aircraft altitudes. Although developed for the avionics industry, this process may be applied by other industrial sectors. This first edition includes the following significant technical changes with respect to the technical specification IEC/TS 62396-2:
- additional information on heavy ion data, neutron and proton data and thermal neutron data;
- updates with regard to neutron sources: additional radiation simulators;
- addition of the Anita spallation neutron source;
- additional information on whole system and equipment testing;
- comparison between accelerator based neutron sources.
General Information
Relations
Standards Content (Sample)
IEC 62396-2 ®
Edition 1.0 2012-09
INTERNATIONAL
STANDARD
colour
inside
Process management for avionics – Atmospheric radiation effects –
Part 2: Guidelines for single event effects testing for avionics systems
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.
IEC Central Office Tel.: +41 22 919 02 11
3, rue de Varembé Fax: +41 22 919 03 00
CH-1211 Geneva 20 info@iec.ch
Switzerland www.iec.ch
About the IEC
The International Electrotechnical Commission (IEC) is the leading global organization that prepares and publishes
International Standards for all electrical, electronic and related technologies.
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.
Useful links:
IEC publications search - www.iec.ch/searchpub Electropedia - www.electropedia.org
The advanced search enables you to find IEC publications The world's leading online dictionary of electronic and
by a variety of criteria (reference number, text, technical electrical terms containing more than 30 000 terms and
committee,…). definitions in English and French, with equivalent terms in
It also gives information on projects, replaced and additional languages. Also known as the International
withdrawn publications. Electrotechnical Vocabulary (IEV) on-line.
IEC Just Published - webstore.iec.ch/justpublished Customer Service Centre - webstore.iec.ch/csc
Stay up to date on all new IEC publications. Just Published If you wish to give us your feedback on this publication
details all new publications released. Available on-line and or need further assistance, please contact the
also once a month by email. Customer Service Centre: csc@iec.ch.
IEC 62396-2 ®
Edition 1.0 2012-09
INTERNATIONAL
STANDARD
colour
inside
Process management for avionics – Atmospheric radiation effects –
Part 2: Guidelines for single event effects testing for avionics systems
INTERNATIONAL
ELECTROTECHNICAL
COMMISSION
PRICE CODE
W
ICS 03.100.50; 31.020; 49.060 ISBN 978-2-83220-389-7
– 2 – 62396-2 © IEC:2012(E)
CONTENTS
FOREWORD . 4
INTRODUCTION . 6
1 Scope . 7
2 Normative references . 7
3 Terms and definitions . 7
4 Abbreviations used in the document . 7
5 Obtaining SEE data . 9
5.1 Types of SEE data . 9
5.2 Use of existing SEE data . 9
5.2.1 General . 9
5.2.2 Heavy ion data . 9
5.2.3 Neutron and proton data . 10
5.2.4 Thermal neutron data . 10
5.3 Deciding to perform dedicated SEE tests . 10
6 Availability of existing SEE data for avionics applications . 11
6.1 Variability of SEE data. 11
6.2 Types of existing SEE data that may be used . 11
6.2.1 General . 11
6.2.2 Sources of data, proprietary versus published data. 12
6.2.3 Data based on the use of different sources . 13
6.2.4 Ground level versus avionics applications. 19
6.3 Sources of existing data . 20
7 Considerations for SEE testing . 21
7.1 General . 21
7.2 Selection of hardware to be tested . 22
7.3 Selection of test method . 22
7.4 Selection of facility providing energetic particles . 23
7.4.1 Radiation sources . 23
7.4.2 Spallation neutron source . 23
7.4.3 Monoenergetic and quasi-monoenergetic beam sources . 24
7.4.4 Thermal neutron sources . 25
7.4.5 Whole system and equipment testing . 25
8 Converting test results to avionics SEE rates . 26
8.1 General . 26
8.2 Use of spallation neutron source . 27
8.3 Use of SEU cross-section curve over energy . 27
8.4 Measured SEU rates for different accelerator based neutron sources . 30
8.5 Influence of upper neutron energy on the accuracy of calculated SEE rates;
verification and compensation . 30
Annex A (informative) Sources of SEE data published before 2000 . 32
Bibliography . 33
Figure 1 – Comparison of Los Alamos, TRIUMF and ANITA neutron spectra with
terrestrial / avionics neutron spectra (JESD-89A and IEC 62396-1) . 15
Figure 2 – Variation of high energy neutron SEU cross-section per bit as a function of
device feature size for SRAM and SRAM arrays in FPGA and microprocessors . 17
62396-2 © IEC:2012(E) – 3 –
Figure 3 – Percentage fraction of SEU rate from atmospheric neutrons contributed by
neutrons with E < 10 MeV . 18
Figure 4 – Comparison of mono-energetic SEU cross-sections with Weibull and piece-
wise linear fits . 29
Table 1 – Sources of existing data (published after 2000) . 20
Table 2 – Spectral distribution of neutron energies . 30
Table A.1 – Sources of existing SEE data published before 2000 . 32
– 4 – 62396-2 © IEC:2012(E)
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 itself does not provide any attestation of conformity. Independent certification bodies provide conformity
assessment services and, in some areas, access to IEC marks of conformity. IEC is not responsible for any
services carried out by independent certification bodies.
