Reply
 
LinkBack Thread Tools Search this Thread Display Modes
  #1   Report Post  
Old June 28th 05, 11:53 PM
Roy Lewallen
 
Posts: n/a
Default SWR - wtf?

I don't see the original posting here on rec.radio.amateur, but there
are a few misconceptions in the followups which should be addressed.

Lancer wrote:
On Tue, 28 Jun 2005 19:13:35 GMT, james wrote:


On Tue, 28 Jun 2005 18:31:58 GMT, Lancer wrote:


To have the measured SWR change with coax length, means you have
current flowing on the outside of the coax. Your coax then becomes
part of the antenna, so changing its length is changing the antenna
length. This would change the feedpoint impedance and the SWR.


That's correct, except that coax loss will also cause the SWR to change
with coax length. Loss will cause the SWR at the antenna (load) to
always be greater than at the transmitter (source).


Unless the line is carrying common mode currents that affect antenna
impedance, changing coax length won't change the SWR, even if the
antenna isn't matched.


Again correct except for overlooking the effect of coax loss.

But there's a real problem in communicating this. If you hook a 50 ohm
SWR meter to the input of a 75 ohm, 300 ohm, or line of any impedance
other than 50 ohms, the meter reading won't be the SWR on the
transmission line. That can mislead people into thinking that the SWR is
changing with line length when it actually isn't.


********

BS

Common mode currents on the shield of coaxial cables do not alter the
feed impedance. Repeat ofter me. Common mode currents on the shield of
coaxial cables do not alter the feed impedance.


Why repeat it if it isn't true? The explanation given by Lancer was
correct. If you change the length of the antenna, the feedpoint
impedance will change. When you have common mode current flowing on the
feedline, the feedline is part of the antenna; changing its length is
changing the antenna's length.


The feed impedance of an antenna is solely determined by its physical
length and any load impedances within the antenna structure. Load
impedances can be stray capacitance with ground via metal objects
within the near field of the antenna or even a building.


You have to realize that a radiating feedline (one carrying common mode
current) IS part of the antenna structure.


The "Magic" of an electrical halfwave transmission line is at a
precise frequency, the reflection of the load to the transmistter is
equal to the characteristic impedance of the transmission line
irregardless of what impedance it is terminated with.


This is true only of a lossless line. If the load impedance isn't far
from the line's characteristic impedance (i.e., the line's SWR is low),
a small amount of loss won't make much difference. However, if the line
SWR is high, even a small amount of loss can make a major change in the
impedance seen at the line's input. The effect is to skew the impedance
toward the line's Z0.

Other lengths
have the load impedance reflected back and transformed by the length
of the coax. The coax then acts as a transformer. It will either step
up or step down the impeadnace of the load depending on the load
itself and the electrical length of the coax.


It's a little more complicated than that. The line doesn't simply
multiply or divide the impedance by a constant, like a transformer --
except in the special case of a quarter electrical wavelength line or
odd multiples thereof. In other cases, the line does transform the
impedance, but in a complex way in which the resistance and reactance
are transformed by different factors. And reactance can be present at a
line's input even when the load is purely resistive. A Smith chart is a
good visual aid in seeing what happens. Assuming a lossless line, the
impedance traverses a circle around the origin. The radius of the circle
corresponds to the line's SWR. With the chart, you can see all the
combinations of R and X which a given line can produce with a given load
by changing its length. Incidentally, loss causes the impedance to
spiral inward toward the origin as the line gets longer, showing how
loss skews the input impedance toward Z0.


All a tuner does is electrically lengthen or shoten the coax by
introducing a lumped LC constant that helps present a resistive load
to the transmitter. The SWR at the feedline does not change. By
placing various different lengths of coax inline, you do the same
thing a tuner does, add a lumped LC constant.


As can be seen from the Smith chart, you can produce only particular
combinations of R and X by changing the length of a line which has a
given load impedance. Unless you're unusually lucky or have planned
things carefully, none of these combinations will result in 50 + j0
ohms, the usual goal, at the line's input. In contrast, a tuner is able
to adjust both R and X to produce, if designed right for the
application, 50 + j0 for a wide range of load impedances.

