I must say that I dont know whether the original 'DISHY' is simply a
dish antenna with an analog amplifier and maybe some mechanical motor
steering, or whether DISHY includes a computer to execute some protocol,
some algorithm.

It's a phased array, not a dish, even if it looks like one. It consists of 100's of fingernail-sized antenna elements that:

• during transmissions, have an individual phase delay added to the signal transmitted from that element, in order to permit transmission of the combined signal from all elements into a particular direction.
• during reception, have an individual phase delay added to the signal collected by that element, before the signals are added to obtain the combined received signal. This allows reception from a particular direction.

Dishy's main direction of transmission / reception is therefore not its surface normal - this simply points to the area of the sky where Dishy expects to see most satellites (a function of geographical latitude and constellation design - essentially straight up in the tropics, and elsewhere in the direction of the 53rd parallel, which corresponds to the predominant orbital inclination in the Starlink fleet). The actual tracking is then done with the phased array without mechanical movement by Dishy.

From what I've seen, Dishy seems to consume more power on receive than on transmit - that's if you actually download stuff. This is somewhat counter-intuitive if you're used to putting link budgets together. But I'd attribute that to a higher degree of digital signal processing required on the receive and demodulation path.

****************************************************************
Dr. Ulrich Speidel

School of Computer Science

Room 303S.594 (City Campus)

The University of Auckland
u.speidel@auckland.ac.nz 
http://www.cs.auckland.ac.nz/~ulrich/
****************************************************************