Why is not all audio equipment using "balanced"?

[earlevel]

I keep reading that balanced has more noise and distortion than single-ended/unbalanced. That does not match my experience as a designer or user.

To change to balanced connections, I think we need to ask ourselves, what problem are we solving?

Look, I have a fair amount of balanced gear here. I have a spool of Mogami Neglex 2534 (bought it decades ago, what a good move) and a bunch of Neutrik XLRs and TRS and TS phone plugs, and RCAs. When I need a connection, I fire up my Weller. My keyboards, guitar, V-Drums, and Stick aren't balanced, but some of my outboard gear is, not all, and of course my mixer is, my monitors can go either way. Pro audio gear is mixed, the most likely to be balances is gear closely associated with mics and preamps (compressors, for instance, and probably tube gear).

But the fact is, only my mics need it, to the preamps. It's all in one room, mainly 20-25 ft cables, shorter interconnects.

Changing routine audio listening gear to balanced i/o is solving a problem that doesn't exist, and making it cost a little more.

I'm not saying I have a problem with it, but if you're going to mount a campaign for change, it probably should be for one you can hear. But, companies will make it, and you can vote with your purchases, for whatever the benefit. For my computer, I bought iLoud MTMs, they have a combo jack and can take TRS or XLR balanced, or TS unbalanced. And a Topping DAC that has RCA or XLR out, so I made a couple of one-meter XLR cables and went balanced. Did I do it because I want to ensure the best sound? No, it just made no sense to make cables with RCA on one end and phone on the other.

I just talking about expectations. Anyone who thinks going XLR for typical listening environments will step them up to a "pro" listening experience doesn't have a good understanding. But the click of inserting an XLR is definitely more satisfying than pushing in an RCA plug.

 

[DonH56]

To change to balanced connections, I think we need to ask ourselves, what problem are we solving?

Look, I have a fair amount of balanced gear here. I have a spool of Mogami Neglex 2534 (bought it decades ago, what a good move) and a bunch of Neutrik XLRs and TRS and TS phone plugs, and RCAs. When I need a connection, I fire up my Weller. My keyboards, guitar, V-Drums, and Stick aren't balanced, but some of my outboard gear is, not all, and of course my mixer is, my monitors can go either way. Pro audio gear is mixed, the most likely to be balances is gear closely associated with mics and preamps (compressors, for instance, and probably tube gear).

But the fact is, only my mics need it, to the preamps. It's all in one room, mainly 20-25 ft cables, shorter interconnects.

Changing routine audio listening gear to balanced i/o is solving a problem that doesn't exist, and making it cost a little more.

I'm not saying I have a problem with it, but if you're going to mount a campaign for change, it probably should be for one you can hear. But, companies will make it, and you can vote with your purchases, for whatever the benefit. For my computer, I bought iLoud MTMs, they have a combo jack and can take TRS or XLR balanced, or TS unbalanced. And a Topping DAC that has RCA or XLR out, so I made a couple of one-meter XLR cables and went balanced. Did I do it because I want to ensure the best sound? No, it just made no sense to make cables with RCA on one end and phone on the other.

I just talking about expectations. Anyone who thinks going XLR for typical listening environments will step them up to a "pro" listening experience doesn't have a good understanding. But the click of inserting an XLR is definitely more satisfying than pushing in an RCA plug.

I am not mounting a campaign to change. I have posted many, many times that it's a non-issue in the consumer world unless you need to break a ground loop or are running very long lines (e.g. whole-house or outdoor runs). And about the myriad of other systems, mainly RF, that are single-ended unbalanced and achieve far greater dynamic range than a typical audio system. I have no desire to step up (or down) to a "pro" listening experience; had plenty of that (live and studio) in the past.

I am questioning the assertion that balanced adds noise and distortion when the circuit theory and practical measurements show the opposite. Even for just balanced and not differential operation. A given balanced component might have higher and noise than a specific single-ended component, but apples-to-apples the balanced design will have the same or better SNR and distortion, sometimes much lower, depending upon the circuit implementation.

 

[earlevel]

I am not mounting a campaign to change. I have posted many, many times that it's a non-issue in the consumer world unless you need to break a ground loop or are running very long lines (e.g. whole-house or outdoor runs). And about the myriad of other systems, mainly RF, that are single-ended unbalanced and achieve far greater dynamic range than a typical audio system. I have no desire to step up (or down) to a "pro" listening experience; had plenty of that (live and studio) in the past.

