I have completely lost track of what constitutes " budget" these days. I really liked the cheapest VDH digital cable; Clearer Audio makes some very good ones; I also had an Oyaide silver cable. These all might be too pricy for you but they are all good buys if they are in your range. Price structure over there is often quite different than here and most of the other cheaper ones I know are US ones that probably have better buy British equivalents. Believe me, digital cables make a difference in sound.

I have one of those [ Wyde Eye] and its not bad but not in the same class as the ones I mentioned. I have a V-DAC and after getting it sold the much more expensive MF TriVista 21 I was using previously. Whether or not the cables SHOULD sound different they in fact DO sound different. Some standard ICs will work well as digital cables. My engineer friend made fun of my sometimes use of VDH TFU as a digital cable but when he measured it it was 75 ohms.

"As a general guide, I buy Belden RG59U cable (precision 75ohm coax, foil and braid screen) for around £1.50 per metre. A couple of phono's and Bob's your uncle. That's all you need for SPDIF - decent screening and the right impedance. SPDIF is not a 100% dc-free encoding system so you get some baseline-wander and ensuing jitter on the recovered data no matter what you do, and troubling ones head about the cable beyond this is pointless - the end performance willl always be dependant on the clock recovery performance in the ensuing DAC."

I agree. I also make leads like that. A decent soundcard can offer improved coax SPDIF over motherboard SPDIF.
I think that the further filtering of the power may assist there. With DACS that use the DIR9001 and the older 1703E, further improvements can usually be obtained by fitting close tolerance polypropylene capacitors and 1% resistors in the "flywheel" (filter) area of the DAC (DIR 9001 - 680 ohms, 4.7nF and 68nF on pin 22) A TCXO can often give a further improvement, but may take a few minutes to "lock in", where you will be aware of a nice sound improvement.
Alex

Hi Stan
No, it's not nonsense. Sometimes cutting corners in a design can result in things like you are reporting.
With proper terminations both ends, the type of 75 ohm coax lead should make very little if any difference.
Some people even go to the trouble of fitting proper 75 ohm BNC connectors at both equipment ends, and at both ends of the cable. My best DAC doesn't have a coax SPDIF transformer at the input, and seems fairly unfussed by the actual value of the input resistor on the PCB. However, it appears to have far better than average "Jitter" rejection.
Regards
Alex

No difference, apart from in the flexibility of the cable. When you get really high in frequency solid core is the norm, as skin effect means that the signal is conducted only in the very outer part of the core anyway. That's why things like satellite cables have single steel cores with copper or silver plating. But SPDIF is way below these signals in frequency.

With grounded-each-end coax set-ups the screen is the most important part of the cable, as, on top of actually acting as a screen, it has to tie the chasses at each end together, any differences appearing as signal at the receiving end. That's why when you look in the catalogues of high performance co-axes for real manufacturers such as Belden, the emphasis is always on the screen, which will usually be foil and braid, foil and double-braid, often silver plated, and some surprisingly large diameters.

But, to emphasise: SPDIF is NOT a 100% dc-free coding system, and there will always be baseline wander on the levels into the data-retrieval circuits if there is AC-coupling anywhere between the driving IC and the receiving IC- this translates to clock-jitter on the received clock, and no amount of fancy cable materials can do anything about this. This means that transformer-coupled (rather than dc-coupled) systems will always have worse jitter, though they do of course remove various grounding problems. For audio quality it all comes down to the subsequent clock de-jittering circuits, with the best using VCXO PLL's.

Also SPDIF contains no data-correction algorithms, so any data errors will NOT be subtle as they will be completely random as to which bits they affect.

Not actually very much. As far as the actual data content goes there is a status bit saying that it IS AES/EBU, which may or may not be ignored by your decoder, and there are some extra formatting/sampling-rate flags. Electrically of course it's different with a balanced pair and a different line impedance - you can buy off-the-shelf baluns to electrically convert one to the other, but with some compromises on signal levels which aren't strictly speaking compatible but will usually work OK, depending on your actual decoder.

But the two systems really just reflect the two different environments - SPDIF is fine if the two bits of kit are close together and it's OK to connect the two bits of kits together via the cable screen (pretty usual for a domestic system) and AES/EBU is for when the two bits of kit can be seperated by many metres with each bit of kit firmly earthed locally (typical of a studio setup). The transformer coupling used in AES/EBU does degrade the jitter of the received signal a bit, but as I've said before this can be removed by the decoder. BUT: studio systems usually have globally distributed clocks (if you think about it they have to because you couldn't have tens or hundreds of bits of gear all with local ADC's running at nominally the same sample rate but not synchronised), and the received data from the AES/EBU link is then usually re-clocked by the distributed master clock, thus making the jitter performance of the link pretty well irrelevant.

I find that all cables used as SPDIF sound different and you just have to pick the one that best suites your setup. My favorite is some old 75ohm cable with yellow connectors on it. At a listening session we put it in because I kept saying it was my favorite over the years and wondered what it would do in that system. It was the best of the evening. But I think it came with an MSX computer 20 years ago or so, so I can't ever buy another :-)

kengale
Say you were using a decent coax cable with proper 75 ohm BNC plugs at both ends, would you agree with Wiki that it would make no discernable difference whether 50 ohm or 75 ohm BNC sockets were used , provided that the frequency was <10MHz ?
Many retailers these days do not even specify whether their BNC sockets are 50 or 75 ohms.
Kind Regards
Alex

kengale
Thanks for the reply. I have seen many reports about improvements after replacing RCA with BNC,
but I have often wondered if the same results may have been obtained by simply replacing the RCA plugs and sockets.
I wouldn't mind $10 for each "gold" RCA socket that I have replaced over the years due to the inner losing tension and going high resistance or o/c.
Kind Regards
Alex

Have you ever actually listened to different cables? Of course not as everyone knows all digital sounds the same. All this blather about dacs is equally silly; discussions of computer audio is even worse. When everything is 1s and 0s there CAN be no difference in sound. Excuse me while I go water my pet unicorn.