| Summary: | The success of the first-generation Camera for Space Exploration (CASPEX) has been a vehicle in solidifying not only the use of Commercial Off The Shelf (COTS) components for space applications but also in extending the 3D Plus stacking technology and processes to the imaging domain. This, as well as the increasing importance of image sensors in space, became two of the main motivations for the development of its direct successor, the 12 Megapixel (MP) Space Camera Module. This study focuses on the evaluation of its most vulnerable components in the context of radiation which have been identified to be the CMOS Image Sensor and the SRAM-based FPGA.
It is important to highlight that 3D Plus and the Centre National d'Études Spatiales (CNES) are in close collaboration in the evaluation of the CASPEX project. Data is therefore shared between the institutions. The dataset from CNES are more geared towards the evaluation of the sensor at higher dose ranges while the data gathered by 3D Plus are intended to be on the qualification and evaluation at the lower doses. These two datasets complement and give more insight on the behavior of the device.
The sensor showed few current over-consumptions, no Latch-up events, and frequent Single Event Functional Interrupts (SEFIs) when exposed to Heavy Ions ranging from 1.5 to 48.5 MeV•cm2/mg of Linear Energy Transfer (LET). Gamma exposure of up to 150 krad increases the dark current by 70x while exposure to an equivalent Proton dose of up to 300 krad with 49.7 MeV increases the dark current by a factor of 6500. 3D Plus evaluation and qualification tests add that at 40krad of gamma exposure, the dark current is increased by 3.2x and 500x from a Proton dose of up to 30 krad.
The evaluation of the Xilinx 20nm SRAM-based FPGA references the studies done on its larger commercial and radiation-tolerant counterparts. While Global Triple Modular Redundancy (GTMR) has historically been the preferred choice, a study done specifically on the Kintex UltraScale XCKU040 suggests that race conditions reduce the effectivity of the GTMR to NoTMR levels at around 4 MeV•cm2/mg of LET. Distributed TMR (DTMR) should now be the choice for mitigation especially for more complex systems. This study implemented both the GTMR and DTMR on 15 instances of the VexRiscV processor and observed about 10x improvement on the CRAM sensitivity and 24x improvement on the Mean Time To Failure (MTTF) under a Geosynchronous (GEO) mission. In comparison, the GTMR and DTMR implementations on the full FPGA has an improvement factor of 9 on CRAM sensitivity and 120 on the MTTF.
With the results from this study, the data shared by the CNES, and the results from the references show that both the CMOS image sensor and the SRAM-based FPGA are expected to remain functional for space applications. Further analysis is needed with the mission specification and when the FPGA design intended for the 12MP CASPEX is made available.
|
|---|