It requires at least two adjustable components to achieve an impedance
match from an arbitrary load impedance, because there are two separate
quantities, R and X or impedance magnitude and phase, which have to be
adjusted. Changing the line length is only one adjustment, so it can't
be guaranteed to provide a match. If you could also change the line's
Z0, for example, or the length of a stub, you'd have two adjustments and
you could guarantee a match providing you have enough adjustment range.


james



So thats all my tuner does, lengthen or shorten the coax?

Are you sure about that?


Rest assured, that's not all it does.

Roy Lewallen, W7EL
  #2   Report Post  
Old June 29th 05, 01:51 AM
K7ITM
 
Posts: n/a
Default

(I've snipped parts of Roy's original posting, indicated by ..., that I
hope are not particularly relevant to my added comments.)

Roy Lewallen wrote:
I don't see the original posting here on rec.radio.amateur, but there
are a few misconceptions in the followups which should be addressed.

....


Unless the line is carrying common mode currents that affect antenna
impedance, changing coax length won't change the SWR, even if the
antenna isn't matched.


Again correct except for overlooking the effect of coax loss.

But there's a real problem in communicating this. If you hook a 50 ohm
SWR meter to the input of a 75 ohm, 300 ohm, or line of any impedance
other than 50 ohms, the meter reading won't be the SWR on the
transmission line. That can mislead people into thinking that the SWR is
changing with line length when it actually isn't.


In addition, most hams (and other non-professionals -- and even many
professionals) don't bother to check that their SWR meter is properly
calibrated to the impedance they think it is. Most are nominally 50
ohms, but they can be built for any practical line impedance. Checking
calibration is not all that difficult, if you take the time to do it.
In addition, your nominally 50 ohm line (or 75 or whatever) can have an
actual impedance 10% or more from the nominal value. If you have
properly calibrated your meter to 50 ohms, and your line is 60 ohms,
you would read 1.2:1 SWR when your line is actually 1:1. And if the
SWR on the 60 ohm line is 1.2:1, that 50 ohm SWR meter can read
anything between 1:1 and 1.44:1, depending on the line length and its
load. Finally, though you may have checked that the meter to reads 1:1
with a 50 ohm load and infinity to 1 with a short or open load, the
construction of inexpensive meters may cause them to have significant
errors at other load impedances.

....



The "Magic" of an electrical halfwave transmission line is at a
precise frequency, the reflection of the load to the transmistter is
equal to the characteristic impedance of the transmission line
irregardless of what impedance it is terminated with.


This is true only of a lossless line. If the load impedance isn't far
from the line's characteristic impedance (i.e., the line's SWR is low),
a small amount of loss won't make much difference. However, if the line
SWR is high, even a small amount of loss can make a major change in the
impedance seen at the line's input. The effect is to skew the impedance
toward the line's Z0.


The piece that Roy quoted is so outrageous that I can easily believe he
didn't read it right, but I've re-read it several times, and it keeps
coming out the same: the "magical" halfwave line does NOT reflect an
impedance to the source (transmitter) equal to the LINE impedance as
the quoted section says, but it reflects the LOAD impedance (altered by
line loss as Roy says).

....

about tuners, Roy went on to write:

It requires at least two adjustable components to achieve an impedance
match from an arbitrary load impedance, because there are two separate
quantities, R and X or impedance magnitude and phase, which have to be
adjusted. Changing the line length is only one adjustment, so it can't
be guaranteed to provide a match. If you could also change the line's
Z0, for example, or the length of a stub, you'd have two adjustments and
you could guarantee a match providing you have enough adjustment range.


In addition, two adjustable components in a particular configuration,
even if they are infinitely adjustable (and reasonably close to
lossless!!--a very tall order!) won't necessarily give you the ability
to transform any arbitrary impedance to 50 ohms. There may be whole
practical areas of the complex impedance plane left untransformable.
Also, the efficiency of a particular tuner topology for a given load
impedance may be very good or may be terrible, when using practical
components in the tuner. To reiterate what Roy wrote, it's important
to use the right topology for the job you need to do.

Cheers,
Tom




james



So thats all my tuner does, lengthen or shorten the coax?

Are you sure about that?


Rest assured, that's not all it does.

Roy Lewallen, W7EL


  #3   Report Post  
Old June 29th 05, 02:11 AM
Roy Lewallen
 
Posts: n/a
Default

Thanks to Tom for the comments and additions.

. . .


[I've lost track of who said this:]

The "Magic" of an electrical halfwave transmission line is at a
precise frequency, the reflection of the load to the transmistter is
equal to the characteristic impedance of the transmission line
irregardless of what impedance it is terminated with.



[Roy:]

This is true only of a lossless line. If the load impedance isn't far
from the line's characteristic impedance (i.e., the line's SWR is low),
a small amount of loss won't make much difference. However, if the line
SWR is high, even a small amount of loss can make a major change in the
impedance seen at the line's input. The effect is to skew the impedance
toward the line's Z0.



[Tom:]

The piece that Roy quoted is so outrageous that I can easily believe he
didn't read it right, but I've re-read it several times, and it keeps
coming out the same: the "magical" halfwave line does NOT reflect an
impedance to the source (transmitter) equal to the LINE impedance as
the quoted section says, but it reflects the LOAD impedance (altered by
line loss as Roy says).
. . .


Wow, I certainly read that (top quote) too quickly. Tom is absolutely
right, as written it's very wrong, and I misread it. I retract my
statement about it's being "true only of a lossless line" -- of course
it's not true at all, but works as Tom says.

Roy Lewallen, W7EL
  #4   Report Post  
Old June 29th 05, 04:49 AM
Frank Gilliland
 
Posts: n/a
Default

On 28 Jun 2005 17:51:10 -0700, "K7ITM" wrote in
. com:

snip
But there's a real problem in communicating this. If you hook a 50 ohm
SWR meter to the input of a 75 ohm, 300 ohm, or line of any impedance
other than 50 ohms, the meter reading won't be the SWR on the
transmission line. That can mislead people into thinking that the SWR is
changing with line length when it actually isn't.


In addition, most hams (and other non-professionals -- and even many
professionals) don't bother to check that their SWR meter is properly
calibrated to the impedance they think it is. Most are nominally 50
ohms, but they can be built for any practical line impedance. Checking
calibration is not all that difficult, if you take the time to do it.
In addition, your nominally 50 ohm line (or 75 or whatever) can have an
actual impedance 10% or more from the nominal value. If you have
properly calibrated your meter to 50 ohms, and your line is 60 ohms,
you would read 1.2:1 SWR when your line is actually 1:1. And if the
SWR on the 60 ohm line is 1.2:1, that 50 ohm SWR meter can read
anything between 1:1 and 1.44:1, depending on the line length and its
load. Finally, though you may have checked that the meter to reads 1:1
with a 50 ohm load and infinity to 1 with a short or open load, the
construction of inexpensive meters may cause them to have significant
errors at other load impedances.



Impedance matching of an SWR meter is generally unimportant since most
SWR meters used for HF have a directional coupler that is much shorter
than the operating wavelength. Regardless, I'm not a big fan of SWR
meters -- they are good for detecting a major malfunction but that's
about it. Antenna tuning/matching is best done with a field strength
meter.







----== 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 =----
  #5   Report Post  
Old June 29th 05, 05:17 AM
Cecil Moore
 
Posts: n/a
Default

Frank Gilliland wrote:
Impedance matching of an SWR meter is generally unimportant since most
SWR meters used for HF have a directional coupler that is much shorter
than the operating wavelength.


Point is that they are usually calibrated for Z0=50 ohms
and are in error when used in Z0 environments differing
from Z0=50 ohms, e.g. Z0=75 ohms.
--
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 =---


  #6   Report Post  
Old June 29th 05, 05:53 AM
Buck
 
Posts: n/a
Default

On Tue, 28 Jun 2005 15:53:03 -0700, Roy Lewallen
wrote:

To have the measured SWR change with coax length, means you have
current flowing on the outside of the coax. Your coax then becomes
part of the antenna, so changing its length is changing the antenna
length. This would change the feedpoint impedance and the SWR.


