The advantages to balanced operation are only gained when it is a "truly balanced" circuit, i.e., there are two separate, but identical, signal paths. One for the positive signal and one for the inverted signal. When the two are combined the noise is canceled out. Many audio components have balanced connectors but are not "truly balanced". Your example of the McCormack amp is indicative of it not being "truly balanced". Think of these amps as having balanced to RCA adapters, similar to the Cardas (among others), built in.
26 responses Add your response
Yes it is better but RCA is good enough for most audiophile applications.
I use balanced XLR as I have active speakers and that is the connection they accept.
I don't even remember the brand of cable I use. I picked the cables up at a musical instruments store...they are used for microphones, guitars and with studio grade gear.
"Assuming fully balanced equipment"
That's the key right there. If the equipment (but principally the amp and preamp) has fully balanced circuit topology, then 'yes" balanced is better -- and for all the reasons you mention, and one other very important one you didn't:
You're giving the circuits more (voltage, usually) to work with (resulting in better signal to noise values) and conversely, you are engaging the circuit(s) to their full design capacity (resulting in lower distortion levels.)
So although the advantages you mention are real, they all have to do with the differences between the cables themselves (SE vs. BAL) whereas my focus has more to do with what the equipment can do with a balanced signal vs. a single-ended signal. The reason Steve McCormack said what he did, is that the DNA 500 uses the balanced input "as is" (the amp is a fully balanced design BTW) and has (I presume) alternate input circuits that "convert" SE input signals into balanced form in order to take full advantage of the amp's balanced design. This is a nice feature, but notwithstanding his remarks, I'd feed it (the amp) a balanced signal if I had one available.
Eldartford, when you have twice the signal, you can run the device with half the gain. That lowers the noise floor considerably. As for the ICs themselves, it's not so much that they resist noise pick-up any better than a SE IC. It's just that the noise in one signal conductor is 180 deg. out of phase with the noise in the other signal conductor, so it self cancels.
Pmotz, the McCormack DNA 500 has fully balanced circuit topology from input to output. Specs are on their website:
Could not agree more with Eldartford, see this link which gives a "balanced" explanation (forgive the pun)
The article explains that it is all about less external noise pick up in the circuits and wires themselves. This should not be a problem for most audiophile applications. In most cases there will be no difference between balanced and unbalanced in an audiophile environment (provided manufacturers have used similar high quality components and an appropriate design topology).
Pro studio noise problems however, are a common nuisance, which is why pros prefer balanced; studios are jam packed with all kinds of portable gear, including the artist's personal instruments, and preferred rental gear, all of which leads to a much noisier electrical environment that changes on a weekly basis. Pros do not prefer balanced circuit design because the sound is intrinsically better because it isn't....it is simply to reduce electrical noise pick up.
It all depends on the circuits on both ends and the quality of interconnects. Many assume that balanced interconnects bearing a certain "name" will perform the same or better than their single-ended counterparts from the same manufacturer and bearing the same name. All I can say is that they are sharing perhaps materials and design concepts, but electrically the two interconnects are very different. It is not a given that the balanced cables are superior!
Now to the circuits, and the easiest way to demonstrate the incoclusiveness of any such recommendation is by way of a real-life example... The equipment on hand was a $15K Spectral preamp and about the same cost of a dCS DAC. Both units were capable of single-ended or balanced connections (they are fully balanced internally) and both can also invert phase, a fact which will play a role in this story... Have patience...
As it turned out, and that was very consistent, the single-ended interconnects produced the best sound with both units in a non-inverting phase mode. This was odd, so we tried inverting phase on both. Low and behold - now the balanced connection was superior.
The only sane explanation is that the inverting and non-inverting modes in one or both units have enough circuit differences to cause such an anomaly.
This ambiguity took place between a $30K worth of hardware, if that little detail escaped any of you that are still reading... Go figure!
Well, I read Mr. Mackie's article, and for starters, he apparently is unaware that "ground" in a SE circuit is "common ground" or "chassis ground" whereas "signal ground" in a BAL circuit is separate from the common/chassis ground, and therefore presents a different (and quieter) reference for the signal. If you look closely at an XLR connector, you will see it has two separate ground connections: signal ground (usually pin 1) and common/chassis ground provided by the bayonet connection on the barrel of the XLR.
This is another way balanced circuitry provides cleaner signal processing. I'm sorry, but I believe Mr. Mackie's article contains misleading information which in the end compromises his conclusion(s). Just my opinion of course ;~))
Are you sure we are not misconstruing balanced input/outputs and complemetry topology? The magic of cancelling noise that is picked up in balanced interconnects occurs in the balanced input stage that converts the balanced signal to a single ended one.
From that point on the amplifier can be of complementry design.
Thanks guys, this is the type of interesting discussion from experienced folks I was hoping this thread would generate. I am in the process of moving equipment around which prompted getting some 3m ics so I can shorten speaker cable runs. I am going to try some sonoran single-ended ics, which Robert at Star Sound assures me will not have RF of EMI problems. I am still considering different equipment, which may include a single ended preamp, so I was encouraged with how good the single ended connections sounded on my Muse front end, since the flexibility of not going balanced opens up more preamp possibilities. From what I have researched and heard so far, balanced connectors have their roots in pro audio and are not necessarily better for home audio, and my listening so far has shown that to be true in some ways.
There is a lot of literature and it can be confusing as not everyone connects things in the same way or even grounds components in the same manner.
Here is another link
XLR balanced connections usually have Pin 1 as both cable shield and chassis ground. Pin 2 and 3 carry the signals. The connector is just a connector and protector.... it is most often not electrically connected to the cable wires even if it connects to the chassis ground of the component when the connection is made.
