Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 62396-4: Atmospheric Radiation Effects in Avionics — Managing Single Event Effects in High Voltage Electronics
Design Guidance for High Voltage Aircraft Electronics Subject to Atmospheric Neutron Radiation
1. Scope and Context of IEC 62396-4
IEC 62396-4 is the fourth part of the IEC 62396 series on process management for avionics — atmospheric radiation effects. This part specifically addresses the design of high voltage aircraft electronics managing potential single event effects (SEE). Published by IEC Technical Committee 107 (Process management for avionics), the standard provides essential guidance for avionics system designers, electronic equipment manufacturers, and component suppliers dealing with semiconductor devices operating at voltages nominally above 200 V in aircraft operating at altitudes up to 60,000 ft (18.3 km).
The fundamental problem addressed by this standard is that high voltage semiconductor devices — power MOSFETs, IGBTs, power diodes — operating in the atmospheric neutron environment at aircraft altitudes are susceptible to destructive single event effects. Atmospheric neutrons, generated by cosmic ray interactions with the upper atmosphere, have sufficient energy to initiate nuclear reactions within the silicon lattice of semiconductor devices. The resulting charge deposition can trigger catastrophic failure mechanisms that do not occur at sea level due to the significantly lower neutron flux.
At typical aircraft cruising altitudes (30,000-40,000 ft), the atmospheric neutron flux is approximately 300 times higher than at sea level. This means a component that has been qualified for terrestrial applications may fail at a rate 300 times higher when used in avionics — a critical consideration that IEC 62396-4 explicitly addresses through its design guidance and testing requirements.
2. Single Event Effects in High Voltage Devices
| Effect | Acronym | Description | Affected Devices | Failure Mechanism |
|---|---|---|---|---|
| Single Event Burnout | SEB | Destructive failure caused by charge deposition triggering parasitic BJT conduction in power MOSFETs | N-channel power MOSFETs, IGBTs, bipolar power transistors, power diodes | Charge deposition forward-biases the body-source junction; if drain bias exceeds local breakdown voltage of parasitic bipolar elements, avalanching leads to thermal runaway and burnout |
| Single Event Gate Rupture | SEGR | Destructive failure of the gate oxide due to transient plasma filament created by ionising particle strike | N-channel and P-channel power MOSFETs, superjunction MOSFETs, SiC MOSFETs | Transient plasma filament causes localised increase in oxide field; if field exceeds dielectric strength, oxide breaks down leading to gate leakage and eventual rupture |
| Single Event Upset | SEU | Non-destructive change of state in memory or register elements | SRAM, registers, configuration memory | Charge collection at a storage node flips the logic state; can be corrected by rewrite or reset |
| Single Event Latch-up | SEL | Destructive regenerative current path triggered by charge deposition in CMOS structures | CMOS ICs, mixed-signal ICs | Charge deposition triggers parasitic SCR structure; if current exceeds device rating, permanent damage occurs |
Single Event Burnout (SEB) is the dominant threat for high voltage aircraft electronics under atmospheric neutron exposure. Experimental data from WNR (white neutron source) testing shows that 400 V and 500 V MOSFETs exhibit SEB cross sections ranging from 10-8 to 10-6 cm2 at rated voltage, depending on device technology and manufacturer. For an aircraft flying at 40,000 ft for 10 hours, the predicted SEB failure rate for a 500 V MOSFET operating at 80% of rated voltage can range from 10-7 to 10-5 failures per device per flight hour — a non-negligible risk for safety-critical systems.
3. Testing and Qualification Methodology
IEC 62396-4 establishes a comprehensive methodology for quantifying single event effects in high voltage avionics components. A critical distinction made by the standard is between test data from heavy ion sources and data from neutron/proton sources:
- Heavy ion data is not relevant for atmospheric radiation qualification. Heavy ion testing uses particles with linear energy transfer (LET) values much higher than those produced by atmospheric neutron reactions within silicon. While heavy ion testing is appropriate for space applications, it grossly overestimates the SEB susceptibility for avionics.
- High energy neutron and proton testing is the appropriate methodology. The standard recommends using spallation neutron sources (such as LANSCE at Los Alamos or the WNR facility) with neutron energies spanning from 1 MeV to several hundred MeV, covering the atmospheric neutron energy spectrum. Accelerated proton testing at energies of 50-200 MeV can also be used as a proxy, with appropriate conversion factors.
