Back To Basics - Antenna Reciprocity
J. Harvey
> Antennas are completely reciprocal. They have
> the same characteristics receiving as transmitting.
> They do both at the same time because they
> don`t know which mode they are supposed to fill.
"David Ryeburn"
> If you run time backwards, what's the difference
> between transmission and reception?
Is there a 'reactive field' surrounding a receive antenna ? We've just had
a discussion on the relative phase of the E and H fields near a transmitting
antenna. It was generally agreed that there existed a reactive field close
to the transmitting antenna where the E and H fields were not "related" by
377 + zero*j ohms.
What about the 'near field' ? If an antenna were transmitting, the
'wavefront' would be very much non-planer. If the antenna were receiving,
then the 'wavefront' would be much closer to planer (assuming that the
signal source was far away). Even given identical signal strengths (very
weak transmitting), the difference would be obvious to a skilled observer.
What about the fact that a transmitting antenna might radiate 'from the
middle' with passive elements equidistant experiencing the same phase,
whereas a receive antenna will experience the waves passing from front to
back and those passive elements might be out of phase ?
If you were to run a film of a transmitting antenna backwards (time
reversal), people would rightfully ask why all the RF in the neighbourhood
was converging on the antenna like that. I can't imagine flashlights
sucking light, let alone recharging their batteries, even if time runs
backwards. [The only reason that the Universe doesn't run backwards is
because of the warning written on the side of non-rechargeable batteries !]
Tea cups break, but they don't spontaneously reassemble.
Three laws of thermodynamics (as slightly extended by Yours Truly):
1) You can't win.
2) You can't break even.
3) You can't quit.
4) You can't run time backwards.
Does anyone have a simple proof (or even an indication) that antennas must
be reciprocal ?
Has anyone ever made an antenna that wasn't ?
Are all of the quantifiable characteristics the exactly same ? Even given
that the fields in the vicinity are different (reactive versus 377+0*j ohms,
and planer versus non-planer) ?
Richard Clark
Hi Jeff,
Once again, you have to separate the effects from the characteristics
that cause the effects. The near field's characteristics of impedance
and geometry with respect to its surroundings are analogs of a lens in
that the presence of this discontinuity in space distorts it to same
purpose: reciprocity.
Example of reciprocity is most striking in the yagi. If reciprocity
failed, it would be most apparent for this particular design.
If you can stretch your imagination (speaking to others than Jeff now,
as his is most malleable), search Scientific American's archives about
passive sonar (sometime in the past two years). They in fact rebuild
the source of a radiation (at sound levels) observed through the
detection of a random field of detectors replaying their detected same
sound, in time synchrony - backwards. That is, each detector has its
own re-radiator that when properly sequenced and playing the sound in
reverse, all such detectors/re-radiators add at the origin of the
initial source reconstructed.
73's
Richard Clark, KB7QHC
J. Harvey
> Example of reciprocity is most striking in the yagi.
> If reciprocity failed, it would be most apparent for
> this particular design.
I had thought about a Yagi antenna while drafting the posting. It certainly
provides comfort in the face of doubts...
> ...Scientific American's archives about passive sonar
> (sometime in the past two years). They in fact rebuild
> the source of a radiation (at sound levels) observed
> through the detection of a random field of detectors
> replaying their detected same sound, in time synchrony
> - backwards. That is, each detector has its own
> re-radiator that when properly sequenced and playing
> the sound in reverse, all such detectors/re-radiators
> add at the origin of the initial source reconstructed.
If the original noise source at the origin was (for example) an exploding
stick of dynamite near the ocean surface and it was weakly detected by an
array of hydrophones scattered around at some significant distance, having
those same hydrophones replay the same small signal that they so weakly
detected will obviously not recreate the effect of a stick of dynamite back
at the origin.
Even having those hydrophones pay the ultimate price and literally explode
with the largest possible amplitude signal, those signals arriving back at
the origin will add up in phase only in proportion to the number of
hyrophones. Use a hundred exploding hydrophones and you'll be only 20dB
above the sounds of a single distant exploding hydrophone. This would still
be far less than a locally exploding stick of dynamite.
Things can add up in phase - that is true. This doesn't mean that time can
run backwards.
