Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!think.com!eplunix!raoul
From: raoul@eplunix.UUCP (Nico Garcia)
Newsgroups: sci.electronics
Subject: Re: Impedence matching...
Message-ID: <1051@eplunix.UUCP>
Date: 29 Mar 91 22:30:32 GMT
References: <1991Mar29.150314.23177@e2big.mko.dec.com> <2532@umriscc.isc.umr.edu>
Organization: Eaton-Peabody Lab, Boston, MA
Lines: 36

In article <2532@umriscc.isc.umr.edu>, robf@mcs213k.cs.umr.edu (Rob Fugina) writes:
> I'm confused as to the importance of impedance matching...  Since I may
> not know the whole scope of this subject, I'll say I'm referring to things
> like stereo line-level, headphone, microphone, antenna inputs and outputs.

There are two factors here: one is a problem of power dissipation, the
other is reflection and transmission problems.

For a given voltage source, with a series resistance before the load, the
maximum power dissipation in the load occurs when both the load and the
line have the same resistance. Try the equations yourself if you don't
believe me.

However, real circuits have parallel capacitance and series inductance, the
characteristics of a transmission line.  These yield an impedance of
sqrt(L/C), which describes the relationship between high frequency voltage
and current or between the transmission line's electric and magnetic fields.
It doesn't *dissipate* power, but it does describe how a transmission line
carries power/signal to your target.  Now, if you connect this circuit to
another one with a different impedance, it has a different relationship, but
they both have to match current and voltage at the junction. Nature cleverly
matches this boundary condition by reflecting another signal back down the
original circuit.  This reflection can bounce back and forth along a real
circuit, causing serious signal grief in the process. The only way to avoid
it completely for a wide range of frequencies is to match these impedances.
You can avoid this for a pre-specified set of frequencies by matching the
length of an intermediate transmission line to fractions of the wavelengths,
but this is a serious pain in the keister. It's much easier to use all 50
ohm or 75 ohm cable and devices.

Does that answer your question?

-- 
			Nico Garcia
			Designs by Geniuses for use by Idiots
			eplunix!cirl!raoul@eddie.mit.edu