Hi all,

I have a pretty technical question here:

I have a 10 meter run through my house of 5 cables, which (as I now know after cutting into them) are 3 digital-rated coax used for component video, coupled with 2 regular RCA audio cables. The exact cable is here: http://www.monoprice.com/products/product.asp?c_id=102&cp_id=10235&cs_id=1023501&p_id=2858&seq=1&format=4#feedback

What I want to do is this- use the 3 component video cables for component video (no brainer), but also use one of the RCA audio cables as the preout connection to my subwoofer, and the other for digital audio S/PDIF return from the TV back to the receiver.

I had to cut one end of the bundle to run the cables, so before I start crimping and wasting expensive connectors, is this even feasible? I have more faith in the the subwoofer working than the digital audio, but I figured I'd see if you guys have tried anything like this. I've done it on shorter runs, but 10 meters is a whole new ballgame.

Thanks in advance!

The 3 component cables are 75 ohm rated, which should work. The subwoofer pre-out is usually line level, which will work with any regular rca cable. However, if i am not mistaken, spdif digital coax also uses 75 ohm rated rca cables... So i am not sure if the remaining cable would be optimum for digital coax. It may work if the run were short, like a few feet, otherwise you may want to run one more cable for just the spdif.

Edit: nevermind, looking at that link, it looks like to save on processing they gave all 5 cables a 75 ohm rating. If thats the case you should be good.

It appears that all 5 cables in that bundle are the same, they describe it as RG6 type cable, which will be 75ohm. This should be fine for digital audio. And absolutely fine for the subwoofer. Additionally, even if it's not 75ohm cable, realistically there's a very good chance it would work anyway, particularly with SPDIF which isn't enormous bandwidth anyway. Consider almost everyone uses SPDIF over coax with RCAs, and RCAs are never 75ohm connectors anyway. I doubt you will have any problems as long as you terminate the cables properly.

Whether it's 75 ohm rated or not, I'm not sure, but I know for a fact that the two audio cables are *not* the same as the component cables.

They are typical RCA. The insulation is extremely thick. There is a single very small white wire core with a thin ground wrap around it. I discovered this when I had to cut one end of the bundle.

hmm, so they aren't coax at all? Then they may not be 75ohm. This is absolutely fine for subwoofer, but not so much for digital audio. However, if you can't pull more wire, you're sort of screwed, so you might as well try it out and see if it works at that distance. Hopefully it'll work fine. You won't be wasting any connectors since if you're soldering onto it, you can always desolder it and reuse the connector for something else.

Whether it's 75 ohm rated or not, I'm not sure, but I know for a fact that the two audio cables are *not* the same as the component cables.


Well eyewitness accounts are not to be triffled with. The Monoprice.com product write up seems to be a little optomistic.


They are typical RCA. The insulation is extremely thick. There is a single very small white wire core with a thin ground wrap around it. I discovered this when I had to cut one end of the bundle.

The impedance of normal audio cables is more like 110 ohms, which is not *that* much difference than 75 ohms. If massive power levels were involved, then we might worry about parts getting hot, but that is definately not the case here!


However, the fact that the cables are different from RG/6 doesn't mean that they definately aren't 75 ohms. It does raise some serious doubts.

More to the point, over 35 feet, any possible impedance mismatch won't make a difference for the bandwidth of the signal that you are trying to transmit because its maximum frequency is too low, given that the signal is TV sound.

The SP/DIF audio output of the TV, if you look at it with a scope, will probably look like a sine wave whose frequency is no more than about 3MHz. The wavelength of 3 MHz is over 320 feet, so your cable is less than 1/8 of a wavelength long.

For the impedance of a cable to matter, it must be much more than 1/8 of a wavelength long at the highest frequency transmitted, even by the pickiest of engineering standards.

1/4 of a wavelength is a more practical standard, so not only do you barely make it, there's actually a safety margin.

The SP/DIF audio output of the TV, if you look at it with a scope, will probably look like a sine wave whose frequency is no more than about 3MHz. Do they use sin waves for spdif? Thought they used square/rectangle waves for it.

The impedance of normal audio cables is more like 110 ohms, which is not *that* much difference than 75 ohms. If massive power levels were involved, then we might worry about parts getting hot, but that is definately not the case here!


The impedance of twisted pair audio cable made for AES/EBU transmission is 110 Ohms.
The impedance of an"RCA" audio cable of unknown design is unknown. There's a big difference between 110 Ohm twisted pairs and 75 Ohm coax and you can't really replace one with the other.
Impedance matching is used to ensure maximum power transfer in high power RF applications. In low level signalling, impedance matching ensures that there are no reflections.

An SPDIF signal is a square wave.

Do they use sin waves for spdif? Thought they used square/rectangle waves for it.

That's a common misconception. If you look at the wave with a scope, its very rounded. The rounding is often enforced by an output transformer that is very lossy at high frequencies. This is done to help the product meet FCC Part 15 regulations.

There's no reason to smooth the light intensity signals on optical interconnects. However the LED-based line drivers are very capacitive, so they tend to triangularize the signals that drive them. The photo transistors used to receive the signal tend to sharpen them back up.