| Test Parameter | Recommended Value / Methodology | Rationale |
|---|---|---|
| Neutron energy range | 1 MeV to 800 MeV (spallation source) | Covers the full atmospheric neutron spectrum at aircraft altitudes; thermal neutrons (<1 eV) also considered for SEB from 10B capture reactions |
| Proton energy range | 50-200 MeV (mono-energetic or broad spectrum) | Protons at these energies produce secondary particles through nuclear reactions, mimicking neutron-induced SEE reasonably well; conversion factor from proton to neutron cross-section typically ranges from 0.5 to 2.0 |
| Device voltage bias | Test at 60%, 75%, 90%, and 100% of rated VDS (for MOSFETs) or VCE (for IGBTs) | SEB cross-section is strongly voltage-dependent; testing at multiple bias points allows extrapolation to operating conditions |
| Gate bias (for SEGR evaluation) | Test with gate at 0 V and at rated VGS (or VGE for IGBTs) | SEGR susceptibility increases with gate bias; worst-case condition must be evaluated |
| Temperature | Room temperature (25 °C) and maximum rated junction temperature (typically 125-150 °C for power devices) | SEB cross-section generally decreases with increasing temperature (negative temperature coefficient for impact ionisation); room temperature is the conservative test condition |
| Fluence | Target: 5 x 109 n/cm2 (En > 10 MeV) for neutrons; 1 x 1010 p/cm2 for protons | Sufficient to observe statistically significant number of events; corresponds to approximately 10,000 flight hours at 40,000 ft for neutron testing |
| Event detection | Continuous monitoring of drain current (IDS); event threshold: 10% increase above pre-irradiation level; post-event verification by electrical characterisation | SEB is destructive and detected by abrupt current increase; SEGR detected by gate leakage current > 100 nA at rated VGS after irradiation |
A key engineering contribution of IEC 62396-4 is the introduction of the EPICS (Events per Current Stress) methodology for characterising SEB and SEGR susceptibility. Rather than relying solely on destructive testing, EPICS plots the number of events versus applied voltage stress level, allowing designers to select a safe operating voltage margin. For example, EPICS data for a 1200 V diode may show zero events at 675 V (56% of rated voltage) but rapidly increasing events above 900 V (75%), enabling a data-driven derating recommendation.
4. Engineering Design Insights for Mitigation
4.1 Voltage Derating Strategy
The most effective mitigation against SEB is voltage derating — operating the device at a fraction of its rated voltage such that the peak internal electric field remains below the threshold for parasitic BJT triggering. Based on empirical data compiled in IEC 62396-4, the recommended derating factors for atmospheric neutron environments are:
- Power MOSFETs (planar): Derate to 60-70% of rated VDS for SEB-free operation at aircraft altitudes. Superjunction MOSFETs may require deeper derating to 50-60% due to their higher internal field peaks.
- IGBTs: Derate to 65-75% of rated VCES. The presence of an internal heterojunction (NPT vs PT vs Trench FS) affects SEB sensitivity; trench field-stop IGBTs generally offer better SEB resistance.
- SiC MOSFETs: Early data suggests that SiC devices may have higher SEB thresholds than equivalently rated Si devices due to the wider bandgap and higher critical field, but the standard advises caution until more atmospheric neutron data becomes available.
4.2 Alternative Semiconductor Materials
IEC 62396-4 dedicates a clause to alternative semiconductor materials for high voltage avionics. Silicon carbide (SiC) and gallium nitride (GaN) offer wider bandgaps and higher critical electric fields compared to silicon, which may provide inherent SEB immunity at equivalent voltage ratings. However, the standard notes a paucity of atmospheric neutron test data for wide-bandgap devices and recommends that qualification programmes include dedicated neutron testing rather than relying on theoretical immunity.
A subtle but critical failure mechanism addressed in IEC 62396-4 is SEB induced by thermal neutrons through 10-Boron (10B) capture reactions. Many semiconductor devices use boron-containing materials (BPSG dielectrics, boron-doped silicon). When a thermal neutron is captured by a 10B nucleus, the reaction produces a lithium-7 nucleus and an alpha particle, releasing 2.8 MeV of energy — sufficient to trigger SEB in a biased high-voltage device. This mechanism is particularly insidious because thermal neutrons penetrate standard shielding easily. The standard requires evaluation of 10B content in device materials for SEB-critical applications.
5. Frequently Asked Questions
Q1: Can IEC 62396-4 test data be used to qualify devices for space applications?
No. IEC 62396-4 is specifically for the atmospheric radiation environment encountered by aircraft. The space radiation environment includes trapped protons (Van Allen belts), solar particle events, and galactic cosmic rays with much higher energy and LET values. Heavy ion testing per MIL-STD-750 and JEDEC standards is required for space applications. Using atmospheric neutron test data for space qualification would grossly underestimate the in-orbit failure rate.
Q2: How does the standard address the effect of device packaging on SEE susceptibility?
The standard acknowledges that packaging materials affect SEE rates through two mechanisms: (1) boron-containing underfill or moulding compounds can increase thermal neutron SEB rates, and (2) heavy metal lid materials (e.g., Kovar) may produce secondary particles when struck by high-energy neutrons. The standard recommends that the total SEE rate calculation include a packaging contribution factor of 1.2-1.5 for packaged devices compared to bare die, based on empirical studies.
Q3: What is the recommended safety margin for voltage derating?
IEC 62396-4 recommends a minimum safety margin of 20% between the operating voltage and the voltage at which the first SEB event is observed during accelerated testing. For example, if testing at a fluence of 5 x 109 n/cm2 shows the first SEB at 75% of rated voltage, the maximum recommended operating voltage is 60% of rated voltage (75% x 0.8 = 60%). This margin accounts for device-to-device variation, temperature effects, and statistical uncertainty in the accelerated test data.
Q4: Does IEC 62396-4 cover single event effects in passive components?
Primarily, the standard focuses on active semiconductor devices. However, it notes that high-voltage ceramic capacitors (MLCCs) used in DC-link and snubber circuits may experience radiation-induced dielectric breakdown in the atmospheric neutron environment. The standard recommends that designers consult the relevant component manufacturers for SEE data on passive components and apply derating factors consistent with the avionics application requirements.
