1、BSI Standards PublicationProcess management for avionics Atmospheric radiation effectsPart 4: Design of high voltage aircraft electronics managing potential single event effectsBS IEC 62396-4:2013National forewordThis British Standard is the UK implementation of IEC 62396-4:2013. Itsupersedes DD IEC
2、/TS 62396-4:2008 which is withdrawn.The UK participation in its preparation was entrusted to TechnicalCommittee GEL/107, Process management for avionics.A list of organizations represented on this committee can be obtained onrequest to its secretary.This publication does not purport to include all t
3、he necessary provisions ofa contract. Users are responsible for its correct application. The British Standards Institution 2013.Published by BSI Standards Limited 2013ISBN 978 0 580 81643 7ICS 03.100.50; 31.020; 49.060Compliance with a British Standard cannot confer immunity fromlegal obligations.Th
4、is British Standard was published under the authority of theStandards Policy and Strategy Committee on 31 October 2013.Amendments/corrigenda issued since publicationDate Text affectedBRITISH STANDARDBS IEC 62396-4:2013IEC 62396-4 Edition 1.0 2013-09 INTERNATIONAL STANDARD Process management for avio
5、nics Atmospheric radiation effects Part 4: Design of high voltage aircraft electronics managing potential single event effects INTERNATIONAL ELECTROTECHNICAL COMMISSION R ICS 03.100.50; 31.020; 49.060 PRICE CODE ISBN 978-2-8322-1094-9 Registered trademark of the International Electrotechnical Commis
6、sion Warning! Make sure that you obtained this publication from an authorized distributor. 2 62396-4 IEC:2013(E) CONTENTS INTRODUCTION . 5 1 Scope . 6 2 Normative references . 6 3 Terms and definitions . 6 4 Potential high voltage single event effects 6 5 Quantifying single event burnout in avionics
7、 for high voltage devices . 8 6 Relevant SEB data and applying it to avionics 9 6.1 SEB data from heavy ion testing is not relevant . 9 6.2 SEB data from high energy neutron and proton testing 9 6.3 Calculating the SEB rate at aircraft altitudes . 12 6.4 Measurement of high voltage component radiati
8、on characteristics, EPICS 12 6.5 Single event burnout due to thermal neutrons 14 6.6 Alternative semiconductor materials to silicon . 15 7 Conclusion . 15 Bibliography 17 Figure 1 SEB cross sections measured in 400 V and 500 V MOSFETs for WNR neutron and proton beams 10 Figure 2 SEB cross sections m
9、easured in 1 000 V MOSFETs and 1 200 V IGBTs with WNR neutron and 200 MeV proton beams . 11 Figure 3 Measurement of radiation event charge and current . 13 Figure 4 EPICS plot of 1 200 V diode numbers of events at currents taken at different applied voltages for a neutron fluence of approximately 3,
10、5 109neutrons per cm2measured at energies greater than 10 MeV . 14 Figure 5 EPICS plot of 1 200 V diode numbers of events at currents taken at 675 V (56 %) and 900 V (75 %) applied voltage (stress) demonstrating the difference between low and high voltage stress Fluence as per Figure 4 14 BS IEC 623
11、96-4:201362396-4 IEC:2013(E) 5 INTRODUCTION This industry-wide international standard provides guidance and requirements to design high voltage aircraft electronics for electronic equipment and avionics systems. It is intended for avionics system designers, electronic equipment manufacturers, compon
12、ent manufacturers and their customers to manage the single event effects produced in semiconductor devices operating at high voltage (nominally above 200 V) by atmospheric radiation. It expands on the information and guidance provided in IEC 62396-1:2012. The internal elements of semiconductor devic
13、es operating at high applied voltage will be subject to high voltage stress. The incident radiation causes ionisation charge within the device, and the high voltage stress may cause a large increase (avalanche) in this charge, which may be destructive. Within this part of IEC 62396 two effects are c
14、onsidered: single event burnout (SEB), and single event gate rupture (SEGR). BS IEC 62396-4:2013 6 62396-4 IEC:2013(E) PROCESS MANAGEMENT FOR AVIONICS ATMOSPHERIC RADIATION EFFECTS Part 4: Design of high voltage aircraft electronics managing potential single event effects 1 Scope This part of IEC 62
15、396 provides guidance on atmospheric radiation effects and their management on high voltage (nominally above 200 V) avionics electronics used in aircraft operating at altitudes up to 60 000 ft (18,3 km). This part of IEC 62396 defines the effects of that environment on high voltage electronics and p
16、rovides design considerations for the accommodation of those effects within avionics systems. This part of IEC 62396 provides technical data and methodology for aerospace equipment manufacturers and designers to standardise their approach to single event effects on high voltage avionics by providing
17、 guidance, leading to a standard methodology. Details are given of the types of single event effects relevant to the operation of high voltage avionics electronics, methods of quantifying those effects, appropriate methods to provide design and methodology to demonstrate the suitability of the elect
