I'm trying to figure out how radio frequency "matching stubs" work. In order to fully understand the problem, I need to know how the "curve of equal SWR" looks like.

I did a few plots, and it looks like it's a circle, or something close.

Whenever I approach the matter analytically, the equations explode and become very ugly.

I hope there's an easy way to prove that the following equations describes a circle, or ellipse on the complex plane:

$$VSWR = \frac{1+|\Gamma|}{1-|\Gamma|} $$

where VSWR is a constant, and $\Gamma$ (capital gamma) is the reflection coefficient:

$$\Gamma = \frac{Z_{load}-Z_{line}}{Z_{load}+Z_{line}}$$

$Z_{line}$ is a constant, normally a real number, actually equals to 50 Ohms in my specific case. (However, it would be interesting to see what happens if this can be a complex value, but this is really unimportant right now).

$Z_{load}$ is the variable, I suspect that these can be found on a circle or ellipse on the complex plane.

So can you girls/guys see an easy way to prove that $Z_{load}$ values are on a circle or ellipse? I do not seek absolute precise mathematical proof (it would be nice though), I just want to be "reasonably sure" that this is ture.

Thank you very much,



Your variable $Z_{load}$ does indeed lie on a circle, provided that $VSWR>1$.

I'll use more typical mathematical names for the unknowns, since most here are not into radio frequency terminology. So let $$z=Z_{load},\qquad z_0=Z_{line}, \qquad C=VSWR$$ Then your equation is given by $${1+\Bigl|{z-z_0\over z+z_0}\Bigr| \over 1-\Bigl|{z-z_0\over z+z_0}\Bigr|}=C$$ A little algebra turns this into $$\Bigl|{z-z_0\over z+z_0}\Bigr|={C-1\over C+1}$$ The map $w=\phi(z)={z-z_0\over z+z_0}$ is a linear fractional transformation, and both $\phi$ and $\phi^{-1}$ will map circles to circles (where some circles may degenerate into lines). Since the curve you are interested in consists of exactly those $z$ which are mapped to the circle $|w|={C-1\over C+1}$, your curve will be a circle.

  • $\begingroup$ Thank you! VSWR is indeed > 1 in my case. If VSWR = 1, then it is a "perfect match" in rf terminology, and there's no need to use matching stubs (or any kind of matching).In this case the "curve" is a single point, since $\Gamma$ - the reflection coefficien - is zero (no reflection at all). If the match is perfect, $z_o = z$, and all rf energy is transferred to the terminating load, no reflection indeed. I'll review the solutions and decide wich one to accept. Thanks again! $\endgroup$ – netom Nov 20 '12 at 10:00
  • $\begingroup$ A more complete answer, nice 8) $\endgroup$ – Alexei Averchenko Nov 20 '12 at 10:14

Your equation uniquely determines $|\Gamma|$, which is real positive, and complex numbers with length $|\Gamma|$ form a circle of radius $|\Gamma|$.

The details: let $x = |\Gamma|$. We have the equation of the form $$C = \frac{1 + x}{1 - x}.$$ When $x \neq 1$ it is equivalent to $C - Cx = 1 + x$, or $(1 + C)x = C - 1$, or $$x = \frac{C - 1}{C + 1}.$$ It is positive whenever $C > 1$. When $x = 1$, the equation makes no sense, but we can still ask what happens when $x$ is arbitrarily close to $1$. $\lim_{x \to 1} \frac{1 + x}{1 - x}$ does not exist because it approaches $+\infty$ or $-\infty$ depending on where you are approaching $1$ from. You can still say it's the infinity in the projective line sense (in that it's infinitely far away from any real number), but I doubt it's useful in your case.

  • $\begingroup$ Good observation with $x = 1$. $|\Gamma|$ is the so-called "reflection coefficient", and it determines how much energy is reflected from a poorly matched load at the end of a transmission line (like an antenna tied to a coaxial cable). When $|\Gamma| = 1$, it means that all power is being reflected, and we say that the SWR is "infinite" (yes, in the projective sense as you've said), which is a terrible situation. :) In this case, no matching stub will help, so we do not have to worry about that. Thanks again, I'll review the answers and choose one to accept. $\endgroup$ – netom Nov 20 '12 at 10:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.