Antennas 101
 

Following below are some thoughts of mine in response to an email sent to
me.

There are some good engineers reading this NG, and I am asking for comments
on what I wrote -- corrections where seen necessary, and any other thoughts.

Thanks.

RF

+ + + + +

----- Original Message -----
The term "accept power" is interesting. To me, it's just a measure
of the input impedance of the antenna. If the resistance (radiation
plus loss resistance) is zero, you're not going to get it to absorb
power no matter what you do.
____________

Yes. And the input conditions depend on the ability of the radiator to
generate EM fields. No current can enter and "flow through" a radiator if
it doesn't have some place to go. If a radiator is not electrically long
enough to allow differential current to exist along its length, it cannot
generate EM fields. It is the di/dt along the radiator length that
generates those fields.

Adding a matching network at the antenna input doesn't change the instrinsic
ability of an antenna to radiate. That is determined by the factors
described in the paragraph above. A matching network can permit the tx to
increase its rated, safe output power while driving that poor antenna, but
any extra power available at the antenna input because of that will be
subject to the same poor radiation efficiency as if the matcher wasn't used.
And much of any added power from the tx may get dissipated in lossy output
system components other than the antenna, rather than being radiated. The
antenna itself will still have the same directivity/gain that it had before
the matching network was added.

Improving the ability of a poor antenna to generate EM fields per unit of
source power is possible in a limited way only by increasing its electrical
length. "Capacity hats" and inductances incorporated into the radiating
structure can be used, as examples. This also raises the antenna radiation
resistance and reduces the reactance at the antenna input -- making it
easier to match into, and reducing system losses.

Someone gave an example of a pager antenna or something that had a
rated gain that was below isotropic. It'd be interesting to see where the
losses were.

This is covered in my comments above, I believe.

//

Richard Fry wrote:

... I am asking for comments on what I wrote -- ...

Comment: good stuff, and a couple of additional comments.

The term "accept power" is interesting. To me, it's just a measure
of the input impedance of the antenna. If the resistance (radiation
plus loss resistance) is zero, you're not going to get it to absorb
power no matter what you do.

Yes. And the input conditions depend on the ability of the radiator to
generate EM fields. No current can enter and "flow through" a radiator if
it doesn't have some place to go. If a radiator is not electrically long
enough to allow differential current to exist along its length, it cannot
generate EM fields. It is the di/dt along the radiator length that
generates those fields.

Adding a matching network at the antenna input doesn't change the
instrinsic
ability of an antenna to radiate. That is determined by the factors
described in the paragraph above. A matching network can permit the tx to
increase its rated, safe output power while driving that poor antenna, but
any extra power available at the antenna input because of that will be
subject to the same poor radiation efficiency as if the matcher wasn't
used.
And much of any added power from the tx may get dissipated in lossy output
system components other than the antenna, rather than being radiated.

The feedline has a certain efficiency. A matching network at the antenna
has a certain efficiency. If the matching network at the antenna is much
more efficient than the feedline, then it may be a reasonable solution.
An SGC-230 tuner mounted at the base of a 22 foot vertical makes a
reasonable antenna for 40m-10m operation. Feeding that same antenna
through coax from an SGC-230 in the shack makes for a pretty poor
antenna on most bands except 30m. Same antenna, same tuner, different
configuration = wildly different results.

The
antenna itself will still have the same directivity/gain that it had before
the matching network was added.

Yes, but the SGC-230 at the base of the antenna may be operating at
50% efficiency while, if you moved the tuner back to the shack, the
feedline might be operating at 10% efficiency. The choice is clear
(if there are no thieves around. :-)

Improving the ability of a poor antenna to generate EM fields per unit of
source power is possible in a limited way only by increasing its electrical
length. "Capacity hats" and inductances incorporated into the radiating
structure can be used, as examples. This also raises the antenna radiation
resistance and reduces the reactance at the antenna input -- making it
easier to match into, and reducing system losses.

Someone gave an example of a pager antenna or something that had a
rated gain that was below isotropic. It'd be interesting to see where the
losses were.

This is covered in my comments above, I believe.

--
73, Cecil http://www.qsl.net/w5dxp

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Richard Fry wrote:
Following below are some thoughts of mine in response to an email sent
to me.

There are some good engineers reading this NG, and I am asking for
comments on what I wrote -- corrections where seen necessary, and any
other thoughts.

Thanks.

RF

+ + + + +

----- Original Message -----

The term "accept power" is interesting. To me, it's just a measure
of the input impedance of the antenna. If the resistance (radiation
plus loss resistance) is zero, you're not going to get it to absorb
power no matter what you do.

