going balanced/symetrical - what benefits?

[Pan]

Michael,

that wouldn't be true for an impedance balanced solution right?


/Peter

[caesar148]

Balanced circuitry has no place in consumer audio. It is all BS.

Balanced audio has two benefits:

1) Reduced common mode noise in cables. This application is seen most commonly in microphone. Due to the low signal level and long cables, balanced/differential signaling has its place in audio recording and public addressing systems.

In consumer electronics the interconnect cables are short (less than 2 m), well shielded and the signal levels are considerably larger than microphone signals. Unless of course you have the CD player at one end of the room and the preamp or amplifier at the other end, but why would you do such a thing?

Common mode noise cancellation only works if the two differential signals are kept close together. In differential cables the wires are twisted together to provide a tight coupling. Having them apart as in differential electronics defeats the purpose.

2) Balanced signaling quadruples the output power of amplifiers because of the doubling of voltage. But what good is power if it does not have the current capability? It's like a Ferrari running on bicycle tires. I question the need for such power anyway. Who needs a 1000 watt amplifier? We are talking about consumer audio here not professional audio.

Here are the cons of differential signaling/electronics:

1) Increased noise. We have two sets of electronic circuits working in differential mode. This doubles the random thermal noise in the circuit. Differential amplification sums the two signals and we get double the noise. I used to have a power amplifier from a highly reputed company which has a mono switch. In mono mode the background noise was considerable higher. I am not surprise if it went up by 6dB but I have no way to confirm it.

2) Increased distortion. Phase difference between the + and - signals results in signal distortion. No electronic is perfect or perfectly symmetrical. One signal will arrive later or earlier than the other. And the sum of it is distortion! And talking about distortion.....Each set of electronic introduces its own set of asymmetrical distortion and the sum of it is double the distortion!

3) More expensive and complicated electronics. Why pay more for no added benefit?

Balanced audio supporter argue that the greatest benefit is the cancellation of symmetrical nonlinearity. The argument is that the + and - circuitry will produce distortion that is symmetrical so the sum of it is the cancellation of the distortion. The way I see it, the benefit is negligible because the THD of modern audio circuitry is already very good. Can you tell the difference in distortion between 0.01% and 0.001%? Tube amps are no where close to this anyway and they can sound good but this is another topic on its own.

There are some companies that swear by it like Ayre and Balanced Audio. McCormack DNA-500 and the McIntosh MC1201 are balanced amplifiers. Yes they sound good but you pay for extra circuitry that you don't need. I believe in the KISS philosophy: Keep It Simple Stupid!

[pooge]

Balanced is about equal impedances to ground, not equal signal levels. Signal levels have nothing to do with balanced. Equal impedances help cancel common mode noise, not signal.

Equal impedance to ground does not require double the circuitry.

If balancing is done with transformers, you can break the ground connection between components. The complete and utter elimination of hum is a definite benefit.

[jupiterjune]

Balanced circuitry has no place in consumer audio. It is all BS.

Hey, Ceasar-
If you are ever in New Jersey, stop by my house. You can listen to the noise that comes through the ground with single-ended circuitry. It can be clearly heard from my listening chair.
With true balanced circuitry, poof, noise is only audible about 5 inches from the speaker.

Common mode noise cancellation only works if the two differential signals are kept close together. In differential cables the wires are twisted together to provide a tight coupling. Having them apart as in differential electronics defeats the purpose.

Actually, I disagree. The true noise reduction occurs when the common signal (ie ground noise) is fed to a differential pair. The noise simply can't be amplified. Take a look at the Son of Zen article by Nelson Pass for a good explanation of the mechanism that causes this. It took quite a while before it 'clicked' in my head, but once you see it, it is clear.

JJ

[Steve Eddy]

Originally posted by jupiterjune
Actually, I disagree. The true noise reduction occurs when the common signal (ie ground noise) is fed to a differential pair. The noise simply can't be amplified.

Well, I don't think it's too far a stretch to say that reducing the amount of noise that's picked up in the first place should be considered "noise reduction."

I mean, you've really got two things going on here. You've got the cables where the noise would be picked up and then you've got the differential input which rejects what noise is picked up. Even if you have a great differential input, its performance can only be compromised if you're not doing what you can to reduce the amount of noise that's picked up in the first place.

se

[Gordy]

Originally posted by jupiterjune

...You can listen to the noise that comes through the ground with single-ended circuitry.

Originally posted by caesar148

2) Increased distortion. Phase difference between the + and - signals results in signal distortion. No electronic is perfect or perfectly symmetrical. One signal will arrive later or earlier than the other. And the sum of it is distortion! And talking about distortion.....Each set of electronic introduces its own set of asymmetrical distortion and the sum of it is double the distortion!

Both balanced and unbalanced topologies have their advantages and disadvantages. Within a professional environment there may be no option, however within a domestic environment it is different and there is a clear choice. Ultimately there are key deciding factors in your choice, and these keys have been noted by jupiterjune and caesar as shown above.

With unbalanced design the key factor is the dependance on a signal referenced to ground. In many situations / installations the 'ground quality' can be very poor, as alluded to by jupiterjune, and this can lead to noise, hum-loops and dynamic distortions not measurable under static / laboratory conditions. The more ground circuits that are connected together, the worse it is likely to be.

With balanced design the key factor is the dependance upon exactly matched positive and negative phases, which calls for precisely matched electronics. If the phases are not matched, as noted by caesar, the signal will simply be distorted away from the original. Personally I believe that this is one of the reasons why much commercial domestic equipment sounds poor when used in balanced mode... the commercial constraints of the design do not allow for precisely mathed + / - phases and hence the signal is not faithful to the original.

So when you consider the question of balanced vs unbalanced in a domestic environment think hard about precisely matching + / - phases for balanced circuits, and optimising your ground distribution for unbalanced working.