What are the differences? Is one better than the other? Always? Sometimes? The CD player I'm purchasing has the facility for XLR, but I was told by my friend who heard the player, that it sounded better with single ended. His feeling is single ended usually wins in coherence and musicality. Feedback, as always, would be appreciated. Thanks in advance,
Balanced lines use three wires with the outside wire being a shield around the inside two. The inside two wires are thus shielded from external RF.
Unbalanced audio lines use just two wires, there is no shield. The signalpasses down the two wires unshielded from outside RF.

XLR is also recommended in long(er) interconnects

The unbalanced input impedance of the preamplifier is at worst case 10 kOhms. In balanced mode, the input impedance is higher, with a differential impedance of at least 20 kOhms.

I made it a point to use all XLR in my setup..some systems often respond better to Balanced configuration

I replaced my XLR with RCA (3.5m) and I got a slight buzz..

take a peak at http://www.balanced.com/faq/balanced.html

its a great page

I am not an expert in this area, but here's what I know.

In a single-ended setup, there is one wire which caries alle the audio signal, and the hield which is usually tied to ground. The shield is there to keep RF noise out. This solution is not 100% perfect; on longer runs of cable, for example, you could still get some noise on the signal line, with no way to get rid of it.

In a balanced setup, there are a number of signals on the cable. I believe it works something like this: Hot pin carries the signal, and the cold pin carries the inverse of the signal. If there is noise on the line, it will offset the hot and cold evenly, and therefor you can get rid of it on the other end of the line, but I have forgotten exactly how. In essence, the system also does not need a ground, so it helps with ground loops. (Like I said, not an expert ;)

On truly balanced equipment, many people seem to agree that balanced cables work best. I have a Mark Levinson pre and power amp; they are interconnected with balanced cables, and the result sounds a bit more pleasing than when I use single ended cables.

On the other hand, my Sony SCD-1 is connected to the preamp with balanced cables, but it does not sound quite as pleasing as it does with single ended cables. I am told this is because the Sony is internally not truly balanced, rather they use a converter to change the single ended signal to a balanced one.

In my opinion, the best thing to do is to listen to either setup and see what's best. I did not do that with my SCD-1 since I assumed balanced was better. Wrong choice.

All other things being equal balanced connections in a truly balanced system should be better.
Balanced connections have two signals, a positive and a negative signal, which are inverted versions of each other. They carry the signal currents. At the destination, these are subtracted from each other to create the signal. Any noise that is the same on both cables will be subtracted-out. There is also usually a ground that does not carry signal current - it's function is to force the grounds in the two components to the same potential. The ground conductor can be a wire or an overall shield surrounding the other two signal conductors.

Unbalanced or Single-Ended connections have a signal and a ground RETURN. Both of these carry current and the signal is driven at the source with the return current flowing back to the source on the ground conductor. The ground conductor can be a wire or a coaxial shield, depending on the design.

XLR's are superior is a number of ways. First, they are locking, so they won't pull loose. Second, they have male and female, so you can't accidently plug them in backwards. Third, they use a two signal system comprising an uninverted and an inverted signal plus a third wire ground, which carries no signal current. This "balanced" or "differential" signalling system allows the receiving component to have better noise rejection than with Single-Ended (SE) signalling. This is possible because the receiving component subtracts the + and - signals to get the signal that is used in the component. If there is any noise that is common to both + and -, this will be eliminated by the subtraction process. This is referred to as "common-mode" noise, which can include 60 Hz hum, Schott noise, RFI etc...

The thing that you need to be aware of is that all designers do not have good designs for Balanced inputs and outputs. To determine which of SE and Balanced sounds better, you need to listen to both on a given component and choose the one that sounds best.
Better coherence and musicality with SE rather than XLRs? Thoughts?
I am with Warrenh. Even though Audioengr's points are valid, the last one is the biggest concern with balanced systems, the impossibility of designing a truly balanced system.

A system that is balanced from beginning to end is actually two chains of amplification. From source to preamp to power amp you have one string of amplifiers amplifying the inverted signal and a different string of amplifiers amplifying the non-inverted. These two signals are combined at the speaker. Even though they should be the same, with tolerances in components it is impossible to build two identical amplifiers.

It is highly unlikely that both signals will be treated exactly the same throughout this process so the reconstructed signal at the speakers is distorted. Yes, there are distortions in single ended systems like in all systems, but this one introduced by the balanced systems is not present in a single ended system.

