Jst FYI - a reply from Charles Hansen (Ayre) :
<< Can I use a XLR-RCA digital cable? >>
First of all, please note that *everything* makes a sonic difference. The question comes down to how much of a difference, and is it worth it to try and improve the situation.
If the transport follows the AES/EBU spec, it will have a balanced output with 110 ohms from hot (+) to cold (-). The original AES/EBU spec was for a voltage of 10 volts peak-to-peak. I believe this has been changed to a lower level in later revisions of the spec, but it is still higher than the S/PDIF spec. If the DAC follows the S/PDIF spec, it will have an input impedance of 75 ohms, with a peak-to-peak signal level of 0.75 volts.
One of the problems with separate transports and DACs is that almost none of them follow the specs. For example, your DAC has an RCA input. There is *no such thing* as an RCA jack with an impedance of 75 ohms. Here's another example. Probably more than 90% of all consumer audio DACs use a Crystal (Cirrus Logic) receiver chip. This chip has Schmitt trigger inputs. The helps avoid the possibility of false triggering, but the input impedance of the chip changes as the signal transitions from high to low.
So basically you are entering a morass when you are trying to get everything optimized.
The purpose of impedance matching is to avoid reflections. Here's how this works. If you stand 50 feet away from a large flat surface and clap your hands, you will hear an echo (reflection). This is because your hands vibrate the air molecules, which propagate through the air. The air has a characteristic impedance that is the ratio of the mass of the air to the stiffness of the air. When the sound wave hits the flat surface (typically a stone wall or a brick building), the flat surface has a different characteristic impedance as its mass to stiffness ratio is different from that of the air. So you hear an echo reflected back to you.
In the case of digital audio, any time there is a change in impedance, a reflection is generated. It is exceedingly difficult to create a pathway that doesn't have any reflections. There is the output impedance of the digital chip, then the characteristic impedance of the PCB traces, then there are often wires to the connector that have a different impedance, then the connector on the transport will have another impedance, the connector on the cable will have another impedance, the cable will have another impedance, then another connector on the other end of the cable, then the connector on the DAC, then (probably) some wires to the PCB, then the traces on the PCB, then the digital chips that receive the signal.
Each time the signal passes one of these transitions, it will generate a reflection. Sometimes the reflection is positive, and sometimes it is negative. The problem with these reflections is that they can affect the timing of when the DAC detects the change in state from a "0" to a "1". This is known as jitter.
The funny thing is that the reflection has to travel in both directions for it to cause a problem. In other words, it has to be reflected back towards the transport and then back to the DAC again to cause problems. If either end of the signal chain has the correct impedance, the reflection will be absorbed and not cause a problem.
In the specific case of an Ayre transport, we make sure that everything is as close to 110 ohms as possible. This helps a lot because if anything downstream is not perfect, the reflection gets absorbed by the transport and not re-reflected back to the DAC, where it would cause problems. (Please note that if you are using an Ayre as a transport that you will need to use a specific adapter, as the pinout on our digital XLR output is slightly non-standard.)
In the case of other transports, in theory it would be beneficial to use impedance matching transformers. But in practice they usually cause at least as many problems as they solve. I personally would not bother with them.
One of the best tricks you can do to avoid these issues is to use a long digital cable, say 20' to 30' long. In this case any reflections due to any impedance mismatches anywhere along the line will come back to the DAC at a time where they won't cause any problems with the signal.
To top it off, both the S/PDIF and AES/EBU interfaces *inherently* add data-related jitter no matter what you do. So in my opinion, this entire topic falls into an area of diminishing returns. About all you can really do in the real world is to try various cables. If you don't hear any differences, then just buy the cheapest one. If you do hear differences, then buy the one that sounds best to you (assuming you can afford it).
But worrying about *why* they sound different is (unfortunately) going to be pretty much a pointless endeavor. There are many other things that affect the sound of a digital cable besides impedances and/or reflections.
Good luck in your search for good sound.