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.
International Standard IEC 62396-2 has been prepared by IEC technical committee 107:
Process management for avionics.
This standard cancels and replaces IEC/TS 62396-2 published in 2008. This first edition
constitutes a technical revision.
This first edition includes the following significant technical changes with respect to the
technical specification IEC/TS 62396-2.
a) Clause 5 information expanded including additional information in sections on heavy ion
data, neutron and proton data and thermal neutron data.
b) The neutron sources Clause 6 has been updated, Figure 1 now contains data on
additional radiation simulators, and Figure 2 contains more recent data with results for
feature sizes below 100 nm. A new Figure 3 contains data on low energy neutron
(< 10 MeV) SEU percentage fraction.
62396-2 © IEC:2012(E) – 5 –
c) The sources of existing data (radiation SEE data) table has been split in to two tables: one
for post 2000 sources and the other for pre 2000 sources which is now in Annex A.
d) The Anita spallation neutron source has been added to Clause 7.
e) A new subclause, 7.4.5, has been added on whole system and equipment testing.
f) A new subclause, 8.4, provides a comparison between accelerator based neutron sources.
g) A new subclause, 8.5, compares the influence of upper neutron energy for neutron
sources.
The text of this standard is based on the following documents:
FDIS Report on voting
107/186/FDIS 107/192/RVD
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 in the IEC 62396 series, published 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 stability 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
• reconfirmed,
• withdrawn,
• replaced by a revised edition, or
• amended.
A bilingual edition of this document may be issued at a later date.
IMPORTANT – The 'colour inside' logo on the cover page of this publication indicates
that it contains colours which are considered to be useful for the correct
understanding of its contents. Users should therefore print this document using a
colour printer.
– 6 – 62396-2 © IEC:2012(E)
INTRODUCTION
This industry-wide international standard 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 62396-1.
Guidance is provided on the use of existing single event effects (SEE) data, sources of data
and the types of accelerated radiation sources used. Where SEE data is not available
considerations for testing are introduced including suitable radiation sources for providing
avionics SEE data. The conversion of data obtained from differing radiation sources into
avionics SEE rates is detailed.
62396-2 © IEC:2012(E) – 7 –
PROCESS MANAGEMENT FOR AVIONICS –
ATMOSPHERIC RADIATION EFFECTS –
Part 2: Guidelines for single event effects
testing for avionics systems
1 Scope
This part of IEC 62396 aims to provide guidance related to the testing of microelectronic
devices for purposes of measuring their susceptibility to single event effects (SEE) induced by
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 atmospheric neutrons at aircraft altitudes.
Although developed for the avionics industry, this process may be applied by other industrial
sectors.
2 Normative references
The following documents, in whole or in part, are normatively referenced in this document and
are indispensable for its application. For dated references, only the edition cited applies. For
undated references, the latest edition of the referenced document (including any
amendments) applies.
IEC 62396-1:2012, Process management for avionics – Atmospheric radiation effects – Part 1:
Accommodation of atmospheric radiation effects via single event effects within avionics
electronic equipment
IEC/TS 62396-3, Process management for avionics – Atmospheric radiation effects – Part 3:
Optimising system design to accommodate the single event effects (SEE) of atmospheric
radiation
IEC/TS 62396-4, Process management for avionics – Atmospheric radiation effects – Part 4:
Guidelines for designing with high voltage aircraft electronics and potential single event
effects
IEC/TS 62396-5, Process management for avionics – Atmospheric radiation effects – Part 5:
Guidelines for assessing thermal neutron fluxes and effects in avionics systems
3 Terms and definitions
For the purpose of this document, the terms and definitions given in IEC 62396-1 apply.
