In order to aim an antenna at Earth, how do unmanned spacecraft find their own orientation (xyz position in space, what direction it is flying and which direction the antenna is pointing)? Are they using a 'camera' to look at stars? Do they need a telescope lens and long tubes to get good image magnification and angular resolution?

How can they move/steer/rotate themselves? Are they using reaction wheels, which run on un-limited electrical power or using limited-supply gas jet thrusters?


1 Answer 1


Spacecraft use a star tracker to find their orientation (attitude).

Here's a random example:

Terma star tracker

Dimensions and Mass:

  • Camera 120 by 120 by 33 mm, 1.0 kg
    (note: optics protrude 58 mm inside baffle)
  • Processor 245 by 165 by 29 mm, 1.2 kg

Baffle Examples:

  • 30º (sun exclusion), Ø234 by 346 mm, 800 g
  • 45º (sun exclusion), Ø167 by 203 mm, 470 g
  • 60º (sun exclusion), Ø160 by 211 mm, 530 g

So the camera and lens are tiny, they just need a small sunshield (baffle) to work correctly.

- <1 arc-seconds RMS pitch, yaw
- <5 arc-seconds RMS roll

Spacecraft use thrusters, reaction wheels, or both to control their attitude. Thrusters have a finite amount of propellant, but can be more reliable than reaction wheels. For long-term missions (including deep-space missions like New Horizons), thrusters are commonly chosen.

Edit: Accuracy is much better than is needed for pointing the antenna.

  • 1
    $\begingroup$ Note that the angular resolution doesn't have to be particularly good for this to work, because radio antennas don't have to be pointed perfectly accurately. Early space probes used similar star trackers with a small number of individual photocells instead of a modern CCD array. $\endgroup$ Commented Jun 13, 2016 at 14:41
  • $\begingroup$ Is 'typical' high gain antenna for these space crafts have beam width of 1 degree or so at 10GHz? This example is 1 degree, en.wikipedia.org/wiki/… If 'eye' only provide basic accuracy, do they move antenna is a 'sweeping pattern" while listening to constant beacon from Earth? Same as setting roof SatTV antenna by human hand sweeping around an initial rough predicted orientation? $\endgroup$
    – EEd
    Commented Jun 13, 2016 at 21:03
  • $\begingroup$ The star tracker is far more accurate than 1 degree, I've added some data to my answer. 1 arc second is 1/3600 degrees, so it's not necessary to do a sweep. $\endgroup$
    – Hobbes
    Commented Jun 14, 2016 at 6:09
  • $\begingroup$ Do you need one, two or three star trackers at xyz plane of craft? Are these typically electrically or otherwise cooled, so that it can do long exposure and see a lot of darker stars? How long exposure is needed to see enough number of stars, given the lens is small and gather limited light? When the craft is near Earth, how warm can the craft get as a result of sun shine? When far from Earth, it that only a few degree K? For typical big mission camera and small stat tracker, are thermal dimension changes large enough, requiring “motorized focus change” mechanism? $\endgroup$
    – EEd
    Commented Jun 16, 2016 at 13:07
  • $\begingroup$ You only need one star tracker. As far as I know, star trackers have no active cooling. The update rate of the linked example is given as 4 Hz, so exposure is less than 250 ms. $\endgroup$
    – Hobbes
    Commented Jun 16, 2016 at 13:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.