In article <4096738e$1@news.cadence.com>,
Louis Scheffer <lou@cadence.com> wrote:

private groups, it may be hard to prevent.  For example, there is 
already a company that will send a message of your choosing from a 
radio telescope in Russia.

Yes.  I'm aware of that.  The person involved with the science 
message (rather than the vanity messages) for this sometimes posts
on sci.astro.seti.  Note it is run by a North American company
and, at least for the original transmissions, used a radio telescope
in the Ukraine, not in Russia.  The total transmission time is 
limited, precisely because of the cost of transmit tubes that
I mentioned.

This seems intuitively reasonable, but is not true.  In a phased array,
replacing modules with smaller cheaper modules saves power.  This is
because the EIRP goes as N^2, so for constant EIRP, the power per module 

No.  There is nothing magical.  It scales with the area in the same way
as for a dish reflector (which is just a phased array with infinite
numbers of elements, but excited by a free space signal, not locally
at each element - think how you would compute the gain by integrating
over the surface area).  This should be obvious from the reciprocity
principle and that gain scales with capture area.  You can also get it
from books like the ARRL handbook, where the array gain for having too
stacked yagis, is 3dB, not 6dB (actually, in my copy, it is a throw away
line about expecting 2 to 3dB, as though it was too obvious to state).
I suspect where your confusion arises is that gain scales as the square
of the diameter, but you are talking about the scaling with area.

In fact, unless the array elements are far enough apart (outside
each others' "capture areas") the gain won't even scale as the number
of elements, but rather as the effective area covered by them).

goes down faster than the number of modules goes up.  

For example, imagine a transmitter with 1M modules of 1 watt each.  

Note that even for reception, the 1 square km array uses a hybrid
configuration with directional antennas at each array node.  That's
because the current break even in a receive configuration occurs when
some of the gain is done by passive reflectors.  Arrays are becoming more
cost effective for reception because the electronics cost is reducing,
and passive reflector costs scale more like the cube of the diamter than
as the area.

The 1 square km array is too costly to build at the moment, and
only exists as a small number of elements as proof of concept.

Incidentally, it would be more or less impossible to keep a million
element array operating on all 1 million elements; you would need
elements with enormous MTBFs.

If transmitters become cheaper, you can replace this with 10M 
modules of 10mw each, for a total power savings of a factor of 10.  

Based on false premise, see above.

True, but no rational SETI transmitter would use vacuum tubes any more,
since they don't last very long and represent a significant cost.

Current planetary radar uses vacuum tubes, as does the Evaptoria
transmitter for Cosmic Call.  I believe that strategic defence radars
do, as I understand that Arecibo uses second hand strategic defence
radar transmit tubes.