Of course, I hope you realize I was speaking in a wide generalization that included the greater audio community not even weighing in on this thread. I was making no assumption about your inclinations.

I am questioning the assertion that balanced adds noise and distortion when the circuit theory and practical measurements show the opposite. Even for just balanced and not differential operation. A given balanced component might have higher and noise than a specific single-ended component, but apples-to-apples the balanced design will have the same or lower noise and distortion, sometimes much lower, depending upon the circuit implementation.

Let's set aside "and practical measurement", because of course you can have an unbalanced design that measures better than a particular balanced one, and vice versa—like you said, "depending on circuit implementation", or "better gear will sound and measure better".

But don't forget that if manufacturers find they need to have balanced outputs to attract some buyers, in addition to their unbalanced design, the balanced outputs will have some level of circuitry over the unbalanced. Whether it be op-amp, transistor, or transformer based. You might say, no, they should have a design that inherently balanced out, and have the unbalanced be derived from that, but it won't be that way, not with budget gear for sure.

So, we're really talking about the circuit theory part, and assume solid execution. There is a limit with noise floor (where thermal and shot noise or simply our limits of hearing take over), and there are limits to hearing distortion. I'd rather just say that there is no barrier in getting to that point with single ended, so I'm back to...what problem are we trying to solve with balanced?

If someone can show me why you can't get there with single ended, but you can with balanced, that's a different story. But I kind of find this like those who want 64-bit float bussing in DAWs, because it's better than 32-bit float bussing. Even though both will put the same 24 bits out to the DAC for listening.

 

[JohnYang1997]

I keep reading that balanced has more noise and distortion than single-ended/unbalanced. That does not match my experience as a designer or user.

I only see the opposite. Everyone is saying balanced is automatically lower noise.
There's a difference between using a single ended circuit with 3 wire connection with matching output impedance on the balanced signal lines and differential circuits with two identical circuits driving the two lines.
The best case scenario, you lose 3dB noise performance in ideal circuits using differential circuits.
It's part of the reason why the last two headphone amplifiers of mine are designed to be single ended output. They don't have the somewhat awkward and noisy single ended to balanced conversion in A90. (it's still less than half the noise than THX in single ended)
There are ways to push it to the absolute limit to achieve less than 100nV noise 20kbw in a fully differential circuit. But it will be very inefficient.

 

[JohnYang1997]

I am not mounting a campaign to change. I have posted many, many times that it's a non-issue in the consumer world unless you need to break a ground loop or are running very long lines (e.g. whole-house or outdoor runs). And about the myriad of other systems, mainly RF, that are single-ended unbalanced and achieve far greater dynamic range than a typical audio system. I have no desire to step up (or down) to a "pro" listening experience; had plenty of that (live and studio) in the past.

I am questioning the assertion that balanced adds noise and distortion when the circuit theory and practical measurements show the opposite. Even for just balanced and not differential operation. A given balanced component might have higher and noise than a specific single-ended component, but apples-to-apples the balanced design will have the same or better SNR and distortion, sometimes much lower, depending upon the circuit implementation.

True apples to apples will show single ended design will have less input referred noise.

 

[boXem]

Issue is the confusion between "balanced", that is a question of impedance, and "differential" that is a question of topology.

 

[GXAlan]

It's part of the reason why the last two headphone amplifiers of mine are designed to be single ended output. They don't have the somewhat awkward and noisy single ended to balanced conversion in A90. (it's still less than half the noise than THX in single ended)

This. John Yang has pulled off some of the highest performing designs tested at ASR and delivered them at affordable prices. If he says single ended is fine, you have your answer.

 

[solderdude]

Single ended is fine until you run into groundloop issues.
Those can often be solved without having to resort to buying gear with balanced connections as well.
When one already has balanced connections I see no reason not to use them.

Balanced (differential) headphone outputs from portable gear makes sense as one can have 6dB more headroom without having to resort to internal DCDC converters in said device.

 

[earlevel]

Issue is the confusion between "balanced", that is a question of impedance, and "differential" that is a question of topology.

I like that distinction. But also we need to be pragmatic—equipment is referred to as having balanced or unbalanced input and outputs. There are other assumptions that people make about "balanced", including what type of connector it might have, or the nominal voltage swing. And we may further look at whether an output is "pseudo balanced" with a resistor, or whether both sides are buffer, etc. But it probably won't help the consumer get too specific, short of the spec sheet at least.

