Balanced Interconnect Recommendations

If the circuit topology of your components are not fully differentially balanced from input to output then there is zero sonic benefit of using balanced ICs.  In fact you could even make the sound worse in some applications such as taking a balanced signal from a CD and sending through an unbalanced amp which creates unnecessary conversions that may degrade sonics. 

But if you take a balanced signal and connect it to a preamplifier that is truly balanced - it keeps the signal balanced from input to output, then the preamplifier’s balanced output can then be connected with balanced cables to a power amplifier, whose input circuitry is also balanced. In this case using balanced connections has theoretical advantages over unbalanced connections. These include the benefits (in theory) of lower noise and randomization of DAC non-linearity plus balanced connection noise immunity.  Whether you notice the difference varies from system to system and person to person.  

Well said. Beautiful. 

I concur with 3_easy_payments.
The AES balanced standard was made for recording studios. Its' implementation in home stereo has only become available in recent decades. Unfortunately, not all manufacturers design equipment using the balanced standard according to the standard.
One of the benefits of using XLR's with true balanced equipment is that cable design should not make a significant difference in sound quality.
You don't have to take my word. PM Ralph Karsten (atmasphere) and get his take on this.
Bob

If the circuit topology of your components are not fully differentially balanced from input to output then there is zero sonic benefit of using balanced ICs.

While this belief is held by many audiophiles, and I believe it is true in the majority of cases, I would not consider it to be a hard and fast rule. It is certainly possible, IMO, that a balanced connection between components having unbalanced internal signal paths (i.e., components that are not "fully balanced") may sound better than an unbalanced interconnection in some or many cases.

Why? There are a number of potential reasons, depending on the designs of the specific components being connected as well as on the cable that is being used:

1) When the interconnection is changed from RCA to XLR or vice-versa the configuration of the interface circuits that are being used in the two interconnected components is being changed;

2) Susceptibility to ground loop effects is being changed. Ground loop effects can cause a multitude of subtle sonic effects, not just hum. And those effects will tend to have little if any predictability, as they depend on factors that are usually unknown, such as the internal grounding configurations of the two components, and on whether each component connects the shields of the cables to circuit ground or to chassis, and on how circuit ground and chassis are interconnected in each component. Both technical factors and my perception of anecdotal evidence that has reported over the years suggest that balanced interconnections are usually less susceptible to ground loop effects than unbalanced interconnections;

3) In many cases input and output impedances will be different;

4) In many cases overall system gain will be increased by 6 db with a balanced connection;

5) Sensitivity to cable differences will be reduced or eliminated in the case of balanced interconnections if the interface circuits in the equipment conform to the AES48 balanced standard Ralph (Atmasphere) has often referred to;

6) And speaking more generally the basic operating principles of the two kinds of interfaces are different.

7) And of course the cable is different.

***None of that has anything to do with whether or not the internal signal paths of the two components are balanced.***

And a lot of that doesn’t have much if any relation to the lengths of the cables or to the noise environment either.

As is often the case in audio, the only way to know for sure in a specific system is to try it both ways.

Regards,
-- Al

@almarg   So your contention is that the benefit of common-mode rejection inherent to an XLR connection may provide an advantage over single-ended connections in some cases?  I'm not sure common-mode rejection is the most important factor to be weighing when selecting cables but I can't disagree with you 100% either (only 99.9% ;-)).  For instance if you're talking about a short length and a low-gain signal then a single-ended circuit may very well outperform balanced regardless of common-mode rejection.  This discussion feels more theoretical than practical and it sounds like we both agree that the vast majority of the time the user is much better off in terms of both cost and performance to only go with XLR when engaged with  fully balanced circuit.  And agreed, if you don't mind spending the money and investing the time, sure it never hurts to try both ways.  I just think there are higher odds pursuits to improve audio quality.