Polarization conversion
Dario Lopez
I've seen some very informative posts here and I'm hoping this
question will generate some more.
Does anyone out there know how to convert antenna field measurements
to linearly polarized components as well as circularly polarized
components?
I have a file with complex field data (E theta & E phi : real and
imaginary) and have been searching for a method to do the conversions
barring using poincare sphere concepts (they are non-intuitive as far
as I'm concerned. The linear conversion is pretty straight forward:
E_theta = sqrt(Re_E_theta^2 + Im_E_Theta^2)
E_theta_phase = atan(Im_E_Theta/Re_E_Theta)ącorrection factor for
quadrant placement
Same goes for E_Phi.
Does anyone know how to do it for circluar polarization?
Regards
Dario
Reg Edwards
Thanks. But how do you recognise an informative post ?
Ed Price
"Reg Edwards" <g4fgq...@ZZZbtinternet.com> wrote in message
news:c92mu2$4k$1...@hercules.btinternet.com...
> > I've seen some very informative posts here
>
> Thanks. But how do you recognise an informative post ?
>
>
Any post with an equation that you agree with.
Ed
wb6wsn
Roy Lewallen
Right Circ E = 0.5 * (Eh - j Ev)
where Eh and Ev are complex horizontal (phi) and vertical (theta)
linearly polarized E field components respectively. Left Circ E and
Right Circ E are of course also complex.
You can find this in just about any antenna text. A text would probably
be a good investment if you're going to be digging into theory at this
level.
Roy Lewallen, W7EL
Dario Lopez
But for those equations to be valid circular represenations,
don't Eh & Ev need to be equal? As they stand, substituting "random"
values for Eh & Ev will yield elliptically polarized waves.
Dario
Roy Lewallen <w7...@eznec.com> wrote in message news:<10bbf10...@corp.supernews.com>...
Roy Lewallen
All waves are elliptically polarized. Linear and circular are two
special cases of eliptically polarized waves. For analytical and
computational convenience, you can split any (elliptically polarized,
TEM) waves into two orthogonal components. One choice of two components
is vertical and horizontal linear. Another, equally valid, choice is
left and right circular. You can convert between one and the other with
the equations I provided.
To be purely linearly polarized, Eh and Ev have to be in phase or 180
degrees out of phase. To be purely circular, Eh and Ev have to be equal
in magnitude, AND in phase quadrature. There is no combination of Eh and
Ev except both zero which result in the left or right circular
components to be zero. That is, a purely linearly polarized wave can
still be split into left and right circular components, and neither of
those components will be zero unless the field itself is zero. Likewise,
a purely circularly polarized field can be split into non-zero vertical
and horizontal linear components.
Yuri Blanarovich
>a purely circularly polarized field can be split into non-zero vertical
>and horizontal linear components.
>
>Roy Lewallen, W7EL
Sooo,
we can say that "slanted" 45 deg (circular) polarization as produced with full
wave square shaped quad loop fed in a corner has vertical and horizontal
components that are (typically) 3 dB down from the maximum in 45 deg plane?
According to modeling software, which shows vertical and horizontal components
of slanted polarization, the radiation pattern is a composite of both, with
antenna responding to either V or H polarized waves (with 3 dB down from
slanted) and according to pattern "belonging" to each (V or H) polarization. Is
anything wrong with this statement?
Can we then say that "slanted" polarization antenna has practically "dual" (V
and H) polarization properties with 3 dB down from slanted orientation?
Advantage being fuller radiation pattern (minimized nulls) and polarization
"diversity" at a cost of 3 dB from the "ideal" slanted orientation.
One "Guru" on his web page claims that there is no such thing as dual
polarization.
The "problem" seems to be in semantics. I see nothing wrong calling it "dual"
polarization, because it produces combination patterns "belonging" to either V
or H polarized antennas, (with 3 dB down from ideal slanted) and fuller pattern
than either of V or H alone.
It ain't so, am I wrong?
Yuri Blanarovich
www.K3BU.us
www.computeradio.us - home of "Dream Radio One"
Roy Lewallen
>>Likewise,
>>a purely circularly polarized field can be split into non-zero vertical
>>and horizontal linear components.
>>
>>Roy Lewallen, W7EL
>
>
> Sooo,
> we can say that "slanted" 45 deg (circular) polarization as produced with full
> wave square shaped quad loop fed in a corner has vertical and horizontal
> components that are (typically) 3 dB down from the maximum in 45 deg plane?
You're confusing 45 degree linear polarization with circular
polarization. Even though both have equal horizontal and vertical
components, they're not the same. The difference is that in a linearly
polarized wave, Eh and Ev are in phase. In a circularly polarized wave,
they're in phase quadrature (90 degrees out of phase with each other).
That makes a large difference. Other relative phase angles result in
elliptical polarization with differing axial ratios. In a purely
linearly polarized wave, the amplitude of the field goes from zero to
maximum, back to zero and maximum again, and back to zero each cycle,
and its physical direction of polarization stays constant. In a
circularly polarized wave, the magnitude of the field stays constant,
but its physical direction of polarization rotates through a full circle
each cycle.
> According to modeling software, which shows vertical and horizontal components
> of slanted polarization, the radiation pattern is a composite of both, with
> antenna responding to either V or H polarized waves (with 3 dB down from
> slanted) and according to pattern "belonging" to each (V or H) polarization. Is
> anything wrong with this statement?
No.
> Can we then say that "slanted" polarization antenna has practically "dual" (V
> and H) polarization properties with 3 dB down from slanted orientation?
Yes.
> Advantage being fuller radiation pattern (minimized nulls) and polarization
> "diversity" at a cost of 3 dB from the "ideal" slanted orientation.
