NEC-LIST: standards, was: calculating antenna bandwidth

At 04:40 PM 4/2/2002 -0500, Steven Best wrote:
>Hello All:
>
>Sorry that this post is off topic.
>
>This discussion regarding antenna bandwidth is very interesting as there
>are quite a number of different "antenna bandwidth" definitions used
>within the antenna community - both academic and industry. I think this
>is an important topic because there seems to be some discrepancy in the
>IEEE literature on this issue. There also seems to be some discrepancy
>in defining antenna bandwidth and what is acceptable in terms of real
>world antenna performance.

I don't think it's really off topic (hey, at least we're not talking about
hyperlight propagation enhancers, or some such). One of the real powers of
numerical methods is that it allows you to evaluate a design over a wide
range of design parameters or requirements spaces without having to
actually "cut metal", or derive a suitable analytical expression.

I've been engaged in some phased array design, for which mutual impedances
are very important. Sure, there are some nice equations out there, based
on analysis, that give Zm for various combinations of dipoles, in echelon,
collinear, broadside, etc. They're real useful to get an idea of the
general magnitude of the values. But, any practical antenna starts to
deviate pretty quickly from the idealized "half wave dipole" case (like,
shortening the elements to reduce the reactive component of the self
impedance), and then the analysis doesn't necessarily hold up (or, more
properly, I'm too lazy to rederive everything for the "off nominal"
situation, when I've broken some key assumption made by Schelkunoff, or
King, or whoever).

It's much faster just to crank NEC through its paces and "measure" the Zm
of the various elements. Sure, it's not a perfect model, but, I can also
(quickly) derive a feel for how sensitive the model is by varying the
parameters a small amount and seeing if I get huge variations in
results. This is even more useful, because it directly relates to an
important engineering issue: manufacturability and usability. The best
design in the world (performance wise) is useless if it requires
excessively tight tolerances and constraints on enviroment.

Steve brings up an excellent point that the antenna world (particularly the
commercial antenna world) tends to have a bit of elasticity and vagueness
when it comes to conceptually simple things like "bandwidth" or "front to
back ratio". Call it "specsmanship" if you will, but the fact of the
matter is, a concept like bandwidth is much fuzzier when applied to an
antenna than to, say, a bandpass filter (and even the filter might have a
bunch of qualifiers: passband ripple, shape factor, ultimate rejection, etc.)

Unfortunately, antennas are almost always part of a system (passive radar
reflectors might be the exception), and the rest of the system has a huge
effect on the operating performance of the antenna. Maybe my voltage
source transmitter is connected right to the feedpoint of the antenna (very
common these days...not much coax between my cellphone and its antenna), so
reflected power, per se, isn't a concern, but radiated power is. Maybe I've
got a LF receiver using a high impedance voltage probe antenna (or a zero
impedance current loop), for which mismatch (in the VSWR sense) is irrelevant.

I think that it would be worthwhile to standardize on some definitions...
IEEE sort of jumps around the issue by using that "range over which XYZ
meets some predefined performance standards", without going any further.

What would a reasonable set of basic "metrics" be? How would one define them?
Overall radiation efficiency bandwidth (power in to power radiated ratio?)
boresight efficiency bandwidth (for directional antennas, power in to power
on boresight.. many antennas have patterns that change substantially as the
frequency changes)

At least there is a standard reference for gain, the isotropic radiator,
which has a nice theoretical, easy to understand, definition. Of course,
issues arise with gain vs directivity, and with whether you count loss in
the antenna, etc. but at least there IS an ideal to compare against. No
such thing exists for bandwidth, since the performance can be measured in
so many ways (5:1 or 2:1 or 1.1:1 VSWR, just to bring up a few.)

Jim Lux
Spacecraft Telecommunications Equipment Section
Jet Propulsion Laboratory
4800 Oak Grove Road, Mail Stop 161-213
Pasadena CA 91109

818/354-2075, fax 818/393-6875

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