____________

Yes. And the input conditions depend on the ability of the radiator to
generate EM fields. No current can enter and "flow through" a radiator if
it doesn't have some place to go. If a radiator is not electrically long
enough to allow differential current to exist along its length, it cannot
generate EM fields. It is the di/dt along the radiator length that
generates those fields.


Actually, you can charge up anything to produce a field. If the field is
static, though, it won't radiate. Also, short radiators work just fine
if you excite them with frequencies whose wavelengths are on the same
order of magnitude. Finally, if you can keep the losses down, radiators
that are small compared to a wavelength of the frequency you want to
radiate, work well, too, though you may have to be satisfied
with narrow bandwidths and make heroic efforts to feed them. Finally,
you might be more accurate if you substituted 'dq/dt' with 'di/dt',
but even that won't guarantee radiation: transmission lines have
plenty of charge acceleration, but are designed specifically to
not radiate. You might want to rethink your ideas. Try not to
be so reductionist. You don't want to end up like Cecil.
73,
Tom Donaly, KA6RUH





Adding a matching network at the antenna input doesn't change the
instrinsic
ability of an antenna to radiate. That is determined by the factors
described in the paragraph above. A matching network can permit the tx to
increase its rated, safe output power while driving that poor antenna, but
any extra power available at the antenna input because of that will be
subject to the same poor radiation efficiency as if the matcher wasn't
used.
And much of any added power from the tx may get dissipated in lossy output
system components other than the antenna, rather than being radiated. The
antenna itself will still have the same directivity/gain that it had before
the matching network was added.

Improving the ability of a poor antenna to generate EM fields per unit of
source power is possible in a limited way only by increasing its electrical
length. "Capacity hats" and inductances incorporated into the radiating
structure can be used, as examples. This also raises the antenna radiation
resistance and reduces the reactance at the antenna input -- making it
easier to match into, and reducing system losses.

Someone gave an example of a pager antenna or something that had a
rated gain that was below isotropic. It'd be interesting to see where the
losses were.


This is covered in my comments above, I believe.

//

Tom Donaly wrote:
You don't want to end up like Cecil.

My end is down, not up, thank you.
--
73, Cecil http://www.qsl.net/w5dxp

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Tom Donaly wrote:
You don't want to end up like Cecil.


My end is down, not up, thank you.


They have a pill for that, now, don't they? :^)



Ed

Ed wrote:

w5dxp wrote:
My end is down, not up, thank you.

They have a pill for that, now, don't they? :^)

If you want your end up, be my guest. I personally like
mine down in a Lazy Boy, where it belongs.
--
73, Cecil http://www.qsl.net/w5dxp


----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 100,000 Newsgroups
---= East/West-Coast Server Farms - Total Privacy via Encryption =---

Richard Fry wrote:
Following below are some thoughts of mine in response to an email sent
to me.

There are some good engineers reading this NG, and I am asking for
comments on what I wrote -- corrections where seen necessary, and any
other thoughts.

Thanks.

RF

+ + + + +

----- Original Message -----

The term "accept power" is interesting. To me, it's just a measure
of the input impedance of the antenna. If the resistance (radiation
plus loss resistance) is zero, you're not going to get it to absorb
power no matter what you do.

____________

Yes. And the input conditions depend on the ability of the radiator to
generate EM fields.

"Ability" could be a misleading term. If 100 watts is fed to a very
short dipole, it will generate just as big a total field as a half wave
dipole fed 100 watts (neglecting conductor loss). So for a given power
input, it's just as able to generate an EM field as the longer dipole.
It does require a lot more current. You might consider that to mean it's
less "able" to generate a field, but if so, I think you should elaborate
in order to avoid perpetuating misconceptions.

No current can enter and "flow through" a radiator if
it doesn't have some place to go.

I'm not sure what that means. A radiator is a circuit. Any current which
flows in one terminal must flow out the other. That's true for any
antenna of any kind.

If a radiator is not electrically long
enough to allow differential current to exist along its length, it cannot
generate EM fields.

There is no such radiator. There's no threshold below which current
can't exist along its length.

It is the di/dt along the radiator length that
generates those fields.

Yep, and a short antenna has a much higher I and therefore di/dt for a
given power input than a long one. The result is the same total field.

Adding a matching network at the antenna input doesn't change the
instrinsic
ability of an antenna to radiate.

Correct.

That is determined by the factors
described in the paragraph above.

