On 9/11/2012 1:28 AM, Tom W8JI wrote:
but the numbers are so small at HF for normal good cables they are
meaningless.
Not quite. Some measured Vf numbers for Commscope 3227 (#10 solid
copper center, dimensions and shield like LMR400) are:0.8 at 194.5 kHz,
0.816 at about 900 kHz, 0.829 at 1.8 MHz, 0.836 at 4 MHz, 0.841 at 10
MHz, 0.845 at 53 MHz where I stopped measuring, and Vf was STILL
increasing with frequency, although not by much. I'm reading those
numbers from the graph I plotted of the data, which is part of
http://audiosystemsgroup.com/Coax-Stubs.pdf
The MEASURED difference in Vf between 1.8 MHz and 53 MHz was 1.93%,
which equates to an error of 3.5 electrical degrees for a 180 degree
stub. The numerical error that Tom pointed out is more than twice as
large -- about 5%, or 9 electrical degrees. The difference between 160M
and 80M corresponds to a 30 kHz error in a quarter wave stub cut for
160M. This sort of error DOES matter if you're trimming a quarter-wave
or half-wave stub to reject harmonics or sub-harmonics of your operating
frequency in a contest station -- the bottom 70 kHz of 40M is only a 1%
change in frequency, but if you've built your stubs from beefy coax,
they're narrow enough that the deepest part of their null misses the
intended target! More measured data for real stubs in that pdf.
None of these errors is very large, but if you spend some time modeling
real antennas in software like SimSmith, you will see that the length of
a 75 ohm line can do some very NICE things to broadband and/or improve
the match of an antenna if it is the "right" length, or degrade the
match if it is even a bit different.
BTW -- SimSmith is FREE modeling software written by AE6TY that runs in
Java.
http://www.ae6ty.com/Smith_Charts.html
You need to feed it impedance data, either from an NEC design, or Vector
Impedance measurements of a real antenna. It is VERY well written, VERY
user friendly, quite powerful, and as near as I can tell, seems to be
giving good answers.
How good are the answers? Starting with measured data using the DG8SAQ
VNWA, over the past few months, I've done a major rework of my antenna
matching system, replacing a collection of antenna tuners with stubs and
matching sections, all switched with Top Ten relay boxes. My antenna
farm includes fan dipoles for 80 and 40, three phased verticals for
160M, a 40M wire Yagi, a 40M dipole, monoband Yagis for 20, 15, and 10M,
and a SteppIR. All of those antennas are now matched to 1.5:1 or better
over the contesting segments of all bands. All using SimSmith for
design, and the results agreed with the designs.
73, Jim K9YC
_______________________________________________
UR RST IS ... ... ..9 QSB QSB - hw? BK
|