On Sat, 26 Mar 2016 18:31:40 -0400, rickman wrote:

Whatever you do with the antenna should work in both directions, no?

No. The gain can be quite different between tx and rx. The usual
symptom of a badly designed reflector antenna is that it can hear/see
wireless access points at considerable distances, but can't connect to
very many of them.

Let's take a parabolic dish reflector as an example. The general
assumption is that the transmit and receive gains are identical.
That's true only if ALL the RF produced by the feed hits the dish and
does NOT spill over the edges of the reflector. See Fig 6.0-1 at:
http://www.w1ghz.org/antbook/chap6-0.pdf
http://www.w1ghz.org/antbook/contents.htm
Any transmit RF that goes over the edge is lost and useless.

You can also screw things up by using an excessively high gain antenna
for the dish feed. Such antennas have a narrow illumination angle,
producing a "spot" on the surface of the dish. That reverses the
tx/rx situation. In transmit, all the RF produced by the narrow feed
hits a spot on the dish, and is reflected in the desired direction.
Nothing is lost. However, in receive, the feed only sees RF coming
from the spot. The rest of the dish is wasted as any rx RF that hits
outside of the spot, will not be "seen" by the feed.

Putting a USB dongle at the focus of a dish reflector is the common
way to obtain some additional gain. Everything that hits the
reflector is reflected towards the feed. So in receive, the gain
calculations are fairly close to the theoretical maximum values.
However, in transmit, much of the RF emitted by the USB dongle goes in
directions that do NOT hit the reflector. Some of these might
accidentally go in the desired direction, but most of the RF is lost.
What's left hits the dish reflector, goes in the right direction, and
if you're lucky, provides some gain. I posted some calcs for this in
alt.internet.wireless a few years ago, but of course, I can't find
them. I'll try again later.

For the tx gain to equal the rx gain, something will need to be done
about the tx RF going in random directions. The basic idea is to
match the "illumination angle" of the feed with that of the dish.
That's done by making the feed directional and setting the beamwidth
to match the f/D ratio of the dish. See:
http://www.w1ghz.org/antbook/chap4.pdf
Putting a reflector about 1/2 wave behind the USB dongle will help
some, but a properly matched feed will work much better.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558