 

[Interference]

What people are missing is that balanced (dual ended) amplifiers, all other factors excluded, have more noise and distortion than unbalanced (single ended). Where balanced wins is in common mode noise rejection.

Soooo...under what conditions is it better to go balanced? When the noise induced on a cable outweighs the added noise and distortion of a balanced amplifier. That is long lines, but in particular long low-level lines, such as mic cables. With mics it's a no-brainer. Cables are almost always at least 20 foot, often much longer (especially in the of days, you wanted to be away from R-R tape machines, tube gear with fans, and the guys running them).

I am not sure if you mean speaker amps or line stages when you talk about "amplifiers". Are added noise and distortion a real issue with today's engineering standards? It seems to me this is one (small) of many factor contributing to the final performance of a device.

Not to mention it is possible to have balanced line receivers in devices with internal SE operation.

What shall we do with all those modern DAC chips that natively output differential signals? Sometimes you need to add an additional line stage in order to add a single-ended output to them, so what is the point of having it at the source instead of having it at the receiver?

For most home listening situations, balanced is a potential (very slight) degradation, with a modest boost to the price tag and possible problem with panel space.

The boost to the price tag is quite relative and depends on the context. Engineering / BOM is only a part of the final bill, especially in audio where there is huge variability in price tags relative to performance.

However you did not address the main point I raised about grounding. In my (direct and indirect) experience, balanced prevents a lot of headaches there, while with many "audiophile" brands it has become impossible to predict whether a combination of earthed devices with unbalanced interconnects will have issues or not.

You can put RCA stereo ins and outs in less space than two mono XLRs, and your cables won't extend out the back as far, and the connectors won't take as much space internally either. I have a PS Audio IVH preamp here, 20 RCA jacks take up half the back panel—try that with XLR. TRS phone jacks need less space than XLR, but more depth. (BTW, I kind of hate RCAs for other reasons, just talking about the practicality of consumer electronics, which is typically connected and disconnected often.)

That is a fair point but personally I wouldn't know what to do with a preamp with that many connectors. Many people just need a digital source to hook up to a power amp (or pair of powered speakers).

 

[DonH56]

Maybe the debate is actual noise floor, perhaps input-referred, vs. SNR. I have designed differential circuits that exhibit the same noise floor (absolute and input-referred) single-ended or differential. The SNR is higher in differential mode, of course, by the usual 3 dB. But many active (and some passive) circuits will indeed have a higher noise floor even though SNR is better. But note my experience is mostly in the RF world; I have been largely just a user in the audio realm (my audio designs were done years ago and not commercialized).

There is a lot of confusion in the audio world (not yours @JohnYang1997 ) about what constitutes a "balanced" connection as @boXem | audio said. Balanced does not mean differential, and for that matter differential does not always mean balanced, though it usually is. And there are many ways to achieve "balanced" and "differential" operation. All with their own pros and cons, natch, usually trading CMRR (noise rejection) for complexity (cost), noise, power, and all that jazz.

In any event the next step would be to start comparing circuit designs, I suppose, and I don't see any point in continuing down that rabbit hole.

 

[earlevel]

I am not sure if you mean speaker amps or line stages when you talk about "amplifiers". Are added noise and distortion a real issue with today's engineering standards? It seems to me this is one (small) of many factor contributing to the final performance of a device.

Not to mention it is possible to have balanced line receivers in devices with internal SE operation.

What shall we do with all those modern DAC chips that natively output differential signals? Sometimes you need to add an additional line stage in order to add a single-ended output to them, so what is the point of having it at the source instead of having it at the receiver?

All good points, I'm lumping them together because I want to limit how much further down the rabbit hole I descend. Yes, I do mean line stages, and really most people are concerned with whether the box is checked and how they are going to connect devices. Either the device performs well or not. No one buys a Pultec EQP-1A because it has a pair of XLRs for i/o—though it may be reassuring. But beyond that, no one scrutinizes the circuit to make sure it's a good "balanced" implementation.

Similarly, with a DAC, you're probably buying it for other reasons and features—OK, maybe if you have two that are in the running for purchase, and one has XLRs and that's convenient because your powered speakers have XLR inputs, you'll go for that one. But if they both do, both sound stellar, price difference is not a consideration, both have XLRs out...how many people at that point are going to scrutinize the circuitry? I'm not saying it isn't worth knowing or talking about, just that for the vast majority of the market it's a non-issue. For some, "pro" connectors will give them a good feeling that the manufacturer went the distance, for others, it's just "I need my connectors to match".

