• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Benefits of balanced dacs, amps and headphones?

Joined
Mar 3, 2018
Messages
3
Likes
1
#1
Hi Audio Science Review, first post, nice to meet you.

There are a few questions that I've wondered for a while and this seems like a good place to ask. With regards to the balanced headphone craze, people have made claims that I've been wondering if are technically valid. Having never heard balanced headphones or possessing an engineering degree I can't begin to evaluate. First is that there is a noticeable improvement in sound quality, soundstage, etc with a balanced headphone amplifier/connection vs a standard TRS connection with a shared ground. Second is that a balanced dac is a better/more suitable source for a balanced headphone amplifier.

Does having dedicated grounds in the dac and/or the amp reduce crosstalk or have any other measurable effect other than doubling the current? Is the increased current the only reason for these reported sound improvements? Would eliminating crosstalk even be desirable because that could make headphones sound even less like speakers, where there is always sound from both sources at both ears?

If I understand correctly the way that 3 pin interfaces between a balanced dac and amp work is by transmitting both a positive and negative of the same signal. Here people also claim that using a source component with a dual-differential dac (dual mono) is better than using phase inverters. Is that true as well?

In either case, are antiphase outputs intrinsically better because they clean up the signal by subtracting away distortion from the output devices? Would that even be audible? Would feeding an ordinary dac to a balanced headphone amp then increase noise/reduce the sound quality?

As you see, many claims have been made about the benefits of balanced dacs, amps and headphones but I really have no idea if any of them are true. Thank you for any clarification.
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,016
Likes
3,593
Location
Monument, CO
#2
This thread might help: https://www.audiosciencereview.com/...d-balanced-unbalanced-and-all-that-jazz.1352/

Unless you have severe noise problems or a ground loop balanced makes little difference in consumer installations IME.

Many DAC IC outputs are intrinsically differential (balanced) with an output buffer to convert to single-ended. Balanced offers greater common-mode noise rejection and often better power-supply noise rejection but in practice rarely makes much if any of an audible difference. Output impedance is generally higher for balanced audio circuits than for single-ended, a minus for a headphone amp. The power levels are so low for headphones that it is hard for me to imagine the difference in 2 vs. 3 wires will make any sort of real difference.

FWIWFM - Don
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,426
Likes
523
#3
This thread might help: https://www.audiosciencereview.com/...d-balanced-unbalanced-and-all-that-jazz.1352/

Unless you have severe noise problems or a ground loop balanced makes little difference in consumer installations IME.

Many DAC IC outputs are intrinsically differential (balanced) with an output buffer to convert to single-ended. Balanced offers greater common-mode noise rejection and often better power-supply noise rejection but in practice rarely makes much if any of an audible difference. Output impedance is generally higher for balanced audio circuits than for single-ended, a minus for a headphone amp. The power levels are so low for headphones that it is hard for me to imagine the difference in 2 vs. 3 wires will make any sort of real difference.

FWIWFM - Don
I agree that for short runs of a few meters, balanced probably does not make much difference in practice. Theoretically it might if there are added conversions to/from non-differential circuits, but even there I doubt it contributes much.

I do feel better about driving my surround and back channel amps via 10-meter balanced XLR cables, not that I made a comparison though. I suspect all my amps are non-differential, however.

But, I take great pleasure in using XLR connectors, and I have eliminated all RCAs, which I hate, from my system altogether. That alone is worth it to me. But, I am also pleased with the absolute dead silence with my ear to the speakers with no signal. I doubt unbalanced could have achieved that.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,671
Likes
25,665
Location
Seattle Area
#6
Welcome to the forum. As Don explained we had a discussion about this earlier. I plan to perform objective testing in the context of headphone amps to see if we can get some data to add to it.
 

Sythrix

Senior Member
Joined
Mar 8, 2018
Messages
307
Likes
229
#7
Also new here... Also wondering about this. Basically I want a DAC that is capable of doing pretty much anything and is future proof... but only if it makes sense to have. Especially concerning balanced, which I've heard no end to counterpoints on.

I'm actually kind of annoyed that the D50 doesn't have balanced. I mean it's not like the home audio crowd would have complained about an extra inch or two of width for the jacks, and I was really hoping for a dedicated DAC from them with that option.

I guess it doesn't matter too much with very few options to even buy from Topping in the US. I don't really want to order from China, as I would want the ability to quickly return it if the need arose. Not waiting around for weeks, or months.
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,016
Likes
3,593
Location
Monument, CO
#9
The usual solution to reducing panel area is to use TRS connectors (look like a standard stereo headphone plug) for balanced outputs. They don't take much more room than RCAs on the panel, but do extend much deeper inside the case, so not always an option.

I have 50' XLR cables to my rear subs but also ran 30' RCA cables without problem. Ground loops can be tricky; my first two subs hummed with a ground loop when they were on outlets 6' apart on the same circuit. The rear ones, further away and on a different circuit, were OK. But different amps so may be other differences.

It may be worth reminded folk that using XLRs does not necessarily break a ground loop; they allow the option for breaking a loop by lifting the shield at one end, but that is something usually done later, and depends on the circuits at both ends.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,426
Likes
523
#12
I doubt you'll hear anything different as long as the cable run is "short" and you don't have a hair dryer runnning in the immediate viscinity :)
Yes, my longest cable is 10 meters as I said. I once tested an unbalanced, home brew, bulk Belden interconnect with a heavy shield at about that length. I did not like the result at all. I have shunned all unbalanced cables with their RCA connectors ever since.

But, the balanced cables can also help eliminate major and minor ground loop problems stemming from gear being on different AC circuits, as mine are. Thus, the reassuring total silence from my speakers with no signal.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,134
#13
When adding headphones to the discussion, we have to be careful not to mix the idea of a balanced interconnect system with a balanced power output stage.
Seems like Mr. Siau and Herr Reim are in disagreement?

