Twisted Pair vs. Coax

[EMT]

In reading the specs for analog audio cables, I see many cables with twisted pair geometry, typically with one of the pair for signal, the other ground and the shield connected to ground at only one end of the cable. However, I've also read that coaxial construction is suitable for analog audio.

Is there any benefit in using one over the other for singled-end RCA audio connections? Comments from you technical experts would be greatly appreciated.

[JonMarsh]

You'll find a lot of opinions about this. I'm sure Lex can chime in with some insights, also.

In general, for audio, even single ended connections, I prefer a twist pair or twisted quad approach (quad using an interleave braiding, similar to Kimber, which cuts inductance and increases CMRR), with an external shield which is only grounded at one end. If a sheild is doing it's job, and the cable is routed in a "EMI dirty" area, there is the potential to have induced sheild current in the shield which then shows up as a relative voltage difference on the input hot. For Video, with a typical maximum S/N of 55-65 dB, this isn't a big deal, and the characteristic impedance uniformity of coax makes it the choice. For Audio, S/N is more important, and I think the benifts of not having the shield connected at both ends as the low side signal connection are clear.

As Obi Wan would say, "trust your feelings, Luke", and trust what you hear in your system when comparing cables.... theory only goes so far.

Best regards,

Jon




Earth First!
_______________________________
We'll screw up the other planets later....

[Lex]

Good to see you EMT.

Great post Jon, as usual. you do know it gets rather boring, all those quality posts. Just once, I'd like to see you say, well, I just don't know.

As to my explanation on this topic. I'd rather see it covered at "the cable forum", www.thecableforum.com, as that site just isn't cooking at all yet.

But it's also part of what we do here, so the reason I think you hear more positives about twisted pair, other than possible "number" differences by two wires being twisted, is that the best audio cables have no shielding at all, it can simply sounds better. But as the next best thing, is the shielding connected just at the source. Of course, do that with coax would be difficult, unless your running 2 runs of coax, as some of my designs have 2 runs of RG-59. So, twisted pair with shielding lends itself well to that design of connecting the shield only at the source. That means any noise picked up, basically falls off the signal at the source. So, from that perspective, twisted pair is a preferred design over standard single run coax.

Lex

[RedStep]

Signal traveling on cable is influenced by two major factors, internal and external. Major fundamental internal factors are those of electrostatic/electromagnetic field interplay and electron migration. I believe the theory behind twisted cable is that it benefits in this area, where coax would not!




RedStep
We are truly a product of the decisions we make

[Lex]

Yeah, unless the coax is twisted. :LOL:

Lex