VSWR doesn't matter?

[Richard Clark]

As Tim Williams alludes, it depends on the transmitter design.

Hi Bob,

No quantifiable answer I see. It's not unexpected, everyone who knows
what it isn't has never been able to say what it is. It seems like
the stock answer you give the cop who asks if you know the speed
limit.
"No. But I wasn't speeding!"

The dependency here started with a conventional Ham transmitter, one
so ordinary as to be a commodity. The design is not so exotic as to
elude a very simple value - except for those who know it isn't 50
Ohms.

73's
Richard Clark, KB7QHC

 

[Tim Williams]

Of course, that was a tongue-in-cheek posting.
But if you could design a Thevenin equivalent
source with a 0.1 ohm source impedance, wouldn't
the efficiency calculate out to be pretty high?

Class D rules. (Using MOSFETs, the Thevenin equivalent is quite easy to
spot, too!)

Tim

 

[Richard Fry]

"Roy Lewallen" wrote

The problem is that the idea of "reflected energy" turning the plates hot
is so easy to understand, that people aren't willing to abandon it simply
because it isn't true.

_____________

But reflected energy/power does exist.

For an easy example, such reflections are evident in the picture seen on an
analog TV receiver when the match between the transmit antenna and the
transmission connected to it is bad enough.

In analog TV transmit systems with a typical 500+ foot length transmission
line from the tx to the antenna, a 5% reflection from a far-end mismatch can
be quite visible, showing as a "ghost" image that is offset from the main
image as related to the round-trip propagation time of the transmission
line.

RF

 

[Dan Bloomquist]

billcalley wrote:

We are all told that VSWR doesn't matter when using low loss
transmission lines, since the RF energy will travel from the
transmitter up to the mismatched antenna, where a certain amount of
this RF energy will reflect back towards the transmitter; after which
the RF will then reflect back up to the antenna -- where the energy is
eventually radiated after bouncing back and forth between the
transmitter and antenna.

As pointed out, VSWR does matter. A lot of bouncing means you heat the
transmission line with the power instead of radiating the power. 'Doesn't
matter', really means it can be tolerated if need be.

I understand the concept, but what I don't
quite understand is why the reflected RF energy isn't simply absorbed
by the 50 ohm output of the transmitter after the first reflection?
For the RF to bounce back and forth, wouldn't the transmitter's
impedance have to be very, very high (or low) when the reflected RF
energy hit its output stages? I know I'm missing something vital
here...

Here is what you are missing. In the case of the output, (real/resistive
component of the transmitter), seeing the reflected wave, it is _not_
reflecting that power back up the transmission line as you think it is. It
would go back to that real impedance and heat the transmitter. Here is
what is done with a miss match in the real world.

trans-output -> match -> line -> antenna

The 'match' is where the magic happens. All the energy coming down the
line that got reflected from the antenna 'sees' the 'trans-output ->
match' as a perfect reflector and gets bounced back[*]. On the other side
of the match is the trans->output. There the trans->output sees a perfect
impedance, (technically, the conjugate of the trans->output), so that all
the power travels through the match toward the antenna.

The magic is that when the match is tuned, both of the above conditions
are satisfied.

*The reflected wave sees a purely reactive reflector not just because of
the network but also because of the output power of the transmitter.
Without transmitter power the impedance as seen from the load will
dramatically change.

Best, Dan.

Saying that SWR doesnt matter is a rather broad statement(like saying never
or always) but I have know of antenna systems having an SWR of 30:1 and his
was normal. The feedline was balanced line made of 1 inch copper. Of course
an SWR lie this on coax could be fatal to coax and equipment. A more common
example of this is the 1/4 wl matching section on a J-pole antenna. It
matches 50 ohms to a few Kohms so an SWR of 60: 1 or so would not be unusal
here.S oas long as the feedline can handle the current and voltage peaks
without much los it doesnt matter much as long as the source impedance is
matched to the impedance at the input to the transmission line.Im sure there
is a practical limit though.

