Inveterate meddling #492: Amp output Z and speaker cables

Bored of the holy wars on whether or not cables affect sound we present one way in which a poor choice of speaker cables (or amplifier!) can significantly spoil the delivered frequency response in-room.

Using nothing but basic Physics here's a rather graphic illustration of the effect of cable resistance based on a simple model of an 8ohm, single driver speaker, like this:

The amplifier is a perfect voltage source; the speaker cable is R1, a pure resistance; the 'dB' probe is the response at the speaker +ve terminal. The speaker model comprises the voicecoil DC resistance and inductance, and some lumped terms which approximate the reactance of a driver in a sealed box. So, step the resistance of the cable from 0.1ohms to 5.6ohms (= the driver's DC resistance) and unsurprisingly you get up to 6dB of ripple:

Of course in practice this would be greatly complicated by the effect of the crossover, and more drivers; and by motional-impedance reflected from cabinet resonances at LF.

Observations:

  • 150mR in this case meets Quad's criterion of 5% of speaker impedance and yields a ripple in response of 0.2dB or less. Good sensible engineering. So keep your cables either or reasonable gauge or short or both - unless you want a particular effect (I'm looking at you, DNM...)
  • The principle effect of series resistance is to emphasise the bass peak / tuning of the speaker.
  • It doesn't matter whether the added reistance is from the cable, or inherent in the amplifier's output impedance.'Valve amps give great bass' anyone? Well, here's a very strong reason for much of that perception, especially when you realise some valve amps have source Z as high as 8ohms.
  • Conversely, a total source Z very much below below 0.1ohms - while nice to have - can make for little further improvement, so don't get too hung up on 'damping factor'. There are far more significant things to worry about.

    As a friend put it nicely, "...Not the whole story explaining amplifier differences by any means, but it pays to eliminate to obvious before embracing, how can I say this, more faith based explanations. "


    © 2009 the twisted pair

  • 25.02.10 Page added