I've been having some trouble with the SPDIF audio out (coaxial) on the wifes PC, which we like to use with our home theater. Currently it's connected to our Yamaha RX-V1400 amplifier via a 12 foot long RCA cable, and due to this the sound tends to cut out randomly for a couple of seconds. It happens about one to three times per hour or so of playback. Obviously the cables quality is questionable for this type of usage, but mostly I think it's the length. For example, a 10 foot cable (which is almost too short to reach) still cuts out just as often, but the drops themselves were shorter. Because I'm out of free TOSLink connections on the amp, I've been considering rearranging all of our home theater audio cables so that we can use the optical out on her motherboard instead. Hence why I'm curious about the maximum length one can use with your average optical audio connection. Most cables sold are generally short, usually 6 feet long or there abouts, and that has me wondering if there was some practical reason that longer cables are harder to find, maybe because of some limitation I've never heard of perhaps. I have seen a three meter one available from TigerDirect, but what I really need is one that is 3.66 meters (12 feet) idealy so that it reaches. Thanks everyone, and I hope that wasn't too long winded lol.

Shouldn't be a problem.

Are you sure it's the length of the cable that's causing the trouble? Twelve feet is not particularly long, and I would certainly not expect it to interfere with an otherwise functional cable. Personally, I'd be looking at the stream from the PC to see if there's a processing hiccup or something that causes the Yamaha to unlock temporarily.

Coaxial digital connections are good to 10m(33') to get this you must be using a properly terminated 75 Ohm cable. A standard RCA cable will not do as it will have the wrong impedance and will probably not be terminated in the correct fashion. In the UK such cables can be got very cheaply from studiospares.com, I suspect in the US something like http://www.sweetwater.com/store/detail/DRA504/ is what you want.

TOSLINK is limited by the quality of the plastic and the bend radius of the cable (bending the cable causes internal reflections which damage the signal). Generally TOSLINK goes up to 15m with 5mm square cable, however under ideal conditions it can go to 30m or more, however this is way outside specification. Decent TOSLINK cables are generally about three times the price of the same length of terminated 75 Ohm coax.

This sounds like the problem I'm having with my reciever (Harman Kardon AVR130), but this has nothing to do with the source.

Sometimes the sounds drops out, appearently with spikes or noise from the power - f.ex. switching the lights or other electrical equipment - But only when I use the digital connections. It looks like an "initialization" of the digital signal, since it also happens with my DVD player everytime it changes chapter (or changes in menus). I had the same problem with another soundcard, everytime it started playing a new song in a playlist. For some reason this card and my DVD player unlocks the signal for a short time.

If you think that your problems are cable related, then the first thing to check is the bend radius - make the cable turn corners as gently as possible (a radius of under 100mm is likely to cause problems) and make sure that the cable isn't being crushed or kinked anywhere along it's length.

Fiber is more delicate than most people think.

How about reading his post correctly? He's using RCA with his problems

I use 10m El-Cheapo RCA cable to connect my computer to my Yamaha RXV 2600.
Usually no dropouts on the connection. However, there is a drop out when I swich
on the neon light nearby the computer. So I'd first check the proper shielding
from RF sources nearby the cable.

Darn!

My advice still stands, you need to straighten the cables so the sharp edges of the 1s don't get stuck in the kinks.

Seriously, though, I think hawkeye_p's suggestion is the best one yet. Switching to a cable with better shielding could be a good idea.

Did no one read my post? You need a 75 Ohm low loss coax cable terminated with RCA connectors. A standard RCA cable meant for analog signals shielded or not has the wrong impedance. It will work for short runs but not for long ones.

No, I was busy flaming cabbagerat

LOL!

Thanks for the posts everyone. I had found a wiki on optical cables, but it wasn't half as educational as this thread is. I know I shouldn't be using a plain RCA cable, and I'm willing to bet it is indeed the impedance as chelgrian says. That is of course if RCA analog cables have more than a proper 75-ohm one does. Back when we had a Sound Blaster Audigy 2 in that PC, hooked up via the same 12 foot RCA cable (plus a mono mini-plug adapter), the problem was even worse and we were lucky to get a signal at all. I had attributed it to weak output from the Audigy card at the time, and is why I ended up using the onboard sound instead. This was originally with a Gigabyte SINXP1394, which was a really fine motherboard and worked quite well. Unfortunately I accidentally damaged it a few months ago while trying to replace the chipset cooler, and thus it was replaced with an ASUS P4P800-E Deluxe which is now the one suffering from the occasional drop outs I mentioned. So it seems that digital out components can vary quite a bit themselves too. Anyways, I think I'll try buying a proper cable, and thankfully length won't be an issue with either type based on what I've leanred from you folks. Thanks again all, I very much appreciate it!

