While classically, orbital satellites were massive, tough to launch, and extremely expensive (a few $Bs), the current (and rapidly accelerating) trend has swung decidedly towards using relatively low-cost (a few $M) and easy to launch constellations of single or multi-U CubeSats (1U = 10x10x10 cm3) to cost-effectively address the plethora of emerging needs. These days, this has been increasingly supported by commercial space ventures (e.g., SpaceX, BlueOrigin et al., vs. the old gang—NASA and DoD), which are proliferating rapidly.
As appealing as space is for visioning fun new science and slick applications, it remains a decidedly unfriendly place to visit. Space is the quintessential “extreme environment,” bathed in intense radiation from both our Sun (high energy electrons and protons trapped by the Earth’s magnetosphere in radiation belts) and the cosmos (GeV energy galactic cosmic rays from supernovae). By way of level setting, a satellite in the most benign Earth orbit, Low Earth Orbit (LEO – 160-1000 km up from the surface), experiences 100,000 rad of ionizing radiation dose over mission life. In comparison, 500 rad will do a person in! That is, we are asking a lot of our electronics in such systems, and given the extreme cost constraints of launch weight, adding a few inches of lead shielding is not the ideal solution! In addition, it is mighty chilly in space (2.73 K = -455°F, the cosmic background), and when the sunlight shines on you, it gets uncomfortably warm, very quickly (e.g., on the surface of the Moon, from -180°C to +120°C from darkness to light, within a few moments). Yep, space is a tough place to do business.
As I have long argued [1], SiGe HBT BiCMOS technology provides a unique solution for many of the needs of these emerging space systems, including: 1) extreme levels of performance (multi-hundred GHz) with the SiGe HBT and high integration levels with on-board CMOS, for realizing compelling system functionality/unit volume, at low cost; 2) the rapid improvement of all electronic circuit relevant performance metrics with cooling, with operational capability down into the mK quantum regime (SiGe HBTs love chilly weather!); 3) the ability to operate robustly up to 150-200°C, with modest performance loss; 4) the ability to operate robustly over wide temperature ranges (in principle from mK to 150-200°C); 5) built-in robustness to multi-Mrad total ionizing dose radiation; and 6) built-in heavy ion induced latchup immunity (read: those pesky GeV cosmic rays).
Long ago (1990s), the notion of creating a low-cost Si-based electronic + photonic integrated circuit (EPIC) “superchip” was envisioned (Soref), which brought together advanced SiGe HBTs (analog, RF-mmW), CMOS (digital), and Si integrated photonics (with the possible exception of a laser, which could be flipped onto the die worse case). In essence, EPICs are a low-cost, high-yielding, reliable, highly integrated Si platform for putting electrons and light into the same conversation! Clearly this represents a paradigm shift to business as usual. Now, with even more compelling system functionality/unit volume, at low cost. Such an EPIC superchip could in principle satisfy all-comers-of-new-needs. While photonics has long been used in space (think solar cells, imagers), EPICs are new to that space game, but possess great potential for the emergent needs in this new vision of CubeSat/SmallSat driven space systems, including, thing like: LIDAR (spacecraft-to-spacecraft positioning); deep space and within-constellation optical communications (huge data rate improvement); and on-spacecraft high bandwidth data transport (think data center in the sky for instruments that spew out tons of data that need to get back home quickly). This field of EPICs in space is only a few years old, but already much has been learned, and results look very encouraging.
In this invited talk, I will highlight the current status and the future trends of using SiGe electronic and photonics in space systems.
[1] J.D. Cressler, Proc. IEEE, vol. 93, pp. 1559-1582, 2005.