Scientific American is just the sort of breathless rag to leap from one to
the other. I dropped my subscription because they were getting so 'over the
top'. One example is how Scientific American (and others) went crazy over
the Eve Hypothesis.
www.google.com/search?q=eve+hypothesis
3rd from the top.
David or Jo Anne Ryeburn
<JM002.Harve...@Sympatico.ca> wrote:
> "David Ryeburn"
> > If you run time backwards, what's the difference
> > between transmission and reception?
>
> Is there a 'reactive field' surrounding a receive antenna ?
Yes, I would think there would be. At most (only when the receiving
antenna is conjugately matched [heaven help us! -- don't want to start
*that* thread up again]) 50% of the incoming energy will be deployed
elsewhere (into whatever we have for a receiver) and at least 50% of the
incoming energy will be re-radiated. The latter process will involve
"reactive fields" as well as a radiation field.
> What about the 'near field' ? If an antenna were transmitting, the
> 'wavefront' would be very much non-planer. If the antenna were receiving,
> then the 'wavefront' would be much closer to planer (assuming that the
> signal source was far away). Even given identical signal strengths (very
> weak transmitting), the difference would be obvious to a skilled observer.
I thought about the non-planarity after I posted. If there were *lots* of
such receiving antennas, each doing their appropriate jobs, then I would
think the nearly planar field would be reconstituted by their re-radiated
fields. If there were *lots* of such transmitting antennas, appropriately
located and fed with appropriate power and phase, a nearly planar field
could be synthesized (at least locally). I don't know if this is a
satisfactory answer or not.
> What about the fact that a transmitting antenna might radiate 'from the
> middle' with passive elements equidistant experiencing the same phase,
> whereas a receive antenna will experience the waves passing from front to
> back and those passive elements might be out of phase ?
What if there were more than one distant transmitting antenna,
appropriately synchronized, so that the passive elements *did* receive
in-phase signals? It isn't the receiving antenna's fault that the time
reversal seems only be working in one spatial direction. Again I don't
know if this is a satisfactory answer.
> If you were to run a film of a transmitting antenna backwards (time
> reversal), people would rightfully ask why all the RF in the neighbourhood
> was converging on the antenna like that. I can't imagine flashlights
> sucking light, let alone recharging their batteries, even if time runs
> backwards. [The only reason that the Universe doesn't run backwards is
> because of the warning written on the side of non-rechargeable batteries !]
> Tea cups break, but they don't spontaneously reassemble.
>
> Three laws of thermodynamics (as slightly extended by Yours Truly):
> 1) You can't win.
> 2) You can't break even.
> 3) You can't quit.
> 4) You can't run time backwards.
You're getting closer to convincing me. Maybe I should stick to
mathematics and not dabble in physics.
> Does anyone have a simple proof (or even an indication) that antennas must
> be reciprocal ?
I'd like to see it too.
> Has anyone ever made an antenna that wasn't ?
Again I'd want to hear about such an antenna too.
> Are all of the quantifiable characteristics the exactly same ? Even given
> that the fields in the vicinity are different (reactive versus 377+0*j ohms,
> and planer versus non-planer) ?
These two things I tried (perhaps unsuccessfully) to deal with. But
nevertheless I'm now not convinced what I first said made sense.
David
--
David or Jo Anne Ryeburn
rye...@sfu.caz
To send e-mail, remove the letter "z" from this address.
Richard Harrison
"Does anyone have a simple proof (or even an indication) that antennas
must be reciprocal?"
Yes. 99 and 44/100ths% of the time or more.
Countless measurements and observations confirm reciprocity. Antenna
characteristics have been determined by transmitting from many locations
into the antenna, and by receiving at many locations from the same
antenna. Patterns and gains are identical whichever method is used.
Far-field and near-field complications are unnecessary to consider in
reciprocity. The antenna conductor has a current in it and a voltage
across, no matter how it got there. It may have come out of a feedpoint
or off the air. It causes radiation the same way, which means production
of induction fields and all the rest, weak or strong, depending on the
signal.
Best regards, Richard Harrison, KB5WZI
CAM
> Does anyone have a simple proof (or even an indication) that antennas must
> be reciprocal ?