The boost to the price tag is quite relative and depends on the context. Engineering / BOM is only a part of the final bill, especially in audio where there is huge variability in price tags relative to performance.

Again, the point I made was, "what problem are we trying to solve?" And if the answer is, "there isn't a problem, doing this doesn't affect the sound at all", then any added cost is...? Maybe that added cost gives comfort to the buyer, makes more sales for the manufacturer—fine, I'm just trying to stick with actual problems. And if you add $0.50 to the BOM, either the price goes up more than that, or the profit goes down more than that. There should be a reason.

However you did not address the main point I raised about grounding. In my (direct and indirect) experience, balanced prevents a lot of headaches there, while with many "audiophile" brands it has become impossible to predict whether a combination of earthed devices with unbalanced interconnects will have issues or not.

OK, I really didn't want to step in that. There can be a number of things involved, and they start with home wiring. I had bad ground loops when I moved in here, and a detached rec room became my office/studio. I had to take audio measures to stop ground loops, but the ultimately solution was to tear out the AC wiring, put a grounded sub panel in, heavier gauge wiring from the main panel across the yard, better wiring in metal conduit in the walls (it was wood paneled with no insulation—had to tear off the panelling, exposing all electrical, to insulate and put up sound deadening board and drywall anyway). So I went from ground loops and getting buzzed on my lip from a dynamic mic while standing barefoot on the tile floor in the '80s, to never hearing a ground loop again, doesn't matter how many different wall outlets gear is attached to.

Also, ground loops are a completely different thing from guarding against AC hum from creeping in due to proximity of audio cabled and AC lines.

That is a fair point but personally I wouldn't know what to do with a preamp with that many connectors. Many people just need a digital source to hook up to a power amp (or pair of powered speakers).

Yes, there are many devices for which panel space is not a consideration. Like I said, I'm not crazy about RCA jacks. They aren't the easiest cables to build by hand, and a sturdy hand-soldered plug is relatively long and exerts a lot of force, when using pro-diameter cable, on the feeble RCA jacks. Many years ago, I ordered replacements of the banked jacks before they had cracked fairly badly over the years from movement, reconnection—they aren't made well for that, and is why they are absent from pro gear. But they are generally adequate for home audio.

 

[AnalogSteph]

I can only see balanced connections becoming more prevalent in the future.

From an equipment design POV, unbalanced connections come with real inconveniences in the power supply department. All but one component of your entire hi-fi setup must remain floating (IEC Class II, with corresponding build standards), or else you will be introducing ground loop action.

That's easy enough for anything that can be powered from small transformers, but with the several hundred VA required for an integrated or power amp you'll be looking at mains leakage approaching problematic levels. That means using a mains transformer with a shield winding, often a custom job. Yes, you can say to hell or rather PE with leakage currents and make the amp a boring IEC Class I affair with grounded chassis, but there goes your one permitted ground connection. Introduce a radio tuner with external antenna, a satellite receiver, broadband cable or an ordinary PC on top, and you'll have fun chasing down ground loops.

And then there's the elephant in the room, the now-ubiquitous switch-mode power supplies introduced everywhere as mandated efficiency requirements are going up. Those generally have a lot of mains leakage, potentially quite dirty too. Not something you want to have going down your audio ground - cables with low-resistance shielding might keep its effect minimal on the outside, but inside of devices you never really know. Again it would be nice to just flush it down the protective earth route (as is standard for higher power units and for a lot of computer and networking equipment in general).

That being said, unbalanced connections are likely to prevail wherever they get the job done fine. Cheap and easy is cheap and easy after all.

I had bad ground loops when I moved in here, and a detached rec room became my office/studio. I had to take audio measures to stop ground loops, but the ultimately solution was to tear out the AC wiring, put a grounded sub panel in, heavier gauge wiring from the main panel across the yard, better wiring in metal conduit in the walls (it was wood paneled with no insulation—had to tear off the panelling, exposing all electrical, to insulate and put up sound deadening board and drywall anyway). So I went from ground loops and getting buzzed on my lip from a dynamic mic while standing barefoot on the tile floor in the '80s, to never hearing a ground loop again, doesn't matter how many different wall outlets gear is attached to.