18、ronics for the application. 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 d
19、ocument (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 3 Terms and definitions For the purposes of this document,
20、the terms and definitions given in IEC 62396-1:2012 apply. 4 Potential high voltage single event effects An N-channel power MOSFET can have two different types of destructive effects induced by the deposition of charge from a single energetic particle, single event burnout (SEB) and single event gat
21、e rupture (SEGR). Different tests performed on several devices show that is difficult to induce SEB in P-channel MOSFET 1, 21. In addition to this kind of power MOSFET, other power devices, such as insulated gate bipolar transistors (IGBTs), bipolar power transistors and diodes, which have large app
22、lied voltage biases and high internal electric fields, are susceptible to SEB. In SEB, the penetration of the source-body-drain region by the deposited charge can forward bias the thin body region under the source. If the bias applied to the drain exceeds the local _ 1 Numbers in square brackets ref
23、er to the Bibliography. BS IEC 62396-4:201362396-4 IEC:2013(E) 7 breakdown voltage of the parasitic bipolar elements, the single event induced pulse initiates avalanching in the drain depletion region that eventually leads to destructive burnout SEB. SEB can be induced by heavy ions, high energy pro
24、tons 3 and high energy neutrons 4. SEGR applies to N- and P-channel MOSFETs. It is explained via the transient plasma filament created by the energy deposition track when the MOSFET is struck through the thin gate oxide region. As a result of this transient track filament, there is a localized incre
25、ase in the oxide field which can cause the oxide to break down, leading first to gate leakage and finally to gate rupture. The SEGR failure mechanism has been widely studied by heavy ion testing and effects have been identified on different devices with various levels of sensitivity 2. For the time
26、being, experiments show also that SEGR induced by heavy ions is more an issue for space systems, and guidance for heavy ion SEGR testing is available 5. As a consequence of the atmospheric neutrons, SEB is the major threat to high voltage electronics. There remains a paucity of data on the question
27、of neutron-induced single event gate rupture (SEGR) in power devices. In the late 1990s one study looked for, but did not find, SEGR in 500 V power MOSFETs during accelerated spallation neutron testing 1. Shortly afterwards, however, dielectric breakdown was observed in 60 V power MOSFETs during 44
28、MeV and 200 MeV proton irradiation 6. As the gate ruptures in these devices were almost certainly caused by charge deposition from recoil ions, rather than by direct ionisation from the very low LET protons, sensitivity to neutrons was implied. Data published more recently show more direct evidence
29、of neutron-induced SEGR in devices rated at 1 kV. Hands et al. observed significant gate damage to a 1 kV power MOSFET at a spallation neutron facility, with a dependence on gate bias consistent with SEGR 7. Griffoni et al. tested a variety of devices, including IGBTs, SiC MOSFETs and superjunction
30、(SJ) MOSFETs in quasi-monoenergetic neutron environments, and observed SEGR only in the SJ MOSFETs 8. Interestingly, in this latter case the SEGR failure rate was sometimes higher than the SEB failure rate, though no dependency on gate bias condition was investigated to characterise the relative sus
31、ceptibilities. These results demonstrate that fast neutrons (and protons) are very capable of causing damage to the gate regions of power devices and, where conditions are right, this damage can lead to dielectric breakdown and catastrophic failure. Therefore this failure mode should be considered a
32、nd, where appropriate, quantified during accelerated testing of HV devices. Although at the outset this threat to the power system in an aircraft from SEB from the atmospheric neutrons may appear to be remote or even far-fetched, the experience of breakdowns in the high voltage electronics on electr
33、ic trains in Europe before 1995 shows that SEB can be real and has happened in the field. In that case, European and Japanese manufacturers of high voltage semiconductors noticed that some of their devices were undergoing burnout failures in the field during normal operation of newly developed train
34、 engines 9, 10. The diodes and GTO thyristors (gate turn-off thyristors) used on the trains were rated at 4 500 V, and were normally operated at 50 % to 60 % of rated voltage. They were designed for terrestrial use for 35 years, so when the failures first appeared in the field after only a few month
35、s, this was puzzling. The failure mode was investigated in great detail and eventually a set of experiments was carried out at three different locations (salt mine, top-floor laboratory and basement); the results convinced the investigators that the cause of the failures was the cosmic ray neutrons.