I'm surprised that digital coax runs out of steam that quickly. Granted, I've never run a digital coax more than a few meters, but 10m seems a bit short. I've had component video runs far longer than that, and I would guess that they operate on a higher frequency when carrying HD.

We did use naught but RG6Q for all of those runs, of course.

As have I (using RGB as I'm in the UK and don't have any truck with this weird component video thing...). The thing is that the video is an analog signal noise on the line will affect the picture quality but in an instantaneous and variable fashion.

S/PDIF on the other hand is trasmitting a digital signal along a coax cable with precious little error detection and correction. You either have a perfect reception or you don't and if you don't you get a dropout until the receiver can obtain synchronisation with the signal again after the error.

Again professional equipment uses a standard called AES/EBU which is identical to S/PDIF as far as the protocol goes but uses balanced signalling at the electrical layer to achieve better noise rejection and therefore longer cable runs.

IIRC, pumping digital signal (square waves) through electrical cables will cause extreme distortion, due to the LF part of the spectrum arriving earlier than the HF part (or vice versa, can't quite remember now, been up all night trying to get an IP from the ISP).

...and I just thought cables were cables

So how do you find what impedance a cable should have? A normal "thick" RCA cable is not nessecarily enough?

Hmm but that explains the dropouts I get from time to time (although they are VERY rare) - my thin analog cables.

@odyssey: Just so you don't get it wrong (you know, the moronic audiophile opinion that different cables have different colorization, warmerization, whathaveyouization)...

Cables are just cables. Keep the signals under 20 kHz and all is well. Push the signals over 100 kHz, and things start to get messy. Push it over 1 MHz, and things get real agley. Waaaay agley.

I don't know the bitrate of SPDIF. But even if it is only 100 kbaud, it most probably is square wave, and the harmonics goes waaaay up there. Each n-th harmonics gets delayed (or accelerated. I still don't remember which) differently, and the signal is totally ruined.

Like others have said, analog audio cables are pretty robust. It's worth a few dollars to get a decently constructed cable, but you don't need to break the bank.

Where it gets trickier is with video and digital, especially on long runs. Because both require the signal to arrive in a precisely timed fashion they can easily be upset with a poor quality cable. Video tends to get color distortion, and digital may start to "drop out". I was thinking purely in terms of frequency response, but as pepoluan pointed out the digital signal is a square wave that could easily be made unsuable if the cable introduces time-base distortion.

chelgrian: Component is actually the format used to store video on a DVD. Why the US got component and the UK and others got SCART/RGB is beyond me. Obviously uncompressed video should be better, but I don't know how much difference it makes when the player would have to convert the signal.

Here's a Wiki link that can surely explain it better than I can. In fact I'm going to brush up myself.

So we can keep our simple $1=£1 parity without the official exchange rate getting in the way.

Maybe I'm a bit thick, but I don't see how the two are related.

DVD uses a YCrCb encoding. There are two reasons for this there are several reasons for this. The first is that it is the standard mostly used in broadcast TV, this is for historical reason as it maintains compatability with black and white TVs. The second reason is that human eyes are much less sensitive chrominance than to luminance encoders can subsample the chrominance channels to achieve higher compression ratios.

Most output devices are fudamentally RGB, so the difference is really where the conversion is performed, in europe it gets performed in the source device in the states it's done in the output device. All the digital standards DVI etc are RGB based so the difference between the EU and the states disappears

Yup, strange indeed. Even ARCnet managed to pull 2.5Mbps on coax over couple of hundred meters. Considering more recent history 10BASE2 Ethernet did a good job too - 10Mbps over 150-180 meters. I forgot which used 50 Ohm and which 75 Ohm coax but take your pick. Makes you wonder why you need expensive optical cables for digital audio transfer.

Not having a 75 ohm cable may introduce enough time-base distortion to cause dropouts. I never would have expected it to settle in so quickly, but poor shielding could hurt a lot too.

As far as optical goes, you need a decent cable to avoid internal reflections and for durability. I've pinched optical cables just to see what they can take and it's more than you'd suspect. Fine dust doesn't bother tham as much as many would have you believe. This was all with Audioquest cables, which I think are pretty well constructed. Do you need a decent optical cable? I'd say yes, but you don't need to spend $300 on it either.