What's better, one conductor or two conductors for an RCA interconnect?

All those examples you gave Kijanki are differential connections where the impedance in each leg is the same. EMI induces a current. The equal impedance means those equal and opposing currents generate equal and opposing voltages which cancel.

The op stated RCA which is not a differential connection. Ground and signal connections have different impedance hence the induced voltages from the equal and opposing currents do not cancel out.

Twisted pairs makes total sense in balanced audio connections. It will be better than non twisted for single endednbut coax is likely to be superior.

Kijanki ... You need to go back and review a bit. Twisted pair is good for magnetic and perhaps low frequency EMI. Coax is superior for RFI (with proper RF shield) and is in general better for electrostatic conduction as you have no field differential between the shield and the inner conductor. Even for magnetic coax can be very good.

Interesting @crustycoot .  That journalist no doubt discovered what was already being used industrially and in the scientific community no doubt for decades prior.

@tcotruvo, coaxial cable is also the cable of choice in almost any scientific application. My understanding is it is better for electrostatic noise rejection but there may be circumstances where shielded twisted pair is better for magnetic. I am going off memory here. It depends on the shape of the noise field. Primary reason for not using co-axial is it is difficult to work with and maintain its properties.

I would expect you would have to have pretty long runs before cable capacitance became important. I am sure there are some bad tube products that could have issues.

I am not ignoring that it is a coaxial cable because it is not a co-axial cable -- the axis of the conductors is not consistent in a twisted pair with shield. Much of the benefit of a coax is the symmetrical arrangement of the shield w.r.t. the center conductor.

A shield around a twisted pair provide electrostatic shielding, but the question was RCA, and the two conductors (as noted by Kijanki) do not see the same electrical impedance which will defeat the benefit of twisting (at least for magnetic I/F) which works where both conductors see the same impedance hence induced voltages cancel. There is probably a reason why oscilloscopes probes, which are single ended, use coaxial cable and not shielded twisted pair.

Single ended RCA cable, the ground is connected on both sides, there is no choice. If there is an additional shield, typically is it better to only connect it on one side.

You are looking at this more theoretically than practically. Excuse me as my expertise is physics (and managing EEs) not day to day EE and a lot of time around semiconductor processing equipment, some of it quite noisy while trying to do precision measurements. We are not passing a reference point. We are connecting the reference point on two pieces of equipment and making them the same. The impedance on the ground connection is not the same as the signal connection. An equal induced current on the ground creates a small voltage only related to the ground circuit. Induced current on the signal creates a voltage a factor of the impedance . You are making the assumption there is not a relevant secondary path for ground, for instance totally disconnected or a large value resistor. How large is a large value?  How large does that resistor need to be to ground so that it is not there?  Poked around enough in tube amps to know that resistor may not be very large at all. What if there is a capacitor?  10 nanofarads? That is only 16K resistance to ground at 1KHZ.

I have a PhD in materials science and close to 2 decades in semiconductors, semiconductor processing, and associate equipment, including forays into process measurement and control ... well exposed to signals in noisy environment. Doing much of the same in batteries now.  I understand the overall issues quite fine and could probably derive a lot from first principles. I understand your argument, I don't agree with all of it and I think you have poorly argued some points specifically as it may apply to real world applications including equipment ground connections and how they relate to loop resistance or impedance. Given the indeterminate ground connection and how that related to the loop resistance all due to the signal connector, I believe you are making erroneous conclusions about what would be common mode noise injection.

A speaker connection is balanced from the standpoint of the speaker, but if you induce noise on the speaker cables, it is not balanced at the amplifier input hence why the amplifier circuits for sensors and bridges, both often floating, are still fully differential circuits.

As I clearly said, from the speaker standpoint, the speaker itself is balanced, however, if you are considering noise on the speaker cables induced back into the amplifier, it most definitely is not balanced because the impedance on either cable is not the same as the other. That should have been clear when I discussed a similar circuit such as a bridge or other similar sensor (or consider it a phono cartridge), that while not grounded itself, is potentially grounded at the phono stage end. Someone else on this thread, if I am not mistaken, makes a differential input phono stage?

If you have two AC connected pieces of equipment, then the ground on either side is never floating even if one piece of equipment has no ground connection due to parasitics. Now obviously those will be lower than if there is a direct ground connection. I was quite clear this would be dependent on implementation. 100K is pretty similar to the loop resistance of single ended RCA connection. I threw a question at some EEs. They said more likely the connection would be a capacitor, at least in the test equipment they develop(ed). I go back to my example of 10nF being 16K at 1kHz. This does null benefits of twisted pair and since the question was twisted pair or coax, it negates potential advantage of coax and swings it towards coax.

Virtually every resource I could find, where they discuss single ended and differential connections agrees, co-ax for single ended, twisted pair for differential. Blue Jeans which appears to get technical direction from a former Belden Engineer agrees. His most recent RCA cable is not a twisted configuration as well. More a modified co-ax.

p.s. bringing up KHz or kHz is pedantic. It is like complaining about spelling / typos in a technical discussion to deflect from the content.

I absolutely guarantee your fellow engineers will write it as KHz. I just did a quick scan of emails from my engineers and I can find many instances of it. It is probably why I do it. Most of these are MSEE and some PhD. I am working on engineers from 3 continents and who knows how many different backgrounds.

w.r.t speaker wires feeding back into the amplifier, or a bridge or phono catridge, you keep saying loop. If there is a ground connection. it is not a loop. This is the mistake you keep making. It is not a loop. It is not about the induced currents being different. It is about where they go once the meet the single ended piece of equipment, i.e. the amplifier, the phono-amp, the bridge amp (which would never be single ended). At the amplifier the resistance, I guess ideally impedance, must be the same so that the induced voltages are the same, or they do not cancel. If you have a ground connection, or simply a difference in termination, as any single ended piece of equipment will be, then you will have an induced voltage difference from common mode injected currents.