A lot of studio gear is IEC Class I and requires grounded outlets, which your previous installation seemingly didn't have yet. Results under such circumstances are unpredictable, but ground loop hell is definitely included. Equipment ground would be floating around somewhere between 0 V and mains voltage, with varying amounts of mains leakage current travelling back and forth between units. I imagine it would have been wreaking havoc on any electric guitar as well (your body is one big capacitor stuck at or near ground potential, now imagine your hand near unshielded high-impedance wiring inside a guitar when the guitar amp ground level is substantially elevated).

On some YT comment section there actually was a guy complaining about ground loop hell in his studio recently. Needless to say, having the electrical installation checked out was one of the things I recommended.

 

[earlevel]

I can only see balanced connections becoming more prevalent in the future.

I guess that will bring some complication and confusion with voltage levels too (pro vs consumer).

 

[escksu]

I guess that will bring some complication and confusion with voltage levels too (pro vs consumer).

IMHO, in the future. analg audio connections will become less prevalent. Digital connections esp. USB/HDMI will become popular instead. Wireless too will become popular.

 

[earlevel]

IMHO, in the future. analg audio connections will become less prevalent. Digital connections esp. USB/HDMI will become popular instead. Wireless too will become popular.

I think that's true too. I'll make some assumptions here, so anyone can feel free to correct me, but I think the market for component home listening systems is smaller than it was in the '80s, say. Not in dollars, I mean that at one time having a nice stereo in the living room was an important entertainment feature, even for young people (in their bedrooms for teens). Now, it's earbuds or headphones on a smart phone. Or it's Bluetooth speakers, likely just a single. but certainly it's not multiple components like a cassette deck and turntable and FM tuner plugged into a preamp, into an amp (or maybe an integrated amp or receiver to cut down on some of the connections), into stereo speakers.

And when I say, "what problem are we trying to solve?", I don't mean noise susceptibility or ground loops, I mean...good grief, if we weren't compelled to move to pro-standards audio levels and balanced connections in the '80s, exactly what is going to drive that now?? I agree there are advantages, but we had more need in that past, and it didn't happen then either.

And, if the answer is that we have cleaner stuff now, predominantly due to digital...well, we have a lot less gear to connect, a lot less for those bad analog things to creep in, and we will more likely just extend the digital path further. Right now, the only analog paths are the cables to my speakers. And we'll just move more towards having the conversion at the speakers, just as we have in putting the amps in the speakers.

So, yeah, I agree—we'll probably just get more digital rather than adapt technology from the '50s that we never thought was compelling enough to adopt for home up to this point.

 

[KSTR]

I mean...good grief, if we weren't compelled to move to pro-standards audio levels and balanced connections in the '80s, exactly what is going to drive that now??

Circumstances have changed. In the 80'ies, SMPS were almost non-existent in audio gear as well as in the myriad of gadgets with wall-warts and in all the typical home appliances of the day.
Today, unbalanced signal cables between devices usually are impacted with much higher and much nastier supply leakage currents of various origins and therefore the chances are higher that the input signal to a device gets polluted with hum/buzz/transients, generating an error voltage drop along the cable shield (or GND wire), and general RF levels flying around in the air are also many magnitudes higher, making proper shielding more important than in the 80'ies.

Judging from the corresponding threads on forums around the world, a lot of people have real (actually audible) issues that go away once balanced inputs are used and/or when a system cabling is used that diverts the error current away from the signal reference ground connection.

And for measurement and SINAD scores, your -110dB'ish SINAD of your DAC may drop significantly down into -80dB realms at the signal cable end by the point it enters the speaker/phones amplifier. Oftentimes this is still inaudible but you get the point, system performance almost always degrades with unbalanced connection no matter how good the components measure in isolated and idealized test setups (which is OK and should be this way as we are interested in what the device itself is capable of, at least initially).

From a system philosophy point of view, cables should never be the weakest link for signal integrity but with the higher noise levels of today unbalanced connections are more prone to fail that goal on a regular basis. Balanced can fail there as well but usually it doesn't in any competent design.

 

[tecnogadget]

There is a lot of things in this hobby that only comes to sense when applied to Professional Audio (PA systems, long distance cabling, audio interfaces and mixers, arena events and such), and Recording Studios. Mean it balanced connections...the need for 24bits digital audio, etc.
But are not really needed to pull out a reference sound system at your home. Not because it exists you need it.

 

[MCH]

Yeah i don't see my kid plugging cables 20 years from now, i would say most people aren't doing it anymore since ages

 

[escksu]

Yeah i don't see my kid plugging cables 20 years from now, i would say most people aren't doing it anymore since ages

This I do agree. Most devices will be connected via wireless.