On 28 Apr 2006 09:42:16 -0700, "K7ITM" wrote:

-- If you are unaware of the errors at low frequencies, it may well not
make much difference whether you are using the Smith chart or not.
Those errors may still bite you.

-- If your computer isn't powered up and you have a quick calc to do,
the "small computer program" may very well far exceed the Smith chart
in user time!

-- Though a program may provide more accurate calculations for the
given input values than a Smith chart, it would be unusual that you
KNOW the impedances, line lengths, inductances, etc. that you are
working with to the accuracy to which they can be plotted on a
decent-size Smith chart. It's easy to plot things to within 0.01
reflection coefficient units on such a chart, and it's very likely that
the "50 ohm" coax you are using will be further away from 50 ohms than
that. To quote some famous guru, 'Have you never heard of the word
"approximation"?'

-- For those of us who appreciate that sort of graphical
representation, a Smith chart can be a wonderful visualization tool.
To me, that's its strongest point. Forget using it to DO the
calculations; let a computer program or a vector network analyzer do
the calcs or make the measurements. It's still worthwhile to me to
have the results presented on the Smith chart overlay of complex
reflection coefficient. (Note that a Smith chart is 'just' grid lines
on top of a linear graph of reflection coefficient referenced to the
normalization impedance you're using.) And of course it's useful for
far more than just transmission lines.

For those who always have an appropriate calculating engine at hand and
don't get anything out of a graphical representation of the results,
toss the Smith chart out the window. But don't assume it's not still
_very_ useful to those who see things easily from the graphics. Can
YOU look at a table of 100 values of complex reflection coefficient
versus frequency for a component and summarize them in your mind as
rapidly as you can by looking at the same values plotted on a Smith
chart? If you can, that's great. I can't. Thank you, Mr. Smith.

Right on.

And another handy feature is the ability to see what elements of a
matching network are doing. Lines of constant Q drawn on the chart
can also keep one out of trouble.

I remember one thread about 3 years ago wherein someone (he knows who
he is) stated that fewer matching elements -always- provide a lower
loss solution than more elements do. A few minutes with a Smith chart
with some constant Q circles and I found an example where that is not
the case.