On Fri, 11 Aug 95 16:47:41 PDT aa2du@netcom.com wrote:
I am forwarding this for WS1A who accidently landed this
somewhere else:
>On Tue, 8 Aug 1995 17:55:34 -0400 Rob Hummel WS1A wrote:
>>Dear Group:
>>
>>I have watched with discomfort the discussions relating to antenna
>>performance and the various gain claims of manufacturers. (Including the
>>negative gain phenomenon associated with coax cable.) It's sometimes tough
>>to watch the stuff that goes back in forth about antennas, gain, and
>>performance. I (hesitatingly) have decided to throw my .02 in. (I'm not mad
>>or anything, I just want to see some more messages with formulas in them.)
>>
>>I believe that those without vested interests in antenna sales genuinely
>>mean well and are simply passing on their impressions and experiences. Even
>>so, the amount of plain wrong information about antennas is distressing.
>>Antenna gain is not subjective. It is a finite, measurable quantity. At HF
>>frequencies, real gain measurement is both possible and practical. As such,
>>I have to conclue that the sole reason antenna companies don't perform these
>>measurements (and thus print realistic gain numbers) is to save cost and to
>>make their antennas look better than they are by publishing theoretical
>>numbers based on models.
>>
>>To determine the true power gain (not directivity) of an antenna, two things
>>are required:
>>
>>First, you must have three (3) antennas capable of operating at the test
>>frequency. NONE of these antennas need to be a calibrated standard. ALL
>>THREE of them can be unknowns. By taking measurements between all three
>>combinations of antennas and solving the three resulting gain equations
>>simultaneously, the absolute gain of all three antennas can be established.
>>In antenna engineering circles, this is called simply the Three-Antenna
>>Method.
>>
>>Second, you must be able to support the antennas at the desired height at
>>the desired separation. In measurements for certain antennas (such as those
>>on satellites) you must sweat and toil to place the antennas high enough to
>>elimate ground effects because you are interested in the FREE-SPACE pattern
>>and gain. In ham radio, you AREN'T interested in that because your antenna
>>will never be in free space. Instead, you should measure the antenna at the
>>height at which it will be used.
>>
>>The separation requirement is a function of the far-field distance. Being in
>>the far-field simply means that the /r, /r^3 and /r^4 terms of energy
>>distribution have fallen off enough that the /r^2 term dominates the
>>measurement. This is the standard radiation on the surface of a sphere stuff
>>that everyone uses for the long-distance performance of an antenna.
>>
>>The minimum distance at which you can assume that you're in the far-field of
>>an antenna, by convention, is established as:
>>
>>2 * D^2
>>-------
>> L
>>
>>Where D = the diameter of the antenna's aperture and L is wavelength. To
>>determine D, we can relate the antenna's presumed gain to an effective
>>aperture using the following formula:
>>
>>A = G * L^2
>> -------
>> 4 * PI
>>
>>Assume that A represents a circular aperature, find D, substitute into the
>>far-field formula and you get:
>>
>>2 * L * G
>>--------- = far-field distance
>> PI
>>
>>So, for example, let's say you want to compare the gain of two 20m antennas
>>with presumed gains of 10dBi. The far-field distance would be about 450
>>feet. Build two towers 450' apart and make your measurements.
>>
>>In my opinion, this is entirely do-able. And someone should do it to put all
>>this "my gain is bigger than your gain" nonsense to rest. Why publications
>>like QST and CQ continue to publish idiocy such as "The antenna seemed to
>>perform well, working Europe with no problem" I cannot understand. The
>>reasons antenna companies continue to publish modeled (instead of measured)
>>pattern and gain, I do understand: greed and possibly ignorance. But for
>>those on this reflector, gain, directivity, and antenna measurements are
>>non-mysterious enough that we should be able to settle these issues with
>>information.
>>
>>And, finally, trashing "dBi" as a measurement is a silly thing for a ham to
>>do. A dbi is no more and no less than a convenient abbreviation for power
>>per unit area over the volume of a sphere. To say, "This antenna has 4.5 dB
>>more gain than a dipole," begs the question: In what direction? Typically,
>>you mean at the peak of the pattern ONLY, then you make relative
>>measurements elsewhere. You're not, for example, showing the antenna pattern
>>relative to the dipole at every point in space. The dipole, as used by hams,
>>is simply a mythical 2.1dBi (depending on your source) theoretical
>>DIRECTIVITY (not gain) number. I'll bet that guys who complain you can't
>>build an isotrope don't have any perfect lossless dipoles in their shack
>>either.
>>
>>My CV:
>> Ham for 21 years.
>> Degree from Georgia Tech in Electromagnetics and Antenna Design under Ed
>>Joy (near-field pioneer).
>> Antenna engineer at Sanders Associates. VLF to millimeter wave antenna
>>design, contruction, and test. Far-field measurements, antenna range and
>>anechoic chamber design.
>> Sattelite antenna engineer at TRW in LA. Worked on far-field and nearfield
>>measurements of ground and satellite antennas, including TDRSS and GPSS.
>>Performed three-antenna far-field gain calibration of gain standards at
>>Capistrano test facility. Near-field gain calibrations at Redondo Beach
>>facility.
>> Antenna designer and range installer of far-field/near-field/compact
>>antenna ranges at Millitech Corp. Included microwave through quasi-optical
>>ground, aircraft and satellite system design and measurement.
>> Now semi-retired, but still desinging and measuring my antennas in the
>>real world.
>> <Rob Hummel>
>> <Internet: rhummel@top.monad.net>
>> <MCI: 371-5998>
>>
>>
>>--
>>Submissions: yccc@eng.pko.dec.com
>>Administrative requests: yccc-request@eng.pko.dec.com
>>Problems: reisert@eng.pko.dec.com
>>
>
>=================================================
>J.P. Kleinhaus, AA2DU ARRL CAC hudson Div. Rep.
>E-mail: aa2du@netcom.com
>Compu$erve: 74660,2606
>
>It's not a bug...It's a feature!
>=================================================
>
>
J.P. Kleinhaus, AA2DU ARRL CAC hudson Div. Rep.
E-mail: aa2du@netcom.com
Compu$erve: 74660,2606
It's not a bug...It's a feature!
|