The balancing act

August 24, 2019
 by Paul McGowan

There’s been an ongoing debate between balanced and unbalanced cables in HiFi as long as I have been involved. And, that’s a long time.

I remember spirited debates with Audio Research founder Bill Johnson about it. AR equipment was a long hold out in adding balanced to their products but over time they gave in, and I believe it to be a good thing.

Balanced interconnects sound better than single-ended do. I know, that’s perhaps too strong of an opinion, but I have yet to have any prove me wrong.

There’s little dispute of the technical advantages when balanced is done right: noise and distortion rejection, 6dB more signal, separation of signal conductors from shield duties.

What many folks perhaps don’t think about when mentally dissecting a single-ended cable is that there are only two conductors inside, and one of them is the outer shield. This asymmetry of design, where the hot lead is a solid core, stranded, or other construction, and the return lead is made of aluminum foil or braided/tinned copper, is not ideal. Rather, you’d want both conductors in a properly designed cable to be identical while the outer shield is separate and distinct from the conductors. That construction is only available in a balanced cable.

My question is a simple one. Why do we manufacturers bother keeping the single-ended RCA connectors on our equipment at all?

The answer in PS Audio’s case is likely the same as others. Compatability.

But it’s a shame.

Subscribe to Paul's Posts

31 comments on “The balancing act”

  1. You answered the question, Paul. In the UK there is great fondness for small integrated amplifiers and it is not practical to have XLR inputs as they take up too much space. Otherwise, I cannot see any benefit for RCA and my last system was balanced only.

    The next question is, given all the benefits come from the design of the cables, proven by the fact that recording studios use very long balanced mic cables made basic materials, how do manufacturers get away with selling expensive XLR cables made from exotic materials? All my balanced cables came from the local pro audio store.

  2. Isn’t the question rather on equipment side than on the side of the undoubtable advantage of balanced cabling?

    So that the question would be: is it better to spend the effort in a balanced component design or leave it unbalanced and put the effort into even more sophisticated circuits and parts?

    My guess is, up to a really high price level where both is applied, no one really checked this out, but decided for an own philosophy and called it best.

    Or did any developer try both designs, balanced and unbalanced, optimised to a certain price level of let’s say 6k and see what sounds best?

    Where is the price point, from which balanced design really makes sense when compared sincerely? 3k, 6k. 10k, 20k?

  3. I’m a convinced user of balanced cables for years.
    And I heard many over the years, so I know manufacturers of expensive cables do not "get away" with expensive cables.
    They (the cables) sound damn good. And the materials they use are expensive. Voila.
    But can a cable be worth 2,000 dollar ? Can a car be worth 60,000 doillar ? Most people will spend 60k on a car if they have the money without a second thought. And find 600 dollar for a cable too expensive. Now, who’s crazy ?
    It all depends on WHAT is HOW important in your life. Everyone can answer this question for himself. I did, although I sometimes envy non audiophiles : ignorance is bliss. (and can save you a lot of money).
    Sound quality is not number one on the priority list of a studio, so they use run-of-the-mill cables.
    Besides, very long high end cables would be far too expensive, even for a studio.
    I only wonder if there is a downside…?
    Why do some high end manucturers (e.g. Spectral) still make their entry level amp(s) unbalanced ?
    And talking Spectral, we’re talking somewhere between 15k and 20k for their "entry level".

  4. Maybe it’s time for a new, low-cost, balanced interface. Even an RCA connector is pretty big.

    We’ve seen how a USB connector has become smaller (and fully reversible) over time. Shouldn’t be beyond the wit of man to create a similar small form factor audio connector. (Yes, such a beast may not accommodate "high end" super thick cables naturally – but should work for a very large percentage of the high-volume, low cost market)

  5. I do think balanced connections are better from an engineering point of view but I wonder just how much better than a correctly made unbalanced cable. Shields should not be connected at both ends, if you use two identical signal conductors and then connect the shield at one end of the cable you should get good results. In the test and measurement world balanced cables are not used for the most part and they seem to work pretty good.

  6. I once had an Emotiva differential balanced preamp and 4 balanced Emotiva monoblocks driving my Magnepans. All were connected with balanced cables. It sounded quite nice. I now have a single ended preamp/stereo amp (my current setup) combo that is WAY past the sound level of the Emotiva balanced system. When I asked the designer of my current combo why he doesn’t make balanced equipment, he said if the single-ended circuit is properly designed he finds no significant advantage to being fully balanced.

    I always thought the advantage of a balanced circuit is that inevitable noise is cancelled out, but that is only true if the circuit of every component in the chain (source to amp) is differentially balanced. There are some components that have XLR connections, but not a balanced circuit. It’s my understanding there is no noise cancelling aspect to this type of scenario. It seems like a properly designed single-ended circuit with very little noise, would not have much noise to cancel out. That leaves cancellation of external noise, which seems to be addressed with proper shielding of the wire. The thing I did like about the XLR connections, however, is that they are less fragile compared to RCA. All things being equal, I would prefer an XLR connection because of the quality of the connection mechanics.