4 Abbreviations used in the document
ANITA Atmospheric-like Neutrons from thIck TArget (TSL, Sweden)
BL1A, BL1B, BL2C Beam line designations at the TRIUMF facility (Canada)
BPSG Borophosphosilicate glass
CMOS Complementary metal oxide semiconductor
COTS Commercial off-the-shelf
D-D Deuterium-deuterium
– 8 – 62396-2 © IEC:2012(E)
DRAM Dynamic random access memory
D-T Deuterium-tritium
DUT Device under test
E Energy
EEPROM Electrically erasable programmable read only memory
EPROM Electrically programmable read only memory
ESA European Space Agency
eV Electron volt
FIT Failures in time (failures in 10 hours)
FPGA Field programmable gate array
GeV Giga electron volt
GNEIS Gatchina Neutron Spectrometer (Russia)
GSFC Goddard Space Flight Center
GV Giga volt (rigidity unit)
IBM International Business Machines
IC Integrated circuit
ICE Irradiation of Chips and Electronics
IEEE Trans. Nucl. Sci. IEEE Transactions on Nuclear Science
IUCF Indiana University Cyclotron Facility (USA)
JEDEC JEDEC Solid State Technology Association
JESD JEDEC standard
JPL Jet Propulsion Laboratory
LANSCE Los Alamos Neutron Science Center (USA)
LET Linear energy transfer
LETth Linear energy transfer threshold
MBU Multiple bit upset (in the same word)
MCU Multiple Cell Upset
MeV Mega electron volt
NASA National Aeronautical and Space Agency
PIF Proton Irradiation Facility (TRIUMF, Canada)
PNPI Petersburg Nuclear Physics Institute (Russia)
PSG Phosphosilicate glass
QMN Quasi-monoenergetic neutrons
RADECS Radiations, effets sur les composants et systèmes.
RAM Random access memory
RCNP Research Center of Nuclear Physics (Osaka, Japan)
RVC Result of voting (IEC)
SBU Single Bit Upset
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 latchup
SEP Solar energetic particles
SER Soft error rate
SET Single event transient
SEU Single event upset
SHE Single event induced hard error
62396-2 © IEC:2012(E) – 9 –
SRAM Static random access memory
SW Software
TID Total ionizing dose
TNF TRIUMF neutron facility (TRIUMF, Canada)
TRIUMF Tri-University Meson Facility (Canada)
TSL Theodor Svedberg Laboratory (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.
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
5.2.1 General
The simplest solution is to find previous SEE data on a specific IC device. Data may be
available on SEE caused by heavy ions, protons, high-energy neutrons, or thermal neutrons.
Heavy-ion data is normally only applicable to space applications, where direct ionization by
the primary cosmic ray flux is of concern. However, heavy ion data can be useful for
screening purposes, as described in 5.2.2. Proton data is usually also gathered for space
applications, where primary cosmic rays and trapped particles are of concern. However, high-
energy protons provide a good proxy for neutrons in SEE measurements, as they undergo
very similar nuclear interactions with device materials. Therefore, both existing neutron data
and existing proton data may be applicable to the evaluation of SEE rates in a device of
interest, as described in section 5.2.3. Low-energy (“thermal”) neutrons can also cause SEE
in some devices but such data is only available on a very small number of devices (see
section 5.2.4) and it involves neutron interactions with boron-10 rather than silicon.
5.2.2 Heavy ion data
An important resource that can be utilized to eliminate devices is the results from heavy ion
SEE testing carried out to support space programs (~80 % of the devices tested for space
applications are tested only with heavy ions). 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. Heavy ion testing is characterized by the LET (linear energy transfer) of the ions
to which the ICs are exposed. The LET is the energy that can be deposited per unit path
length, divided by the density (units of MeV⋅cm /mg). With neutron SEE, secondary particles
or recoils created by the neutron interactions act as heavy ions, and the highest possible LET
of neutron-induced recoils in silicon is ~15 MeV⋅cm /mg [1, 2] . Thus, any device tested with
heavy ions that has a LET threshold > 15 MeV⋅cm /mg will be immune from neutron-induced
SEE. In a recent paper summarizing SEE testing at NASA-GSFC [3], 21 ICs of various types
were tested with only heavy ions and eight of them (~40 %) had LET thresholds
> 15 MeV⋅cm /mg for diverse SEE effects.
However, for the rare commercial SRAMs that are susceptible to SEL from heavy ions [4], this
susceptibility can be increased due to the presence of small amounts of high Z materials
___________
Numbers in square brackets refer to the Bibliography.
– 10 – 62396-2 © IEC:2012(E)
within the IC, e.g., tungsten plugs, because higher Z recoils are created which can cause SEE
reactions due to their higher values of LET. The high Z materials also lead to higher proton
and neutron SEL cross-sections due to the neutron/proton reactions producing these recoils
with higher LET and energy. Therefore heavy ion SEL cross-sections need to be examined
carefully for applicability to proton-neutron SEL susceptibility caused by embedded high Z
materials in the SRAMs. A suggested conservative value of LET threshold above which a
device can be considered immune from SEL induced by neutrons is 40 MeV⋅cm /mg [4].
However, this caution does not apply to the primary rationale given above for eliminating
some devices from consideration for neutron SEE sensitivity based on heavy ion SEE testing,
since only some devices incorporate these higher Z materials and the limitation applies to
SEL.
