Audio over Cat5/5e/6 cables

Originally Posted by lonjim2

Just wondering, if anyone has an opinion on the topic -- the other day a co-worker was telling me that some installers are now running Cat6 cable in the walls for line level audio distribution. I know the twists in CatX cable are of a very high quality, but is there a case for it actually being superior to the standard audio cable being used out there? Is there any more of an advantage to using Cat6 over the older Cat5/5e to carry line audio?

I know the twists in CatX cable are of a very high quality, but is there a case for it actually being superior to the standard audio cable being used out there?

Originally Posted by lonjim2

Just wondering, if anyone has an opinion on the topic -- the other day a co-worker was telling me that some installers are now running Cat6 cable in the walls for line level audio distribution. I know the twists in CatX cable are of a very high quality, but is there a case for it actually being superior to the standard audio cable being used out there? Is there any more of an advantage to using Cat6 over the older Cat5/5e to carry line audio?

The other thing is the common mode noise rejection inherent in twisted pair configuration is very very good.

Also, I wouldn't use anything shielded (cat6) unless absolutely necessary as the shielding can be coupled with by the audio signals thus coloring sound.

Originally Posted by duvetyne

Twisted conductors don't reject noise....this is a common myth, often repeated in these fora.



wow! I mean, it doesn't even make sense.

Originally Posted by Ugly1

You're clearly not an EE are you? Suffice to say you may want to do a little research before making such statements.

Also, I wouldn't use anything shielded (cat6) unless absolutely necessary as the shielding can be coupled with by the audio signals thus coloring sound.

Originally Posted by jarrod1937

Twisted pairs don't magically reject noise.

Generally wires are twisted together to induce noise equally so that noise may be rejected ONLY in balanced and/or differential setups.

Non-balanced lines need shielding or else interference will be introduced into a line level signal.

And i too feel this makes no sense.

Originally Posted by Ugly1

Your right it isn't magic it's physics. I take it by your comments you haven't studied it much either.



Not sure what you are saying here but sounds somewhere out in left field.

As I said earlier the twisted pair wire configuration can be used when reduced antenna effect to ambient noise fields is desirable. Antenna effect is related to the geometry and composition of the antenna and ambient noise fields character. These two things alone determine the amount of energy recieved by a given antenna in a given noise field. This noise energy will mix with any signals being conducted on these lines regardless of their nature. One type of antenna is known as a current loop antenna. Loop (for short) antennae gain is dependent on a number of factors one of which includes the loop area as seen by whatever fields the loop is located within. The act of twisting the wires reduces loop area significantly lowering antenna gain by detuning the antenna to most of the frequencies present in typical ambient noise fields. As you might imagine the current loops found in audio systems may also act as antennae.

If you are truly interested in learning more about this try a google search for "current loop antenna effect".

I take it you meant to say lines carrying balanced signals. Differential signals traveling on those same lines would see the same noise energy as single ended signal assuming all else equal, The one difference is that differential signals swing both ways instead of one yielding twice the signal to noise ratio for a given signal level and noise energy level which have been mixed on a signal carrying conductor.

Shielding can be used to reduce noise field intensity on either single ended or differential signal carrying conductors. Signal type matters not. Of course, in practice shielding is imperfect and can only reduce noise field intensity not eliminate it.

Signal wires encapsulated in shielding will experience both capacitive and inductive coupling with the shield conductor....again it's physics. These parasitics will draw energy off from the signal which gets attenuated in eddy currents etc resulting in distortion.

If you truly would like to learn more about this you may want to investigate what is known as inductive and/or capacitive coupling. They are two mechanisms of the observed effect that some of the energy from a signal on a conductor will be transferred to nearby conductors. Any time you remove energy from a signal it is altered aka distorted. The operation of transformers and coupling capacitors is based on these effects being real. Energy really does go into one coil of a transformer and out from the other even though high levels of electrical isolation between the coils may exist. Energy truly does go into one series capacitor lead and out the other and yet there is usually a very highly non conductive dielectric seperating the plates.

Originally Posted by ChrisWiggles

Um, no, thats the reason wires are twisted together, it is so that any noise induced on the conductors are as close to equal as possible, and in a balanced connection that noise cancels out. Without the twist, the induced noise is not equal and won't fully cancel.

That has nothing to do with it.

Try googling twisted pair instead.

A balanaced line (which isn't necessarily differential and doesn't need to be),

No, again you're missing the entire point. A balanaced line (which isn't necessarily differential and doesn't need to be), ensures that any noise induced on both conductors travels back to ground and cancels out because the impedance to ground in a balanced line is the same.

Your differential signal twice the SNR explanation is bologna, that has nothing to do with it at all.

but you really NEED to have shielding to protect an unbalanced line, whereas a balanced line protects itself, and is aided by the twist geometry.

Of course capacitance is real. What does that have to do with noise rejection? Nothing.

Originally Posted by duvetyne

Twisted conductors don't reject noise....this is a common myth, often repeated in these fora.

Originally Posted by jarrod1937

Duvetyne is correct. Twisted pairs don't magically reject noise.

Generally wires are twisted together to induce noise equally so that noise may be rejected ONLY in balanced and/or differential setups.

This allows the signal to be much more immune to noise, sometimes not even requiring shielding depending on the application.

Non-balanced lines need shielding or else interference will be introduced into a line level signal.

The benefits of twisted pair and balanced inputs, which go hand-in-hand, are well known, both theoretically and practically.

