"Richard Clark" wrote

Tell me, please, this is the FIRST time you have read our master
fence sitter (in the American historical context otherwise known as a
Mugwump: someone who has his mug on one side of the fence and
his wump on the other).
________________

No, which is the prime reason I chose to confront him.

- RF

On Mon, 17 May 2004 15:02:19 -0500, "Richard Fry"
wrote:

"Reg Edwards" wrote
No need to make measurements. If you are unable to make a simple
calculation
and answer the question then you are not qualified to continue the
discussion.
__________________

How convenient. Why _would_ you want to make measurements, and prove
yourself wrong?

Ah Richard!

Confrontation? Feeble at best when I doubt you could even get his
recipe for RF Mud. ;-)

Poor Reggie can quote Dead White Scientists like Lord Kelvinator with
high dudgeon and still ignore the tenet of the plagiarized message.

73's
Richard Clark, KB7QHC

Reg Edwards wrote:
Richard,

A radio amateur, by an easy mistake, uses a 75-ohm plug and socket in a
50-ohm coaxial transmission system. The total length of the plug plus
socket is 1"

As a result of the mismatch what is the SWR produced on the 50-ohm line at 2
MHz. At 30 MHz? At 150 MHz?

Is the amateur, or anyone else, likely to be aware of any difference in
performance?


The practical answer for amateurs is somewhere between the two extreme
positions that Richard and Reg are taking.

Richard quotes a case where even very small impedance bumps do matter;
but it's in full-quality TV broadcasting, not amateur radio.

Reg, on the other hand, wants to dumb it down too far. There *are* cases
in amateur radio where small impedance bumps are at least noticeable.

At 2MHz or 30MHz, the effect is so small that no amateur would notice
it. Even using professional test equipment, you'd be hard-pressed to
measure the effect of a single connector of the wrong impedance.

At 150(144)MHz, even a single connector is noticeable... but that's not
the problem.

The real problem is that if people believe a simple slogan like
"connector impedances don't matter", they will probably go ahead and use
*several* mismatched connectors, at various places along the line.

Then they start to find bewildering problems at 144MHz and above, such
as indicated SWR and power output values that vary according to the
length of the coax jumpers that they use. It still may not matter in
terms of the contacts they can make, but they are completely unable to
understand what is happening - and that *does* matter!

(What is happening, by the way, is that the lengths of the line sections
between the mismatched connectors will determine how the small
reflections from each one combine together. If you're lucky with the
line lengths, they may tend to cancel; if you're not, they may tend to
add... and usually it's somewhere in between.)


No need to make measurements.

In this particular case, that's true. When the impedance bump is small,
it is easier and more accurate to calculate the effect than to measure
it.


--
73 from Ian G3SEK
'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

Let's do what nobody has ever done before, not even in the ARRL handbooks or
in Terman, and get an idea of the magnitudes involved.

Examine two cases over a range of frequencies.

Case (1). In a 50-ohm system, use of a poor connector having an impedance
deviating 10 percent from its nominal value of 50 ohms.

Case (2). Making the mistake of using a 75-ohm connector in a 50-ohm
system.

In both cases the connector, plug and socket, is 1" (25.4mm) long.

We first calculate the input impedance of a 75-ohm transmission line, 1"
long, terminated with 50 ohms. Zin will not be very much different from 50
ohms.

We then calculate the SWR on a 50-ohm line which is terminated by the
afore-mentioned input impedance.

RESULTS of calculation

MHz SWR Case 1 SWR Case 2
------ ----------------- ----------------
2 1.0002 1.0009
30 1.0028 1.0146
150 1.014 1.073
300 1.029 1.145
1000 1.105 1.524

It is seen that results do not become significant to a radio amateur, and
almost everybody else, until he has made the serious mistake of using the
wrong impedance connector, and the frequency has risen to 1000 MHz for which
he hasn't an SWR meter anyway.

Below 300 MHz the results are submerged well beneath the uncertainty of an
SWR meter. Now we can take a balanced view of the situation.
----
Reg, G4FGQ

"Reg Edwards" wrote in message
...
Let's do what nobody has ever done before, not even in the ARRL handbooks
or
in Terman, and get an idea of the magnitudes involved.

Examine two cases over a range of frequencies.

Case (1). In a 50-ohm system, use of a poor connector having an impedance
deviating 10 percent from its nominal value of 50 ohms.

Case (2). Making the mistake of using a 75-ohm connector in a 50-ohm
system.

In both cases the connector, plug and socket, is 1" (25.4mm) long.

We first calculate the input impedance of a 75-ohm transmission line, 1"
long, terminated with 50 ohms. Zin will not be very much different from
50
ohms.

We then calculate the SWR on a 50-ohm line which is terminated by the
afore-mentioned input impedance.

RESULTS of calculation

MHz SWR Case 1 SWR Case 2
------ ----------------- ----------------
2 1.0002 1.0009
30 1.0028 1.0146
150 1.014 1.073
300 1.029 1.145
1000 1.105 1.524

It is seen that results do not become significant to a radio amateur, and
almost everybody else, until he has made the serious mistake of using the
wrong impedance connector, and the frequency has risen to 1000 MHz for
which
he hasn't an SWR meter anyway.

