space@mit-mc (02/18/85)
From: Dale.Amon@CMU-RI-FAS Why hasn't anybody considered building an all electronic telescope? Why even bother with a mirror? A large flat plate of CCD must certainly be cheaper than the expensive optics and aligning techniques now being used. I cannot think of a good reason why it wouldn't work, and I'm curious if anyone out there can see a problem. For those to whom the idea is not immediately obvious, the main mirror would be replaced with a flat plate of photomultipliers impinging on CCD. About 3-4 years ago Scientific American ran an article on some of the new technology for making the zone plates better and cheaper. With this approach we can count photons and arrival times immediately. The scope would consist of nothing but a tube to prevent extraneous light from getting in, the 'objective' plate and a computer with humungously high speed and enormously large storage capabilities. The only thing I'm not sure of is how to get a good spectral resolution. What is really desired is a vector of length 4 identifying each impinging photon over a useful delta-f with <x,y,t,f> Even with the above computing requirements, it MUST be cheaper than forming a few tons of glass to within a 1/4 wavelength of a complex 3-d figuring curve!!! If anyone decides to build one, please name it after me, or at least send me the check for the difference between the electronics and the optics...
ethan@utastro.UUCP (Ethan Vishniac) (02/18/85)
> From: Dale.Amon@CMU-RI-FAS > > Why hasn't anybody considered building an all electronic telescope? Why > even bother with a mirror? A large flat plate of CCD must certainly be > cheaper than the expensive optics and aligning techniques now being used. > > For those to whom the idea is not immediately obvious, the main mirror would > be replaced with a flat plate of photomultipliers impinging on CCD. About > 3-4 years ago Scientific American ran an article on some of the new > technology for making the zone plates better and cheaper. > > With this approach we can count photons and arrival times immediately. The > scope would consist of nothing but a tube to prevent extraneous light from > getting in, the 'objective' plate and a computer with humungously high speed > and enormously large storage capabilities. > > The only thing I'm not sure of is how to get a good spectral resolution. > What is really desired is a vector of length 4 identifying each impinging > photon over a useful delta-f with <x,y,t,f> > > Even with the above computing requirements, it MUST be cheaper than forming > a few tons of glass to within a 1/4 wavelength of a complex 3-d figuring > curve!!! > *** REPLACE THIS TELESCOPE WITH YOUR CCD *** Hmm........ As a theorist I'm probably not the person to answer this, but here goes... The problem is not obtaining reasonably spectral resolution, that can be done with filters if necessary. The problem is that the scheme as presented here has no means of distinguishing between photons coming from slightly different directions. It has no *angular* resolution. This is the whole point of the optics, to create an image, to separate out photons according to their wavevectors. A CCD will simply record whether or not a photon has struck its surface. In order to make a telescope like this work one would have to restrict its field of view to the angular resolution desired, i.e. less than an arc second. Note that I haven't mentioned (yet) the difficulty involved in covering an area comparable to a large telescope mirror with a coherent CCD array. "Don't argue with a fool. Ethan Vishniac Borrow his money." {charm,ut-sally,ut-ngp,noao}!utastro!ethan Department of Astronomy University of Texas Austin, Texas 78712 *Anyone who wants to claim these opinions is welcome to them*
space@mit-mc (02/19/85)
From: dual!ihnp4!iheds!hon I don't believe you would get any magnification with your single flat plat. When you record the photons with the CCD you lose the phase information. The mirror preserves the phase and allows the magnified image from correlated photons to be cast at the focus where it can then be causght by a CC. Herb Norton Bell Labs
space@mit-mc (02/23/85)
From: <dual!qantel!hplabs!tektronix!carlc> I don't buy your electronic telescope scheme. A mirror "knows" more than just the position (x,y,z) of a light ray impinging on it, it knows the *direction* as well. Without direction information, how do you know where in the sky a ray came from? In other words, how does your scheme *focus* the light? --Carl