Hi Jimmie,
Keep in mind I'm answering in the context of the op's post. And the
theoretical SWR on a stub is infinite. The point of the stub at the
antenna is to keep the SWR on the transmission line in a reasonable
range, to make a match if you will. To put high SWR on the feedline
instead of matching at the antenna isn't a great idea in my book.

OTOH.
I finally did some sidebanding a couple of months ago. (First time on
HF.) I got my hands on an old swan 500c. After changing the 6je6's and
supply caps, I had to find out what it was like to get on the air. I ran
outside and hung a wire between the lab and the shop. 40-50 feet. Put a
couple of alligator clips on the end of a chunk of rg-58 and into the
window. I started looking for the antenna through the trans-match with
an antenna bridge. The tuning was very sharp, lots of Q. I don't know if
I could have found it without the bridge I was willing to tolerate
the miss match to get on the air.

Well, it worked out. I made some great QSLs across the mid west and into
northern CA. I live in Vernon AZ. I'm pleased this turned out to be as
great a radio location as I thought. It shouldn't be long before I get a
beam on a tower. By then I'll look to match at the antenna and keep the
SWR off the feed line as much as possible.

Best, Dan.

 

[Dan Bloomquist]

Yes! All that matters to the transmitter is the impedance it sees. It
doesn't know or care that you've mathematically separated the delivered
power into "forward" and "reverse" components. It doesn't know or care
what the SWR is on the transmission line connected to it, or even if a
transmission line is connected at all.

Well, without a line, you don't have a real component to tune into.
Drawing arcs on a smith chart from an open line with capacitors and
coils will only get you to another purely reactive point.

Best, Dan.

 

[[email protected]]

Unfortunately, you'd be learning the wrong lesson.

The cherry color is due to the transmitter being loaded with an
impedance it's not designed for, causing the final to run at low
efficiency. You can disconnect the antenna and replace it with a lumped
RC or RL impedance of the same value and get exactly the same result.
Alternatively, you can attach any combination of load and transmission
line which give the same impedance, resulting in a wide variation of
"reflected energy", and get exactly the same result. All that counts is
the impedance seen by the transmitter, not the VSWR on the line or the
"reflected power".

The problem is that the idea of "reflected energy" turning the plates
hot is so easy to understand, that people aren't willing to abandon it
simply because it isn't true.

See http://eznec.com/misc/Food_for_thought.pdf for more.

Roy Lewallen, W7EL

The fact that any transmission line and antenna combination can be
replaced with an RLC lumped load at the transmitter output and the
transmitter can't tell the difference is something that a lot of
hams seem to have a problem understanding.

What I've never understood is why so many hams have a problem with
the concept of equivalent circuits only when antennas and transmission
lines are involved.

 

[Jimmie D]

Hi Jimmie,
Keep in mind I'm answering in the context of the op's post. And the
theoretical SWR on a stub is infinite. The point of the stub at the
antenna is to keep the SWR on the transmission line in a reasonable range,
to make a match if you will. To put high SWR on the feedline instead of
matching at the antenna isn't a great idea in my book.

Sure yoiu can, that stub is a transmission line. It would matter if it s a
1/4 wl long or 21 1/4 wl long. If it is designed to handle the current and
voltage peaks it can transmit power with low loss when a high VSWR is
present. Its just that most people dont make there feedlines out of inch
copper tubing. Even with 450 ohm ladder line 10:1 VSWR is very acceptable.

 

[[email protected]]

Hi Jim,

Would it be fair to say there are a number of Hams (no need to go into
proportionality, could be equal number) who have difficulties of
understanding with going from lumped, equivalent circuits to antennas
and transmission lines?

The two perspectives are not exclusionary nor mutually incompatible,
only the arguers are.

I'd have to say that as soon as a circuit contains a radiator or a
transmission line the arm waving begins.

