Cubesat with 8 arcsecond pointing accuracy - is that typical, or exceptional?

Question

I just read in Satnews Daily that:

A 3U CubeSat funded by NASA and led by the University of Colorado’s Laboratory for Atmospheric and Space Physics, MinXSS requires precision pointing to carry out its scientific objectives. Integrating an array of high-performance miniaturized space sensors and actuators, including BCT’s Nano Star Tracker and Micro Reaction Wheels, the XACT unit provides state-of-the-art attitude control for small spacecraft like MinXSS. On-orbit telemetry from the XACT unit and independent spacecraft instrumentation indicates 1-sigma cross-axis pointing error better than 8 arcseconds, which is less than the width of the Empire State Building as viewed from Los Angeles. (my emphasis)

8 arcseconds is only about 0.002 degrees! Of course a "big sat" like a major telescope with high magnification (or high precision) can do much better, I thought that in general 1U to 3U cubesats were in the 0.1 to 10 degree range - mostly determined by a balance between cost and requirements.

Is this some kind of record for a 3U cubesat?

note: the MinXSS (and this and ArXiv) generally points at the sun to measure low energy X-ray spectra. So when it is making measurements it is not rotating with respect to the stars, as an earth-pointing spacecraft would have to do.

---

edit: I figured out where the need for high pointing accuracy might come from.

above: GIF made from frames of the video AR1226 M2.5 Flare and CME, June 7, 2011 05:00-13:00 AIA 211/193/171 Composite

The X-ray detector inside the Amptek X123 (and here) is a single silicon PIN diode, about 500 microns thick, with areas between 6 and 25 square millimeters shown as standard on the website. An optional pinhole collimator is also available.

The collimator is actually two pinholes spaced about 37 millimeters apart, and they come standard down to 25 microns in diameter. So it's possible that the FWHM could be of the order of 0.04 degrees configured off-the-shelf, and even narrower if a larger distance between pinholes was chosen. They may be able to point at the corona or a CME specifically and record a soft X-ray spectrum.

Answer 1

I'm one of the leads on MinXSS. Yes, that is unusually high pointing accuracy for a CubeSat but it's probably not a record. It's about to become commonplace because we were just the first to fly the Blue Canyon Technologies XACT attitude determination and control system (ADCS). There are already lots of them out there ready to be flown, including two at JPL that will be flying to Mars independently with the InSight mission in about a year.

We needed better pointing on MinXSS than is typical for a CubeSat because the Sun is half a degree across and we needed to keep it in our field of view of ±4 degrees. Obviously, 8 arcsec is way better than that, so it's overkill for us but BCT is just down the street from us and we have a strong partnership with them and we had to focus our development efforts elsewhere to ensure success. ADCS is hard, so we were grateful that they were just coming out with the XACT product right at the time we needed it.

We are measuring soft x-ray spectra from the sun, so we don't actually image the sun at all. We just need to make sure that it's in our field of view and we're okay. Now that we've demonstrated the XACT works even better than the specification sheet says, we're going to be flying it on lots of future missions that require higher pointing accuracy than was needed on MinXSS.

Answer 2

I'm not sure whether it's a record, but that's definitely an unusually fine pointing accuracy for a cubesat. Your range of 0.1 to 10 degrees is indeed fairly typical. It's mostly about requirements - because cubesats are physically small, they can't have very high gain antennas that need to be pointed (5 degrees is typically sufficient even for a high-end cubesat downlink such as Planet Labs' 200MBps X-band transmitter).

Likewise the limited aperture means they don't tend to have telescopes or other instruments with very narrow fields of view << 1 degree. So mostly there's no necessity to point better than a degree or so, which is readily achievable with low-cost magnetometers, sun sensors, MEMS gyroscopes, magnetorquers and simple reaction wheels.