In article <v2stc.4694$Vf.184315@news000.worldonline.dk>,
"Jesus loves you <John15.13@Heaven>" <> (BOGUS ADDRESS) wrote:

fx. 1,42 GHz +/- 5000 Hz.

Note that the input to S@H is 1.42GHz +/- 1.25 MHz and the data sent to
the client has already been separated into multiple sub-channels.  Even
+/- 5000Hz is considered too wideband for an artificial signal confirmation
and the S@H processing (because of the 1 bit quantisation) will ignore
signals that wide.

The bandwidth analysed is as low as about 0.07Hz.  A 350MHz PII can
reduce ~10kHz by 107 seconds to this frequency resolution in just
over 1 second, so analyzing large numbers of narrow band channels is
not difficult.  Typical amateur SETI systems use sound cards and analyse
to about 10Hz, which is limited more by receiver stability than processing
power.

By this You mean, that noise from 1,42 GHz -5000 Hz to +5000 Hz is greather
that, noise from just -50 Hz to +50 Hz ...

100 times greater.

TerraWatt ?

Do You mean by this 1E12 W?

Yes.

Arecibo 1E13W * 1E6 (as output) = 1E19W

The  1E13 watts plus is effective isotropic power; it includes the antenna
gain.  The feed point power is only 1MW (1E6 W).

But ... if they send out the signal with 1E19W, what then ?

But then You also get's the noise from all these directions!

No.  In any one beam direction, a phased array behaves the same as
a big dish.  The noise voltages sum non-coherently, so they form
a gaussian statistical distribution with a mean of zero, and, more
importantly, a standard deviation that scales as the square root of
the number of elements.  The signal sums coherently, as the number
of elements.  The voltage signal to noise ratio therefore scales
as the square root of the number of elemnts and the power signal
to ratio as the number of elements.

That's exactly the same as a dish in a single direction.  Each patch
on the dish surface produces a contribution that is phased by different
path length the rays through it take and is added up, passively, at
the feedpoint.  In other words, to calculate the gain of a dish, you
actually have to treat it as though it was an array, with very small
element.

On the other hand, lots of individual dishes only get one unit of
signal and one unit of noise, so they do have a poor signal to noise
ratio.

Do You have a web-side ?

http://www.google.com

The SETI League, which I've already mentioned, has started assembling
a network of small dishes that are precisely intended to detect small
duration signals from any direction.  <http://www.setileague.org/>