Re: NEC-LIST: cross field antennas

Bonjour,

I've first heard of this thing at the IEE International Broadcast
Conference IBC'97 in Amsterdam. I had only attended the [large] trade
show, the number of papers and the technical level of the conference
being very far from being worth the rather dear registration. Buying
the (thin) proceedings for 75 UKP would be about as far as I'd go, on
the grounds that since I was there, I'd just as well know what's going
on, but just in case.

When I returned home, I tried to find out more about this, as it
really sounded way too good to be true, and piqued my interest. I went
to the library, and looked up the references (see the reference list
at the end, which complements the one by Mr. Belrose.)

I couldn't reconcile the claims which were made with the design which
was presented. One just can't get both high bandwidth and high gain
simultaneously from a small antenna. If the input impedance of the
so-called E and H terminals vary very much depending on their
excitation, whis would indicate that might also be a large circulating
current in the matching circuit and the between two ports.

If the antenna works, then the implication would be that there is
something which is overlooked in Schelkunoff's equivalence theorems,
and in all these theorems on superdirective antennas, and those
fundamental theorems relating dimensions, Q bandwidth, and gain, which
I had studied and thought to understand. I could look up in my
library, and throw in a pile of references in several languages, but
it "ain't" worth it...

The papers published in Wireless World did not appear to the entire
backing of the magazine's editors - each page had the heading
"Hypothesis", and the title "CFA: Working assumption" carries some of
the interrogation this raises. The authors try to address some of the
issues, but basically say that "it works - believe me".

To wit, a quotation from [2]:

<<
Crossed field antennas look to have major advantages over
conventional antennas in terms of size, efficiency, and lower working
voltages. The CFA is not a resonant antenna as its structure is
substantially smaller than the radiated wavelength, and it is low Q
and broad banded. Fundamental differences to conventional wire
antennas mean it is not surprising that the CFA has attracted
hostility from some experts. However, the fact that it works,
indicates its basis is credible. Furthermore the theory may be said to
have passed the most severe test of a new hypothesis in that it can be
used to design new devices. Poynting vector synthesis has universal
application as a design method for compact,
efficient. electrically-small antennas. with demonstrated successful
application at MF and HF, and experiments af VHF and LF. The technique
could provide the solution to the difficult problem of efficient ELF
radiation.
>>

No radiation patterns or other conventional antenna measurements are
presented in any of the papers I found.

In [5], there is the curious statement:

<<
It is thought that the high efficiencies measured are due to the
absence of the "missing power" integrated from the field strength
graph by Kraus in Ref.4. We calculate that the quarter wave monopole
never exceeded 38% efficiency.
>>

What do they mean here about "efficiency" ?

I had borrowed the first edition of Kraus from a friend to check this,
but couldn't see which figure there's referring to. I can't find
anything either right now in the second edition which I own. Could
this be an attempt to compare the "CFA" to an an electrically short
dipole of similar dimensions, with losses?

Has any learned reader of this mailing list ever heard about such a
38% theoretical efficiency for a dipole???? (This is a serious
question, not just a rethorical one.)

But the field strength and installation records can't just be
dismissed off-hand. Unless this whole thing is a hoax, which is
difficult to consider if there *really* is a dozen stations on the air
connected to this "thing". This is really the part which really
bothers me. How can they really measure such field strengths? I
suppose they use some sort of calibrated commercial instrument, and
that this part is less fishy than the rest.

But yet... There are measurements performed supposedly with 30kW of
drive, with a statement of the improvement over a lambda/4
monopole. (Where are these monopoles relative to the CFA? The values
are in "dB". What kind of dB is that? dWuV? dBW? dBWm^(-2) ?
dB-pifométriques? How many measurements did they make?

The description of the "barrel shape CFA" states that the synthesis of
"H component of the field" is achieved by the displacement current
flowing between the "D plates". But they seem to ignore that an
eventual displacement current flowing between the plates is exactly
balanced by that carried by the feed wire (right in the center)
connecting the upper plate to the source, placed somewhere below the
ground plane, thus rendering extremely unlikely that a significant H
field could be induced "at some distance" from the structure,
interacting with the "synthesized E field"...

Finally, ref. [2] contains the following assertion :

<<
Input impedances :

When the CFA is operating, input impedances seen at the two separate
input ports are moderately valued, and contain resistance, which is
itself a cogent evidence of real radiation.
>>

I'd could make the same statement about a dummy load, but I don't
think this is a "cogent evidence" of radiation... (There used to be a
scam in the late 1970's in Ham Radio magazines, which advertised "a
high speed fully automatic antenna tuner", sold by a company with a PO
box address in Florida, which could really tune *anything*. Seducive
stuff. The QST staff opened one, and found an attenuator inside... An
X ray showed that there was also a piece of scrap PC board potted with
the rest...)

My feeling is that *if* they really measured some encouraging result,
(field strength and input impedance), the ground might play a
significant role, especially since they are basically doing away with
the usual set of radials. What is the ground conductivity in the lush
and wet Nile valley? How about just outside the valley in the desert?
Could the ground be resistive enough to lower enough the Q of the
antenne to make it look good? (All this good electricity used to heat
the Egyptian earth?) And the feedlines might also play a part.

IMVHO, this whole thing isn't very encouraging. (I initially put in a
stronger judgment, but I fear the moderator wouldn't let it
through). The WW paper [1] was followed by an ad from "Hately
Engineering" for a meter high SW antenna. I'd say that a scale model
made with a pineapple tin can, some copper foil, and a two port VANA
could place the issue to rest in a day... And rereading VE2CV's
contribution, I see this was done already.

Alexandre

References :

[1] F.C. Kabbary, M.C. Hately, B.C. Stewart, "Maxwell's Equations and
    the Crossed-Field Antenna", Electronic & Wireless World, March
    1989, pp. 216-218

[2] F.C. Kabbary, M.C. Hately, B.C. Stewart, "CFA: working
    assumption?", Electronic & Wireless World, December 1990,
    pp. 1094-1099

A critical rebuttal:

[3] Letter by K. Donaldson (Sevenoaks), Electronic & Wireless World,
    October 1992, pp. 837-838

[4] Reply by the authors, Electronic & Wireless World, December 1992,
    p. 1007

The title of the paper from the IBC was:

[5] 'Extremely small high power MW broadcasting antennas', F M Kabbary
    and M Khattab (Egyptian Radio & TV Union, Egypt), M C Hately
    (Hately Antenna Technology, UK), IBC'97 record, page 421-426

ps: I also have the preceding papers in electronic form.
Received on Wed Feb 10 1999 - 01:57:45 EST