Balanced Outputs
Comments
-
Another advantage with respect to fully differential designs (as in balanced from input to out not the quasi balanced type with just a summer/inverter added at the input/output ) is that the common mode rejection will also have a benificial effect of like noise cancelling any distortion common to both sides of the circuit.An excellent example of a true balanced design is Pass's Super Symmetrical circuit used in his X,XA series of amplifiers.Heres an informative read on the subject.http://www.passlabs.com/pdfs/articles/susy.pdf
-
And here is an article that goes into this very issue and really does a good job of presenting the other side of the coin from your argument, FTGV.
http://www.stereophile.com/features/335/index.html
If you notice on page 2 of the article, the illustration of the common balance input, you will notice that the CMRR stage output is single ended. Converting this back to balance at the output illustrated (or to drive the rest of any piece of equipment), the circuit then using a high input of a op-amp and the low input on the other side. This alone is a sonic defect, as op-amps are sonically different between the inverting or non-inverting inputs.
The argument that it simply sounds better being balanced is a incomplete statement. Why would it sound better? Often being balanced includes additional circuits stages which is an additional burden. Or the penalty of requiring extremely precise component matching when fully balanced. (Please read the link before you think it automatically has to be superior when "fully balanced"!) Answers why to use balanced inconnect include the highly desirable and very underrated seperating of the grounds between devices. Or known as current loops getting interconnected.
I most certainly not saying the balanced or unbalanced topology or interconnect is better or worse, just that the reasons why are clear as to advantages and disadvantages, and circuit techniques are often are far more important than just balanced or not. I'm also not saying there aren't superior circuit techniques on balanced inputs. They aren't so common though...
CJA so called science type proudly says... "I do realize that I would fool myself all the time, about listening conclusions and many other observations, if I did listen before buying. That’s why I don’t, I bought all of my current gear based on technical parameters alone, such as specs and measurements."
More amazing Internet Science Pink Panther wisdom..."My DAC has since been upgraded from Mark Levinson to Topping." -
The true "fully balanced" topology I'm refering to is where the positive and inverted halves of the signal each have their own individual signal paths.Unlike the opamp circuit you site each half is identical from input to output and in the case of Pass and Bryston consist of matched discrete transistors not IC opamps.Does the fact that they are fully balanced make them superior to amps that aren't?Not necesarily but it is highly likely this class of amp will sound better when fed a balanced signal than one that is single ended.
As for extra circuit complexity of fully balanced design ,while there is a complete duplication thus twice the amount of components each half of the signal only sees it's half of the circuit.In the "quasi balanced' units that use opamps for input summing or output invertion, the inverted half of the signal will see atleast one extra stage.Also they will not have the common mode distortion canceling properties of the fully balanced units. -
As has been pointed out in the past as well as now, the double voltage inherent in the balanced output/input process can result in a 6dB increase in loudness at the same volume control setting if the balanced input circuitry doesn't have a 6db lower gain, as should be done. If this wasn't done accurately a mere increase in loudness could be confused with some inherent superiority in sound quality, which the balanced connection doesn't possess.
As to the noise floor, the bottom line should be if it's actually audible in use. This is one of the few audio parameters that can actually be evaluated accurately by ear. Set the volume control to max and put your ear a few inches away from the speaker(no program playing, of course). With modern receivers or other amplifying equipment at most a very faint hiss would be heard in most home environments, and this would be inaudible either at normal volume settings or at normal listening distances. If nothing can be heard in moments when the program material is silent, no audible benefit from noise floor reduction results from using a balanced connection which may measure even a few dB quiter with instruments more sensitive than our ears. Professional applications involving long distances and an environment with more sources of interference may benefit from balanced connections, but in most homes(including mine)the ordinary shielded coaxial cables do the job without a problem.