What if the driving impedance of the transmitter is a lot different
than the impedance across the transmission line during receive, e.g.
the transmission line acting like a shorted stub during transmit
and like an open stub during receive?
--
cheers, CAM http://www.qsl.net/w5dxp
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
'Doc
CAM,
---------------
>
> What if the driving impedance of the transmitter is a lot different
> than the impedance across the transmission line during receive, e.g.
> the transmission line acting like a shorted stub during transmit
> and like an open stub during receive?
> --
> cheers, CAM http://www.qsl.net/w5dxp
----------------
But how does it know when to switch?
'Doc
J. Harvey
> What if the driving impedance of the transmitter is
> a lot different than the impedance across the
> transmission line during receive, e.g. the
> transmission line acting like a shorted stub
> during transmit and like an open stub during receive?
Even if the Zo of the transmitter were zero, the E/I ratio would start-out
as the cable Zo for a while (time 'related' to the length of the cable), and
then revert back to the normal antenna SWR mess. The Zo of the transmitter
never enters into it. (?)
You're certainly correct that the receiver might be different.
Not the 'antenna' reciprocity though.
'Doc
> But how does it know when to switch?
PTT switch (or key).
'Doc
" J. Harvey" wrote:
-------------
> 'Doc
> > But how does it know when to switch?
>
> PTT switch (or key).
-------------
Gee, I thought it would be smarter than that...
'Doc
David Robbins
"David or Jo Anne Ryeburn" <rye...@sfu.caz> wrote in message
news:ryeburn-0707...@207.6.82.157...
> In article <j75W8.286$KY6.2...@news20.bellglobal.com>, " J. Harvey"
> <JM002.Harve...@Sympatico.ca> wrote:
>
> > "David Ryeburn"
must
> > be reciprocal ?
>
> I'd like to see it too.
>
Ramo, Whinnery, and Van Duzer 'Fields and Waves in Commumication
Electronics' section 12.31.
and yes, in the second sentence it does assume there is a 'conjugate match'.
and yes, you can have antennas that aren't reciprocal. they one example
they give is when the transmission path includes something like the
ionosphere that does not have 'strictly bilateral properties'. one other
example i have heard of elsewhere that can make antennas non reciprocal is
the introduction of ferrites in microwave systems. (not the same as beads on
coax!)
David Robbins
" J. Harvey" <JM002.Harve...@Sympatico.ca> wrote in message
news:YPdW8.945$c_2.3...@news20.bellglobal.com...
would violate the assumptions that the conjugate match exist and that all
other parts of the system remain the same when exchanging the tx and rx
antennas to prove reciprocity. obviously if you change the external systems
connected to the antennas they will not necessarily perform the same way.
Richard Harrison
"---transmission lines acting like a shorted stub during transmit and
like an open stub during receive."
A receiver can extract maximum energy only when matched to an antenna.
The antenna`s radiation resistance is the source resistance for volts
"generated" in an antenna.
When an antenna is matched to a receiver, 50% of the energy in the
antenna goes to the receiver. The other 50% is reradiated.
There are T/R switches, isolators, directional couplers, etc. As Cecil
noted, receivers and transmitters can have vast differences in
impedance, but inherently, antennas are reciorocal.
CAM
>
> CAM wrote:
> > What if the driving impedance of the transmitter is a lot different
> > than the impedance across the transmission line during receive, e.g.
> > the transmission line acting like a shorted stub during transmit
> > and like an open stub during receive?
> But how does it know when to switch?
With my high school ham rig, I threw the switch(s). Nowadays, there's
usually a T/R switch/relay built in. It just seems to me that the
driving impedance of the transmitter would have to equal the loading
impedance of the receiver for the antenna system to be truly reciprocal.
Just a random thought.
--
cheers, CAM http://www.qsl.net/w5dxp
CAM
>
> "CAM" wrote:
> > What if the driving impedance of the transmitter is
> > a lot different than the impedance across the
> > transmission line during receive, e.g. the
> > transmission line acting like a shorted stub
> > during transmit and like an open stub during receive?
