Edmund <nomail@hotmail.com> hath wroth:
>>Same with RF. A high gain feed will just illuminate a spot. A
>>hemispherical feed (like a USB dongle) will spray RF everywhere except
>>the dish. Only a matched feed will work.
>I don't see any problem with a feed pointing at only
>a part of a dish. The reflection of the dish still will go
>to the point where the disk is aiming to and no energy
>is wasted in a wrong direction.
>Correct me if a am wrong about that.
You're wrong.
http://en.wikipedia.org/wiki/Dish_antenna
In receive, the dish picks up all the signal it can get and sends it
to the feed. If the feed has to wide a beamwidth (such as using a
hemispherical feed that picks up from all directions), the gain will
be the same as with a properly designed feed. However, if the feed is
a spot beam, it will only pickup the percentage of area covered on the
dish by the spot, and therefore have much less gain.
In transmit, the situation is the same. A hemispherical (too wide)
feed will spray much of the RF over the edge of the dish, where it is
wasted. However, you're correct for the spot beam. All the RF
generated by a spot beam will be reflected by the dish and sent to the
remote target. In effect, this combination uses the feed to obtain
its gain, and largely ignores the capture area (effective aperture) of
the dish.
There's quite a bit on the subject of feed design in:
http://www.qsl.net/n1bwt/contents.htm
It might be a bit too techy, but try reading just Chapter 4 (21 pages)
which discusses feed design.
>One other thing I like to know is the influence of a reflector
>on a feed mounted on a dish. As said the reflector in some
>examples are still 110 mm square. While a the focus of a dish
>should be a about 1 mm^2. Does this mean a biquad with such
>a large reflector has to be mounted off focus?
Yes, it has to be mounted as an "offset feed". See above URL.
The problem is that the feed is so large as compared to the effective
area of the dish, that any feed blockage will cause a substantial
decrease in gain. For very small antennas (i.e. 0.6m DBS pizza dish
antennas), this is a serious problem. The dish would need to be about
2m across before the blocking effect of the feed can be ignored.
Please note that the DBS dish was designed to work at 12-13GHz, where
the feed diameters are much smaller. The scale is linear so a 10cm x
10cm biquad feed would be only 18cm x 18cm.
The problem with a biquad feed is that it's to wide and therefore
blocks quite a bit of RF going to the dish. A better choice is a
dipole or 1 element yagi, which is narrow and does NOT block much RF.
This is what's inside one of the MMDS (2.5GHz) dish antennas.
http://home.alphalink.com.au/~vak/galaxy/8.jpg
It's a circuit board 2 element yagi feed. The traces close to the
circuitry are the dipole. The wide trace with the step is the
reflector. No clue what the trace in between does.
>I really don't know anything about antenna's but I do know a little
>about reflection. So I expect since the biquad is much larger is size
>then a single point, it should be mounted off focus.
Finding the optimum location is difficult. All antennas have a phase
center. That's the point where all the RF appears to be coming from
if you measured it from a distance. That's also the point where you
want the focus of the dish to appear.
If you mount the feed off focus, you lose more than gain. There could
be frequency selective cancellations (nulls), boresight errors, and
excessively wide beamwidth. If you have a flashlight with an
adjustable focus (Maglight), you can see the results as you move the
position of the lamp through the reflector focus.
>Maybe a fair distance towards the dish or a little distance away from
>the dish. Hopefully this drawing makes clear what I mean.
>For the ease of the drawing the dish is pointed downwards.
>
>
> I S
> D H
> \ /
> \ /
> \ /
> \ /
> \ /
> \ /========\ / === biquad
> ____\/__________\/___ ___ reflector towards dish
> \ /
> \ /
> \ /
> \ /
> \ /
> ______\/__________ FOCUS
> /\
> \========/
> __\/____\/_____ reflector away from dish
The biquad reflector does NOT capture RF equally over its entire
surface. Grossly over-simplified, the bulk of the signal is captured
in a "ring" around the center coax cable feed at about a 3cm radius.
(I'll see if I can grind out a model with 4NEC2). If the dish focus
happen to hit on this "ring", you'll have maximum gain. If it misses,
the gain will be less. If you move the biquad feed around, you might
be able to see the effect.
Anyway, only with the phase center of the feed at the dish focus, do
all the reflected signals (at any frequency) arrive in phase. If you
move the feed to some other place, you will see some rather strange
effects.
Also, please note that the maximum gain from a 0.6meter dish antenna
is about
:
gain = 9.87 * Dia^2 / wavelength^2 * (feed efficiency)
gain = 9.87 * 600mm^2 / 125mm^2 * 0.4
gain = 91
dBi = 10 log(91) = 19.5dBi
The 40% efficiency (that's the 0.4) is probably optimistic.
>Remember I don't know nothing about antenna's :-)
You're doing fine and asking the right questions.
--
Jeff Liebermann
jeffl@comix.santa-cruz.ca.us
150 Felker St #D
http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
biquad/dish combination 
Linear Mode
