This article looks at the different aspects of circuit design to better understand the tradeoffs of choosing COTS or rad-hard devices.
Performance, reliability, and flight heritage are typically chief concerns when it comes to electronics for space applications. Depending on the mission longevity and profile, designers may consider using commercial-off-the-shelf (COTS) parts in some cases. But COTS electronics are vastly different from radiation-hardened (rad-hard) devices. Rad-hard components such as Si MOSFETs are designed, tested, and verified to perform under the worst operating conditions, such as prolonged exposure to radiation in space.
From a design perspective, it’s important to weigh the unique considerations of using rad-hard Si MOSFETs versus COTS devices based on alternative materials, such as GaN HEMT power devices, in high-reliability space applications. In this article, we’ll look at different aspects of circuit design to better understand the tradeoffs of choosing one or the other.
COTS or not?
With today’s increased commercialization of the space industry, designers face more challenges to balance performance, program cost, mission profile, and risks. This is true even for the traditional space government and public sector players. Hundreds of startup companies, university researchers, and even private citizens now build and launch budget satellites, such as the popular CubeSat design. Typically targeting a low Earth orbit (LEO) and a mission length of months rather than years, these new space missions tend to use rad-tolerant or automotive-qualified COTS electronics to save costs or investigate new technology.
Available at much lower cost points, automotive-grade and COTS electronics meet a baseline of reliability standards and performance for industrial applications but are not designed with radiation robustness in mind. While some COTS parts may show inherent radiation tolerance, they may or may not have been designed for radiation robustness to the same degree that rad-hard components are.
Using COTS electronics introduces a host of unknowns, such as in part homogeneity and consistency across wafer lots and part traceability. To increase confidence levels for space applications, such devices may undergo further tests at additional cost before use, known as up-screening. This also extends to the use of wide-bandgap devices, such as silicon carbide (SiC) and gallium nitride (GaN) transistors. However, even with up-screening, there are no guarantees. Test results may vary, even from the same manufacturer. Or the COTS parts may not perform as needed and survive in radiation conditions. All this adds more risk to the project.
Rad-hard electronics offer traceability to a single wafer lot so that when doing destructive physical analysis or other screening, space designers can be confident of part uniformity and long-term performance, including radiation and reliability in space. Shorter, high-redundancy, sub-year missions and LEO satellites exploring new technologies may certainly benefit from using COTS components. However, for longer-term missions in which “significant risk of failure” is unacceptable, the benchmark for high-reliability electronics remains rad-hard Si.
Space’s radiation challenges
Radiation is pervasive in space and can negatively affect electronics that are without mitigation measures. Space radiation can impact functionality in two primary ways. Radiation that interacts with the die oxide layers can result in long-term, cumulative damage specified as the total ionizing dose. The second impact is that of single-event effects that can result in both recoverable single-event transient and catastrophic failures. A fast, heavy particle striking the gate region when a high voltage is applied can result in a high transient electrical field across the gate oxide, resulting in its rupture. This is known as a single-event gate rupture. A similar event in the drift region can also cause short-circuit between the source and drain. The best case is that it is only a momentary non-destructive short-circuit. At worst, it can result in irreparable damage, known as a single-event burnout.
Using rad-hard electronics protects against such failure mechanisms. As an example, rad-hard Si MOSFETs were originally introduced in the 1980s using design and manufacturing techniques to reduce sensitivity to radiation exposure. Over the years, more robust design, manufacturing know-how, screening, and qualification have evolved to virtually ensure failure-free radiation performance.
Ultimately, whether to use rad-hard or COTS electronics depends on several factors — mission profile, performance parameters, function criticality, cost, and more. In some cases, sacrificing reliability and radiation immunity may be acceptable risks to help meet budget constraints or test new technology in redundant or less critical systems. But when prioritizing reliability, such as for highly critical functions or for longer-term, deep space, or interplanetary missions, rad-hard Si is the clear choice.
Streamlined upgrades are key
In this challenging environment, reuse of proven technology is key to mission reliability. Using flight-proven designs maintains the demonstrated reliability and the expectation for the chances of long-term success. Board layout and circuit optimization are a major design, test, and evaluation investment, especially for high-reliability applications. As an example, after spending significant effort to optimize the trace parasitics of a buck converter (Figure 2), upgrading to a more advanced, next-generation Si MOSFET is far simpler than starting a completely new design with a different technology like GaN. New footprint-compatible, more efficient Si MOSFETs like IR HiRel’s R9 can be dropped in for immediate performance improvements, with much less work needed for design justification and re-qualification.
Rad-hard Si MOSFETs support higher gate ratings (±20 V versus GaN –5 V to 6 V) and have 30- to 200-ns rise times (versus <5 ns for GaN), making them less susceptible to circuit parasitics. Reducing gate-source voltage sensitivity can be an issue for GaN, prompting time-consuming design iterations to optimize board layout. In comparison, Si MOSFETs are relatively forgiving when it comes to layout, making it easier to design circuitry that would survive voltage overshoots caused by parasitic inductances. Latest-generation Si devices also show improvements in die- and package-related parasitics, enabling higher-performance circuitry and efficiency gains without the significant risk tradeoffs of using GaN.