Heavy ion SEE data should not be used for application to the atmospheric neutron
environment for calculation of neutron cross-section, except by scientists and engineers who
have extensive experience in using this kind of data. Unless otherwise stated explicitly, when
SEE data is discussed in the remainder of this international standard, it refers only to single
event testing using a neutron or proton source, not to the results from testing with heavy ions.
5.2.3 Neutron and proton data
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.
5.2.4 Thermal neutron data
There is little data on thermal neutron cross-section. However a number of the spallation
neutron sources including TRIUMF, TSL and ISIS contain a substantial percentage of thermal
neutrons within the high energy beam. Using thermal neutron filters or time of flight it is
possible at such sources to determine thermal neutron cross-section. In addition there are a
number of dedicated thermal neutron sources and these are listed in IEC 62396-1:2012.
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
62396-2 © IEC:2012(E) – 11 –
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
international standard, 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
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. 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 international standard 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 (cm /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 determined by
measurements made with high energy protons can be used as the cross-sections for high
energy atmospheric neutrons. This is far more than an assumption, since it has been
demonstrated by direct measurement in many different devices see [5, 6, 7, 8, 9] and
IEC 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
h] × avionics SEE cross-section [cm per device] (1)
SEE rate per device = 6 000 [n/cm
2.
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 62396-1, this flux is
suitable for devices with feature size above 150 nm. 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 62396-1 and is the nominal flux under the above conditions. For
– 12 – 62396-2 © IEC:2012(E)
devices with feature size below 150 nm the relevant neutron flux will be higher than
2.
6 000 n/cm h because the threshold energy will be lower than 10 MeV, therefore the
threshold energy (and flux) used for estimation must be clearly shown and validation
demonstrated, see IEC 62396-1.
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 [10], 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 (QMN) 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 [7], (feature size
< 0,5 µm), and a factor of 4 times higher for older devices [8]. For some of the very latest
devices, the factor is close to 1. In general, there are a number of spallation neutron facilities
around the world for neutron soft error rate testing, the accuracy of these is considered in
references [11, 12]. Calculation of soft-error rate depends largely on the combination of the
device and the facility to be used. There does need to be some kind of practical threshold
energy to determine the neutron flux, but the threshold cannot be a fixed value and generally
decreases as scaling of device proceeds. The value of “10 MeV” threshold has been used for
devices with geometry above 100 nm, however the threshold energy used for neutron flux
determination must be clearly shown and should be validated with reference to the device
technology.
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];
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 [13, 14, 15, 16],
and examples of JPL reports of SEE testing containing proton SEE test results are given in
[17, 18, 19]. 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
62396-2 © IEC:2012(E) – 13 –
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 13 (13 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 [10] and so most often
–9 2
used.) Thus, FIT×10 /13 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 /13 to obtain the SEE cross-
/bit. Other papers report the FIT value in arbitrary units which allows the
section in cm
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 [6], [20, 21, 22, 23].
Most of the SEE data that has been discussed 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 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 used on manned space missions. This testing is carried out with a beam of protons,
and while it is recorded in NASA-JSC reports, these reports are not widely available,
examples are given in [24, 25, 26]. 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 space
applications (proton tests for low earth orbit spacecraft contractors), and this data is rarely
reported in the open literature. By 2005 the number of user groups had grown to more than 25,
but the ratio of test groups that published their results had not changed much.
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 accelerator-based 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.
– 14 – 62396-2 © IEC:2012(E)
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 (> 1 MeV) and at very low energies (thermal neutron
energy, 0,025 eV). High energy neutrons cause the SEU by the nuclear reaction with device
materials, causing energetic ions to be emitted. It is the energy from these ions which cause
ionization in the semiconductor lattice, leading to the upset. Neutrons with energies above
about 10 MeV are of the greatest concern. This is because many of the interactions which
lead to SEE have threshold energies in the region between 2 MeV and 10 MeV in silicon,
oxygen and other typical device materials [27, 28, 29] because elastic interactions (which
have no threshold energy) are more effective at higher energies, and also because the natural
neutron spectrum has a minimum around 10 MeV [30, 31, 32, 33]. Nonetheless, neutrons
with lower energies (e.g. 3 MeV or below) can cause SEE [28, 34], albeit with much reduced
probability. Estimates of the SEU contribution by neutrons below 10 MeV in electronics
technologies with geometry greater than 0,2 µm are below 10 %, but for lower feature sizes
this fraction increases [28, 35]. This is consistent with measurements made with
monoenergetic neutrons on devices of the mid 1990s (feature sizes above 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 [36]. However, for more recent devices,
especially those with feature sizes less than 0,2 µm and even down to 45 nm, the contribution
of neutrons with energies below 10 MeV is expected to be in the range 8 % to 10 % [28].
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 (QMN) 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 [37]. Figure 1 com
...
Questions, Comments and Discussion
Ask us and Technical Secretary will try to provide an answer. You can facilitate discussion about the standard in here.
Loading comments...