Below 300 MHz the results are submerged well beneath the uncertainty of an
SWR meter. Now we can take a balanced view of the situation.
----
Reg, G4FGQ


Reg,

I suspect your 1 inch length is overly pessimistic. Clearly, for an F
connector, it is more like 1/2 inch. Lastly, if the load at the end is not
50.0, then any small deviation in the feedline could just as well improve
things, as make it worse; this is probably not true for pulses or video.
Think of an antenna tuner. It does nothing to the SWR on the main piece of
line.

Tam/WB2TT

Reg,

Your math example does not illustrate the reality that the impedance
change produced by a ~5/8" OD brass pin inserted radially into the
dielectric space of coaxial transmission line can match that line into an
adjacent termination of fairly high SWR (1.3:1 or so), at frequencies as
low as 54MHz.

As I wrote earlier, this technique is widely and successfully used to match
the main transmission line of broadcast TV and FM stations to the net input
impedance of their antenna (including its input elbows).

- RF

Your math example does not illustrate the reality that the impedance
change produced by a ~5/8" OD brass pin inserted radially into the
dielectric space of coaxial transmission line can match that line into an
adjacent termination of fairly high SWR (1.3:1 or so), at frequencies as
low as 54MHz.

=============================

It's not intended to.

But I'm sure you are right. Try not to worry about it.

Wait a minute here... Seems to me that Reg is talking about two
opposing steps separated by a very short distance, and Richard is
talking about a single step from one impedance to another. Clearly,
Richard's case results in uncancelled echos related to the ratio of
the impedances at the step. In fact, Reg's example results in
significant uncancelled echos if the steps are separated by enough
distance: worst at odd multiples of pi/2 electrical degrees. When the
connector is only perhaps a few electrical degrees long, the steps
nearly cancel. How long is 1", in electrical degrees? Well, at
1000MHz and a v.f. of perhaps 0.7, longer than one might have
anticipated: almost 45 degrees long. So in fact at 1000MHz, a 1"
section of 51.5 ohm connector might introduce a bit more than 1.03:1
SWR. But didn't we start out talking about a much lower frequency,
and a much larger impedance difference? Another example I ran was 4
electrical degrees of connector...perhaps a bit less than an inch at
150MHz...where it's 75 ohms in a 50 ohm system. The swr for that came
out about 1.06:1.

The TV broadcast engineer should worry about things like that. The
typical ham doesn't have equipment calibrated accurately enough that
s/he should be worried about it, and it's unlikely to make any
substantive difference anyway in typical ham work.

Cheers,
Tom

"Richard Fry" wrote in message ...
"Reg Edwards" wrote
Richard, I have no reason to doubt anything you say.
I have no reason to doubt what I have said.
But why imply there's serious disagreement when there is none?
_______________________

Clips from our previous posts (below) -

I don't know what constitutes a "serious disagreement" to you, Reg, but I
think most readers would say that we have opposite conclusions about this
topic.

If you have no reason to doubt what I wrote about this, how can you continue
to support what you wrote? Our statements are mutually exclusive.

YOU: "Provided the mechanical connection is sound, you can use any coaxial
connectors you like, regardless of nominal impedance, at frequencies less
than about 300 MHz without any observed ill effects. A connector less than
1" long of impedance 51.5 ohms in a 50 ohm system will NOT produce an SWR of
1.03:1 or anything anywhere near to it at frequencies less than 1000 MHz."

ME: "This PRACTICAL experience illustrates that an impedance change even in
a 54-60MHz TV channel occurring within a physical space of less than one
inch can produce important and commercially supportable system benefits,
despite your statements quoted above."

- RF

I won't "worry about it," as long as you refrain from posting absolute
statements about this that are demonstrably untrue.

- RF


"Reg Edwards" wrote in message
...

Your math example does not illustrate the reality that the
impedance
change produced by a ~5/8" OD brass pin inserted radially into the
dielectric space of coaxial transmission line can match that line into
an
adjacent termination of fairly high SWR (1.3:1 or so), at frequencies
as
low as 54MHz.

=============================

It's not intended to.

But I'm sure you are right. Try not to worry about it.

Below is my response to an off-list e-mail I received on this topic, which
to some readers may add additional perspective on this topic.

RF

Visit http://rfry.org for FM broadcast RF system papers.
_____________________________

Thanks for your comments.

The reason I chimed in originally was because of Reg's absolute statement
that preserving system Z throughout the transmission line and its
connectors was unimportant below 1GHz (later 300MHz?). He said further
that an impedance change across a 1" length of transmission line was
irrelevant in that same spectrum.

That is demonstrably untrue, as I pointed out by a real-world example taken
from the broadcast industry.

The net terminating Z of a TV/FM transmit antenna with its input elbows
usually is not exactly 50 ohms, even though that is the impedance value
that the hardware was designed to provide. Manufacturing, assembly and
installation issues can and do change it. The main transmission line
normally is closer to 50 ohms across the relevant bandwidth than the
antenna/elbow termination connected to the far end of that line.

The Z-matching hardware I described is effective at optimizing this far-end
match, it _does_ improve the quality of the radiated signal, and it
minimizes the stress on the main transmission line and the transmitter.
And it does so by changing the impedance in a 1" or less length of
transmission line adjacent to the antenna input, regardless of Reg's
beliefs.

Hams operating on the HF bands may not wish or even need to consider this,
as I mentioned in an earlier post to this thread. But that is not license
for Reg or anyone else to write that such disregard is universally
justified.

- RF