Originally posted by twhitehead
Actually, once the technology is in place, we would probably start with as large as possible. We would probably however start with earth orbit satellites with a direct line of sight to each other that doesn't go through too much atmosphere. So I don't think it would go 1km, 10km, .... but rather depend on the placement of the satellites.
You are thinking in terms of satellites around Earth. I am thinking in terms of probes in space far from Earth, the farther the better. For instance, there is a dust cloud around the distance of Neptune I think called Gigenshien, something like that, don't pretend to know the proper spelling, but anyway, ALL telescopes in and around Earth, Hubble, Webb, whatever, are limited by that cloud. If you can get a system of telescopes past Neptune, maybe out to pluto and separated by a few million Km, the full use can be made of the technology, if the same equipment were say, only 100 million Km away from Earth it wouldn't do any better than if it was around Jupiter. Only if you get past Neptune can optics be used down to its own limit.
That is one limitation of very large telescopes near Earth that cannot be overcome no matter what kind of super technology you implement. It clouds the images of stars trying to get images of planets and so forth because the star lights up the cloud and that light, dim as it is, for the best of the best, interferes with imaging objects close to a star.
That said, you can still do a lot of great science if you can get them separated by any distance.
The problem with just jumping to the greatest distance is you have to start small, that's why the multiple mirror scopes on earth are not very far apart, they have to have vacuum tunnels between them, accurate laser beams and so forth. It's a very tricky technology, a lot harder to accomplish than the combinations of radio telescopes that has been going on for 50 years or more. The latest on that front is a Russian craft with an onboard radio telescope has just been launched with the idea of putting it in a way looping orbit that actually goes past the moon and connected to the system of telescopes on earth to extend the effective size to the distance between them so instead of an effective reflector size of 8000 miles or 13000 Km, it would be 300,000 miles or about 500,000 km, expanding the resolution of radio telescopes 40 times what is available to earth bound systems of radio telescopes now. Mind you, there would be little if any increase in the sensitivity of the scope but the resolution will be phenomenal. The radio telescope system is already hundreds of times greater resolution wise than even the Hubble because of the vast overall effective size of the reflectors. They talk about milli and micro arc seconds of resolution where the Hubble can only do about 0.05 arc seconds, phenomenal enough but come back in 50 years...
The problem with radio telescopes is resolution is related to not only just the mirror size but the wavelength as well, so if you build a radio telescope for 1 Ghz, a wavelength of about 1/3 meter and you get X amount of resolution, if you want to use it at 0.5 Ghz, the same dish will now have half the resolution because the wavelength is now about 2/3 meter in size.
So if you go the other way in frequency, 10 Ghz, that same dish now has 10X the resolution of the 1 Ghz frequency because there are ten wavelengths at 10 Ghz for every one wavelength at 1 Ghz.
If you used a radio telescope with a 2 meter dish and you were observing at 2 meters, you would only be able to concentrate one wave at a time so would not get much gain in sensitivity or resolution. But a 2 meter dish at 0.002 meters would now have not one thousand wavelengths but 1000X1000 wavelengths, times .78 if it was a circular shaped dish, or 780,000 waves concentrated to one so you can see as you go up in frequency, or down in wavelength, the same mirror gives you greater resolution.
So a two meter mirror like the Hubble can give that 0.05 arc second resolution but if it was a radio telescope at 1 meter wavelength it could only concentrate about 3 waves into one with a very poor angular resolution, maybe 30 degress or so at best, a far cry from .05 arc seconds.
That's why they need separation of dishes as far apart as possible to surpass the resolution of the Hubble, which now, they do with ease. They can see stuff that the Hubble can't because it is at Ghz frequencies and with much much better resolution than the Hubble could ever think about.
So if you can make two Hubble's separated by a few thousand Km in space, communicating phase data by laser beam, the resolution would far exceed the best radio telescopes.
But for the absolute best use, they need to be out in space past Neptune to avoid the fogging due to the Gegenschein, ( I googled it):
http://en.wikipedia.org/wiki/Gegenschein