  7. You say “compatibility”…. I would have guessed so you don’t isolate anyone …. the balanced or single ended crowd.

    I have a question…. if you are certain that balanced has better sound than single ended… then instead of putting the rca’s on your electronics why not sell adapters ? Adapters will easily allow the use of inferior cables while remaining ‘pure’ to your beliefs.

    I won’t buy equipment that inputs or outputs ONLY in single ended mode anymore. I did use adaptors as I gradually made the switch to balanced all the way through my systems.

  8. You can’t compare a lower quality preamp/amp balanced setup to a better quality integrated unbalanced setup. Plus I’m not sure if Emotiva gear is fully balanced, which would make a big difference.
    From an engineering perspective, balanced cables are a much better design over an unbalanced design especially over longer runs. If you compare all cables, there are length restrictions and balanced cables fare much better. When you compare the specs, balanced cables are 6-10db louder than an unbalanced cable.
    USB is a terrible interface for audio. People can spend hundreds/4 figures on band aids to try to make the usb interface acceptable in the audio world. I2S is a better interface but hardly used.
    I use balanced cables for short or long distances and make sure my gear are fully balanced designed.

  9. Dear Paul,

    The theoretical (and objective) technical advantages do not lead ineluctably to the practical (and subjective) conclusion that “[b]alanced interconnects sound better than single-ended do.”

  10. I’m not getting into the endless debate about which wire is better. But I will make a few observations about coax versus twisted pairs.
    Most electronic instrumentation is connected by coax with BNC connector terminations (this is a much better and more reliable connector than RCA but a lot more expensive until the audiophile industry came up with some absurdly expensive designs for RCA.)

    Coaxial cable can be quad shielded. The best shielded cable I ever saw was thick ethernet in the 1980s. it has a foil shield followed by a braided shield followed by another foil shield, followed by another braided shielded. You can buy twisted pairs that are individually foil shielded with an overall shield as well. In professional installations many coax cables are installed in steel conduit such as EMT or rigid steel conduit. Sometimes twisted pair data cables are installed in conduit as well but not usually. Conduit offers both excellent shielding and physical protection. It also allows for easy replacement and additional installation. I always leave a drag line for that when I spec them out.

    I had an up front seat in the RBOCs that is the local telephone companies desperate efforts to save what has come to be known as the copper network which are twisted pairs originally used for voice only circuits and even though they were multiplexed had relatively limited bandwidth. A 3.3 inch diameter cables has 2300 twisted pairs. The capital investment was huge and the RBOCs stood to lose out to the cable TV companies which used coax offering much wider bandwidth over long distances. The best the RBOCs could do was DSL which just wan’t competitive and had a limitation of one mile from central offices. The best people in the business couldn’t get even one digitized NTSC channel through twisted pair without digital artifacts and they tried everything they could think of. In the end both went the route of fiber optic for long runs. Television still uses mostly coax although much has been converted to HDMI or USB as digital is in and analog is phasing out. I don’t know what the current distance restrictions are but the last one I did in 2002 for Novartis was 250 feet of CAT 5e on a 100 Mbps network. The rest was OC48 fiber optic from Verizon both incoming and between buildings in an approximately 55 building 180 acre campus.

    John Curl once posted on Audio Asylum that he measured for the 7th harmonic of 5 khz an noise level of -120 db for $1 Radio Shack coax versus -135 db for the best wire he could find. If I had any lingering doubts Curl dispelled them, $1 Radio Shack wire was good enough for me and good enough for government work as we say. BTW, Curl is not an electrical engineer by training. He’s a physicist who took some courses in electrical engineering later on. How ironic since for a long time he would not reply to anyone who was not an electrical engineer. Believe me, he made a few mistakes in his postings. For the longest time I loved teasing him about it by referring to them and how he ascribed his own blunders to me and my corrections to himself.

    One thing to know about professional equipment besides the fact that it usually uses balanced cables is that they normally operate at a much higher signal level, about 10 db higher which gives them and inherent advantage in signal to noise ratio.

    Which is better for audiophiles? I haven’t got a clue. My source for coax is the Dollar Store. 🙂

  11. I don’t recall the exact context, but I believe that Mr. Iconoclast Galen Gareis mentioned that, done properly and given the right conditions (no noise problems), RCAs can sound better than XLRs. This was the first time I had heard such a thing.

  12. >>>> What many folks perhaps don’t think about when mentally dissecting a single-ended cable is that there are only two conductors inside, and one of them is the outer shield. <<<

    I do utilize balanced interconnects. Yet I have unbalanced interconnects constructed not as you describe. Q-Audio. They are braided litz. No solid and shield. They are woven in such a way to counter the negative effects. Very quiet and purity of sound. It explains why they are so quiet. Yet the higher output of balanced cables I have learned to appreciate.

