"Alan Peake" wrote in message
...


Dave wrote:
a parabolic reflector fed with a feedhorn. no 'elements', just a
hole in a pipe and a big curved plate. you need to define the
parameters a bit more.

OK, maximum gain for a single frequency, free space, sidelobes and
back-front ratio not important. Not concerned about number of elements -
only minimum total material length. Doesn't need to be rotatable - this is
a purely theoretical exercise.

Parabolic reflector sounds good but it's a bit hard to quantify for the
purposes of minimising total material length. Perhaps one could use a
wire mesh dish. Would that use more or less material than a Yagi? I
would imagine that a phased array radar could use the wire mesh approach
but the same questions would apply as for the parabolic reflector. Same
for corner reflectors.
Arrays of driven elements may be promising but the few such antennae
that I've simulated so far, use more material than Yagis for the same
gain.

Alan


length is not a property of 'material'. mass, volume, their ratio, density,
conductivity, color, hardness, etc, are properties that can be measured.
'theoretically' the best antenna is a conductor from the source to the
receiver. a parabolic reflector can have area and thickness, therefore
volume, but the area is variable depending on how thick or thin you can make
it. any wire can be made into a parabolic reflector by smashing it thin
enough, witness the reflectors used on deep space satellites that are
extremely thin and light. or the metallic coating of a telescope mirror
that may be only a few atoms thick and yet yields tremendous gain. phased
arrays for radar get better as you remove more material from the surface
they are built from, the more holes, the better the pattern can be... so
less is more. arrays of driven elements, like the lpda, while looking
impressive and using lots of material, perform poorly at a single frequency,
but have the advantage of performing equally poorly over a wide range of
frequencies. designing antennas is a game of tradeoffs.... bandwidth for
gain, size for efficiency, gain for size, add in weight or some other
constraint like diameter and length of tubing, or dollars worth of
materials, and you add a whole new dimension. and then you need
'practicality'. as our friend art has found, you can feed parameters into
an optimizer program and let it run wild and get a supergain antenna that
fits in a shoebox, but try to build it and you get an air cooled dummy
load... or something that only induces currents on the support structure or
feedline.

the first step of engineering an antenna is to constrain the design with
practical measures... frequency range, size, weight, wind load area, cost.
then research possible alternative designs. then tweak the possible designs
carefully to see if they can be adjusted for your specific use. but be very
careful, if you suddenly find the tweaked design providing much larger gains
or varying greatly from the starting point, back up and see what has
happened... something is wrong. the most common problem is that someone
takes a standard yagi and puts it into an optimizer and sets it for 'max
gain' at one frequency, with no other constraints. the optimizer chugs
along and the gain goes up, and up, and up, and up!!! but when you look at
the results there are several elements bunched around the driven element and
the feedpoint impedance has gone down to a fraction of an ohm. don't apply
for a patent like art, throw it out and start over with more reasonable
constraints. give it a range of frequencies, constrain the feedpoint
impedance to a useful range, limit the element spacing, the total boom
length, etc, until it gives you something slightly tweaked for your specific
application but not off in left field.