The impedance of twisted pair audio cable made for AES/EBU transmission is 110 Ohms.


Agreed. The history of this is that 110 ohms is typical of standard mic cables, and its also typical of many forms of twisted pair, whether shielded (mic cable) or unshielded (CAT5)


The impedance of an"RCA" audio cable of unknown design is unknown.


But, it is both knowable and also tends to converge in the 50-90 ohm range.


There's a big difference between 110 Ohm twisted pairs and 75 Ohm coax


The major difference being the absence of one insulated conductor.


and you can't really replace one with the other.


Actually, this is done all the time, and generally with good results. Not that I don't very much prefer the balanced interfaces that twisted pair enables and encourages.


Impedance matching is used to ensure maximum power transfer in high power RF applications.


Agreed. If you don't get impedance matching right in high powered applications, reliability suffers. Parts can literally burn up. The heating represents loss of available signal, so its a double loss.


In low level signalling, impedance matching ensures that there are no reflections.


Also agreed, but the importance of reflections is often overstated, particularly in consumer audio. Signal bandwidth (e.g. NTSC video and 2-channel or Dolby audio) has been low, and line lengths have been short.

HDTV is probably the first consumer A/V situation where reflections are actually a serious issue just about all the time.

Reflections in interconnects have long been a serious issue for high quality consumer computer video displays. They aren't much of a problem for 640 x 480 x 25 fps, but they are quite problematical for 1024 x 768 x 75 fps.


An SPDIF signal is a square wave.

Ideally and conceptially, yes. Inside digital audio gear, yes.

On coax stretching between a TV and a surround receiver, SPDIF is probably not a square wave. The main reason is FCC Part 15 compliance, and general practical necessities.

Square waves are not necessary for SPDIF interconnects. They make regulatory compliance more difficult. Receiving them accurately opens the door for additional outside interfering signals.

hmm, so they aren't coax at all?

Why aren't the Monoprice cables coax? They say RG-6 on them.

btw, to the OP I use the same cable from monoprice and I run Digital audio, interconnects (for subs) over them without problems.

Agreed. The history of this is that 110 ohms is typical of standard mic cables, and its also typical of many forms of twisted pair, whether shielded (mic cable) or unshielded (CAT5)

No, traditionally mic cables matched the mic, at 150 to 250 Ohms. Line level audio used 600 Ohms.


But, it is both knowable and also tends to converge in the 50-90 ohm range.

I've read 40 to 70 Ohms. They're not impedance controlled cables, they're not specced.

The major difference being the absence of one insulated conductor.

..and the twists in the wires, the circuit driving the cable, the circuit receiving the signal and the characteristic impedance.

Actually, this is done all the time, and generally with good results.

I disagree that the results are good.

The heating represents loss of available signal, so its a double loss.

What's double about it?

No, traditionally mic cables matched the mic, at 150 to 250 Ohms. Line level audio used 600 Ohms.


Traditional = telephone system.

Studio mics haven't been impedance matched for decades. Modern line level studio gear is generally not-impedance matched, and hasn't been for decades.


I've read 40 to 70 Ohms. They're not impedance controlled cables, they're not specced.


Since they are typically relatively short, characteristic impedance just isn't an issue.

I'm not saying plan and intentially build systems with random cables. I'm saying that if you are between a rock and a hard place like the OP, give what you got a spin because it will probably work brilliantly. And if it doesnt' work brilliantly, you may know where to look first.

I broke my own rules in the past week.

I've got a ca. 400' DMX run at work that we have been extending. It's supposed to be 100% digital-rated 2-conductor shielded cable. This system has always had a few short lengths of regular mic cable with no problems. Under the gun for the holidays, we made a ca. 100' extension including 2 new intelligent fixtures, about half regular mic cables, half proper cables.

We immediately starting having problems with fixtures and dimmers responding in seemingly unpredictable ways. I thought that I had finally sleazed the system over the edge and needed to get the right cables.

But first we did some more tests and found that the problems were actually due to two dimmers in the same pack responding to addresses that were exactly too low by 1.

On a hunch I went up into the attic where the dimmers are installed, and found that others had been up there coincidentally due to a roof leak and apparently brushed a dimmer pack and accidentally set its address down by one. I corrected it, and all was well.



..and the twists in the wires, the circuit driving the cable, the circuit receiving the signal and the characteristic impedance.


Unless the cables are really long, and I gave a common industry standard for calculating the need for impedance-matched cables conservatively. Typically in home audio, impedance variations in digital lines don't matter.


I disagree that the results are good.


They are what they are, but based on decades of working in pro audio and AV, most things work unless you completely miss making a connection, or flagrantly break the rules.


What's double about it?

What I mean is that impedance mismatches in high power environment (big transmitter) can fry things, sometimes creating actual fires. The second loss is that you loose power, and therefore things like transimtting range go down.

For example, decades ago I spent a few years working with high-powered miltary radars. Big enough to have dedicated portable power plants. Impedance mismatches in transmission lines were especially common when I worked on a missile site in the Everglades. Moisture build up inside transmission lines would cause standing waves. This could happen faster than standard PM intervals. If you were on top of things, you'd notice that the radar's range would start slipping. If you missed that, little things like burned up sections of transmission lines might be your wake up call. There were ionization detectors and built-in SWR bridges, but they weren't 100% effective. They were mostly there to protect the output tubes.

