[MadMan007]

Hi. I've read somewhere or another, and probably at multiple places, that digital coax cable length can affect the signal quality due to signal reflection, standing waves or some other explanation. From a long time ago I am familiar with but forgetful about signal reflection and/or ringing in RF (radio frequency) coax cables. The big difference is that RF is in a different frequency range where it may matter more. Then we've got the inherent mismatch in 75ohm S/PDIF coax due to the RCA connector, maybe that contributes to the problem but I don't know.

Now from what I recall it was recommended to either have very short (<9") or moderately long (3ft+) digital coax cables but nothing in between. So, while the theory might be sound like many other audio(phile) theories it may not matter in the real world or may be a misapplication of a theory that is true in other cases - I can see manufacturers wanting to upsell a certain length of cable to longer 3ft+ lengths than what's an impractical <9" length

In other words 'Don't get 1ft or 2ft cables.'

So, is this true in any way that would affect a S/PDIF signal, has anyone tried different lengths of the same cable, or what other solid information do people know about this topic?

 

[IPodPJ]

You are not supposed to exceed 30 feet with a 75 ohm digital cable. In my system, I've found the best length to be 18 feet. The Blue Jeans BNC cable was only $24 and works better than any "premium" digital cable I've heard (up to $1000).

 

[MadMan007]

Ok thanks for input on the maximum. Did you ever try other lengths of the same cable in your system? If 18ft works for you that's great but I wouldn't get an 18ft cable to attach gear in a stack

I'm not really sure if this gets at what I was asking though. The reason I posted in Sound Science and not the Tweaks forum is because I wanted real science answers in addition to subjective ones. Soo the question is, is there a certain range of length where there might be a problem with signal reflection or ringing (this dosn't mean audible 'ringing,' talking about the signal here) or other effect due to the interaction of frequency of the S/PDIF data and the length of a cable as can happen in RF signals?

After a little more thinking and running numbers for frequency/wavelength I'm thinking no, there is no 'bad length range' because the wavelengths at S/PDIF data frequency are much larger than any sane cable length unlike RF applications. Now there may be gear-specific interactions but there's no way to make a rule for that.

 

[henryflower]

You are talking about a digital cable; either it works or it doesn't. Anyone who tells you otherwise is selling something (probably a digital cable).

 

[MadMan007]

Yes, I've never heard that one before

Do you understand what's meant by signal reflections and ringing in circuits? (A digital cable is part of a circuit after all.) Since it's not thrown about very often, even by snake-oily cable makers, I'm going to figure there's not much to this but I'm still curious. No one else has ever read anything like this?

 

[anetode]

Balderdash. If any of that pseudo-scientific tripe is a concern of yours, you really need better things to worry about. There are easy ways of verifying the design limitations of cabling that don't involve internet forums. I suggest a book or AES leaflet, one with math and references.

 

[Budgie]

Reflections occur when the termination impedance is seriously mismatched. The length of the coax only changes the time delay of the reflection.

Even though RCA's aren't usually 75 ohm, they usually are close enough to not cause problems. A more typical cause of reflection would be no termination resistor in the equipment. Usually it's not a problem in digital audio.

 

[SilverCans]

Guys reflection in a signal is real, I've measured it and it can do nasy things, but that was not on peice of audio gear and the frequecy of switching was failry high (out of the audio band) AND the mismatch in impedence was large. Now I do beleive there may be reflections in digital audio transmission but you know what, not a whole lot can be done about it. No matter what someone tries to tell you about 75ohm, no RCA connector on earth can maintain 75ohms, not to mention the connection to the board doesnt maintain that either. All I can say is this, audition a few cables and go with the one you *think* sounds good/best and be done with it.

 

[MadMan007]

Thanks guys, I felt like I was getting some uncalled for aggressiveness in replies without addressing the facts of the matter. I did some brushing up on transmission line theory to make sure I wasn't imagining things completely wrong

It comes down to impedance mismatch and not cable length per se then (even though cable length does add impedance it's orders of magnitude smaller than the impedance levels we're talking about oops length affects capacitance and inductance not impedance (?)) cable length just affects the time delay as Budgie said, that makes sense. So in a coax S/PDIF setup with RCAs (why couldn't they have made BNCs the standard

) we know there is going to be some degree of impedance mismatch, it can't be avoided with RCAs. That means there will be reflections so it's a matter of degree rather than if it happens. I've tried some searches but can't find anything concrete on whether reflections are significant enough to be a factor in S/PDIF transmission. I'm guessing not, or that no one talks about it any more because it can't be helped and people just set up equipment however they can, otherwise we'd hear all about the 'best length' of cable.