 

[Cecil Moore]

The problem is that the idea of "reflected energy" turning the plates
hot is so easy to understand, that people aren't willing to abandon it
simply because it isn't true.

It also isn't true that there is no energy in the reflected wave, that
such beliefs are gobbledegook, and that RF standing wave energy
just sloshes around in a transmission line at less than light speed.
To really understand what is going on, one has to understand
superposition and interference between RF energy waves. You
are on record as not caring to understand reflected energy. Please
don't condemn those of us who are trying to understand.

 

[Cecil Moore]

Yes! All that matters to the transmitter is the impedance it sees. It
doesn't know or care that you've mathematically separated the delivered
power into "forward" and "reverse" components. It doesn't know or care
what the SWR is on the transmission line connected to it, or even if a
transmission line is connected at all.

Think about this - if the transmission line is exactly one-wavelength
long
and lossless, the transmitter sees exactly the same impedance as the
load. At the load, we know reflections occur, but they are same-cycle
reflections so during steady-state with no modulation, exactly the
same
conditions exist at the transmitter as exist at the load if the
transmitter
has the same impedance as the transmission line. So even if we
cannot measure the reflections back into the transmitter, they are
no doubt, there - that is, unless one denies the existence of
reflections
in which case, one needs to explain how standing waves are possible
without reflections in a single-source system.

 

[Richard Clark]

The fact that any transmission line and antenna combination can be
replaced with an RLC lumped load at the transmitter output and the
transmitter can't tell the difference is something that a lot of
hams seem to have a problem understanding.

Hi Jim,

Would it be fair to say there are a number of Hams (no need to go into
proportionality, could be equal number) who have difficulties of
understanding with going from lumped, equivalent circuits to antennas
and transmission lines?

The two perspectives are not exclusionary nor mutually incompatible,
only the arguers are.

73's
Richard Clark, KB7QHC

 

[Richard Clark]

I'd have to say that as soon as a circuit contains a radiator or a
transmission line the arm waving begins.

Amen

 

[Uncle Peter]

No, we are not all told that.
That would cause half the power to be lost as heat in
the output stage. It's only 50ohm once it becomes a moving
wave in the transmission line.

Bob9

In that case...

Half power is only lost when terminated to a 50-ohm load;
i.e. no standing waves. What happens when there's a
mismatch and reflected energy



I'll go stand in a corner...

 

[Jimmie D]

In that case...

Half power is only lost when terminated to a 50-ohm load;
i.e. no standing waves. What happens when there's a
mismatch and reflected energy



I'll go stand in a corner...

A mismatch where, between the feedline and the antenna, feedline and source,
source and impedance seen at the input to the feedline.

Jimmie

 

[Tim Williams]

In analog TV transmit systems with a typical 500+ foot length transmission
line from the tx to the antenna, a 5% reflection from a far-end mismatch can
be quite visible, showing as a "ghost" image that is offset from the main
image as related to the round-trip propagation time of the transmission
line.

I've noticed that, at least in this area, Fox 39 (WQRF) has a ghost of a few
microseconds (I forget what exactly, I've calculated it before). Something
like 500 feet, IIRC.

Tim

 

[Dan Bloomquist]

billcalley wrote:

<snip>

I've been reading the posts on this. One poster said this has been going
on for twenty years! (For the other groups, this thread has life on
rec.radio.amateur.antenna) It doesn't need to be so.

First, there should be no doubt that reflected power on a transmission
line is real. Sure, you can replace the line with a lump but that
doesn't clear up the question for others.

For the next two examples, see page 179:
http://cp.literature.agilent.com/litweb/pdf/54753-97015.pdf
All examples assume the same impedance for source and line.

First example, step into an open line with a Thevenin source. The energy
is divided between the source and the line. Half the energy is moving
down the line and when it returns changes the impedance the source sees
to an open circuit. The energy does not flow back into the source, so,
where did it go? It is stored in the capacitance of the line.

