TWISTS & TERMS

by R. A. GREINER

Recently, I got a letter with some irate comments about an August 1989 article on loudspeaker cables that I wrote for Audio. The letter called me to task for saying that the inductance of cables increases as the two wires in a cable pair are moved further apart. I was taken aback by this comment since this result is very well known to anyone who knows anything about electromagnetics, and I never expected the conclusion to be questioned.

With careful reading of the complaint, however, I realized that the claimant was talking about the "mutual inductance" between two pairs of wires and not about the "self-inductance" of a single two-wire cable. To confuse these two types of inductance is a serious misunderstanding, and it is very important in considering cable applications to recognize the difference.

Self-inductance exists in a single cable that consists of two conductors carrying a current to a load (and back, of course). This is the situation for a regular loudspeaker cable (two wires) that has an amplifier at one end and a loudspeaker at the other. It is this self-inductance that is important when considering the parameters of the cable connecting the amplifier to the load; it is also the self-inductance that was under consideration in the Audio article.

There is a second type of inductance that exists between two cables carrying currents. This is the mutual inductance between the two cables, each of which consists of a pair of wires carrying current, generally to two different loads. This mutual inductance results in an interaction between the currents in the two cables. For example, the signals to the left and right loudspeakers in an ordinary stereo system should be independent of each other. If there is significant mutual inductance between the cables, there will be crosstalk. Since the left- and right-channel cables are usually separated by a distance somewhat greater than the wire separation within the cable, the mutual inductance is usually small. But the mutual inductance between two current-carrying cables might be significant in some cases.

You can easily imagine a system that has two cables carrying the left and right channels in a stereo system, or even more pairs of wires carrying the high- and low-frequency signals in a biamplified system, or a case where the loudspeaker cables and some power-line cables lie parallel to each other. It is informative to consider what happens when several runs of cable are laid down parallel to each other for a considerable distance.

Parallel pairs of wires, when tightly spaced physically, can easily have significant mutual inductance; thus, some signal from one circuit might leak into the other. This interaction, cross coupling, or leakage between the two circuits can have some minor effect on the stereo signal in each or, at worst, might cause some audible crosstalk interference between the signals.

Normally we would hope to have isolation between the signals in the wires carrying the information to the left and right loudspeakers. Because the left and right channels in normal program material are substantially correlated, such crosstalk is usually not serious.

But in a case where the loudspeaker cable interacts with a nearby power line, we might find hum introduced into the loudspeaker cable. While all of this crosstalk and stray pickup can happen, these problems can be minimized by reducing the mutual inductance between the circuits. Fortunately this can be done easily.

First, we must understand the source of mutual inductance. Mutual inductance is the coupling phenomenon that causes currents in one circuit to result from currents in another, nearby circuit. For example, if you have a rather long cable consisting of a pair of wires stretched out in a straight line and then you place a second pair of wires down near and parallel to the first, there will be mutual coupling between the pairs. That is, some of the flux from each pair of wires will cut the current loop of the other. This coupling is specified as the mutual inductance between the pairs. The interaction becomes greater when the wire pairs are closer together and is reduced when they are placed farther apart.

To reduce coupling between parallel wires carrying independent signals, the wires should not be bound together in a bundle. This seems like a rather strict requirement; in fact, straight parallel cables should not even be laid down next to each other.

There is a very easy way to reduce the mutual inductance or mutual interaction between cables. It happens that mutual inductance has a sign (polarity) associated with it. If the wire pairs are alternately interchanged in position, the mutual inductance can be made to cancel almost completely. This is why professional loudspeaker cables are always twisted pairs of wires. It is the reason that telephone companies use twisted pairs to transmit telephone signals and why they interchange wires, from left to right, when running long lines parallel to power lines along a roadway.

Once we recognize the problem, a solution is easy. The rules are quite simple. If cables are run next to each other, they should consist of twisted pairs of wires. All professional loudspeaker cables are designed with just such construction.

But what about the ubiquitous zip cord used by so many consumers? All zip-cord-type cables, that is, cables having two parallel wires molded in a jacket holding them side by side, have the potential for interaction with similar cables laid next to them. What is the solution? The flat parallel zip cords must be twisted in order to reduce the interaction between them as well as to reduce the interference from external fields.

--------

If cables are run next to each other, they should consist of twisted pairs of wires; all pro speaker cables are made this way.

---------

This simple tactic will reduce crosstalk between cables and will also serve to keep out power-line radiation and radio interference of many types. We must realize that we are immersed in a conglomeration of low- and high-frequency radio waves, power-line radiation, and who knows what other electrical interference. Twisted pairs will save the day. It is usually more important to use twisted pairs at low power levels, which is, of course, the reason why professional microphone cable consists of twisted, shielded pairs and why telephone wires are usually twisted pairs as well.

It is for these reasons that I do not recommend wire with zip cord construction, no matter what the gauge, for loudspeaker applications. But I do highly recommend the use of professional-quality, twisted-pair loudspeaker cable. The gauge selected should be large enough so that the cable has a resistance that is less than a small percentage of the lowest resistance presented by the speaker.

(adapted from Audio magazine, Jan. 1992)

= = = =