I'll try a little more to see if I can find anything interesting...

 

[IPodPJ]

Quote:

Originally Posted by MadMan007 /img/forum/go_quote.gif Ok thanks for input on the maximum. Did you ever try other lengths of the same cable in your system? If 18ft works for you that's great but I wouldn't get an 18ft cable to attach gear in a stack

I have no need for the 18 ft. length, I am using it because it sounds the best. I have it coiled up with a velcro strip supplied by BJC. I have tried 1.5 ft, 3 feet, 6 feet, 12 feet, 18 feet and 30 feet.

 

[b0dhi]

Not enough to interfere with digital transmission, but since the clock (which is analog since it's time domain) is also sent with the audio data, it will cause issues. However, that's usually solved by the DAC if it's a decent one since it will reclock. I've heard there are limitations with that though, and implementations differ in their effectiveness. In any case you end up in the murky area of jitter...

tldr; the shorter the cable the better.

 

[pompon]

You can read this article:
spdif
It's depend your cable material, your transport speed "rinse-time" ...

 

[pompon]

Quote:

Originally Posted by IPodPJ /img/forum/go_quote.gif You are not supposed to exceed 30 feet with a 75 ohm digital cable. In my system, I've found the best length to be 18 feet. The Blue Jeans BNC cable was only $24 and works better than any "premium" digital cable I've heard (up to $1000).

What model of Blue JEans you have exactly ?

I have (not blue jeans) but a coax cable with belden 1694a from Ram
http://www.ramelectronics.net/audio-...a/prodCDA.html

When I will get my new DAC (SRC2496 modded), I will give it a second chance ... (when I got the VD David, I dropped the 1694a without any regrets because I was seduced with the dynamic, snap and bass from the VD cable).

Now I know more what I want to reach ... (not anymore sound "in your face". I look for laid back and get details from the "behind").

Between 3 foot and 18 foots ... big improvement ?

 

[SiBurning]

From a very foggy memory... There's some radio frequencies that operates in those ranges. After doing some checking, it looks like most of the amateur and mobile frequencies fall in this range, so it's likely that the ham community is responsible for the guideline to avoid coax cables from 9" to 3'. It makes perfect sense if you're actually attached to an antenna operating at those frequencies.

For reflections, read up here, but it's really NOT an issue in audio. For example, a 100 kHz signal has a 1/4 wave signal length of 750 meters, so just keep your cables under about 1000 feet and you're okay.
http://en.wikipedia.org/wiki/Signal_reflection
http://en.wikipedia.org/wiki/Time-domain_reflectometer

 

[IPodPJ]

Quote:

Originally Posted by pompon /img/forum/go_quote.gif What model of Blue JEans you have exactly ? I have (not blue jeans) but a coax cable with belden 1694a from Ram RCA to RCA connector cable, Canare RCA connectors at both ends, Belden 1694A cable, Black Cable Jacket, SPDIF (S/PDIF) , Composite or Subwoofer When I will get my new DAC (SRC2496 modded), I will give it a second chance ... (when I got the VD David, I dropped the 1694a without any regrets because I was seduced with the dynamic, snap and bass from the VD cable). Now I know more what I want to reach ... (not anymore sound "in your face". I look for laid back and get details from the "behind"). Between 3 foot and 18 foots ... big improvement ?

Yes, my BJC has the Belden 1694A.

The issue of length has nothing to do with tailoring the sound to your taste. Digital cables do not do that, and anyone who tells you they do is blowing smoke up your arse. It is not like using an analog cable to adjust the tone (if that is even possible). I find that with the longer cable I am hearing the details that should be there without suffering from "digititis." The DAC is able to reproduce the audio as it should. With the shorter lengths the audio just didn't sound as real.

For $24, why not try it? You could always return it or sell it if you find it makes no difference.

1.5 feet was okay
3 feet sounded the worst
6 feet was okay
12 feet was a little better than okay
18 feet was the best
30 feet was okay

Perhaps someone who better understands signal reflections can explain why I experienced the above? Steve Nugent?

Edit: It's funny you say you dropped your 1694A when you got your VD David digital cable. Don't get caught up on price, ever -- but especially with digital cables. Use your ears. They need to be made to spec. I had the VD Master LE 2.0 digital cable (which retailed for $1000 and was much higher up in the product line than the VD David) and I sold it in favor of the Cobalt Cable digital interconnect ($80) which I now don't use in favor of my BJC 1694A ($24). If you choose the "premium" cable that isn't made to spec and you like the sound it provides you with, that's fine, just as long as you know it is not transmitting the data the way it's supposed to, nor is the DAC receiving it as it's supposed to.