Originally Posted by Ugly1

On a read back through my original post I now see that I mistakenly did say "common" mode noise rejection and it clearly should have been differential mode noise rejection that I said. For that reason I agree you do have a half a point here. However both duvetyne and jarrod1937 were were both arguing that twisted pairs offers no immunity to any noise at all which clearly isn't true. I made a simple and honest mistake in my original wording which is obviously wrong. Is this what you are taking issue with?

The half point removed is due to your own sloppy wording since wire twisting does nothing in itself to reduce common mode noise only differential noise.



It is unclear what you disagree with. That differential noise can be recieved on a conductor via antenna effect or that twisting reduces it. Please explain what your disagreement is and attempt to explain why you disagree. As it stands now I disagree with your disagreement.



That's probably a good one too but the search I recomended gets the job done when correctly taken in the context of the originl topic and the point I was disagreeing with jarrod1937 on where he was agreeing with duvetynes statement "Twisted conductors don't reject noise". It really helps to keep things in the context they were clearly intended to be taken in.



This turned out to be the interesting search for me. I looked at balanced vs differential and now understand. My wording in response to jarrod1937 was bad, wrong, whatever. However to the original topic and my original post this is irrelevant. I do, however, concede this point.



Actually this time it is you who is missing the point. The original question was regarding twisted pairs. Twisted pairs does increas differential noise immunity which I now see balanced circuits absolutely would be susceptible to just as unbalanced circuits will be. The point made about differential not equal balanced, while correct, is irrelevant to the point being made in my first post to OP and the point I thought I was making in my reply to jarrod1937 which I now understand where that replys weakness lies. Thanks for the one thing, but in this case you may want to rethink the other.



While it may have nothing to do with the topic as I've now learned thanks to your motivating post it absolutely is not balogna. The explanation given about why differential signaling is used and useful is correct.



On the one hand I partly agree in that balanced lines with or without twisted conductors are would appear to be more immune to differential noise than unbalanced by circuit design. There is still something wrong about what you said.

The more correct statement is: Both balanced and unbalanced circuitry carrying either single ended or differential signaling are susceptible to differential noise which is reduced by using twisted conductors. Also all purmutations of balanced and unbalanced circuitry combined with single ended and differential signal types could potentially see reduced noise from shielding. Balanced circuitry can be more immune to differential noise than unbalanced and differential signaling is more immune to differential noise than single ended signals.

None of this says anything about common mode noise which may or may not be dealt with by design as well.



Who said it did? Perhaps you ought to put the statement in the context it was given. Shielding causes its own signal integrity issues just as it can eliminate other types of signal integrity issues and part of it is due to the capacitance between the shielding and signal conductor pairs. It is the signal integrity issues it causes I was pointing out.

In summary: yep you've helped me see some real booboos I made. Thanks for that and I apologise to all who may be reading for any confusion it caused. At the same time I am even surer now than ever that both duvetyne and jarrod1937were completely incorrect and you made a doozy or two yourself.

Originally Posted by mntmst

How about some tests. http://www.adireaudio.com/Files/CarCables.pdf

Originally Posted by Aphasia

I would say that wikipedia is en rather excellent starter source of information in how TP-cables transfers signals and the different things involved.

http://en.wikipedia.org/wiki/Twisted_pair

Twisted pair cabling is a form of wiring in which two conductors (the forward and return conductors of a single circuit) are twisted together for the purposes of canceling out electromagnetic interference (EMI) from external sources;

Originally Posted by duvetyne

Not by most of you , unfortunately.

Differential receivers...look it up sometime.

Originally Posted by Ugly1

First of all the OP's question was related to "standard audio cable " not coax. I never said said one bad thing about coax.

Second coax isn't without it's own set of care and feeding requirements. There are many downsides to coax but to keep things in context I'll just mention one. Since it isn't what I would consider to be truly shielded in the traditional sense, ie uses it's outer conductor as one of the signal current carrying conductors, external fields will couple to the outer conductor. Point is it isn't impervious to radiation either. It can be beat in the area of radiated noise immunity by STP cabling. Note that professionals have universally gravitated to the use STP cables for the most sensitive low level signals such as mic feeds and phono signals and not coax.

Finally what's with this idea of posting listening tests? This is the theory forum. ;-)

Originally Posted by mntmst

Last time I checked "standard audio cable " is coax

So a STP cable is the correct thing to use in those applications.

PS. I don't know what you mean by "listening tests". My link was pure electronics and test instruments.

Originally Posted by Ugly1

See this duvetyne? sound vaguely familiar?

Originally Posted by Ugly1

lol. The fact that you would say this indicates you are reeaallly reaching hard, or you haven't been alive very long. The "standard" has been two or three wire not configured as coax, for many many years.

Originally Posted by jarrod1937

Correct me if i'm wrong but the same article gives the impression that the twisting is only used to keep noise common-mode in order to be able to reject it in balanced lines (like the phones systems mentioned).

This does not apply to the transmission of a line level signal like the op mentioned.

Though the above statement is only in response to the article for which you're citing, arnyk already gave some good reasons for using twisted pairs.

To me it just feels like you're taking that quote out of context though.

Originally Posted by mntmst

Sure. Lets see what Blue jeans says for unbalanced audio.

When did you get your first formal training in electronics.

I had a FCC first radio telephone ticket in 1974.

Originally Posted by Ugly1


A ham license?