>
> Even if the Zo of the transmitter were zero, the E/I ratio would start-out
> as the cable Zo for a while (time 'related' to the length of the cable), and
> then revert back to the normal antenna SWR mess. The Zo of the transmitter
> never enters into it. (?)
If there is an SWR on the line greater than 1:1, the steady-state E/I ratio
will be anything except Z0. A 1/2WL dipole with a 1/2WL shorted stub hanging
from the middle would be resonant on a certain frequency. Would that same 1/2WL
dipole with a 1/2WL open stub hanging from the middle be resonant on that same
frequency? If not, it leads me to believe that the driving impedance
on transmit and the loading impedance on receive need to be the same for
the antenna to have the same characteristics on transmit and receive.
CAM
>
> > <JM002.Harve...@Sympatico.ca> wrote:
> >
> > > "David Ryeburn"
> > > Does anyone have a simple proof (or even an indication) that antennas
> > > must be reciprocal ?
>
> Electronics' section 12.31.
>
> and yes, in the second sentence it does assume there is a 'conjugate match'.
Which probably implies a lossless antenna system.
Gene Fuller
> . . . Nowadays, there's
> usually a T/R switch/relay built in. It just seems to me that the
> driving impedance of the transmitter would have to equal the loading
> impedance of the receiver for the antenna system to be truly reciprocal.
I don't think anyone has seriously claimed that any arbitrary antenna
*system* is reciprocal, truly or otherwise.
Quoting from Kraus "Antennas", 2nd edition, pages 409-410,
*****
This theorem [reciprocity theorem for antennas] as applied to antennas
may be stated as follows:
If an emf is applied to the terminals of an antenna A and the current
measured at the terminals of another antenna B, then an equal current
(in both amplitude and phase) will be obtained at the terminals of
antenna A if the same emf is applied to the terminals of antenna B.
It is assumed that the emfs are of the same frequency and that the media
are linear, passive and also isotropic.
[end quote]
*****
If one tries to add real-world conditions to this ideal system then all
reciprocity bets are off.
As usual, this thread was not arguable in its original form, so it is
quickly drifting to an ill-defined topic that can be argued ad nauseam.
73,
Gene
W4SZ
Crazy George
feedpoint. Most of you are confusing what might be called system
reciprocity with antenna characteristics. A couple of you have realized
this, but not yet expressed it in simple terms. System reciprocity and
antenna reciprocity are not necessarily the same demon.
--
Crazy George
Remove NO and SPAM from return address
J. Harvey <JM002.Harve...@Sympatico.ca> wrote in message
Richard Clark
<JM002.Harve...@Sympatico.ca> wrote:
>
>Even having those hydrophones pay the ultimate price and literally explode
>with the largest possible amplitude signal, those signals arriving back at
>the origin will add up in phase only in proportion to the number of
>hyrophones. Use a hundred exploding hydrophones and you'll be only 20dB
>above the sounds of a single distant exploding hydrophone. This would still
>be far less than a locally exploding stick of dynamite.
>
Hi Jeff,
Certainly if the receiver/transmitter faithfully reproduced the
signal, the result at the origin of the sound would be weak - however
many dB down suggests simply turning up the gain (haven't you seen
Dune?).
I recall that back in the 60's the French experimented with a sonic
weapon. The problem there (and typically French) was that it
scrambled the brains of the operator. C'est ca....
You are right to say this is not running time backwards, that would
suggest the effect of the reversal would lead to an "implosion."
However I might guess that a fish at that site, that survived the
original to only experience the reverse, would think (in fish
thoughts) Deja Vu.
'Doc
Gene,
--------------------
> If one tries to add real-world conditions to this ideal system then all
> reciprocity bets are off.
>
> As usual, this thread was not arguable in its original form, so it is
> quickly drifting to an ill-defined topic that can be argued ad nauseam.
>
> 73,
> Gene
> W4SZ
--------------------
"10-4" the ad nauseum...
'Doc
'Doc
CAM,
Then, answer this. Since when is the feed line
and transmitter/receiver part of the antenna?
'Doc
CAM
> Then, answer this. Since when is the feed line
> and transmitter/receiver part of the antenna?
The reciprocity theorem apparently requires a conjugate match.
