[Top] [All Lists]

[TowerTalk] antenna gain (long)

To: <>
Subject: [TowerTalk] antenna gain (long)
From: (L. B. Cebik)
Date: Sun, 6 Jul 1997 07:31:08 -0400 (EDT)
> W4RNL can give us some input on this aspect of the
> problem. What say L.B.?
Input, yes; insight, maybe not.  But here goes.

The gain estimations for monoband beams are fairly straightforward, in
accord with Lawson et al.  Multiband beams are another beast.  In one
sense, all elements are active all the time in one way or another, but to
varying degrees.  In some designs, small changes on one band may or may
not affect the other bands--and that is a measure of to what degree the
seemingly inactive elements actually play a role.  There are too many
designs out there to make a universal rule.

For those interested, I have posted a small note (actually about 10 KB,
which is why I am not passing along an ASCII copy here--I can if the list
manager thinks it ok/useful) on why I use dBi.  It is not important why I
use dBi (quick answer: because I am a modeler and do not have a range
standard).  Perhaps the most useful parts of the note are some
distinctions I make among dBi (well known standard), dBdI (ideal dipole in
free space--not especially helpful over dBi), dBdR (real dipole at the
same height and of relevantly similar materials to an antenna under test,
with due regard to test elevation angles).  Modeling can make very
effective use of dBi so that all antenna models under any conditions
allowed by the modeling program can be expressed in the same reference.
Range testing, however, is faced by another problem:  it always needs an
on-site reference antenna.  Not all ranges have the same properties, so it
is not usually possible to transform one set of figures to another between
two manufacturers.  And who would not have a bit of hesitation at
accepting the numbers for ABC antennas as tested by a competitor?

Range tests can by backed into modeling software, IF the model is adequate
to the task.  However, the development of adequate models which are then
correlated with range tested antennas and placed on a common footing is a
professional task requiring the investment of time and $ (since the
professional doing the job has to eat, purchase software, etc.).  For
various working purposes, manufacturers often use a collection of models
for the same antenna, some of which are simplified for speed of performing
What If tests, etc.  Modeling to the degree necessary to make reasonably
"authoritative" models adequate for comparative purposes, independent of a
particular set of design goals a manufacturer might have, is not a matter
of simply throwing together a set of dimensions found on an assembly
sketch.  I have a few models on hand of commercial designs and do not
consider my efforts to make them adequate models to be sufficient to make
any results generally public, since they might be taken as
authoritative--which they ain't.  Casual modeling, especally of multiband
or loaded or trap antennas simply will not get the job done.

Alternatively, one can do an independent range test using a protocol that
covers all important aspects of the testing.  The key to the protocol is
a. ensuring a cross check on any significant reading and b. being fair to
all antennas tested.  (Remember, it was the absence of a cross check on
certain readings that allowed operators to misread the situation at 3-mile
Island many years ago.)  The test would employ a standard antenna as a
reference, likely a dipole.  For multiband beam testing, this would
ordinarily be an aluminum tubing dipole not radically different in
diameter than the average element sizes of the antennas being tested.

As has been noted in another posting, I am aware of a project underway to
perform just such tests over a limited range of antennas.  The actual
tests are somewhere down the line, since the project team is taking great
pains to develop an adequate protocol to ensure accuracy and fairness of

If ARRL should undertake a more general project, it would be a fairly
large investment of resources in terms of manpower, equipment, time, and
expertise.  Although finances (according to the most recent annual report)
are tight, perhaps they are the only ones equipped to do the job at
present for all antennas on the market.  If they have to purchase the
antennas--well, go through the complete lines of all manufacturers and add
up the cost of the antennas, even allowing a generous discount.  And all
of this assumes that ARRL has an adequate range site in the hills of
central CT.

I have suggested in the past a truly independent "underwriters lab"
operation for antennas to establish standards of both modeling and range
testing and then to perform tests on all antennas.  You can understand
from the $ involved why such an operation is unlikely in the near future.
Perhaps only an Institute of Antenna Designers and Manufacturers could
make such an operation a reality, and to the best of my knowledge, no such
organization exists.  Remember that the standards set by this hypothetical
organization would not be the only usable standards, but rather a set of
agreed upon standards used by everyone.

Until then, like you, I shall be unable to correlate the performance
figures given by various manufacturers.  The best I can do for myself is
to acquaint myself with what it is reasonable to expect from various types
of antennas in both full  size models and various modifications.  Then I
examine both the claims and techniques of antennas of interest to assess
their reasonableness (as well as basic comprehensibleness).  If you claim
the equivalent of 3-element performance from 2 elements, I am unlikely to
believe you--unless you show me exactly how you got it in a way that I can
replicate in models or in actuality.  If you claim you are getting an
extra dB gain because of induced hyperflux, then you owe me a replicable
account of how to get induced hyperflux--even if you think induced
hyperflux is protected by patent.  Not everything patented is either new
or productive.  For example, I recently heard that someone has managed to
patent a capacitive-hat dipole, an antenna that has been around since
pre-Terman days.

In the absence of a truly independent set of range tests of all antennas
of interest, we are left to making determinations of the reasonableness of
antenna claims based on our own knowledge of antennas.  Until then, all we
can do is to continue learning all we can about antennas.

I hope this ramble is of some use to someone.



L. B. Cebik, W4RNL         /\  /\     *   /  /    /    (Off)(423) 974-7215
1434 High Mesa Drive      /  \/  \/\     ----/\---     (Hm) (423) 938-6335
Knoxville, Tennessee     /\   \   \ \   /  / || /      (FAX)(423) 974-3509
37938-4443     USA      /  \   \   \ \       ||    

FAQ on WWW:     
Administrative requests:

<Prev in Thread] Current Thread [Next in Thread>