Also, digital signals start out being single ended on the disc and must be converted to balanced somewhere along the path. This is another source of distortion.

As for noise rejection, balanced does have the theoretical upper hand on long lengths of cable. But with cable lengths and the amplitude of the signals in a typical system you could not tell any difference.
Like many other topics in the world of high-end audio this one is controversial. Balanced lines have been around for a very long time now in the pro-sound world due to the extraordinarily long cable runs often necessary for studio, stage and performance. With cable runs sometimes in the hundredes of feet noise becomes a real issue. But most audiophiles are rarely using runs that exceed 20 feet. Jeff Rowland was one of the first in high-end audio to use true balanced circuits in his amps and preamps and does so to this day as he feels it to be superior. And his gear is certainly dead quiet, not to mention some of the finest sounding gear available. Balanced Audio Technology has also been very sucessful with their balanced amps and preamps. Be aware, however, that while some manufactures have balanced XLR lines in and out, these may not be true balanced circuits such as those used by Rowland and may offer no sonic advantage and may very well not sound as good as the single ended inputs and outputs on the same piece of gear.

Another very significant point, and herein lies the controversy, is the number of truly excellent and highly regarded amps and preamps that only come single ended. Names such as Convergent Audio Technologies (CAT), Jadis, Brown Electronic Labs (BEL), Audible Illusions, Joule Electra, and Blue Circle Audio, just to name a few, come to mind. I personally have owned many many components over the years and have had systems that were fully balanced (except my turntable), but most that were not. I also have been using 16 to 18 foot interconnects between my preamp and amp for over ten years, and to date the best systems I've put together have been single ended. In the end I believe you will find it is the component itself that makes the difference, not whether it is single ended or balanced. My advice is to choose the components you like the sound of using your ears rather than someone's sales hype. No matter what you say or do there will be those in the balanced camp and those in the single ended camp. Neither is right, neither is wrong, just different approaches to the same end. All of this stuff sounds good, the trick is finding out what your preferences are and working toward finding components that are a synergistic match with each other.
Great posts, but Herman I don't undersatnd how balanced operation results in UNbalanced summation at the speaker...any more than a single-ended design. It's pretty clear that preference is component specific. My EMC-1 MkII CDP, for example, sounds clearly better running balanced.
As my Aleph P and monos are running truly balanced, I can get away with longish UNshielded XLRs (Nordost) with utter quietness and VERY fine low-level detail resolution. Would I have the same result with RCAs? Haven't tried, but don't want to, either...especially with the house breaker panel under the same room. Shades of grey, eh?
Hi Subaruguru, I said they were combined at the speaker. Not that they were unabalanced. The electrical signal that is applied to the speaker terminals is balanced but the motion of the speaker is not.

Each driver can only move in one direction at a time, The cone is either moving in or it is moving out but can't go both ways at once. So in that sense it's motion is single ended.

I agree that there are well designed, great sounding components in each camp. Since I am firmly in the camp that simpler is better, I'm going with single ended. I'm using a passive preamp, my power amps have 2 tubes each, and my speakers only have one driver.
I was using the balanced outputsof my Electrocompaniet EMC-1 into the balanced inputs of my integrated amp (a non-fully balanced design) and it was an improvement over the use of RCA connections. Then I switched to a different preamp (passive) without balanced connections and I think it sounds even better. It seems to depend on many things.
Impossible to generalize, but I would definitely always at least try balanced connection on any gear that is *truly balanced* interally. On gear that is not, but provides for balanced connection anyway, experiment at your option - it's really the only way to know what you'll get.

BTW Audioengr, I take it from your comment about male/female connectors and not being able to hook up things 'backwards' that you are down with the theory of wire directionality - or were you just referring to not being able to accidentally plug an input into an input or an output into an output?
Ernie, the Pass Aleph products aren't truly balanced, are they? I mean, single-ended was sort of the raison d'etre of the Aleph line, wasn't it?
I believe the Pass Aleph line used a single gain stage in a fully balanced configuration.
I believe that Pass uses double single ended stages for each channel to create a true balanced system.
Yup, Alephs are truly balanced, and the EMC-1 likes 'em!
My mistake. The Aleph 3 did not have XLR inputs though. But the others did.
Yes, tht's right, but isn't the Aleph 30 successor balanced too?
Subaruguru, BTW nice system.