For high-switching–frequency applications, GaN’s small, <5-ns rise time may be compelling enough to outweigh its sensitivity to parasitics. However, using switches with extremely small rise/fall times does come at a cost in terms of more design, test, and evaluation time to optimize board layout and careful component choices, along with the need to reduce parasitics (Table 1).
For applications in which linear-mode operation is required, such as pass elements for linear regulators, short-circuit protection, and hot swap/soft start, Si MOSFETs are still the superior, more rugged option. When operating in the presence of drain-source voltage, it is necessary to consider safe operating area (SOA) characteristics. A device such as the 100-V R9 MOSFET from IR HiRel can operate, with the case held at 25˚C, for 100 µs at 50 V and 20 A. By comparison, a GaN transistor with similar voltage and current ratings fares worse, operating at the edge of the 10-µs boundary under the same conditions (green circle in Figure 3).
For load-switching or high-side–switching applications, P-channel Si MOSFETs are an excellent, simple, and reliable choice. Because the gate voltage plus the threshold voltage to turn the device on are lower than the input voltage, the driver circuitry in this application is exceptionally simple and cost-effective compared with N-channel FETs, whether Si or GaN. This is also beneficial in applications in which space is at a premium, such as non-isolated points of load and low-voltage drivers. It should be noted that, currently, there are no commercially available P-channel GaN options for space due to poor performance compared with Si alternatives. While theoretically possible, P-channel GaN devices are not easy to make with low resistivity and crystal defect density.
Thanks to a lower thermal impedance, jc, Si MOSFETs also show a smaller increase in junction temperature when subjected to pulsed power. Compared with an eGaN HEMT, the difference can be as much as 25%.
Transients caused by radiation or battery/load issues often result in switches being immediately engaged/disengaged to protect circuitry. Any series inductance can generate a di/dt-induced voltage spike, leading to avalanche current flow if this surpasses the specific breakdown voltage (Figure 4), which serves as a self-clamp. Provided the switch junction temperature is not exceeded, rugged new-generation Si MOSFETs can recover, returning to normal operation under such conditions.
While there are commercial GaN parts that list higher permissible drain-to-source voltages beyond their absolute max ratings, none are yet available as rad-hard. Due to absence of this self-clamping in GaN, an unabated increase of drain-to-source voltage beyond the rated value could lead to reduced usable lifetime or catastrophic destruction, thus making rad-hard Si the more rugged option. Notably, rad-hard Si MOSFETs up to 650 V, such as Infineon’s latest ESA-qualified PowerMOS devices, are now on the market.
Rad-hard Si MOSFETs
The rad-hard R9 MOSFET series from IR HiRel is the latest generation of Si devices designed explicitly for space-grade electronics’ challenges requiring high reliability, robustness, and traceability. A simple drop-in replacement enables reuse of established, flight-proven designs, delivering system-efficiency improvements with minimal effort and reducing cost per bit in high-throughput satellites. Designers benefit from R9’s compatibility with a wide range of gate drivers and less sensitivity to parasitics, higher-current capability, and better SOA in linear-mode operation than alternative technologies. These Si devices also offer space application designers drop-in performance and packaging improvements over previous-generation rad-hard MOSFETs while maintaining the established and expected levels of traceability and reliability.
Qualified to MIL-PRF-19500 JANS and released direct to DLA’s Qualified Parts List (QPL), R9 MOSFETs are available in multiple package options, including the new SupIR-SMD (Figure 5). SupIR-SMD offers significant improvements to relieve thermally induced stress in the solder joints between the circuit board and package.1 There are currently no GaN options for space that are qualified to industry standards like MIL-PRF-19500 or are available as DLA or ESA QPLs.
Choosing the right components is essential to the success of all space missions, but many factors—like mission profile, budget constraints, risks, and more—influence which parts and technologies will best suit the situation. As the industry and technologies evolve, designers will undoubtedly find use for both COTS and rad-hard components. At this time, however, only rad-hard Si devices have demonstrated flight-proven heritage from decades of use, coupled with well-established quality and reliability standards and a wealth of technological understanding. Furthermore, with rad-hard Si, system designers can be sure that such devices are JANS- and QPL-qualified and can meet TOR requirements for the missions that need these levels of reliability. For the highest levels of confidence and reliability in space applications, rad-hard Si remains the benchmark.
This article was originally published on EEWeb.
Chris Hart is Senior Director of Marketing at IR HiRel, an Infineon Technologies company.
New products & solutions, whitepaper downloads, reference designs, videos
Register, join the conference, and visit the booths for a chance to win great prizes.