This seems like a good place to insert a discussion about a
"shielded loop" being a myth for anything except for deeper
nulls in the directivity pattern: (these were posted about
18 months ago).
------------------------------------------------------------
------------------------------------------------------------
From: Roy Lewallen
Date: Fri, 21 Mar 2003 11:30:17 -0800
Newsgroups: sci.electronics.design,alt.engineering.electrical, rec.radio.amateur.antenna
Subject: Loop Antenna questions
It's fairly easy to see that what Tom wrote (the first paragraph of
quoted material below) is true. "Shielding" a loop simply moves the
feedpoint to the gap, and the outside of the "shield" becomes the loop.
It's no more or less immune to electric fields than any other loop. What
"shielding" does (and all it does) is to increase rejection of common
mode excitation, improving loop balance and the resulting pattern null
in practical installations. The sensitivity of a "shielded" loop to both
electric and magnetic fields is exactly the same as for an "unshielded"
loop.
The myth that "shielding" a loop somehow rejects the electric field
without affecting the magnetic field is one that's been kicking around
the amateur community for decades. I don't suppose it'll ever die.
Another thing that seems to be unknown or overlooked is that a small
(so-called "magnetic") loop antenna is actually more sensitive to an
electric field and less sensitive to a magnetic field than a short
("electric") dipole, at all distances from the antenna unless the
radiation source is very close -- within less than a wavelength.
Roy Lewallen, W7EL
JLB wrote:
The external "metal toroid" is then the antenna, for it is on the
outside of that "toroid" that antenna currents must exist, and not
inside it. The gap is then the feedpoint, to which the internal wire
couples and then acts as a simple feedline. To best reject response
to electric fields, the gap must be properly positioned...
No, the gap is usually opposite the feed point and the sheild is grounded.
It shields the loop from electric fields but it still responds to magnetic
fields. This is common practice and the idea has been around for many
decades. From personal experience I can say that it makes a tremendous
difference in the sensitivity pattern of the loop, even when extremely close
to other conductors (including AC house wiring).
References available on request.
Ok, what are they?
Jim
N8EE
From: Roy Lewallen
Date: Fri, 21 Mar 2003 22:52:21 -0800
Newsgroups: sci.electronics.design,alt.engineering.electrical, rec.radio.amateur.antenna
Subject: Loop Antenna questions
After some reflection, I see that I can't really justify the statement I
made. Thanks for questioning it.
Let me explain what I do know. The following discussion assumes that all
antennas are lossless, to eliminate a factor that would obscure the
subject under discussion.
If you put 100 watts into a very short dipole, the ratio of the E field
it produces to the H field it produces is called the wave impedance. At
a distance of considerably less than a wavelength, or greater, this
ratio is about 377 ohms (in air or free space), and E and H are in
phase. The same thing is true for a very small loop antenna, or any
other small antenna. If you get far enough away, the E/H ratio is the
same for any kind of antenna, although you might have to get a farther
from an antenna that's large in terms of wavelength. Invoking the
reciprocity principle, you find that the relative response of a small
loop to E and H fields is the same as for a small dipole if the field
originates some distance from the antenna. The so-called "magnetic" loop
responds just the same as any other antenna to fields originating beyond
a wavelength or so from the antenna.
What's interesting is the relative strength of radiated E and H fields
close to the antenna. My overstatement was the implication that the
ratio of E to H transmitted fields represent relative "sensitivity" to E
and H fields when receiving. Contrary to what the CFA and EH proponents
claim, it's not possible to generate E and H fields independently. And
contrary to what people who misunderstand "shielded" loops say, it's not
possible to isolate E and H fields. (It is possible to change the
impedance of a wave -- that is, the E/H ratio -- in a region, but it
always reverts back to the impedance of the medium -- 377 ohms for free
space -- within a pretty small distance of the disturbance.) So
sensitivity to an E or H field is somewhat problematic, since you can't
create either one in isolation for a test. Consequently, I'll just
discuss the E and H fields that the antennas produce when transmitting.
I believe that an antenna with high transmit E/H ratio will be
relatively insensitive to a wave with low E/H ratio (which would have to
be created nearby), and vice-versa. But I can't quantify the
relationship. I only know it involves the concept of transfer impedance,
and I haven't given the topic enough study to say more about it. It is
discussed in a number of texts, for those who are interested.
What I've done is look at the E/H ratio radiated by a small loop and a
small dipole at various distances, in the direction of the maximum far
field radiation (in the plane of the loop and broadside to the dipole).
The following analysis was done with EZNEC/4 using its double precision
NEC-4 calculating engine. This was necessary because of problems NEC
codes have with very small loops, and the size of loop I chose was too
small for analysis with NEC-2. However, it could probably be done with
MININEC, or it could be done with an NEC-2 based program like the EZNEC
demo if the loop is made a bit larger. It's vital to set the loss to
zero so that a negative feed point impedance isn't concealed by positive
loss resistance. I've gotten email from a person who was skeptical of
the results and did the calculations analytically. He came to the same
conclusion, which gives me added confidence in the results. The
following calculations were done at 3 MHz (100 meter wavelength).