That's correct, except that coax loss will also cause the SWR to change
with coax length. Loss will cause the SWR at the antenna (load) to
always be greater than at the transmitter (source).



Would the changing of the coax lead to moving the SWR meter to a
different voltage point on the coax?

--
73 for now
Buck
N4PGW
  #7   Report Post  
Old June 29th 05, 06:04 AM
Buck
 
Posts: n/a
Default

On Tue, 28 Jun 2005 20:49:46 -0700, Frank Gilliland
wrote:

I'm not a big fan of SWR
meters -- they are good for detecting a major malfunction but that's
about it. Antenna tuning/matching is best done with a field strength
meter.



A local retired instructor of some sort (military, i believe) has the
same opinion. He doesn't like SWR meters but instead measures all his
antennas by field strength meter.

I used to tune my Swan with one. I found when I used an SWR meter,
the minimum SWR dip was NEVER the maximum field strength reading. I
always had to raise the SWR to about 1.3:1 or so.

Around here, most of us know not to mention the performance of an
antenna to him if we only used an SWR meter or antenna analyzer. His
first question is "How did it do with the FSM?"

I believe he is right. Radios drop power when they don't like the SWR
and raise it when it does.

73
N4PGW

--
73 for now
Buck
N4PGW
  #8   Report Post  
Old June 29th 05, 06:37 AM
Owen
 
Posts: n/a
Default

On Wed, 29 Jun 2005 01:04:59 -0400, Buck wrote:


I used to tune my Swan with one. I found when I used an SWR meter,
the minimum SWR dip was NEVER the maximum field strength reading. I
always had to raise the SWR to about 1.3:1 or so.


The probable reason for maximum power output not coinciding with the
plate current dip is imperfect neutralisation (whether or not the
stage is neutralised). If you have a neutralisation adjustment, you
can get the two to coincide by properly adjusting the neutralisation.

You can use this coincidence as a quite sensitive indication of
optimal neutralisation if you use a digital meter to monitor plate
current, and the power out indicator to monitor RF output power.
Adjust the tuning and loading for rated power into a dummy load, check
the tune cap is peaked for max Po, observe the plate current,
carefully dip the plate current, noting whether the dip was left or
right of max Po. Now tweak the neutralisation until the two coincide.

When it is all done properly, the "dip" of the plate current should be
symmetric, ie it should be as "sharp" approaching from one side as the
other. (Asymetric dips are another symptom of non-omptimal
neutralisation.

Owen
--
  #9   Report Post  
Old June 29th 05, 08:22 AM
Roy Lewallen
 
Posts: n/a
Default

Buck wrote:

Would the changing of the coax lead to moving the SWR meter to a
different voltage point on the coax?


Sort of, but not exactly. Let's take an example. I'll keep the values
purely real to help folks who aren't familiar with complex math, but
keep in mind that these are special cases and a full treatment would be
somewhat more involved. I'll also make all the transmission lines
lossless to simplify things.

An SWR meter really just provides another way of reporting the impedance
it sees. You can verify this by connecting pure resistances of various
values to its output. For example, a (properly calibrated and operating)
50 ohm SWR meter will report 1:1 if you connect its output to a 50 ohm
resistor. If you connect it to either a 25 or 100 ohm resistor, it
reports 2:1. It does this despite the fact that there's no transmission
line at all connected to its output. Some people can put up a huge
smokescreen and waving of hands about reflected waves of one kind or
another, but at the end of the day the SWR meter can't tell the
difference between a resistor and a transmission line terminated with a
load, if the impedances the meter sees are the same. It's sensitive only
to impedance; it has no way of knowing even if a transmission line is
connected to its output, let alone what the transmission line's SWR or
even characteristic impedance is.