Sennheiser are also supporters of balanced headphones and amps. Just audiophoolery?
 
Joined
Mar 3, 2018
Messages
3
Likes
1
#14
Thanks very much Don and svar-hvitt for the great resources. So, balanced connections can be good for eliminating ground noise if you have it, reducing added harmonics and creating better channel separation. Svar-hvitt's second article says that the channel separation is audible, and contributes to the wow factor people report. Noise reduction seems real too, but maybe on too small a scale to be audible? And balanced dacs should be preferably used with balanced amps. That does answer a lot of my question, thanks. One that remains is- is reduced crossfeed desirable? I and many others prefer good speakers to good headphones, and headphones don't have that mixing of channels. I imagine that the added stereo separation with headphones wasn't intended by mixing engineers, so is a modification/degradation of the intended sound.

Yes, amirm, measurements would be great, especially for putting numbers/scale on the noise reduction.
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,016
Likes
3,593
Location
Monument, CO
#15
I did not see the other post so cannot comment on what channel separation might be audible or not. In general channel separation in the electronics is orders of magnitude more than what is provided by the source. You hear both speakers, you hear both headphones, but no room with headphones so I think any argument of whether crossfeed is or is not important is moot. If it is in the mix, you'll hear it; if it wasn't, then whatever your speakers add can be considered distortion as you said.

Balanced connections in and of themselves do not change channel separation. Yes, there can be theoretical improvements, but in practice I find it hard to belive any such would be audible. Which won't stop Marketing and others from claiming it is audible...

Noise reduction comes into a play a couple of ways. First is differential connections offer greater common-mode noise rejection and in interconnects like XLRs the outer shield is not the signal return as it is for unbalanced (RCA) connections. That also means you can break a ground loop without impacting the signal path (assuming both ends of the interconnect are fully differential, not always true since quasi-differential circuits abound that facilitate XLRs but do not necessarily offer all the benefits of fully differential designs). So there is a noise reduction benefit, but whether it is audible depends upon the system. IME the vast majority of the time it is not. But this is audio where everyone is the exception. ;) The second major noise factor is that, as you increase the signal level in a differential design, the noise is uncorrelated so some of it "cancels" or doesn't add in phase, so for every 6 dB (2x) increase in signal the noise only rises by 3 dB (sqrt(2)). So there is a potential 3 dB improvement in SNR -- if you need the extra signal. Most do not, and the additional active devices needed for a differential design add power and such.

All else being equal I would choose a component that had fully differential ("balanced") connections over one that does not. All other things are equal only in extremely rare instances IME and balanced is pretty low on my priority list for my household systems. When I was running live sound it was much more important to use differential connections, but that is a totally different (typically much, much worse) environment. I have had pretty decent systems for decades and very rarely had balanced connections or felt the need for them in the house.

FWIWFM - Don
 

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,918
Likes
3,752
Location
Perth Western Australia
#16
It may be worth reminded folk that using XLRs does not necessarily break a ground loop; they allow the option for breaking a loop by lifting the shield at one end, but that is something usually done later, and depends on the circuits at both ends.
But it does mean that the ground loop isnt on a signal carrying conductor :)
 

Ken Newton

Active Member
Joined
Mar 29, 2016
Messages
190
Likes
40
#17
One of the main issues when discussing this topic is the widespread conflation of the terms; balanced and differential. Differential refers to a signaling method. It has to do with whether the signal is processed (amplified, etc.) in anti-phase (opposing) halves or versions of itself. Balanced, on the other hand, refers to a signal interface method which rejects common-mode noise, such as ground-loop hum.

Balanced interfaces do not depend on whether the signal is processed differentially or single-ended, either type of signal output can be interfaced in a balanced fashion. However, the signal receiving circuit at the far end of a balanced interface should be capable of common-mode rejection, regardless of whether the driver circuit signaling is single-ended or differential.
 
Last edited:

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,426
Likes
523
#18
One that remains is- is reduced crossfeed desirable? I and many others prefer good speakers to good headphones, and headphones don't have that mixing of channels. I imagine that the added stereo separation with headphones wasn't intended by mixing engineers, so is a modification/degradation of the intended sound.
You are correct for most all recordings, unless the original recording was binaural via a dummy head, which is comparatively rare. The Chesky label has been doing that in recent years, but binaural does not play well via speakers. The Baach system also reduces crosstalk from ordinary speakers via proprietary DSP, an advance over the old, mediocre Carver Sonic Holography.

And, no, balanced connections do not reduce added harmonics or create better channel separation.
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,016
Likes
3,593
Location
Monument, CO
#19
But it does mean that the ground loop isnt on a signal carrying conductor :)
Alas, depends on the circuit implementation. Not supposed to be, true, but I've had a loop coupled in via a poor circuit implementation, and of course a lot of conversion cables short shield to signal (-) and thus ruin the balance. But I agree with you for any decent design (most of them IME).

Re. Ken's comments: There is a note at the start of the other thread I linked about differential vs. balanced. My world generally revolved around differential circuits that intrinsically reject common-mode noise by design. So I do tend to intermingle/conflate/screw up/whatever the terminology because I was weaned on differential circuit designs (at the transistor level and on up). Audio has all sorts of balanced circuits that are not fully differential from my point of view but still provide some level of common-mode rejection. I have to keep reminding myself of that... I tend to lump them together as quasi-differential circuits. I also only use balanced when talking audio, because it is prevalent in audio terminology; it's too close to "balun" (which can be balanced, or not) in the RF world.
 
Last edited:
Top Bottom