On Tue, 14 Aug 2007 09:53:31 -0700, K7ITM wrote:

Bingo. It's that ratiometric thing that is a big plus for stability.
In a coupler made of all the same metal, or at least metals that have
nearly equal coefficients of expansion, the ratios stay the same, and
it's the dimensional ratios that establish the coupling and
impedances, not the absolute size. Actually, the change in length
does matter, but if you make the assembly a quarter wave long, the
d(coupling)/d(length) is zero at that point anyway. In any event, I
suppose the thermal coefficient of expansion of metals you'd be most
likely to use is small enough that you'd be fine with a shorter
coupler. There doesn't need to be anything terribly complex about the
geometry of the whole thing, either. It's probably safe to say that
changes in the dielectric constant of air due to air pressure and
humidity aren't going to be significant in this case. ;-)

Cheers,
Tom
Tom, I thought this thread concerned measurement of attenuation in transmission lines. On the 11th I posted a
precedure that involves measuring the line input impedances with the line terminated in both a short circuit
and an open circuit, then plugging the measured data into a BASIC program that outputs the attenuation,
complex Zo, and electrical length. My thoughts were that this procedure gives results with more accuracy and
precision than the procedures discussed before my post appeared.

However, I noticed that my post drew zero response. Is my procedure out-of-line, or out dated?

Walt, W2DU