Building cables with Mogami 2549

[dalbert02]

I often build balanced audio cables at work. We use various sizes of XLR-to-1/4 plug or XLR-to-XLR. It is cheaper/faster/easier to build custom lengths than ordering custom lengths online. I always use Neutrik connectors as they last a long time. Anyway, I had a request today for 2 10-foot lengths of RCA terminated cable. I had a handful of RCA plugs but I wasn't sure how to wire it up or if it made any difference at all. For reference, the Mogami cable has 2 inner conductors with an outside braided copper shield. Some people use both internal conductors connected to the center pin, others use one for the center pin and use the other for "ground" and float the shield, and others float the shield on one side only. Does it make a difference? Take a look at the pic linked. Silly snake oil or credible? https://pixy.org/5306653/

 

[Speedskater]

Connecting the cable 'shield' to the RCA shells at both ends, is more important than how you connect the second central conductor.

 

[Salt]

Connecting the cable 'shield' to the RCA shells at both ends, is more important than how you connect the second central conductor.

No ground loop issues with this?

 

[SIY]

No ground loop issues with this?

No more so than any other single-ended shielded cable.

Nor any less so.

 

[Salt]

No more so than any other single-ended shielded cable.

Nor any less so.

So RCA is messed up no matter what topology the cable is (two strands plus braid, or, as the classic one central strand and outer braid)?
Ok, just a philosophical question, as XLR exists, but according to this I preferred the pseudo symmetrical variation with one strand to hot, one to ground and the braid floating to one end.
To be honest, not on scientifical basic, just a 'feeling' that might go right

 

[dalbert02]

So RCA is messed up no matter what topology the cable is (two strands plus braid, or, as the classic one central strand and outer braid)?
Ok, just a philosophical question, as XLR exists, but according to this I preferred the pseudo symmetrical variation with one strand to hot, one to ground and the braid floating to one end.
To be honest, not on scientifical basic, just a 'feeling' that might go right

So like the way in the drawing linked to... I read of some manufacturers doing it this way and hence why they are considered "directional" cables. The source is grounded and the destination is floated. But then this starts to sound like snake oil so I really don't know.

 

[dalbert02]

Connecting the cable 'shield' to the RCA shells at both ends, is more important than how you connect the second central conductor.

So you disagree with the way that is pictured? Would you have some theory as to why it would be bad (...or less good?)

 

[SIY]

So RCA is messed up no matter what topology the cable is

For shorts runs in home systems, 99% of the time, it's fine. Ten feet is pushing it... but still will likely work fine as long as the outer shield is the return.

Would you have some theory as to why it would be bad (...or less good?)

It will work fine but is more complicated than necessary. The ideal single ended cable is coaxial. The closest you can come to that with the constraints of that particular wire is to use the shield as return, and connect the two inner wires together at the Hot pin.

 

[middlemarch]

I was taught as per the Worlds Best Cables suggestion. Basically the braid shields the overall cable and you then try to shunt whatever it picks up to ground via the shortest ground path available, usually at the source end. For what it's worth, not audio related, my experience is in data transmission of various types. But there is definitely an aspect of black magic involved, so your mileage may vary...

 

[sam_adams]

While the below applies specifically to balanced, differential interconnects; ask yourself what would be different if this were an unbalanced interconnect?

This is what happens when you connect two double-insulated (two prong mains cable) devices:

The below explains some of the concerns of the first issue:

If you are going to make your own unbalanced cables, coaxial cable using either RG-58 or RG-59—with stranded center conductors for flexibility—make the best interconnects. Why? See first bulleted point:

If you wish to use shielded-twisted pair wires for interconnects, consider this:

After all of the above, my real world experience has shown:

• Coaxial cables and shielded-twisted pair cables perform almost exactly the same in my equipment setup.

• Unbalanced, shielded-twisted pair cables—made with either star-quad or single twisted pair cables—show the lowest noise if the shield is connected at both ends.

• The one thing about unbalanced cables of either type that produces the lowest noise, is a shield conductor with the lowest spec'ed resistance. For coaxial it's double braided shield and for twisted pair it's double Ruessen shielding.

If you want to get the best noise performance from interconnects, you have to test and measure to see what works best in your environment.

 

[dalbert02]

For shorts runs in home systems, 99% of the time, it's fine. Ten feet is pushing it... but still will likely work fine as long as the outer shield is the return.

It will work fine but is more complicated than necessary. The ideal single ended cable is coaxial. The closest you can come to that with the constraints of that particular wire is to use the shield as return, and connect the two inner wires together at the Hot p

I'm surprised that is the best option as I would expect very small time variations in the arrival of the signal between the two isolated paths. I would have thought using a single center conductor would be the best choice. Maybe Amir has a sensitive time domain reflectometer or vector network analyzer and can measure these things for us!

 

[dalbert02]

While the below applies specifically to balanced, differential interconnects; ask yourself what would be different if this were an unbalanced interconnect?

That is some good stuff right there! Thank you. Once again I learn I don't even know what I don't know!

 

[sam_adams]

I would expect very small time variations in the arrival of the signal between the two isolated paths.

Which would be measured in fractions of a nanosecond.

 

[dalbert02]

Which would be measured in fractions of a nanosecond.

True. But don't we celebrate 1dB differences between DACs with >120 dB SNR? I once worked on a satellite system that had a variable phase combiner that would combine two Ku-band signals. I was shocked at how easily minuscule changes in cable length (or more accurately electrical length) would drastically change phase synchronization.

 

[sam_adams]

. . . Ku-band signals. I was shocked at how easily minuscule changes in cable length (or more accurately electrical length) would drastically change phase synchronization.

The frequency differences are orders of magnitude different between the Ku-band and the audio band. Nothing that happens there—from a time domain perspective—applies to our area of interest.

Admiral Hopper shows us one nanosecond:

 

[Holmz]

I put them together this way, but using the Mogami and Neutrik:

 

[SIY]

I'm surprised that is the best option as I would expect very small time variations in the arrival of the signal between the two isolated paths.

There's no "two isolated paths." The two inner wires are essentially in parallel.

 

[MaxwellsEq]

Unbalanced wiring with RCAs: always use a high quality coaxial cable. Don't use microphone type cables. These are for balanced/differential working.

 

[Sylvia]

For shorts runs in home systems, 99% of the time, it's fine. Ten feet is pushing it... but still will likely work fine as long as the outer shield is the return.

It will work fine but is more complicated than necessary. The ideal single ended cable is coaxial. The closest you can come to that with the constraints of that particular wire is to use the shield as return, and connect the two inner wires together at the Hot pin.

What would be a short run? Anything below 3 meters?
I'm genuinely asking, no disrespect intended

 

[Holmz]

What would be a short run? Anything below 3 meters?
I'm genuinely asking, no disrespect intended

50’ would be getting up there IME.