On Tue, 08 Nov 2005 23:09:51 GMT, Cecil Moore wrote:

Owen Duffy wrote:
In my 100m of W551 with a 16+j0 load at 30MHz, the loss in one metre
of line nearest the load is over 4%, the good news is that since 75%
of the transmitter power is already lost, the weighted effect of that
4.3% is nearer 1% of tx output.

What the heck is one "metre"? Netscape says that is misspelled and
probably should be corrected to "metro". Why aren't you guys on
the English system?

Met The fundamental base of the metre is the quarter of the
terrestrial meridian, or the distance from the pole to equator, which
has been divided into ten millions of equal parts, one of which is of
the length of the metre.

I think we saw the light before the English, but I think they have a
partial metrication now.

If the loss in each meter is 4%, wouldn't the loss in 100 meters
be 400%? What am I missing?

I did not say "the loss in each meter is 4%", I said "the loss in one
metre of line nearest the load is over 4%".

Firstly, percentage losses on cascaded sections are not additive...
you know that. Losses multiply, dB losses add because adding exponents
is multiply the fundamental quantity.

As I have said before, you seem to be under the misconception that the
overall loss (ie Pin/Pout) per unit length of a transmission line
operating with VSWR1 is constant,

It is not necessarily a constant. It is for a lossless cable, and I
think it probably is for a distortionless cable... but I would have to
check that.

(It is true that the loss per unit length of a transmission line
operating with VSWR=1 is constant.)

We were discussing an example based on Wireman 551 ladder-line. The
dominant factor affecting loss at 30MHz is the series resistance
element. Does it make sense that since in that example, the magnitude
of the current varies by nearly 25:1 along the line, that the I**2*R
loss per unit length along the line is not constant, and will vary by
a factor approaching 625:1?

Owen
--

Owen Duffy wrote:
We were discussing an example based on Wireman 551 ladder-line. The
dominant factor affecting loss at 30MHz is the series resistance
element. Does it make sense that since in that example, the magnitude
of the current varies by nearly 25:1 along the line, that the I**2*R
loss per unit length along the line is not constant, and will vary by
a factor approaching 625:1?

25% of the power is delivered to the load. There are eleven current
maximum points in 100m on 10m. Does that 11% of the feedline really
contribute 59% of the losses? Does the remaining 89% of the feedline
really only contribute 41% of the losses?
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 09 Nov 2005 05:46:00 GMT, Cecil Moore wrote:

Owen Duffy wrote:
We were discussing an example based on Wireman 551 ladder-line. The
dominant factor affecting loss at 30MHz is the series resistance
element. Does it make sense that since in that example, the magnitude
of the current varies by nearly 25:1 along the line, that the I**2*R
loss per unit length along the line is not constant, and will vary by
a factor approaching 625:1?

25% of the power is delivered to the load. There are eleven current
maximum points in 100m on 10m. Does that 11% of the feedline really
contribute 59% of the losses? Does the remaining 89% of the feedline
really only contribute 41% of the losses?

When I have written about loss per unit length, I have implied "loss
at the rate of y per unit length". If you have tried to apply the 4+%
figure to one meter at each maximum, then you are unlikely to get any
meaningful results for a number of reasons. See the graph I just
posted (our posts crossed in the mail so to speak).

I haven't stated it it the post, but it should be obvious that the
rate of attenuation is the slope of the line in the plot referenced in
the post.

Owen
--

On Wed, 09 Nov 2005 03:25:47 GMT, Owen Duffy wrote:



We were discussing an example based on Wireman 551 ladder-line. The
dominant factor affecting loss at 30MHz is the series resistance
element. Does it make sense that since in that example, the magnitude
of the current varies by nearly 25:1 along the line, that the I**2*R
loss per unit length along the line is not constant, and will vary by
a factor approaching 625:1?

I knocked up a quick graph of the attenuation from point x to the load
for this scenario. (I think / hope it is correct!)

http://www.vk1od.net/lost/W551Example.htm

Is the shape of the curve (the cyclic variation over each electrical
half wave diminishing away from the load, and the general shape of the
curve a surprise? The effects plotted here might not be explained by
the ARRL charts.

Owen
--

Owen Duffy wrote:
http://www.vk1od.net/lost/W551Example.htm

Is the shape of the curve (the cyclic variation over each electrical
half wave diminishing away from the load, and the general shape of the
curve a surprise? The effects plotted here might not be explained by
the ARRL charts.

The graph is unclear. What does it mean that 6% loss occurs at
100 neters? Is that 6% loss per meter at the source? There's 4%
loss at 50 meters. Does that mean the average loss per meter is
4%? Where is the 4% loss in the meter closest to the load plotted?
--
73, Cecil http://www.qsl.net/w5dxp

On Wed, 09 Nov 2005 14:46:07 GMT, Cecil Moore wrote:

Owen Duffy wrote:
http://www.vk1od.net/lost/W551Example.htm

Is the shape of the curve (the cyclic variation over each electrical
half wave diminishing away from the load, and the general shape of the
curve a surprise? The effects plotted here might not be explained by
the ARRL charts.

The graph is unclear. What does it mean that 6% loss occurs at
100 neters? Is that 6% loss per meter at the source? There's 4%
loss at 50 meters. Does that mean the average loss per meter is
4%? Where is the 4% loss in the meter closest to the load plotted?

The loss scale is in dB, it is the loss in dB at position x metres
from the load.

If you examine the graph, you will find that the slope of the loss vs
position line is as high as about -22dB/100m at the load, it has a
minimum slope of close to 0dB/100m, and you can see that at large x,
the slope approaches the matched line loss of -1dB/100m.

(You find the -22dB/100m by using a ruler to scale off the slope.
-22dB/100m is -0.22dB/m, or 10**-0.022 which is 0.9506, which
corresponds to a loss of almost 5% in that one metre of line nearest
the load. These aren't mental gymnastics!)

You could calculate an average loss per meter figure, but I don't know
what you could you use it for? The fact that this line is not straight
(as some people seem to assume) means that working with average
numbers is inherently inaccurate.

Owen
--

Owen Duffy wrote:
The loss scale is in dB, it is the loss in dB at position x metres
from the load.

Aha, I see that noted at the bottom now that I scroll down.
It didn't make sense to me if the loss scale was in percent.
So what do I get if I integrate the area under the curve?

Incidentally, I was engaging in fuzzy republican thinking when
I came up with eleven current maximum points in 100 meters of
feedline. Of course, there are twice that, i.e. 22 current maximum
points which can be counted on your graph.

You could calculate an average loss per meter figure, but I don't know
what you could you use it for? The fact that this line is not straight
(as some people seem to assume) means that working with average
numbers is inherently inaccurate.

Owen, most hams are not rocket scientists like you :-) Quite
often, a rule-of-thumb average beats total ignorance. All
measurements contain errors and are inherently inaccurate.
Some of us live with that reality. Some of us rant and rave
about it. Next time you are on your motorcycle, note that
your speedometer is "inherently inaccurate" as is your gas
gauge as are your reflexes. If you are not inherently inaccurate
throwing darts in the local pub, you are a very unusual homo sapien.

I, for one, am satisfied with average losses, presumably averaged
over one half wavelength. The way I came up with that 25:1 limit
on my open-wire SWR is that 600/25 equals 24 ohms and that is an
acceptable impedance to my IC-256PRO's autotuner. Noting that the
losses in 100 ft. of open-wire line running at an SWR of 25:1 are
acceptable was an afterthought.
--
73, Cecil http://www.qsl.net/w5dxp