Digital Cable Question

I wrote in a post last week about breaking my Morrow Digital Cable by accident. Mike Morrow very kindly repaired the cord and mailed it back to me. While I was getting ready to replace the cable into my system, my Wife was looking at it and asked me "why can't you just use a regular interconnect cable for a digital cable? They both have the same connections". I opened my mouth to respond to her, and then realized that I have no idea what differentiates a digital cable from a standard RCA interconnect cable. Can someone please enlighten me?
The main difference is that an unbalanced digital cable is (presumably and hopefully) designed to have a "characteristic impedance" of 75 ohms, within a tight tolerance, while chances are that an analog interconnect will not be.

A digital audio signal has frequency components that extend well into the radio frequency region. At those frequencies, it is necessary that the characteristic impedance of the cable, the output impedance of the component driving the cable, and the input impedance of the destination component, all be the same (within a narrow tolerance). Otherwise VSWR effects, aka "transmission line effects," will result in reflections back and forth along the cable which will degrade waveform quality, causing increased jitter or even misclocking and data corruption.

That said, I would expect that some analog interconnects will work satisfactorily in some systems.

-- Al
In theory they shound be different but the reality maybe some thing else. Some brands, ASI Liveline and Kondo-Japan for example use the same analog cable for digital with great results. I happened to use the Liveline and it`s just a remarkable digital cable. Srajan Ebaen(6 Moons Editor) found this cable better than his long time reference, the Stealth Varidig Sextet(one man`s opinion). As with all things audio, listen and judge yourself,
75 ohms is the primary reason. The impedance match allows less reflections inside the cable and thus less 'jitter'
On another site a comment was made about length. At least one site offers two lengths: 12 inches, or 5 meters!!
Which apparently are the best lengths for less reflections being a problem.
Also 'special' connectors.
Well I happened to have an Audioquest 'true 75 ohm' cable with that sort of 'snap on' connectors (from junk pile at local store $3 it was used for a video cable) except mine is 25' long. Well I stuck it in anyway and HEY! better high frequency for real.
I may just go ahead a get one of those in a five meter length.
I'm using a Clarity Cable XLR as a digital cable and it smokes my JPS Superconductor Q and Morrow DIG4 AES/EBU. Soundstage is more defined and precise.
At least one site offers two lengths: 12 inches, or 5 meters!! Which apparently are the best lengths for less reflections being a problem.
The usual recommendation, if the cable is not very short (less than 12 inches) is 1.5 meters, or about 5 feet. See this paper.

Length beyond 1.5 meters would not provide additional benefit in terms of reflection timing (unless perhaps the output signal of the source component has unusually slow risetimes and falltimes). The additional length could, however, conceivably result in increased noise pickup from ground-loop or rfi/emi effects, thereby worsening jitter. Also, if hi rez formats are involved a 5 meter length could actually cause reflection timing to become more of a problem rather than less.

These length recommendations, btw, are generally intended to further improve the situation where a reasonably good impedance match exists to begin with. A severe mismatch will not necessarily be helped by using a 12 inch or less cable length.

The 1.5 meter length probably would help a severe mismatch in the case of redbook data rates, but not necessarily for hi rez rates. In the case of hi rez data rates, if the impedance mismatch is severe enough, reflections corresponding to a given signal edge (i.e., a given 0 to 1 or 1 to 0 transition) may keep propagating back and forth across the length of the cable enough times to cause distortion on the following signal edge.

-- Al
Thanks for all of your kind replies gents, they were very informative.