Abstract: This document details the results of a cosmic ray radiation test performed on two planar SiC
power MOSFET technologies belonging to the 650V and 1200V voltage classes. The testing
followed JEDEC standard JEP151A, using a 200 MeV proton beam to simulate the effects of
cosmic ray-induced single event burnout (SEB). The testing involved comprehensive pre characterization, real-time monitoring during irradiation, and post-exposure analysis. Key
parameters like Failure In Time (FIT) were computed across different operating voltages
and the results are compared against data from various SiC Mosfets available in public
literature.

9. Test Results
The test results revealed key insights into the radiation tolerance of both the 1200V and
650V SiC MOSFETs. For each tested voltage level, the devices were monitored for failure
signatures, and their corresponding fluence values were recorded. Using this information,
the Failure In Time (FIT) metric was computed. FIT values were then plotted against
applied high voltage (VDS) to investigate trends and voltage sensitivity. These results are
plotted in Figure 8, assuming a neutron flux value n= 1/cm2/h. Error bars represent 2σ
confidence intervals based on sample variation and flux uncertainty. As expected, the FIT
rates increase with applied voltage for both technologies. The 1200V SiC MOSFET has a
higher FIT rate compared to its 650V counterpart, because the former has a higher chip
area compared to the latter. It is important to note the exponential scaling on the y-axis,
which means that FIT rate increases near-exponentially with increase in applied voltage.
Therefore, the applied dc-link voltage in the application has to be within suitable limits to
limit the FIT rate.
As the FIT rate is dependent on the chip size, it is desirable to normalize the FIT rates
according to the active area of the chip for comparability. They have also been normalized
according to their actual breakdown voltage (obtained from pre-characterization).
Furthermore, the results have been normalized to n= 13/cm2/h, which is the reference
neutron flux for New York City at sea level as per JEDEC JEP89 [4]. This is considered, as it
makes it comparable to results from state-of-the-art devices available in literature, as will
be explained in the later part of this paper. The resulting FIT rates are plotted in Figure 9. It
can now be seen that both technologies, though having different breakdown voltages arising
from different doping and thickness of the drift layer, behave similarly. This confirms that
the SEB effects are mainly dictated by how close the working voltage (Vdc) is with respect
to the designed breakdown voltage. This also implies that the drift layer has to be
appropriately dimensioned for a given Vdc, to meet certain FIT targets.
In order to compare how these devices fare compared to other SiC Mosfets available in the
market, the measured results were compared against data available from literature. FIT
rates documented for several competitors in [5] were replotted as-they-are and compared
against the above mentioned results in . It can be confirmed that the measured FIT rates for
both the tested devices are in the same ball park as the FIT rates measured on devices from
other suppliers.

10. Summary and Conclusion
This study evaluated the susceptibility of two types of Silicon Carbide (SiC) planar
MOSFETs—1200V SiC MOSFET and 650V SiC MOSFET—to radiation-induced destructive
effects. The tests were conducted using a 200 MeV proton beam, following JEDEC JEP151A. standard procedures for simulating terrestrial neutron effects. Each device type was tested
across a range of high voltage conditions. All devices were monitored in real-time using a
custom data acquisition system. Failures were detected by tracking sudden drops in VDS
accompanied by spikes in IDS. These failure points were used to compute FIT (Failure In
Time) values normalized to a reference neutron flux of 1 n/cm²/h.
The key findings are as follows:
• Post-irradiation IV curves confirmed irreversible degradation and increased leakage
currents.
• FIT values showed an exponential dependency on the applied drain-source voltage
(VDS), as expected.
• When plotted as function of the ratio of the applied voltage to the designed breakdown
voltage, FIT rates for both technologies match well, as expected from literature.
• The FIT rates for both the tested devices were compared against FIT rates known from
literature for devices from other suppliers. The FIT rates for the tested devices lie in the
same ball park as the data from other suppliers, confirming a comparable level of
sensitivity to cosmic radiation.
The study demonstrates that SiC MOSFETs, while robust under many conditions, remain
vulnerable to SEB when exposed to high-energy protons at elevated voltages. Optimal
selection of devices, especially regarding the designed breakdown voltage, to meet a given
application is crucial to ensure reliable operation.