Balanced vs. unbalanced audio

[jeffjames]

Finally, a source of information that explains how in terms I can understand. Thanks👍

[tahititoutou]

I have worked as a circuit designer. And I approve Paul 200%. 100% for what he said and 100% for what he hasn't.
The "common mode rejection" can be very easily understood using beginner's algebra. There are two wires carrying the signal : let's call them wire A and wire B. Let us say that the signal sent in wire A is S ; thus the signal in wire B is -S. Let us say we have long wires in a noisy environment. This the electromagnetic or electrostatic fields induce in both wires a signal N (for "Noise"). Thus, wire A carries a voltage Va = N + S, and wire B carries Vb = N - S. At the receiving end, the "differential" amplifier amplifies the "difference" between Va and Vb. The "difference" is the result of subtraction. thus the difference is Va - Vb = (N + S) - (N - S) = N - N + S + S = 0 + 2S. The noise has simply disappeared. This is what Paul explained to us.
What he did not is that the noise disappears no matter whether the cable is shielded or not. And, if you carry a signal across a long distance, this is important. Because a shield acts as a capacitor between the signal; and the ground. So in a long distance you lose transient response. Furthermore, a shield blocks a FRACTION of the induced noise whereas subtraction cancels ALL the induced noise. And this is particularly true for small signals and high impedance signals. (This, by the way, is why telephone companies who carry 602-ohm low-level signals across tens or hundreds of kilometres use only balanced lines.)
So, in my system, I use unshielded balanced wires throughout : from cartridge to speakers. I think the most important part is from the cartridge to the preamp.

[RandyLott]

I sometimes work with circuits that amplify picoamps of current. Literally, your measurement will drift by 100% if someone enters the room.

In this case, we use triaxial cables. The center conductor is called the "force" and the first shield is the "guard". These are buffered so that the leakage current will actually be provided by the buffer circuit, thus eliminating leakage from the circuit being tested. The leakage signal is then shielded by the second shield and shunted to ground. Pretty sweet!

[gstanley]

Hi Paul. Nice explanation. I always choose balanced and common mode rejection if available, especially for long runs. Nice to hear someone address the addition level in balanced circuits. I have been convinced that the extra gain has as much if not more effect on quality. Another issue is that most unbalanced interconnects have both ends grounded and the result is a nest of ground loops waiting to happen.

[dell]

My first job out of high school (1966) was being a sound man for the Sheraton Boston hotel, this was a 1, 000 room hotel with good meeting room facilities and a ballroom complex that was substantial.

We made all of our balanced microphone cables (XLR on each end) ourselves in 25 and 50ft lengths, we probably had 2, 000 feet of cables made up. The output of a balanced microphone is 150 ohms and very low level (-56db) so the cable has to be well shielded. The ballroom complex had a projection / sound booth at the rear of the main ballroom above the balcony. There were huge power amps (all tubed) and a mixing board up there and all the microphone socket and speaker wires terminated in that room. There was also a lighting control board next to the sound mixing console

One bugaboo was the ballroom lighting system used SCR based dimmers instead of the customary huge variacs that had been usually used for lighting control. The SCR system worked well but there was switching noise associated with it so the microphone cabling had to be kept well away from any lighting control conduit. Even though the lighting wiring was in steel conduit and the microphone cabling as also in steel conduit if they ran next to each other the hash would be picked up by those beldon shielded microphone cables.

That job gave me good OJT training on balanced systems and noise control.

[SpeakerBuilder]

Many of the more expensive and high quality RCA interconnects use two center leads of the same design to carry the signal, and a separate outer shielding wire that is connected to just one end or the other, so that one component's ground is extended through the cable for RF shielding, but that grounding cable does not carry any signal. The problem is these cables are called balanced, since both signal leads are of the same design (hence the term), yet this is not the same as a true balanced cable that connects to a balanced input, and is distinguished by three connectors vs two. For those of use using components with unbalanced input, this type of "balanced" interconnect cable is the best option. Also, in my systems, I am never running more that a one meter length of interconnect cable, and RF noise from such short runs is negligible.

[turbo]

Hi Paul..really appreciate all these videos you put out...

[CoolDudeClem]

Balanced is great for microphones, a microphone I bought once had an XLR connector on it but the XLR plug was only wired for un-balanced with a standard unbalanced coaxial cable and 6.3mm jack on the end. So I took that off and put a proper cable in to it, then made a preamp that could take advantage of the now balanced audio coming from the microphone and it works so much better, no hum at all, even with a ridiculously long cable! What's annoying is when you need to connect something that only has unbalanced output to something that only has balanced input (or vice versa), fortunately for me I know how to make circuits to convert between the two.

[jolox.]

Very clear, but I have much more to learn before I understand. I'm going to thumbs up and subscribe to support your channel!

[snakeoilaudio]

it is always so funny when people completely forget to talk about the disadvantages once they start selling the stuff ;-)

[pedroseguinot]

Clear as an (IF)GIA Diamond, thank you, Sir.

[andershammer]

True balanced requires twice the circuitry. In my experience balanced helps with noise problems and I would recommend it if you have noise problems but if you don't unbalanced can be fine.

[generalkickass]

Paul I've been wanting to know this subject myself for a long time

[onemoremisfit]

I have some Goertz interconnects that are RCA, with unshielded flat copper ribbon conductors, and also some Alpha Core with twisted pair flat silver conductors, also unshielded RCA. Apparently these use CMR to stay quiet. I also have a Mackie mixer that all the XLR connections are simply impedance balanced, and when one output went dead, I traced it to the resistor that keeps the ground isolated from the signal cold. It's a tiny surface mount component, and I bought a roll of them to get an exact replacement. I have about 499 of them left (lol) in case I ever need another ... meh, the whole roll was about 40 bucks.

[markfischer]

I'm not taking a position one way or the other but I would point out that a lot of electronic instruments use BNC connectors for inputs which are unbalanced using coaxial cable. So does the feed from your cable TV company which also can include your Internet connection not to mention hundreds of high definition TV channels.

To take best advantage of the noise reducing capabilities of balanced cables they are twisted, I think one full twist per foot. This gives the best chance of both conductors getting the same induced noise.

In the war between the cable TV companies who originally used only coaxial cable before fiber optic cable was available and the telephone companies whose so called copper network was twisted pair, the cable TV companies easily won because of the far greater bandwidth capabilities of coax. This does not apply to audio systems because by electrical engineering standards the bandwidth of even high definition signals is miniscule.

[darinbrunet]

Unless you are running long distance cabling or in an electrically noisy environment (such as many pro audio venues), you do not need "balanced" cables/inputs. Also, "balanced" may not be the BEST terminology for this. Differential versus Single-Ended IMO is more appropriate. In most consumer gear, Single-Ended input means fewer parts and lower noise. Yes, balanced has higher noise. Read the excellent white-paper on balanced audio from Rane Corp., one of the granddaddy audio companies and dispell this audio mythos.

[Yootjoob]

For me XLR balanced, the best way to connect your audio equipment.

[hunkychunky]

Hmm I wonder. What about balanced headphone cables? Does the same rules apply?

[arnelcomia]

Thnks, can i modify my 325s grado to balance plug??

[dhpbear]

5:44 - Are you implying that certain manufacturers CLAIM that their products are 'true-balanced' then they are single-ended?