Stephen Lampen of Belden Cable wrote this:
In a message dated 04/06/2007 1:23:59 AM Pacific Daylight Time,
proaudio-request at pgm.com writes:

Over on the Ampex list, we were having a discussion about how long a crummy cable one can get away with using to carry AES/EBU audio date. Thanks to Larry Miller for reminding us that the critical number is the rise time and gave an empirical rule of thumb (1/6 wavelength) for when the cable goes into the transmission line mode and needs to be the correct impedance to avoid getting into trouble.

So I'll bet the AES/EBU (IEC-60958) standard specifies the maximum and minimum rise time for the data waveform. Does anyone have a copy of that standard or know what the rise time is? I'm looking for a simple answer based on real facts to the ever popular question: "Can I use a plain old (not made from special 110 ohm cable) for AES/EBU connections?" - the answer being "Yes, if it's shorter than X feet."

Friends-

Only just saw this ancient email before it fell off my list and I had to comment.

I normally tell people that the critical distance is 1/4 wavelength which then varies depending on the sampling rate (and resultant bandwidth) of a cable. I also have some eye-opening graphs (no pun intended) of 48 kHz audio running on 110 ohm cable and Belden 8451 (both 22 AWG, so we're comparing apples to apples). The result is 50 ft. is probably do-able for 8451. At 100 ft. literally half of the signal is reflected because of the impedance mismatch (8451 =~38 ohms) and it screws up the risetime (clock) because the capacitance is about 3 times as much as the 110 ohm cable (Belden 1696A).

If you go by the 1/4-wave numbers the critical distance (after which the impedance is important to match) is:

44.1 kHz = 5.6448 MHz= 44 ft.
48 kHz = 6.144 MHz = 40 ft.
96 kHz = 12.288 MHz = 20 ft.
192 kHz = 24.576 MHz = 10 ft.

Hope this helps set a limit!
Steve Lampen
Belden

Just today I read something else he said. In effect:
All cable is "Analog" the devices sending and receiveing the signal may be "Digital" but it's analog going down the cable.

Test results:

My house is still a wreck with the move and everything, so here's what I have for my best test thus far:

I hooked up the "crummy" regular old RCA cable (it's definitely not RG-6, I think monoprice is either lying, or their supplier is) between two wall plates, and then connected the TV digital out to one wall plate and my receiver's digital in to the other with two similar "crummy" monoprice RCA cables. I then played audio across this transmission line:

TV -> 6' Monoprice RCA -> WALL PLATE -> 10m Monoprice RCA -> WALL PLATE -> 6' Monoprice RCA -> Receiver

Note, that one end of the 35' cable was butchered due to the fact that Monoprice uses some odd kind of solder (that or it was a weld??) on their connector, and I wasn't able to reclaim it as well as I had hoped.

Result: Audio sounds great. I ran a cable over to the TV, found a digital channel, and whalla! Audio. The Dolby Digital icon on my receiver lit up as well. I only pluged in left and right speakers for this test (which sounded great), but it was a fine enough test for me- I don't expect 2 speakers to work and the other 3.1 not to with a digital stream.

Thanks to everyone who gave me a sliver of hope to even take the time to try to terminate the regular old crummy 10m RCA cable. (Perhaps that's why Monoprice is able to get away with calling them RG6- they work w/ S/PDIF anyway.)

Traditional = telephone system.

Studio mics haven't been impedance matched for decades. Modern line level studio gear is generally not-impedance matched, and hasn't been for decades.


You seem to be responding to the wrong quote.

Since they are typically relatively short, characteristic impedance just isn't an issue.

Nor is the impedance of these cables even known...as mentioned above.

I've got a ca. 400' DMX run at work that we have been extending.

DMX512 only operates at 250K and uses RS422 protocol, which is very robust....not sure why you posted that.

What I mean is that impedance mismatches in high power environment (big transmitter) can fry things, sometimes creating actual fires.

..after disableing all the safety and monitoring circuits in the transmitter...why do you keep bringing up high power RF, it doesn't apply here.

(Perhaps that's why Monoprice is able to get away with calling them RG6- they work w/ S/PDIF anyway.)

RG-6 refers to coaxial cable. Are you saying it's labeled as RG-6 but it's not a coaxial cable?

The video cables are RG6. The audio cables are not.

From Monoprice specs:
Video Cable
Triple heavy-duty super shielded RG-6/U coaxial cables to maximum signal transfer and minimize interference.
Gold plated RCA plugs to improve conductivity.
Color coded and labelled for the luminance (Y) and chrominance (Pr, Pb) Signals.
Solid center conductor: 18 AWG (1*1.02) Super shielding: 120 % aluminum foil + 9*16/0.12 TINNED COPPER Braid
Fully molded construction. Excellent quality.

Audio Cable
26 AWG (10/0.12) Super shielding: 120 % aluminum foil + 48/0.12 TINNED COPPER Braid