  13. From Wikkipedia;
    "A twisted pair can be used as a balanced line, which as part of a balanced circuit can greatly reduce the effect of noise currents induced on the line by coupling of electric or magnetic fields. The idea is that the currents induced in each of the two wires are very nearly equal. The twisting ensures that the two wires are on average the same distance from the interfering source and are affected equally. The noise thus produces a common-mode signal which can be cancelled at the receiver by detecting the difference signal only, the latter being the wanted signal."

    "wisted pair cabling is a type of wiring in which two conductors of a single circuit are twisted together for the purposes of improving electromagnetic compatibility. Compared to a single conductor or an untwisted balanced pair, a twisted pair reduces electromagnetic radiation from the pair and crosstalk between neighboring pairs and improves rejection of external electromagnetic interference. It was invented by Alexander Graham Bell."

    One problem with twisted pairs is relatively high equivalent shunt capacitance in the telegrapher’s equation I’ve referenced here many times. This is probably why the copper network couldn’t handle the bandwidth coax could. That being said twisted pairs can easily handle the bandwidth of audio signals which by comparison to other circuits is a relatively very narrow band. For example, an analog NTSC US color television signal has over 300 times the audible bandwidth of a 20 khz audio signal. RG-6 coax and others like RG-9 coax can handle about a thousand or more multiplexed channels simultaneously.

  14. There is an increasing trend towards balanced equipment and cables. Of course with an increase in price. Now how much better the sound is,is debatable. While in a high resolution systems it would be readily obvious it makes one wonder whether it will be of any practical use in a inferior systems which make up the majority.This is possibly another reason for the continued existence of single ended equipment and cables. Regards.

  15. As everyone knows the balanced cable is most advantageous when used with components that have balanced circuitry. Some people are in fact proclaiming better results with unbalanced cable only because they are unwittingly comparing it with components that also have balanced connectors, but unbalanced circuitry.

  16. This topic has already been treated in this space, however it is convenient to bring it up today given that there are new subscribers.

    In the early 1980s, James Bongiorno introduced the line of Sumo amplifiers, and although at that time the concepts of balanced bridge operation were not new, previous attempts have not been very rewarding due to the lack of understanding of all of the parameters in the applications of these concepts.

    The line of the Sumo that included: the Power, The Half Power, The Gold and the Nine, were practically the only amplifiers for domestic use (and for professional too) 100% completely balanced (from input to output) utlizing push-pull quadrature feedback from EACH side of the speaker load.

    Since fully balanced amplifiers were almost nonexistent at that time, except for some for professional audio, Bongiorno had to design a first stage single-ended audio signal into a fully balanced to ground signal with two opposite phase outputs. (The bandwidth of this stage exceeds 5 MHz.) The output of this stage goes through the balanced input jacks of these amplifiers, where they enter the balanced signals of the source to be used.

    In the early 2000s, amplifiers with XLR connectors appeared on the market, which were only single ended calling for deceit, unsuspecting customers.

    Today in the world, there are serious manufacturers of amplifiers for home audio, 100% balanced, having been the pioneer of this technology James Bongiorno. (RIP)

    There are detractors who say that the amplifiers with single ended inputs sound better than the truly balanced ones without having checked it, I suggest that before issuing an opinion, check yourselves, being the only possible way to do the test is to use a balanced amplifier from the input to the output and at the same time have single inputs with unbalanced to balanced converter circuits such as Sumo amplifiers, and you can see the obvious difference in favor of fully balanced lines and circuits.

    This controversy will always be topical, as more and more manufacturers have adopted the use of fully balanced circuits because they are technically better than single ended.

  17. There’s no question that the DS dac’s balanced out, as I use, is better than the unbalanced. However, people who have made comparisons consider that the current middle of the line Lampi dacs using unbalanced outs sound much better than the DirectStream, especially now that tube rolling has been introduced at that level. I will be finding out very soon. Lampi also makes a true balanced section for another big bill in Euros, but I’ve yet to come across one Lampi owner online that claims it’s better, rather just more useful for longer runs when the other end has balanced inputs.

  18. Hi Paul,

    I’m few days late to this conversation but I agree with you Pa

    ul. The first time I setup a fully balanced setup (my hacked Sony AVR to a Crown Amp – Guitar Center Rental for testing), I noticed a huge difference in sound. Lack of any background noise and the 6dB increase just made a big difference in how the sound was presented. This was part of my preparation while I waited for my PSA Stack (SGCD/M700 x2).

    I also bought another Bose Series 2 EQ box for my Center Channel 901. First thing I did was convert the box into true differential (balanced) single channel output which I tested temporarily with one of the M700’s. Since these earlier EQ boxes were discrete designs, I simply used the right channel for the non-inverting input, and the left channel as the inverting input. The GND pin was tied to chassis GND. It worked and yes there is a difference in sound quality when I compared it to my other Bose EQ box of the same vintage which I adapted to "pseudo" differential operation with OPA627AU opamps on brown dog adapters (see my posts on the "Do It Yourself Page (

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram