>How large an area would you need to cover to accuratly dim a star's output to an array at another star, say 100 light years away
Jupiter-sized is good enough for some 3700 light-years with 2016-era technology [1]. We know because that's one of the current methods of detecting exoplanets [2].
On the other hand Kepler 37b is only slightly bigger than the moon and its 200 l/y away.
Planetary transistion detection isn’t just a question of size (of both the star and the planet) and it’s orbit but also of alignment of the orbit of the planet, the orbit of earth and the two systems.
Now with a predetermined communication array the alignment isn’t a problem more so you can use a light shield at the receptor similar to StarShade which will be used to block the light from the star allowing Kepler to image the planets directly to block the majority of the light from the star which would reduce the required size of your transmitter further.
It’s quite likely that over a distance of even a few 100 light years with a modulator as small as a few 100 km across might be sufficient.
Jupiter-sized is good enough for some 3700 light-years with 2016-era technology [1]. We know because that's one of the current methods of detecting exoplanets [2].
--
[1] http://www.sci-news.com/astronomy/kepler-1647b-circumbinary-...
[2] https://en.wikipedia.org/wiki/Methods_of_detecting_exoplanet...