Second example, step into a shorted line. When the energy returns the
source now sees a short. The energy does not flow back into the source,
so, where did it go? It is stored in the inductance of the line.

So here are two examples where the energy sent down the line do not
return to the source.

Third example. Send a pulse down the line. The Thevenin voltage source
will go to short, as it should, when the pulse falls. The pulse is
reflected from either an open or a short at the end of the line. All the
energy is dissipated in the source impedance when this pulse returns.
That is where the energy goes. And it is obviously the _same_ energy
created at the source.

Sure, non of the cases above represent steady state AC. But they do show
that energy may or may not be returned to the real component of the source.

With the above in mind, it can be shown, (in some part II), that a real
accounting of energy from source to load and back is possible.
Equivalent circuits are just that, the trading of line for lump. But,
and this is really important, the only reason the effective impedance at
the input of a 50 ohm line is not 50 ohms is because of reflected energy.

Best, Dan.

 

[joseph2k]

Unfortunately, you'd be learning the wrong lesson.

The cherry color is due to the transmitter being loaded with an
impedance it's not designed for, causing the final to run at low
efficiency. You can disconnect the antenna and replace it with a lumped
RC or RL impedance of the same value and get exactly the same result.
Alternatively, you can attach any combination of load and transmission
line which give the same impedance, resulting in a wide variation of
"reflected energy", and get exactly the same result. All that counts is
the impedance seen by the transmitter, not the VSWR on the line or the
"reflected power".

The problem is that the idea of "reflected energy" turning the plates
hot is so easy to understand, that people aren't willing to abandon it
simply because it isn't true.

See http://eznec.com/misc/Food_for_thought.pdf for more.

Roy Lewallen, W7EL

Do us a favor, compute the S-vectors for an incandescent lamp with a linear
filament.
Then follow though with the same for a transmitter, transmission line and a
mismatched load.
You will find that is the reflected S-vector that adds heat to the plate.

 

[joseph2k]

"Roy Lewallen" wrote
_____________

But reflected energy/power does exist.

For an easy example, such reflections are evident in the picture seen on
an analog TV receiver when the match between the transmit antenna and the
transmission connected to it is bad enough.

In analog TV transmit systems with a typical 500+ foot length transmission
line from the tx to the antenna, a 5% reflection from a far-end mismatch
can be quite visible, showing as a "ghost" image that is offset from the
main image as related to the round-trip propagation time of the
transmission line.

RF

Poppycock, TV ghosting is caused by multipath length differences. Calculate
the position ratio and the horizontal scan frequency (15750 Hz is close
enough). That gives you the path length difference; it is generally on the
order of miles (= major terrain features).

 

[joseph2k]

I'd have to say that as soon as a circuit contains a radiator or a
transmission line the arm waving begins.

Except for less than 1%. In the 1960's that would have been about 5% to
10%, and handwaving will never include Wes Hayward W7ZOI.

 

[Richard Fry]

Poppycock, TV ghosting is caused by multipath length differences.
Calculate the position ratio and the horizontal scan frequency
(15750 Hz is close enough). That gives you the path length
difference; it is generally on the > order of miles (= major
terrain features).

______________

Analog TV ghosts can be produced within the TV transmit antenna system as
well as by reflections of the transmitted signal in the propagation
environment. I know this from my experience as an RCA Broadcast Field
Engineer, because I've evaluated and corrected many transmit antenna systems
that had been the source of such ghosts.

For example, a reflection from a mismatch between a 1,000 foot long,
air-dielectric transmission line and the TV transmit antenna connected to it
produces a ghost with ~ 2 µs delay from the main image. The active scan
width of an NTSC TV line is about 53 µs, so 2/53 = ~4% of the width of the
screen, or maybe 5% counting overscan. This ghost is easy to see in a
typical TV set/viewing setup.

RF