Re: NEC-LIST:Parabolic Dish Wave Impedance

Kenneth, (All):

Rusch and Potter (Analysis of Reflector Antennas, Academic Press 1970,
Library of Congress #73-97483) page 100 show for a prime focus feed
that the reflected field, (voltage, I *think*) in the feed transmission line is
approx equal to:

                        (G/4Pi)/(lambda/f)

where G is the numeric gain of the feedhorn (usually 4 to 10, +6 to +10 dBi)
Pi is 3.14159.....
lambda is the wavelength and
f is the focal length of the paraboloid.

So, for paraboloids of small wavelength size where f is not many wavelengths
there is a noticeable reflection that shows up in the transmission line, as well
as a modified power pattern across the dish (modified from the freespace
radiation pattern of the feedhorn by multiple scattering). For most practical
cases the reflection and the modified illumination pattern effects are small.

However, if you were to use a little 2 foot dish with f/D = 0.4 at 1 GHz (lambda
= 11.8 inches) it would be pathologic: G of the feedhorn would be about 7
(8.5 dBi), the focal length less than 1 wl and the reflection would be about 0.7,
VSWR nearly 6, mismatch loss 3 dB. In essence, the feed is looking into a
'halfway decent' short. Power flow is not smooth and will vary wildly with
frequency. Further to this, the reflector is likely in the near zone of the horn,
invalidating several performance estimates based on far field assumptions.
Contrast this with an 8 foot dish with f/D of 0.4 up at 2 GHz: the focal length is
now over 6 wl, the reflection is way down to about 0.086, VSWR about 1.19,
mismatch loss negligible. The power is flowing smoothly and one does not
expect much multiple scattering to upset the illumination pattern.

Little dishes (small in wavelengths) just don't act well and are best avoided, for
a number of reasons, especially in any calibration environment. Furthermore, in
a swept-frequency application, large ripples in all performance parameters result
from changing frequency. A 2 foot/1 GHz dish will develop at best, only 13 dBi,
which is definitely in microwave "horn country" or alternately VHF-UHF Yagi or other
"end-fire country". Dish country starts somewhere above +25 dBi as best I know.

See also S. Silver, Rad. Lab Series Vol 12, page 155.

I believe the above is not overstated as it might seem, and hope it helps. Offsetting
the feed will "scan" the dish output beam and likely helps the reflection, but not by
much if the focal length is small in wavelengths.

dan bathker

  At 02:23 PM 2/17/2002 -0800, Kenneth Webb wrote:
>Hello all.
>
>I am considering using a parabolic dish antenna for radiated susceptibility
>testing. The antenna range in question (at this point) is 1-2GHz. The field
>levels obtained thus far indicate a very efficient radiator. The feed horn is
>slightly offset so that the focal point is just below the feed horn.
>Recently, someone asked if the wave impedance of the standard gain horns used
>in this area (and a large reverberation chamber) are the same as the parabolic
>dish reflector.
>
>Without going to my antenna books and NEC 4 program to simulate the impedance,
>I was wondering if someone might have some insight into this question.
>
>Thanks in advance.
>
>Kenneth John Webb
>Principal EMC Design Engineer
>--
>The NEC-List mailing list <nec-list_at_gweep.ca>
>http://www.gweep.ca/mailman/listinfo.cgi/nec-list

Dan Bathker

-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list