36、 Since that time, the manufacturers of these very high voltage devices have been careful in recommending the voltage at which the devices can be operated safely without SEB. In addition, these manufacturers have followed the methodology established by an experienced radiation effects group 1 by carr
37、ying out tests in the WNR beam at Los Alamos National Laboratory to characterize the response of their devices to a simulated high-energy neutron environment. Because the atmospheric neutron flux is higher by about a factor of 300 at aircraft altitudes compared to sea level, it is clear that the sam
38、e effect can occur in high voltage electronics in aircraft. The reason that, as far as is known, such failures have not been experienced previously in the field in aircraft power electronics is that the bus voltage used in aircraft systems has always been low enough to preclude SEB or SEGR. BS IEC 6
39、2396-4:2013 8 62396-4 IEC:2013(E) Generally, the highest voltage used in aircraft power systems has been 270 V, and a practical lower onset limit for most high voltage devices is 300 V. This practical lower limit stems from the fact that with SEB there is a threshold voltage for the effect to occur;
40、 if Vdsis kept below the threshold voltage, there will be no SEB. Thus for 270 V operation, devices rated at 400 V or 500 V would be used, resulting in a situation in which the devices are being operated at a derating factor of 67,5 % and 54 % respectively. Since the devices are being used at 300 V
41、and in fact 600 V has often been mentioned as a practical bus voltage. Thus, in order to preclude SEB from occurring in the high voltage electronics of such advanced avionics systems, a sufficiently low derating factor will have to be used, and the adequacy of the derating factor will have to be dem
42、onstrated through testing. 5 Quantifying single event burnout in avionics for high voltage devices Thus, the problem becomes that avionics vendors are asked to provide systems that will operate at higher voltages, e.g., 600 V, and there has been virtually no guidance for them to use in developing th
43、e designs that will avoid the potential of SEB in the high voltage devices such as power MOSFETs and IGBTs. In reality, the situation with SEB in high voltage electronics is relatively similar to that of single event upset (SEU), in low voltage devices ( 10 MeV, at 40 000 ft (12,2 km) altitude and 4
44、5 latitude. It shall be adjusted for different altitudes and latitudes using the data tables in Annex D of IEC 62396-1:2012. For RAMs especially, a great deal of SEU cross section data has been published, allowing users of the standard to estimate the SEU rate, and some SEU cross section data is als
45、o available for microprocessors and FPGAs. The same Equation (1) shall be used for SEB rates in high voltage devices provided that SEB cross sections are known for specific devices operated at a specified voltage. This part of IEC 62396 recommends the use of Equation (1) for calculating SEB rates ev
46、en though it is recognized that this is conservative. There is very little published data on the SEB cross sections, but the data that does exist 1, 4 suggests that the SEB cross section is significantly reduced at lower neutron energies compared to e.g. 200 MeV. The most suitable facilities for mea
47、suring SEB cross sections are spallation sources with maximum energy above 200 MeV. Thus the minimum neutron energy threshold for calculating the SEB rate (energy at which the SEB cross section is similar to that at high energy, e.g., 200 MeV) is 100 MeV. The available SEB cross section data is docu
48、mented in Clause 6. For avionics applications it should be recognized that assuming the high voltage electronics will be operating at a single voltage is unrealistic. First, the airplane power system is expected to experience power transients and spikes during flight. The transients typically last f
49、or less than 1 s, during which time Vdscould increase from 270 V to 350 V. The cascading power spikes can increase the voltage to even higher levels above nominal, although the duration is much shorter, usually 2 kV) by the microelectronics companies that manufacture these devices. This testing has used the WNR facility as well as other sources of neutrons. The other neutron sources include the quasi mono-energetic neutron beam created by a proton beam on a lithium target (e.g., at t