Where is the conjugate match with just an antenna? How can
you compare transmit to receive without a transmitter and
receiver? Without a transmitter and receiver, what is the
meaning of transmit versus receive "reciprocity"? For a
center-fed antenna, removing the feedline drastically
changes the antenna characteristics so what is the meaning
of "reciprocity" after we remove the feedline?
JGBOYLES
>> Antennas are completely reciprocal. They have
>> the same characteristics receiving as transmitting.
If one considers an rf path as two rather large copper wires (no loss), as does
reciprocal theory, then move the generator and load from one end to the other.
Check the voltage across the load each time it is moved. It will be the same
each time. That is reciprocal. The rf circuit is reciprocal. The effects of
fields, propagation, transmitter power, ect should not enter. If you look at
it as a dc circuit it makes sense, but if you add other stuff?
Reciprocal is defined, so the discussion is about external influences
changing Reciprocity is??.
73 Gary N4AST
J. Harvey
> Quoting from Kraus "Antennas", 2nd edition, pages 409-410,
> *****
> This theorem [reciprocity theorem for antennas] as
>applied to antennas may be stated as follows:
>
> If an emf is applied to the terminals of an antenna A
> and the current measured at the terminals of another
> antenna B, then an equal current (in both amplitude
> and phase) will be obtained at the terminals of
> antenna A if the same emf is applied to the terminals
> of antenna B.
>
> It is assumed that the emfs are of the same frequency
> and that the media are linear, passive and also isotropic.
> [end quote]
> *****
Oh, is THAT (-^) all ? That is a much smaller claim than the ones that I
questioned.
> Antennas are completely reciprocal. They have
> the same characteristics receiving as transmitting.
and
> If you run time backwards, what's the difference
> between transmission and reception?
Thank you Gene (and Kraus). Subject nailed.
Dick Carroll
CAM wrote:
>
For a
> center-fed antenna, removing the feedline drastically
> changes the antenna characteristics so what is the meaning
> of "reciprocity" after we remove the feedline?
It means that the drastically changed antenna is still reciprocal as a
drastically changed antenna.
Dick
CAM
> It means that the drastically changed antenna is still reciprocal as a
> drastically changed antenna.
Bart has reminded that the reciprocity theorem requires a transmitter
and a receiver connected to those antennas.
Dick Carroll
CAM wrote:
>
> Dick Carroll wrote:
> > It means that the drastically changed antenna is still reciprocal as a
> > drastically changed antenna.
>
> Bart has reminded that the reciprocity theorem requires a transmitter
> and a receiver connected to those antennas.
OK. What that means is the separate transmit and receive *systems*
aren't reciprocal unless identically well matched to the antenna.
Richard Harrison
"OK. What that means is the separate transmit and receive "systems"
aren`t reciprocal unless identically well matched to the antenna."
Yeah. A joker earlier asked: "What if I stick a ferrite circulator in
your waveguide?"
Bart Rowlett
CAM wrote:
>
> Dick Carroll wrote:
> > It means that the drastically changed antenna is still reciprocal as a
> > drastically changed antenna.
>
> Bart has reminded that the reciprocity theorem requires a transmitter
> and a receiver connected to those antennas.
In the paper I forwarded, the proof of reciprocity utilized a source
and load but
it certainly is not required for a reciprocal relationship to exist
between two ports.
Reciprocity between two ports in a network exists when the transfer
impedance, or admittance, is the same in one direction as in the
other. In other words, Zji = Zij or Yji = Yij. That's all there
is to it. In the usual matrix representation, reciprocal networks
have Z, Y, or S parameters which are symmetrical about the diagonal.
The transfer impedance between two ports (port 1 input, and port 2
output) is defined as the voltage applied to port 1 divided by the
short circuit current in port 2 with all other ports open circuited.
Various other equivalent relationships exist, such as the
representation used in scattering (S) parameters, and for the very
special, but unnecessary, case where a conjugate match exists
somewhere in the system. They all can be reduced to the basic
definition given above with algebraic manipulation. Any good network
theory textbook will provide the proofs.
bart
wb6hqk
CAM
>
> CAM wrote:
> > Bart has reminded that the reciprocity theorem requires a transmitter
> > and a receiver connected to those antennas.
