[gphall]

by: brucek

A speaker connection is a very low impedance circuit and as such is not a candidate for crosstalk problems. Notice that you don't need any shield on these cables? It takes a very large signal to induce noise in a speaker wire connection and isn't susceptible to outside interference. You could arc weld over a set of speaker cables and not have to worry........

While I agree with you that a speaker connection is a fairly low impedance circuit, there is ALWAYS crosstalk taking place between adjacent transmission lines. The level of crosstalk is dependant upon the specific geometry of the circuit and the material properties that make up that circuit. The geometry aspect of crosstalk manifests as a dependance of the distance between adjacent circuits. The farther away any nearby current carrying conductor, the less crosstalk there will be in the victim. This is because the mutual capacitance and mutual inductance between the two circuits falls of VERY rapidly with distance. For the conduit to be an effective shield, it would need to have a very low impedance connection to the equipment chassis at both ends. This is not going to happen, so conduits reduce crosstalk by forcing a minimum amount of spacing between your quite circuit and any other circuit.

As an example, try this experiment. Using your favorite low impedance circuit, seperate the positve and negative wires and run the positve over the top of a florencent light and the negative under the bottom. NASTY. You haven't changed the impedance of the cicuit, but you sure will have a crapload of noise induced in it.

The key to keeping noise out of audio lines is rarely related to any shielding effects. Instead, it is done by using balanced cables that are in very close proximity to each other and far away from other circuits. Far in this case may be only a few inches in this case.

Think about phone lines. These carry audio signals long and far and right near some of the nastiest power lines you could dream of. They are not low impedance circuits, but they (usually) have very little noise induced on them.

My point is that crosstalk always happens, low impedance , high impedance, doesn't matter. What the impedance of the circuit does is convert the induced currents from nearby changing magnetic fields into a voltage... that's it. A low impedance circuit will not convert those currents into as large a voltage as a higher impedance circuit, but the driver isn't responding to circuit voltage... it's a current (Read Magnetic) responsive transducer.