I don't believe the factors you cited play any role in determining the
ability of an antenna to radiate.

A matching network can permit the tx to
increase its rated, safe output power while driving that poor antenna, but
any extra power available at the antenna input because of that will be
subject to the same poor radiation efficiency as if the matcher wasn't
used.

Why "poor" radiation efficiency? An antenna radiates all the power it's
supplied, less the amount lost due to conductor and insulation
resistance. While it's more difficult to make a short antenna efficient,
it's not impossible. There's no loss of radiation efficiency directly
due to the length of the antenna.

And much of any added power from the tx may get dissipated in lossy output
system components other than the antenna, rather than being radiated.

That's correct. An antenna system which includes a short antenna is
likely to be less efficient because the short antenna's impedance
requires greater transformation than a longer antenna's, and this is
difficult to do with high efficiency. The loss of system efficiency (as
opposed to antenna radiation efficiency) is due to loss in the matching
network components.

The
antenna itself will still have the same directivity/gain that it had before
the matching network was added.


Correct.

Improving the ability of a poor antenna to generate EM fields per unit of
source power is possible in a limited way only by increasing its electrical
length.

Sorry, that's completely wrong. Try modeling a half wave dipole in free
space with the EZNEC demo program, then model a very short dipole, say
0.01 wavelength long. For both, set the wire loss to zero. You'll see
that the gain is only about a half dB different, due to the slightly
fatter shape of the short dipole's lobes. An Average Gain check, which
compares the total power in the field with the power fed from the
source, can be used as a direct measure of the radiation efficiency. It
should be very close to one for both antennas. (Any deviation from one
is due to calculation imprecision in the program.)

These EZNEC results are correct. If you have trouble believing them,
I'll be glad to furnish you with a few references from well known textbooks.

Of course, an actual antenna *will* have wire loss, and the wire
resistance will cause more loss in the short antenna than in a longer
one, due to the higher current for a given power input. It's that, plus
the matching system loss, that causes real short antennas to often
perform poorly -- not something inherently due to the generation of the
field from a short conductor.

"Capacity hats" and inductances incorporated into the radiating
structure can be used, as examples. This also raises the antenna radiation
resistance and reduces the reactance at the antenna input -- making it
easier to match into, and reducing system losses.

Yes. That word "system" is vital.

Someone gave an example of a pager antenna or something that had a
rated gain that was below isotropic. It'd be interesting to see where the
losses were.


This is covered in my comments above, I believe.

//

I hope you find these comments helpful.

Roy Lewallen, W7EL

Roy Lewallen wrote:
A radiator is a circuit.

Most antennas are NOT a circuit. Most antennas are distributed
networks.

Circuit theory doesn't work on distributed network problems.
That's why distributed network analysis was invented.
--
73, Cecil http://www.qsl.net/w5dxp

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

Antennas *are* circuits. Circuits can consist of either lumped or
distributed components, or both.

However, that's not particularly relevant to my point. In the steady
state, there is no test which can be devised that can distinguish an
antenna (including feedline, if desired) from a black box containing
lumped components -- a lumped component circuit(*). This allows us to
simply design such things as matching networks without any consideration
of the real properties of an antenna. Only when transient signals are
involved is other than simple lumped-network analysis necessary. In a
typical antenna and feedline system, steady state is reached much, much
faster than even the fastest CW dit or speech component, so transient
analysis isn't required for these common modes of operation. There are a
few familiar situations where analysis under normal operating conditions
requires consideration of the distributed nature of the system, such as
when doing time domain reflectometry or in some situations involving
television signals, both modulated and baseband. None of these, however,
can be considered steady state so the simplified model isn't applicable.

Cecil, if you feel a need to expound yet more on your theories, please
do so in one of the many threads you've come to dominate already, start
a new one, or concentrate your efforts on your forthcoming QEX article.
I hope you'll let us try and make an objective and hopefully helpful
contribution from time to time on this newsgroup without your constantly
attempting to steer the discussion to your theories.

(*) Furthermore, if we restrict analysis to a single frequency, the
black box needs to contain only two components - a resistor and either a
capacitor or inductor.

Roy Lewallen, W7EL

Cecil Moore wrote:
Roy Lewallen wrote:

A radiator is a circuit.


Most antennas are NOT a circuit. Most antennas are distributed
networks.

Circuit theory doesn't work on distributed network problems.
That's why distributed network analysis was invented.

"Richard Fry" wrote:
... I am asking for comments on what I wrote --
________________

Thanks to all who responded. Now to process those responses.

RF