The "fuller" radiation pattern doesn't necessarily follow. And it still
suffers the disadvantage that a linearly polarized antenna whose
polarization is slanted the opposite way (at right angles to the wave
polarization slant) will encounter much more than 3 dB attenuation.
That's why circular, rather than slanted linear, polarization is often
used for FM broadcasting.
> One "Guru" on his web page claims that there is no such thing as dual
> polarization.
Please re-read what I said in my original posting. We're describing a
single wave by mathematically dividing the field into two orthogonal
components, which we can call "polarizations". We can choose horizontal
and vertical linear, left and right circular, or an infinite number of
other combinations, including right-slant, left-slant. A wave has only
one E field; our description of polarizations is one of convenience. If
I choose left-slant and right-slant, I can declare with complete
accuracy that your wave has a single polarization component. If I choose
instead vertical and horizontal, I find that it has two equal
components. If I choose some other combination, I find it has two
unequal components. All are equally valid descriptions of the single field.
> The "problem" seems to be in semantics. I see nothing wrong calling it "dual"
> polarization, because it produces combination patterns "belonging" to either V
> or H polarized antennas, (with 3 dB down from ideal slanted) and fuller pattern
> than either of V or H alone.
Suit yourself. Arguing about it would surely be good for at least a
couple of hundred postings, providing an extended diversion for the
entertainment challenged.
> It ain't so, am I wrong?
No, you're right. It's dual polarization. And you're also wrong, since
it's also single polarization (left-slant or right-slant linear).
People with a deep interest in this topic might benefit from the new
EZNEC+ program type, which can display the field strength from any
antenna in terms of left and right circular as well as vertical and
Yuri Blanarovich
>polarization. Even though both have equal horizontal and vertical
>components, they're not the same. The difference is that in a linearly
>polarized wave, Eh and Ev are in phase. In a circularly polarized wave,
>they're in phase quadrature (90 degrees out of phase with each other).
>That makes a large difference.
QSL, I plead guilty to confusion and oversimplification, I really meant
"slanted linear" polarization.
>No, you're right. It's dual polarization. And you're also wrong, since
>it's also single polarization (left-slant or right-slant linear).
Well (as belowed Ronnie would say), I feel more right in the quest for design
of antennas that minimize the polarization fading and cover wider vertical
angle range. I have used slanted polarization quads with good results and I am
willing to pay small 3 dB pealty for the better coverage and fewer nulls in the
pattern (in order of 10 - 20 dB). Plays well in the ocean front locations.
After waves go through the ionosphere, who knows what polarization they are.
>People with a deep interest in this topic might benefit from the new
>EZNEC+ program type, which can display the field strength from any
>antenna in terms of left and right circular as well as vertical and
>horizontal linear components.
>
>Roy Lewallen, W7EL
Thanks for the new + one, I have not explored it yet, but looking forward to
use it and play with 2m (hard) models to verify few things, including insulated
wires in the Quad elements.
Yuri, K3BU.us
Dario Lopez
except for the "No" part and the conversion equations. You left out
one part too. The RCP and LCP components of an elliptical wave are
not equal otherwise they would collapse to a purely linear wave.
Here's an example I came up with.
Let E_elliptical = x + 2*j*y (left hand elliptical)
E_elliptical = RCP + LCP
= (a*x - j*a*y) + (b*x + j*b*y) = x + 2*j*y
equating coefficients...
a*x + b*x = x
-j*a*y + j*b*y = 2*j*y ==> -a*(j*y) + b*(j*y) = 2*(j*y)
or
a + b = 1
-a + b = 2
solving for a & b
a = -0.5
b = 1.5
so
RCP = -0.5*x + j*0.5*y
LCP = 1.5*x + j*1.5*y
which clearly isn't
Right Circ E = 0.5 * (Eh - j Ev)
Left Circ E = 0.5 * (Eh + j Ev)
Regards
Dario
Roy Lewallen <w7...@eznec.com> wrote in message news:<10dcik6...@corp.supernews.com>...
Dario Lopez
dar...@yahoo.com (Dario Lopez) wrote in message news:<c92ca4d3.0406...@posting.google.com>...
Jer
Date: Wed, 23 Jun 2004 13:40:40 +0200
List-post: <mailto:anten...@contesting.com>
hello Arie,
thanks for your contribution... your figures come very close to my empirical
data when analysing dipoles (of exactly the same length in exactly the same
place on the same day in bright sunshine all day) for 14 MHz using different
kinds of materials, insulations, thicknesses, diameters etc....
obviously there is a flaw in Roys formula as well... some people erroneously
think that he employs NEC/4 with EZNEC 4.0 when taking wire insulation into
consideration...
regards
Uli, DJ2YA
Roy Lewallen
not use any part of NEC-4 for any calculations. The geometry check and
the wire insulation features perform the same functions as similar NEC-4
features, but they are implemented with independently written code that
is not part of, copied from, or derived from NEC-4. I have been very
careful about this, because NEC-4 is copyrighted software.
EZNEC/4 of all versions including v. 4.0 does use NEC-4 for calculations
including wire insulation. Because the EZNEC geometry check feature
detects more problems than the corresponding NEC-4 routine (SEGCHK), the
EZNEC feature is used for that function by default. However, the user
can choose to use the NEC-4 function, or both, if desired. EZNEC/4 is
sold only to people or companies who have already purchased an NEC-4
license from the copyright holder.
In the tests I've run (by no means a comprehensive group of tests), the
EZNEC insulated wire routine gives answers very similar to those given
by NEC-4. I'd be very interested in any case where they significantly
differ.
Roy Lewallen, W7EL