The Physics of Electricity

06-24-11: Johnnyb53
Ideally they must convey all the various signals (which are minute voltage fluctuations) in time with one another (with the challenge that signal speed varies with frequency), keep the amplitude of each accurate, minimize (ideally prevent) signals from being absorbed by the dielectric to be released back into the signal path later, and have near perfect rise time (which requires frequency response somewhere past 1 GHz).

Johnnyb53, while your well written post is, IMO, completely correct from a qualitative standpoint (i.e., if quantitative considerations are ignored), as I've mentioned in response to similar posts you've made in the past it is incorrect and misleading quantitatively.

Risetime need only be fast enough to not have audible effects within the frequency range of human hearing. As you know, that is nominally considered to be 20kHz, but even if we apply a huge margin to that figure, 1GHz is still thousands of times faster than necessary.

Furthermore, it is thousands of times faster than the bandwidths of the speakers, the source material, the source component, most electronic components, as well as our ears, so even if cable bandwidth were vastly lower than 1GHz, and cable risetime correspondingly slower, it would not affect overall system bandwidth and risetime to more than an infinitesimal degree.

Finally, I would maintain that for reasonable cable lengths the claimed need for time alignment is speculative at best, as even the audible frequency having the slowest propagation speed (20Hz) still propagates through wires at approximately 5 million meters per second (with the propagation speed of higher frequencies approaching the speed of light). See Figure 2 of this reference.

Regards,
-- Al

06-24-11: Almarg
Johnnyb53, while your well written post is, IMO, completely correct from a qualitative standpoint (i.e., if quantitative considerations are ignored), as I've mentioned in response to similar posts you've made in the past it is incorrect and misleading quantitatively.

Qualitative was the nature of the OP's original question. He was unaware of the qualitative functional difference between a refrigerator power cord and interconnects between a preamp and amplifier. You don't have to agree with how important the difference is, but the purpose and functions of the two kinds of wires is fundamentally different.

As for calling me "incorrect," that's rather presumptuous for what is a matter of opinion supported by research on both sides. Some very bright lights in audio base their designs on ultrawide bandwidth, including Soulution, Symphonic Line, Odyssey Audio, Spectral Audio, Kimber Cable, Zu Audio, and many others. If you disagree with them, write a letter and set them straight.

Test reports that chart frequency response and square wave response show that risetime is visibly slower on amps that start rolling off at about 20KHz. For example:

-1dB @ 20KHz, visibly very rounded risetime

wide bandwidth, fast (nearly vertical) risetime

06-24-11: Johnnyb53
Some very bright lights in audio base their designs on ultrawide bandwidth ...

In the case of electronic components having ultrawide bandwidth, such as the Spectral products, that means a few MHz, not anything remotely approaching 1GHz. And the basic rationale for that wide bandwidth is, presumably, to avoid undesirable sonic effects that are within the bandwidth of our hearing, which is far less than 1MHz.

If a cable having a bandwidth in the GHz range sounds different than a cable having a lower bandwidth, then given the bandwidth limitations of the speakers, the source material, the source component, the other electronics in the system, and our hearing, the reason for that sonic difference is not the cable's GHz bandwidth.

Test reports that chart frequency response and square wave response show that risetime is visibly slower on amps that start rolling off at about 20KHz.

Yes, of course. I never said anything to the contrary. As I know you realize, bandwidth and risetime have an inverse relationship. Therefore a frequency rolloff that begins too close to the audible frequency range will result in risetimes that are visibly slow when viewed on a time scale that is meaningful with respect to the bandwidth and risetime of our hearing mechanisms.

Regards,
-- Al

So it would seem, but here is some food for thought. Audio is all about signal transfer. Is it cost-effective to spend crazy money on components and then lose signal speed and coherency to budget wires between the components? Could it be more cost-effective to get spend less on the components and then get better cables so that more signal arrives intact at your speakers?