I fully agree with some comments above that balanced will have a cleaner signal processing as the -ve signal is not connected to the chassis ground as it most often is in RCA. Cleaner signal equals less noise pick up. Balanced clearly has an advantage in this sense.
I think the differences on this thread revolve around whether a more complex design and cleaner signal path is needed for most home audio. This is a matter of opinion, circumstances and personal experience. I think people will always draw the line differently.
The many manufacturers who do not bother with the additional cost of balanced designs and the many audiophiles who still pay huge $$$ for components with only the cheaper RCA connections and unbalanced designs, obviously believe RCA and chassis ground is good enough for the -ve signal. The home audio market may not be educated about the problems of noise or has never suffered significant noise problems with connecting a variety of components, but the market reflects people's opinions in what is built and sold.
My thoughts are if the equipment is truly balanced and allows for a truly balanced connection, run it that way. Why do you want to go through a process of changing it to SE? It's like going from SE to balanced, why go through the added splitter or whatever to get it when it really want sound a bit better, if as good? My point is, keep things simple.
My Calypso preamp takes anything single ended and converts it to balanced internally. Therefore, why would I not want to run balanced interconnects so it wasn't required to do that.
I definitely think there's more to the picture than noise. The up or down conversion is an issue also. I trust manufacturers like Ayre, Theta, Aesthetix and others who say balanced is better with their equipment. I know they have done their homework.
I think Shadorne's comment:
"The home audio market may not be educated about the problems of noise or has never suffered significant noise problems "
is important, because in discussing BAL vs. SE, it's the noise you can't hear (consciously) that contributes to (what I consider) a qualitative difference between the two.
Steve is correct about the location of the noise cancellation -- it's at the input end of the cable. However, components with truly balanced circuit topology do not convert balanced input signals to SE. Rather they convert SE signals to pseudo-balanced signals.
Nsgarch...Twice the signal and twice the noise. The only thing that matters is the signal-to-noise ratio, and this will be exactly the same for single ended or complementary circuitry, assuming the inverted signal circuit path performs the same as the inphase path.
The so called "extra" circuit needed to drive a complementary amplifier from a single ended input is in many cases an illusion. The amplifier usually includes a buffer stage at the input. With balanced input this buffer is the same (inverting or non-inverting)for both sides of the signal.
If you want to use an unbalanced input, one buffer is wired to be inverting and the other non-inverting. No extra circuitry is required.
Another thought I have heard regards phase effects, and the inability of some balanced cable makers to appropriately design cables to resist phase shift, which might be one reason some believe single-ended sounds better for home audio. I heard one of the successes of using bybees is to correct phase shift problems. Any truth to any of that?
McCormack uses a transformer (Jensen I believe) to perform this task in his upgrades. You can read about it on his sight. That is an extra conmponent. The Sony SCD-1 used a circuit to derive balanced and it didn't sound as good as SE. As I said, I've been messing around with this stuff and some sound better balanced, some worse and some sound better SE. There ways to implement this and it's not all about noise in all cases. Some implementations degrade the signal. All you have to do is listen for yourself.
Eldartford... "Twice the signal and twice the noise" is not a correct description of a balanced differential input circuit. Noise adds up when uncorrelated, but the circuit will reject supply noise and RF pickup better than the single-ended circuit.
In addition, the differential circuit cancels out even harmonics, for a lower measured distortion.
All said and done, the single-ended circuit is inherently simpler and - in my opinion - can provide the ultimate sound purity. If noise levels are acceptable (i.e. close to non-existent) then I much prefer the SE amplification. To me, it is the realization of "as simple as possible, but not simpler..."
Serus..."Twice the signal and twice the noise" is more accurate than "Twice the signal (period)" which is the common misconception that I wanted to correct. Complementary circuitry will reject common mode noise on the signal and distortion, for example power supply ripple, assuming that the (+) and (-) rails of the power supply have the same noise.
*Bsic balanced is slightly more expensive for obvious reasons (complementary circuits)
*Balanced is useful because you just have to match circuits -- not components (so easier to implement linearity) + you get common mode rejection.
*Of course, in balanced you get twice the circuit noise
*Single-ended, in "ultimate" implementation, would be extremely expensive -- it would require component matching for each channel and extremely careful ground to ahcieve ultimate linearity. Forget it -- but that would be the best.
Finally, the rca connections are horrible; the xlrs are much better & reliable. No wonder pros use balanced as standard.
>>Finally, the rca connections are horrible; the xlrs are much better & reliable<<
Totally untrue. There are some extremely high quality RCA connectors available just as there are very poor balanced connectors. Many manufacturers use only single ended connections on their amps and preamps because they believe them to provide performance either superior or equal to balanced connections. OTOH some manufacturers such as BAT and Atma-Sphere design their products for balanced connections. As always YMMV.
I use XLR cabling between components and even for runs to my active speakers but I have no illusions - the sound quality is no better than RCA (to my hearing). As some have pointed out, there is indeed better noise rejection and less problems with ground loops/hiss .....especially when using long cable runs. These are very important factors to pros who have different equipment hooked up in their studio each day and cannot afford to waste time trying to trace a hum or ground loop.
In a modest home audio set up, RCA jacks can perform just as good as XLR....and they are cheaper to manufacture! (Although it is sometimes a sad reflection on audiophile equipment consumers that some extremely expensive equipment manufacturers should cut cost by not providing balanced circuits and XLR jacks whilst packaging their items with beautiful brushed metal and expensive finishes....form over function is what sells I guess)
There are some extremely high quality RCA connectors available just as there are very poor balanced connectors. As always YMMVNo, no Audiofeil.
I wasn't relating to mileage or to "sound" -- rather, to the fact that xlr clicks on and that rca are spec'd for a small number of connection-disconnection (after which they may lose some of the mechanical contact they're spec'd for).