>
> OK. What that means is the separate transmit and receive *systems*
> aren't reciprocal unless identically well matched to the antenna.
Actually, my memory was fuzzy on exactly what reciprocity means. It
means in (almost) any RF communications system, if the receiver and
transmitter positions are swapped, the result will be the same.
Tdonaly
Systems" by D. Roy Choudhury:
" If we consider two loops A and B of a
network N and if an ideal voltage source
E in loop A produces a current I in loop
B, then interchanging positions, if an
identical source in loop B produces the
same current I in loop A, the network is
said to be reciprocal.
" The dual is also true.
" A linear network is said to be reciprocal
or bilateral if it remains invariant due to the
interchange of position of cause and effect
in the network."
Later, "A reciprocal network is comprised of linear, time-invariant,
bilateral, passive elements." (This isn't
exactly what he said; I changed it slightly
to correct his language.)
I don't know how the above relates
exactly to antennas. The real-world
atmosphere certainly isn't time-invariant,
but I don't suppose that will stop those
who like to use the simplifications of
network theory to explain all electrical
phenomena.
Tom Donaly KA6RUH
Bill Klocko
> Dick Carroll wrote:
> >
> > CAM wrote:
> > > Bart has reminded that the reciprocity theorem requires a transmitter
> > > and a receiver connected to those antennas.
> >
> > OK. What that means is the separate transmit and receive *systems*
> > aren't reciprocal unless identically well matched to the antenna.
>
> Actually, my memory was fuzzy on exactly what reciprocity means. It
> means in (almost) any RF communications system, if the receiver and
> transmitter positions are swapped, the result will be the same.
In addition to that, if reciprocity holds, several important results
follow. First of all, the transmitting and receiving patterns of an
antenna are identical. Then, since directivity depends only on the
shape of the power pattern, the directivity of the antenna when
transmitting is the same as its directivity when receiving. If the
terminal impedance of a transmitting antenna is Zt, then the load
impedance required for max power transfer when receiving is the
complex conjugate Zt*. 73,
Bill N3WK
J. Harvey
> In addition to that, if reciprocity holds, several important
> results follow. First of all, the transmitting and receiving
> patterns of an antenna are identical. Then, since directivity
> depends only on the shape of the power pattern, the
> directivity of the antenna when transmitting is the same
> as its directivity when receiving. If the terminal impedance
> of a transmitting antenna is Zt, then the load impedance
> required for max power transfer when receiving is the
> complex conjugate Zt*.
I like the way that you laid that out. Each naturally follows the previous.
The only part that I've yet to understand is why a 'structure' (part of an
otherwise ordinary antenna) couldn't be arranged that would distinguish
between a *spherical* wavefront being emitted by the transmitting antenna,
and a *planer* wavefront arriving at that very same antenna in when in
receive mode. I can't think of such a structure, but the wavefront shapes
are very different. It 'seems' possible. If such a structure were
possible, then one could make a receive-only, or a transmit-only,
(non-reciprocal) antenna.
Obviously the burden of proof is on the 'inventor' (sic) to present an
example... sigh... Suggestions welcomed.
Andy Cowley
>
> " J. Harvey" wrote:
> > Does anyone have a simple proof (or even an indication) that antennas must
> > be reciprocal ?
>
> than the impedance across the transmission line during receive, e.g.
> the transmission line acting like a shorted stub during transmit
> and like an open stub during receive?
Then you are using a _different _ antenna system for receive and
transmit and reciprocity is irrelevant.
vy 73
Andy, M3ABC
Andy Cowley
1885. When Heinrich Hertz started it all? ;-)
The distinction between antenna and antenna system
seems arbitrary unless ideal components are used,
hence Hertz and not Maxwell.
Unless of course you can prevent a real feedline
from radiating at all?
vy 73
Andy, M3ABC
Peter O. Brackett
>"Richard Harrison" <richard...@webtv.net> wrote in message
news:18728-3D...@storefull-2395.public.lawson.webtv.net...
[snip]
Or... indeed an electronic circulator in your coax or twinlead.
--Peter K1PO
Indialantic By-the-Sea, FL.