Here's what happens. A small loop is sometimes called a "low impedance"
antenna because of the low E/H ratio close to the antenna. This might be
adequate justification for calling it a "magnetic" antenna as amateurs
have taken to doing, but the term unfortunately conveys the mistaken
impression that the antenna somehow responds only to or more strongly to
magnetic fields coming far from the antenna, which isn't true. But very
close to the antenna, the E/H ratio is indeed lower than the 377 ohm
value this ratio always has far from the antenna. In exactly
complementary fashion, a small dipole is a "high impedance" antenna,
having an E/H ratio higher than 377 ohms very close to the antenna. Here
are the values of the magnitude of E/H for a 2 meter (1/50 wavelength)
circumference octagonal loop and a 2 meter long dipole at 3 MHz. The
distance is measured from the center of the loop or dipole, and the
direction is in the plane of the loop or broadside to the dipole.
Distance (m) E/H (ohms)
Dipole Loop
2 2630 63.9
4 1360 110
6 862 169
8 597 241
10 440 326
12 345 414
14 291 490
16 265 538
18 257 555
20 258 552
22 264 538
24 273 522
26 281 506
28 290 491
30 297 478
50 342 415
100 367 386
The curious thing is that, while the dipole field impedance is high
close to the dipole, and low close to the loop, in both cases it
overshoots 377 ohms by a considerable amount, reaching a peak (lowest Z
for the dipole and highest for the loop) at about 18 meters, then
asymptotically approaches 377 ohms from the "wrong side" beyond that. At
all distances of 12 meters (about 1/8 wavelength) and beyond, the E/H
ratio of the loop is actually higher than that of the dipole! That was
the basis for my original statement. Whatever might be said about the
"sensitivity" of a "magnetic" loop to E and H fields, it certainly can't
be said to be more sensitive to H fields, or less sensitive to E fields,
than a small dipole, at any distance greater than about 1/8 wavelength.
The exact sizes of the loop and dipole aren't important. The near fields
at these distances are essentially equal for all small dipoles and for
all small loops for a given input power, provided that they're much
smaller than a wavelength in any dimension.
Roy Lewallen, W7EL
Reg Edwards wrote:
Roy says -
- - - - a small
(so-called "magnetic") loop antenna is actually more sensitive to an
electric field and less sensitive to a magnetic field than a short
("electric") dipole - - - -
-----------------------------------------------------
Roy,
What are your units of "sensitivity"?
How are things to be measured (even hypothetically) and compared?
In any case, the statement is meaningless unless loop diameter and dipole
length are incorporated.
Explanation please.
A simple learned reference will not be of much use since 99.9 percent of
readers will have no chance of ever getting their hands on it.
Yours, Reg, G4FGQ
From: Roy Lewallen
Date: Sat, 22 Mar 2003 13:15:46 -0800
Newsgroups: sci.electronics.design,alt.engineering.electrical, rec.radio.amateur.antenna
Subject: Loop Antenna questions
Richard Clark wrote:
If anything, your work contributes to the notion of these "notable"
differences rather than eliminates them. You forcefully detail how
different the short dipole is from a small loop to exactly the same
degree as offered by CFA/EH proponents as that being its boon - even
if they invert the rationale of what constitutes the sensitivity.
This is to say, that if they are entirely wrong about a magnetic
antenna being sensitive to the magnetic field (something you reject
and offer compelling numbers as evidence); they do have a point about
the magnitude of difference (they are not entirely wedded to the
principle as they are the effect - if in fact that effect is not
another figment of wish fulfillment).
My understanding of the CFA/EH antenna is undeniably vague, but I think
it goes something like this: The E and H fields are created
independently, one by a capacitor-like part of the antenna and the other
by a coil-like part of the antenna. This, as I said and as many others
have pointed out, can't be done. Creation of a time-varying E field
always results in a time-varying H field, and vice-versa. The E and H
fields thus "independently" created are done so in the ratio of 377
ohms. This being the ratio of the the components in the far field, the
direct synthesis results in the lack of a near field. This lack of near
field is credited somehow with the magical properties of these antennas
(high efficiency despite small size) by mechanisms that have never been
clear to me. At any rate, these wonderful properties have never been
demonstrated with either believable measurements or modeling despite
more than a decade of hype, so it sure looks like they're fiction also.
The EH antenna further distinguishes itself by requiring a magic
inductor whose current at one terminal is shifted in phase 90 degrees
from the current at the other terminal, adding yet another layer of
impossibility.
If I've misinterpreted the explanations for how these things are
supposed to work, I apologize -- it wasn't intentional.
The argument about being less sensitive to neighborhood noise for
magnetic antennas, small loops, or loops in general, seems to be
justified by your data supporting very different E/H ratios out to
about 10M, which for me is easily the most disruptive region to any
antenna in a noisy neighborhood and makes that case in some sense.
It might. Remember, though, that this is constant in terms of
wavelength, so the range drops by a factor of two at twice the
frequency. (But on the other hand, the range will be greater at 160
meters.) Also, note that the difference isn't very great even at a
distance of 10 meters. You have to get very close indeed to see a truly
major difference. I suspect that major differences in interference
rejection are as likely to be due to differences in balance (causing
better or worse nulls) or polarization.
It seems to me that the difference applies only to situations where
that first 10M is significant, and to what degree coupling of
interference within that region becomes meaningful.
Yes, that's the conclusion I reach.
Roy Lewallen, W7EL
|