Let’s go back to our speaker connection. I have common mode current that goes towards the speaker. Both see the same impedance at the speaker (it is a floating speaker after all). The voltage on both sides of the speaker, for argument/illustration raises the same amount, hence no change in voltage across the speaker. If the current goes the other way, one side sees the impedance of the amplifier. The other (ground) sees the the impedance of the amplifier, but also sees a parasitic path through through that ground and out to "somewhere". Now you have a differential voltage caused by the common mode noise.

This is common mode. There is still the superior symmetric coaxial structure for rejecting the generation of differential currents from electrostatic fields and even for magnetic fields, the coaxial structure provides rejection.

Let’s approach this from an engineering perspective. My search, though limited, resulted in a consensus that for single ended connections, which we are talking, and which you cannot hand wave away a ground connection that is a resistance/impedance significant with respect to the termination resistance/impedance, that coaxial cables are superior for rejecting external EMI. If I am wrong, and what you say is true, there should be ample evidence of this on the web one would prefer with measurements. Here are some measurements done by an engineer who posts on ASR (not Amir/a mod). I also have another posts from a cable company where they talk about most RCA connections being fully grounded (I know I saw that in the past when playing around with tube equipment). That throws out that big resistor to ground in all cases argument and makes the case strongly for my system dependent argument.

Your thinking is flawed @kijanki. You have an erroneous understanding of how noise induces (or does not) induce current into these circuits and hence have drawn incorrect conclusions about what the result will be. If the noise is flowing in a loop, through the speaker (or cartridge), it is not common mode noise, it is differential.

Let us change the problem to a phono cartridge and a pre-amp. By your understanding, even if one side of the phono amp input is grounded, the system is "differential", and hence will receive the full benefits of twisted pair wiring. That is wrong. I noted that @atmasphere who supported you above uses a true differential input on the phono inputs to his preamplifier (for improved noise rejection).

The flaw in your logic is that with common mode noise, the current does not (have to) flow through the turntable cartridge (or speaker as the case may be). A simple case is electrostatic coupling which capacitively couples a current onto both wires (common mode). No wire "loop" is required to induce current as the parasitic capacitors forms the circuit. However, since you have one side of the phono pre-amp "grounded" the current in that wire flows into the ground (or some portion of it depending on whether hard ground, resistor, or capacitor. The current in the other wire flows into the loading network, offset by whatever current did not flow to ground (from the other side) but also flowed into the loading network. The result is of course a noise voltage. Coax has superior electrostatic noise rejection compared to even shielded twisted pair. Without a proper fully differential system, there is no benefit of twisted pair.

For audio I would assume we can ignore RF antenna coupling onto both wires (common mode), or at least we can ignore the effects which could be similar to electrostatic coupling. If we include them, again, the single ended connection defeats any benefit of twisted pair wiring as above, while the twisted pair has less rejection than coaxial, even shielded.

For magnetic EMI coupling, there would need to be a differential current generated at the point of interference as a complete electrical loop, for practical purposes, would need to exist to have a magnetically induced current. A twisted pair does, obviously, reject magnetic coupling of EMI, but so does a coax (through arguably a different mechanism) and it does it rather well. In most practical cases, because the differential current from magnetic interference happens at the point of interference and not at the load/receiver, the balanced connection does not matter for magnetic interference. (note the experiment link I provided that showed in that case superior rejection of magnetic interference with coax).

So, @jumia, I go back to my original statement, still true, explained in great detail, apparently in agreement with others knowledgeable about the subject, that typically coax will be the superior cable for single ended connections.

@jumia , this is not touting my horn, but questioning the competence of many cable companies, but I expect few of them, based on the technical quality of their communication, have the understanding to either refute or agree with what I write. Some of them I am sure do have that understanding.

It is often best to connect the shield on only one end. Is it always best?

@jumia

A shielded RCA cable (not coax) is 2 conductors, one for ground, and one for signal, plus a 3rd "conductor" which is the shield. The shield will extend the full length of the cable, but only connect on one end. The shield is effectively a Faraday cage preventing RF energy from getting into cables inside of it. Yes, the shield is potentially an antenna. However, if it is connected on only one end, any RF energy should be shunted harmlessly to ground.

If you connect the shield on both ends, and the RF energy generates a large current in the shield, then it could induce a difference in ground voltage which will (or could) show up in the signal on a single ended system. That would also require the stars to align such you had a huge amount of RF and that your audio system converts the RF into something audible. More likely is an electrostatic conduction path which the shield will provide some protection for, or not and/or a magnetic path for which it may provide protection or not. There is more complexity than this and it would take far too much space to put here so I found you a good link. (and RF is obviously both electric and magnetic field).

Most fields have a lot of lore, pass down through the ages, repeated, reinforced, and wrong. Just look at this hobby! Electrical Engineering is no different (and nor is physics). What works in one situation may not work in another, hence rules of thumb often cause bruised thumbs.

This particular link takes a very hard line on cable shielding (not system level noise). From a practical stand point as an end user, you cannot ignore system level noise in order to perfect shielding, but your takeaway should be that unless everything is designed properly (his examples of mil-spec and FAA), then there is no 1 correct way for every instance, only a correct way for your particular instance. I know most of this paper will likely be over your head, but I think perusing it will give you a feel for what the issues:

https://www.emcstandards.co.uk/cable-shield-grounded-at-one-end-only