How is a space probe's antenna directed toward Earth?

I wonder how is it possible to direct the probe's antenna from enormous distances (like Pluto) toward the Earth. Because if even an extremely small inclination in antenna direction emerge, it will make a huge miss (of target Earth), having in mind such long distance. How is this problem solved, or maybe I misunderstood something ?

Also, does for example, Mars orbiter have to change antenna position over time in order to retain direct visibility to Earth ?

• Related question (about the antenna position of the receiver on Earth): space.stackexchange.com/questions/9803/… – Hobbes Jul 11 '15 at 8:50
• The emission beamwidth is also a specific angle (dependent on the relationship between antenna size and transmission frequency or wavelength). For large antennas and high frequencies (upper UHF and higher, which is what the DSN uses) this can quite easily become fractions of a degree, but over tens of AU, that's still quite a lot of margin. You'll want to maintain as good a lock as possible to maximize the useful signal strength received, but it's not a total disaster if you are slightly off: the mission's radio link budget is going to have to include some margin for error and variance anyway. – user Jul 11 '15 at 12:28
• Also, the beam width isn't a binary on/off either, but a gradual reduction of signal strength. You pick a value, say, -30 dB to peak power, and calculate the specific antenna's beam width at the specific intended frequency to that power ratio, then state that as the transmission beam width. If you are slightly outside of that, you might receive at -32 dB to peak power, whereas if you are slightly inside of it, the value might be -28 dB. – user Jul 11 '15 at 12:36

See this answer for how accurately the antenna must be pointed. That accuracy is not a function of the distance. It is only a function of the size of the antenna and the wavelength of the transmission. The antenna on New Horizons has the same required pointing accuracy at Pluto as it did at Jupiter. What changes is the received power at Earth, and therefore the maximum data rate. That goes down with the square of the distance.

The size of the antenna and the wavelength determine the tightest possible angle that the beam can be focused to. Once you have that angle, you just need to keep Earth somewhere in there. The angular size of the Earth at the distance of the spacecraft is generally much, much smaller than the beam width of the antenna. So only a tiny fraction of the power from the antenna falls on the Earth. An even tinier fraction falls on the receiving antenna on the Earth.

Antennas are pointed using star trackers and an on-board model of the position of Earth relative to the spacecraft on its trajectory. Star trackers are cameras that accurately determine the attitude of the spacecraft by matching star field patterns. Just a few bright stars in the field of view will do the trick. Typical accuracies are a few arcseconds. Ball has a 0.1 arcsecond tracker.

Yes, Mars Orbiters and all spacecraft have to change the pointing of their antennas over time to track the Earth. Both the Earth and the spacecraft are moving in inertial space (i.e. relative to the stars the star tracker is tracking), so the direction to point the antenna is always changing.

• Does the probe get any help from Earth? You could set the DSN antenna to broadcast a strong signal that the probe can receive so it knows it's pointing in the right direction, and can finetune its aim based on signal strength; does that happen in practice? – Hobbes Jul 11 '15 at 17:07
• Yes, the probe gets help from Earth in that the onboard model of where the Earth is has to get updated periodically. – Mark Adler Jul 11 '15 at 17:41
• In theory, yes, you could have multiple off-center feeds from a high-gain antenna to allow you to fine tune the pointing of the antenna to point at a broadcasting station. However I am not aware of a deep-space spacecraft that does that. The pointing approach I described has been sufficient in all cases. – Mark Adler Jul 11 '15 at 17:44
• "At a Pluto-like distance, the high gain antenna boresight, mispointed by 0.2 degrees, would miss Earth by over 10 million miles," (Source). "The HGA provides better than 42 dBic gain within 0.3° of the +Y axis. The medium gain antenna allows communication at larger angles between the +Y axis and Earth (up to 4°), and specifically allows commands to be received by the spacecraft at ranges up to 50 AU."(Source). – mins Aug 5 '15 at 17:14
• @mins: That is useful data. You should turn your comment into an answer. – Mark Adler Aug 5 '15 at 22:01