Re: Gain measurement using the Precision beamwidth method

Louis Botha wrote:
=========================================================
I came upon the term Precision beamwidth method for gain
measurements (In a product specification). I have never heard of such a
method and would like to find out exactly what it is. None of my books or
colleages know it.

If anyone has more information please tell me what it is or give a
reference.
=========================================================

You will find a very good explanation of the method, together with proper
mention of its limitations (mostly relating to the efficiency of the antenna)
in:

Terman & Pettit, "Electronic Measurements" 2nd Edition, McGraw-Hill Book
Company, New York, 1952, pp433-442.

Also discussed is an interesting method of "Absolute Measurement of Gain of
Microwave Antennas" by the "reflection method."

As I was mostly unaware of this method until I looked through the book to
answer your question, I will quote an excerpt which explains the principle
for everyone's benefit.

The illustration, Fig 10-18 on page 440 shows an oscillator, an attenuator, a
slotted-line, impedance matching, and then a microwave parabolic antenna of
diameter D. At a distance R/2 from the antenna is placed a large, plane,
reflective sheet of linear extent "h". At a distance R/2 from the back of the
reflector, drawn dotted, is an image antenna.

To quote now from the text, " Another method of measuring the absolute power
gain is illustrated in Fig. 10-18. (1) This can be called the reflection
method. Here the antenna under test is first carefully matched to its
feedline, which includes a slotted section so that in the absence of the
reflector no standing waves are present on the slotted line. A reflector
consisting of a conducting surface, flat to better than lambda/16, is then
placed at right angles to the axis of the antenna as shown. This surface
reflects the radiation that strikes it, producing a reflected wave directed
backward in the general direction of the antenna. Part of the reflected wave
is intercepted by the antenna, and enters the feed line of this antenna to
produce a backward-traveling wave. The energy represented by this
backward-traveling wave is the same as the energy that would be absorbed in
the absence of the reflector by an antenna identical with the transmitting
antenna and representing its image with respect to the reflector, as shown
dotted in Fig. 10-18. The magnitude of this reflcted power that enters the
transmitting antenna can be determined from the standing-wave ratio that is
produced by the presence of the reflector, as measured by the slotted
section. To prevent the oscillator from being affected by the backward wave,
an attenuator A is introduced between the oscillator and antenna. The antenna
side of this attenuator must also present a good impedance match to the
slotted line, so that a backward-traveling wave from the slotted line into A
will be absorbed without reflection.

The quantitative relations that exist in Fig. 10-18 are

           G= (4 pi R) x (S-1) / (lambda) x (S+1)

where G= gain of antenna compared with isotropic radiator
           R= spacing between antenna and its image
           S= voltage standing wave ratio
           lambda= wavelength

The distance R in Fig. 10-18 should be no greater than necessary, since the
greater the distance the more nearly will the standing-wave ratio that is to
be determined approach unity, with the resulting loss in the accuracy of the
determination. At the same time, the distance S between the radiating antenna
and its image should be not less than about 2(D^2)/lambda, where D is the
aperture of the antenna under test. Otherwise the errors from inadequate
spacing will exceed the improved accuracy resulting from having a larger
standing-wave ratio to measure.

It is also necessary that the reflector be large enough to intercept most of
the main beam. Since the total width of this beam is of the order of 2
lambda / D radians, then for a square mirror the height h must be not less
than R lambda / D. For R=2(D^2)/lambda, the required height is hence 2D. "

The description goes on to deal with some ways of reducing errors in the
measurement, but what I quoted above is the real meat of the method.

I hope this is interesting, and perhaps useful, to the members of the
nec-list.

Jacob Z. Schanker, P.E.
Received on Fri Oct 04 1996 - 13:59:00 EDT