Difference in sound between copper and silver digital cables?

@almarg - I too wondered about this aspect of digital cables design.

It all started about 5 years ago when KLE Innovaitons contacted me about their products.

I built a very good analogue cable and wondered  how it would perform in the digital realm - so I simply tried it and  it worked great.

I contacted Keith Louie Eichman and we exchanged several emails on the subject. But although he assured me that his RCA's are perfect for digital application because of their impedance, he would not reveal the actual impedance. So I am unable to comment on this aspect in any further detail.

Apparently - they work extremely well with 50, 75 and even 110 ohm applications.

After a lot of reading from the web - could it be possible that the impedance of the RCA only has to be greater than 50,75,101 ohms? 

Perhaps it is the design of the actual RCA ?
 i.e.the neutral connection is very small compared to other RCA's 

I have tried a few different KLE Innovations RCA's in digital applications  and they all worked extremely well.

Sorry I could not provide more details - Steve

... could it be possible that the impedance of the RCA only has to be greater than 50,75,101 ohms?

@williewonka

Hi Steve,

Thanks for providing the additional background. But no, for accurate minimally distorted transmission of digital waveforms, the "characteristic impedance" of a digital cable, the output impedance of the component providing the digital signal, and the input impedance of the component receiving the signal, should all be very close to being exactly the same.

In saying that, I should emphasize two things. First, this has nothing to due with accurate conveyance of the digital bits, which figures to be perfect in any half-way reasonable interconnection. Second, the distortion of the waveform that is received by the destination component does not by any means have any **direct** relation to distortion that may result in the sound that is eventually heard, as it might in the case of transmission of analog signals. But what may happen is that the frequency components corresponding to that distortion of the digital waveform, which will be well into the RF region, probably at several tens of MHz or even more, may find their way past the receiving circuit (via grounds, power supplies, stray capacitances, etc.) and contribute to timing jitter at the point of D/A conversion, or to intermodulation or AM demodulation or other such effects at analog circuit points further downstream. All of this being very much dependent in unpredictable ways on the particular components, on the normally unspecified "risetimes" and "falltimes" of the signal provided by the source component (i.e., the amount of time required for the signal to transition between its two voltage states), and on the length and several other characteristics of the particular cable.

Ground loop effects between interconnected digital components can also be cable-sensitive, btw, potentially resulting in low-level high frequency noise being injected into the DAC or other receiving component, with consequences similar to those I’ve described above.

All of this is completely different than the potential effects of cables conveying analog audio signals, due to the vastly higher frequency components of digital audio signals. In the latter case bit rates are already well into the RF region (roughly between 1.5 and 10 MHz or so, depending on whether it is redbook data, 24/192 data, etc.), where imprecise impedance matching can degrade waveform quality due to signal reflection effects. But the frequency content of the risetimes and falltimes of all of those signals are at even much higher frequencies, at several tens of MHz even for redbook data as I said.

One thing that follows from all of this, IMO, is that any similarity that may be observed between the sonic effects of various metals and various geometries as used in analog audio cables vs. as used in digital audio cables is entirely coincidental.

And, finally, I would approach with a great deal of skepticism any claims that a specific wire type can simultaneously be optimal for 50 ohm, 75 ohm, and 110 ohm digital applications. And I feel pretty certain that nearly all other experienced digital circuit designers would agree with me. While it may not make any difference in some systems, or it might even make a difference in some systems that seems to be subjectively preferable, why do what amounts to introducing a known design flaw into the system?

Best regards,
-- Al

I have experienced very large differences between digital cables and I've tried many. I've paid less attention to materials/metallurgy when it comes to digital cables more to overall design. I've had silver, Kimber AGDL,  mixed materials, Cerious Technologies, and copper, Nirvana, Acrolink, Jorma.


As an overall comment over the years I've gravitated from all or mostly silver cables to all copper now. I prize warmth and musicality over excessive detail. The Jorma digital cable I use, hyper-pure copper, gives me a warm rich organic sound with plenty of detail.


My advice is to experiment buy some used digital cables and keep at it til you find one you love. 

@almarg - Al, thanks as always for you complete and thorough descriptions and reasonings.

Your point about... 

one "Wire Type" being optimal for 50, 75, and 110 applications.

It was the RCA connectors I was referring to as being suitable for those impedances and not the wire type. But I do understand that according to conventional wisdom,  the impedance of the RCA should match the cable - so I too, wondered how one connector could be used for all three impedances.

However. Keith Louie Eichman has been delving into the minutia of the science of cable and connector design for over 30 years and has far more knowledge than myself, so I have no reason not to believe his assertions about his RCA connectors. 

The only tools I have to test their performance is my ears and to me they sounded better in every way.

The geometry of my cables is a signal wire inside a helix coiled neutral wire and unfortunately I have no idea what the impedance of that cable might be.  But perhaps it is the cable geometry that is performing "the magic" here? 

My I first attempt at making a cable for SPDIF purposes used some wire from a piece of CAT6 Ethernet cable and Copper Harmony RCA's using the Helix geometry. The signal wire was a single CAT6 wire and the Helix neutral was a twisted pair.

My "point of reference" for the comparison was a commercially available 75 ohm 1 meter cable specifically designed for digital transfers from Van den Hul, that had the requisite 75 Ohm RCA's. This was a conventional Coax geometry.

The results even surprised myself - that simple little CAT6 cable provided significantly better performance than the Van den Hul product.

I do not claim that my Helix geometry, with the KLEI RCA's,  would work for any length of cable, since my interconnects, as tested, were only 3 ft long.

The set of cables I finally ended up keeping in my system were only 18 inches long. Which also seemed strange since, from what I have read on this forum,  a digital cable should be over a certain length in order to work at it's best..

But I did compare the 3ft to the 18 inch cable and found no difference

Others have tried my Helix design as a SPDIF interconnect on longer cables,up to 3 meters, and they claim to have had similar success. 

There are many out there that do not believe the claims I have made regarding the Helix Cables and the Harmony RCA's in my many postings on this forum. 

And - if I hadn't tried them for myself, perhaps I wouldn't believe those claims either.

But there are now dozens of people from around the world that have tried them and reported significant improvements in sound quality.
- So it is not just my ears that is hearing the improvement

Regards - Steve

Jorma digital gets rave reviews but at $1700 it's way out of my price range. What common or boutique brands with a balance of richness and detail can be had for ~$300? Notwithstanding the implementation, does silver coated copper give you the best of both?