> On Sun, 28 Feb 2010, Mike Morris WA6ILQ wrote:
> > 2) the pair that was receiving the problem was a high
> impedance load
> > or an unbalanced load (i.e. one side grounded). Use an
> ungrounded 600
> > ohm winding from a transformer on each end of each pair.
>
> The nominal impedance of a copper pair from CAT5 is 110-ohms,
> not 600.
Actually, unless something has changed or I'm totally losing it, balanced
Ethernet over Cat3 or higher is 100 ohms. AES3 (aka AES/EBU) digital audio
over balanced lines is 110 ohms. IIRC, the tolerance for Ethernet is +/-
15% and for AES it's +/- 20%. AES/EBU can be run over Cat5/5e/6/7 cabling,
typically with excellent results.
Back to the analog world. Unless you're going long distances, the
characteristic transmission line impedance doesn't have a very big effect at
audio frequencies. It's not until you get up to a significant fraction of a
wavelength that it starts to "act like a transmission line". At small
fractions of a wavelength, there's no "room" for standing waves to form, so
current and voltage are, for all intents and purposes, always in-phase. At
20 kHz, one wavelength is somewhere around 10 miles (ignoring VF).
That's not to say that there's no advantage to using "the right cable" at
audio frequencies. Twist rate, capacitance (both between conductors as well
as to the shield), gauge/resistive loss, etc. all have an effect, good and
bad, even down in the AF range.
--- Jeff WN3A
