[Discuss-gnuradio] amateur radiotelescope/SETI receiver using SDR?

[Pete Heist]

Ignoring SDR for just a moment I've been trying to
design a relatively low cost home brew amateur
radiotelescope / SETI receiver. Having read
information online I understand most of the basics but
I'm a first-timer when it comes to this stuff. I have
a comp eng background with good software experience
and some signal processing theory but no practical
experience with microwave receivers or amateur radio
so bear with me if some of what I'm suggesting is
ludicrous.

I was reading slashdot this weekend and saw the
article about the GNU Radio project, so immediately I
thought wow, maybe there's an opportunity to make the
creation of my radiotelescope either cheaper, better,
more of a pain in the ass or even all of the above. I
guess an advantage would be that you can improve some
of the receiver functions in software rather than
hardware. SETI folks look to scan more bandwidth at a
time to "increase the chances". Perhaps you could do
this through software improvements rather than
expensive hardware upgrades. On the flip side your
processor has less time to spend looking for candidate
signals now that it's busy performing more of the
receiver functions. Scanning through the digest of the
list I've begun thinking it's "crazy" or prohibitively
expensive at this point since there aren't any posts
about it.

I wanted to propose a simple design and request some
feedback to see if I'm heading in the right direction
and if this is even feasible. Basically it would be
this diagram: 

http://www.setileague.org/hardware/blkdiag.htm

except the receiver component is replaced with a
downconverter and an ADC suitable for use with GNU
Radio. My main questions are:

1) What would be an appropriate downconverter and ADC
combination for this unit? If I can get a
downconverter that takes 1.4GHz and makes an IF of
144MHz or 28-30MHz range am I even in the right
ballpark for use with a low cost ADC board? At $1300
The DAS4020/12 board that's supported by GNU Radio is
not exactly cheap and it only samples up to 20MHz, so
is this not even usable for this application? Can I
convert to a lower frequency and get a cheaper board?
What's the tradeoff here, bandwidth?

2) What processing needs to happen in the GNU Radio
side? It seems like "all we have to do" in the case of
SETI is do an FFT on the signal and scan for
interesting signals (nevermind that it's anybody's
guess as to how to do that). How much bandwidth might
I expect to be able to scan with a dual PIII-450? I
feel like both processors could easily be kept busy
trying to do a real-time FFT on 12-bit samples at
20MHz, but maybe I'm wrong.

Finally, any other thoughts or ideas on this would be
appreciated. In general GNU Radio looks like a really
interesting project.

Pete

__________________________________________________
Do You Yahoo!?
Yahoo! Tax Center - online filing with TurboTax
http://taxes.yahoo.com/