ESA GNC Conference Papers Repository

Over a decade ago ESA successfully instigated the development of CMOS-based image sensors for optical navigation. This resulted in imaging-only sensor chips which by now are well established in star trackers and sun sensors [1]. Then, in 2004 a technology demonstrator emerged, integrating modest digital signal processing with the sensor on a single chip [2]. While not meant to be a true product, this proof of concept was convincing enough to make it to in-flight demonstration on UK-DMC-2 [3]. From 2008 on ESA has been planning the next generation of navigation sensors, again following two roads: one aimed at electro-optical performance, the other at bringing more integration in support of a future class of miniature optical heads. On this second road lies the Faint Star: a system-on-chip combining improved performance, with the bulk of pixel-level and object-level processing required for star and sun sensing, as well as applications such as visual monitoring and rover/lander guidance. Faint Star is easy to integrate, and robust to operate, using full radiation hardening against total ionising dose (TiD), single event upset and latch-up (SEU/LU). This paper looks at the Faint Star's initial requirements, and at the winding road taken to harmonise these with today's industrial reality. A lot of changes had to be introduced, resulting in the design and functionalities presented in detail here. Even so the chip's physical, mechanical and electro-optical aspects are largely neglected in this article. Currently the Faint Star is in the detailed design stage. Samples will be manufactured and characterised through 2015. Then industry will receive samples for evaluation. In a second phase a (possible) design update will be done, followed with a full evaluation/qualification campaign.