Line-level stereo audio over CAT5? Length: 5m / 16ft

What are the pitfalls if I were to send my line-level UNbalanced stereo audio over the CAT5 network cable?

(I just need other wires of that Cat-5 to send some DC voltage, seems to make sense to then use the network cable for both audio and constant voltage... ?)
 
Last edited:
Nope, regular unbalanced L/R stereo signal + ground.
If the driving impedance is low it still should be fine. I have some mastering clients who use CAT-5 interconnect for balanced lines. In broadcast its common.

Look for shielded CAT-5. Ground one end of the shield to provide an electrostatic drain. Observe pairing.
You would probably be better off using CAT-5 stranded patch cord cable rather than solid riser cable.
Harvested USB cables also provide a source of 4C plus shield though the ones I've used aren't as tightly twisted as CAT-5/6.
 
Do you already have the cat5 cable? if no, check out beldens website. if they don't make what you need, I'd reasses your needs (ie they have everything under the sun). the trick is finding a place to buy in small quantities. I've also learned a lot by reading bluejeanscable.com "papers"
 
If it doesn't have 4 pairs its not CAT-5 cable.
If the source and receiving end impedances are balanced, a twisted pair with no shield should work fine.

Jan
He edited his original post to state that the source and termination were unbalanced.

Like I said earlier a shielded, stranded CAT-5 patch cable with the ends cut off is one option.
Amazon has lots of listings for overall-shielded and stranded conductor CAT-5/6.

The OP mentioned that it also needed to carry DC.
Use one pair for Left Hot/Gnd, second pair for Right Hot/Gnd, third pair for bipolar +/- and on the fourth pair both conductors for DC 0V.
Tie the shield drain wire to chassis ground at one end.

Most likely over a short run it being shielded might not make a measurable difference owing to tight twist and low source Z but shielded CAT-5 is easy enough to find at Amazon et al.
 
  • Like
Reactions: jan.didden
Tie the shield drain wire to chassis ground at one end.
This is really clear and great info, thanks again! I was agonizing over both ends grounding vs one end.
But here's the question of what really is the Ground. Is it the chassis? The power supply's ground? Are they the same?
For instance, should the "- (minus/ground)" of the power supply be connected to the chassis on each device, respectfully?
Like preamps, sound amps, headphone amps, etc.

I find it hard to understand what are the best practices. It'd be best not have interference / buzz to begin with, instead of spending some days trying to chase down the source of the problems after they occur 🙂
Your opinion: best anti-interference/buzz/loop design practices and tips that you found practical and proven?
 
Last edited:
Ideally, the signal ground would be part of the two-wire signal, running inside the screened part.

The screen should be grounded at one side only (generally the sending side); you don't want to ground it at both sides, to prevent ground currents in the screen that could induce noise in the signal wires.
Ground it at the sending signal ground or the power supply ground, should make little or no difference but you can test.

If you have an additional wire or pair available you could use that to connect the two chassis to equalize stray voltages without inducing noise.
This additional wire/pair should be outside the screen!

Yes, Muncy and Whitlock are the go-to guys here.

Jan
 
  • Like
Reactions: mo7ies
The screen should be grounded at one side only
Question: what about the ground of unbalanced L/R stereo channels?
We have two pairs of wires:
L signal, L GND
R signal, R GND
The L/R GND are interconnected at the source by default, as it's an unbalanced stereo signal source (preamp.)
Obviously, one of them (say, L GND) should also be connected to the signal receiver (power amp) to form a sound signal circuit. But what about R GND - should I also connect it on the power amp side, or not? (If connected: wouldn't that create multiple paths for the ground L/R with all the negative consequences?)
 
In theory you create a loop, that is correct. But you can assume that the two gsignal grounds at the sending and receiving side have the same level as they will be connected very close or at the same point. So there will not much danger of a ground loop.
OTOH if the two twisted pairs each have their signal grounds connected on both sides that will help screening.
So in this described case I would ground both signal grounds at both sides, but NOT the chassis-connected screen.

Jan