Now put a half wavelength piece of 50 ohm coax between the SWR meter and
those resistors. The SWR meter will still see the same impedances as
before, so it'll report the same SWRs. Now, though, there really is a
transmission line connected to its output. And because the meter is a 50
ohm meter and the line has a 50 ohm Z0, the SWR meter reading is the
same as the actual SWR on the line. When the load is 50 ohms, the line's
SWR is 1:1 and the meter sees 50 ohms so it reports 1:1. When the load
is 25 ohms, the line's SWR is 2:1, and the meter sees 25 ohms and
reports 2:1. When the load is 100 ohms, the line's SWR is 2:1, and the
meter sees 100 ohms and reports 2:1.

Next experiment: Connect the SWR meter through a *quarter* wavelength of
50 ohm line to a 100 ohm load. Now the impedance looking into the line
is 25 ohms instead of 100. But the SWR meter reads 2:1 when it sees 25
ohms as well as 100, so it still reads 2:1, which is also still the SWR
on the 50 ohm line. You can change the length of the 50 ohm line all you
want and, if it's lossless, the line's actual SWR stays the same -- but
the impedance at the input end of the line changes. For a 100 ohm load,
when the line is any even number of half wavelengths long, the input Z
is 100 ohms. When the line is any odd number of quarter wavelengths
long, the input Z is 25 ohms. At other lengths, the impedance is both
resistive and reactive, but the line's SWR is always 2:1. And the SWR
meter interprets all these possible impedances as 2:1, and that's what
it reads. The line SWR doesn't change as you change its length, and the
SWR meter reading doesn't change, either.

Now instead of a 50 ohm line, let's connect a half wavelength 100 ohm
line to the output of the same 50 ohm SWR meter and hook that to a 50
ohm resistive load. The line's actual SWR is 2:1 and, just like any
lossless line, the SWR stays the same regardless of its length. If the
transmission line is an even number of half wavelengths long we'll have
50 ohms at the input and the SWR meter will read 1:1, since it's a 50
ohm meter and interprets 50 ohms as 1:1. If we change the line length to
a quarter wavelength, the input impedance will be 200 ohms, which the 50
ohm SWR meter will interpret and report as 4:1. So by changing the line
length from a half to a quarter wavelength we've changed the SWR meter
reading from 1:1 to 4:1, even though the line's actual SWR was 2:1 all
along. That's what I was talking about. The SWR meter makes assumptions
about the SWR on the line from the impedances it sees. The line
transforms the load impedance in a different way than a 50 ohm line
would. The SWR meter then assumes an incorrect SWR value for the line,
and this incorrect value changes as the line length changes.

The same thing happens if the line has a 50 ohm characteristic impedance
and the meter is designed for some other Z0. The lesson is that an SWR
meter shows the actual SWR on a transmission line connected to its
output only if the SWR meter is designed for the same Z0 as the line.

Too often, people say "The SWR is. . .", but really mean "The SWR meter
reading is. . .". As you've seen, the two can often be very different.
When you see the SWR reading changing as you change the line length, it
doesn't necessarily mean that the line's SWR is actually changing.

Remember, in the preceding discussion I've assumed for simplicity that
all lines were lossless. In the real world, no line is, so the actual
line SWR will always be higher at the load than the source (unless of
course it's 1:1 at the load).

Roy Lewallen, W7EL
  #10   Report Post  
Old June 29th 05, 09:32 AM
Reg Edwards
 
Posts: n/a
Default

Roy, to cut things short, why don't you just say SWR meters don't
measure SWR on anything. All they do is indicate whether or not the
transmitter is terminated with its correct load resistance. So they
are quite useful.

They won't even tell you what the load resistance actually is unless
the load is exactly correct.

Stop fooling and confusing yourselves. The solution to everybody's
problems is simple - just change the name of the thing to TLI.
(Transmitter Loading Indicator).
----
Reg, G4FGQ


Reply
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules

Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT +1. The time now is 11:35 AM.

Powered by vBulletin® Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 RadioBanter.
The comments are property of their posters.
 

About Us

"It's about Radio"

 

Copyright © 2017