Single Ended, Balanced & Differential - What it Means

One of the most common points of confusion, and questions raised, relates to the differences between, and meanings of, the terms “single ended”, “balanced” and “differential” operation. And, further to that, what the qualifier “fully” (as in “fully balanced”) means in relation to these terms:

Single Ended

A single-ended configuration uses two signal connections, one for + and one for ground (or signal return). It is the most common connection used to headphones, between components, and in the internals of audio gear, by far.

Balanced

A balanced configuration still uses two signal connections, but requires that both lines have the same impedance (not nearly the same, actually the same). How this is achieved doesn’t matter (i.e. it could be from differential signal lines off a source or from a single or a single signal line and a resistor on the ground/return line). This, combined with the twisting of the lines, is what’s required to get the noise rejection from balanced operation (CMRR).

Differential

A differential configuration, which is still a two-wire connection, drives one line in + phase and the other in - phase. This doubles the potential difference between the signal lines, which is what yields the higher (double) peak-to-peak voltage of differential operation (and thus the higher signal level).

Typically a differential amplifier (balanced or not) requires twice as many amplification circuits (one each, per phase, per channel, so four total) as a non-differential design (one per channel). This means that for an otherwise identical amplification circuit, a differential implementation will require twice the amplification circuitry.

One implication of which is that for a given level of performance, a differential amplifier either has to be more expensive than a non-differential version, or corners have to be cut in terms of the bill-of-materials (or saved elsewhere). So, in general, single-ended amplifiers will tend to offer better “bang for the buck”.

Note: Being balanced does not require differential operation. Nor does being differential in operation require a balanced configuration!

This is why you can have a single-ended amplifier, but still get the benefits (noise rejection) of a balanced connection. It is also why you can have an otherwise single-ended DAC or amplifier offer a balanced output.

Fully Balanced & Balanced Differential

Sometimes we see the term “fully balanced” or “balanced differential”. In theory these terms are actually describing different things.

Fully Balanced

“Fully balanced” should be taken to mean the entire design, from input to output, is using a balanced configuration.

This isn’t super useful, as the benefits from balanced operation are really in the realm of improved CMRR (common mode rejection ratio) and are most beneficial over long cable runs … not the very short paths on a PCB.

So, more often than not, “fully balanced” actually means “balanced differential from end-to-end”.

Balanced Differential

Formally, this is a configuration that uses a differential topology from input to output (i.e. no phase splitters on the input, no summers on the output) and uses purely balanced transmission lines from end-to-end.

More commonly it just means something is fully differential, and has balanced inputs and outputs (because balanced transmission lines do not matter as much on the PCB).

Power

It’s not being “balanced” that yields more power vs. “single-ended”, it’s being differential. You can have balanced connections and the same power output as the single-ended connections (see things like the Phonitor X).

However, for the connection to a headphone you need to have a balanced (four-wire, no shared ground) connection to have differential drive (but you can have non-differential drive on a 4-wire connection … the ground wires are separate but are still ground).

With a differential connection you’ll see double the potential difference (voltage) in the signal, which doubles the volume, and then you have twice as much power available from the amplifiers (since there are twice as many of them, with pairs operating in opposite phase), giving a doubling of current capacity. Thus yielding a 4x increase in the output power.

Notes

  • Being “dual mono” says nothing at all about whether something is balanced OR differential. There is no required interplay between those factors.

  • Balanced headphone drive, whether differential or not, typically yields better stereo separation and the reduced crosstalk often means that low-level details (resolution) is improved too.

28 Likes

Thanks for posting this kind of info @torq, it can save hours of searching through wrong answers and irrelevant info!

2 Likes

Great write-up with some helpful info. And to add to that regarding balanced DACs, Mike Moffat of Schiit Audio recently mentioned a simple rule:

“There are two kinds of balanced ways to build DACs; analog - (take one DAC per channel and balance it in analog) – and hard way (take two DACs per channel and balance it in digital).”

4 Likes

This is all great information to ha e at hand. Thank you.

1 Like

I wasn’t thinking any amount in particular. I just know many people (whether justified or not) look to the chip when making a decision. $100 difference wouldn’t have changed my mind but others might disagree.

If Grace/drop managed even a sideways move equal the DM+ for $150 it’s impressive.
I had an original DM but preferred the Matrix mini-I. with the dual AD1955 chip.
Both original units (DM + M mini-i) have very limited USB, so I’ve been looking for quite some time.
As I use some half decent universal disc spinners a lot of the time and foobar does fairly decent conversion with SOX, it hasn’t been a pressing emergency.
I considered a decent USB converter but never got around to it.
I’ll keep my eye out for reviews that are sure to come in before long. I suppose part of the price difference from the DM+ Matrix is obviously the included Headphone amp in the one box unit.

BTW I’ve read similar explanations of balanced configurations in DACs and amps but would need more knowledge about electronics for any of them to be truly helpful.
I do know the amp I use is a dual mono balanced configuration, with separate toroidal transformers for each channel; with the “guts” actually looking like two separate amps. But as I said, I don’t know anything about electronics.
I just listen and buy/keep what I can afford and like the best.
And I do realize it’s a much different approach than that of those who are true hobbyists with plenty of spare cash.

1 Like

Dual-mono construction in amplifiers is generally done to reduce crosstalk/improve separation. But an amplifier can be dual-mono and still be single-ended.

My WA234 MkII MONO, for example, is literally two separate chassis, with independent volume controls and power sockets, one for each channel. It is still a single-ended amplifier internally.

In other words, something being dual-mono tells you nothing about whether it is balanced or not. That’s a separate thing and one is not dependent on the other.

A fully balanced/differential amplifier needs four amplification circuits, one for each phase (+/-) of each channel. In many cases, even in “dual mono” amplifiers, that winds up being a pair of stereo amplifiers, with each channel driven out of phase with the other. It doesn’t have to be, but it often is. Some amplifiers are effectively “quad mono”, and have four mono-amplifiers, one for each phase (+/-) of each channel.

But you can have balanced or differential input or output without being a balanced/differential amplifier, either via splitter/summer circuits or via transformers. What you won’t get in that case is the increase in power that comes with having the additional amplifier circuits operating in opposite phase.

In DACs, dual-mono construction, or more commonly just using two DAC chips in mono-mode, also reduces crosstalk and improves separation. But in most cases it is done to allow the IC to use both its internal conversion implementations to work as one, which can improve dynamic range, signal-to-noise and reduce overall distortion/increase precision.

7 Likes

I appreciate the detailed reply.

I imagine that (as with anything else) a practical working knowledge of electronics would create an image in one’s mind as to the “route” the audio signal takes on its path; from where it is generated by a source, through a particular configuration to the appropriate output terminal (be it balanced/unbalanced…whatever).

Without a working understanding of things like ICs, transistors, buffers or circuitry in general, explanations are pretty much expressed in what amounts to a foreign language to someone like myself.
I started a sound engineering course a couple of times but was unable to get far either time, due to unforeseeable, unfortunate circumstances.

I think to truly grasp what you’ve explained here, I’d need a visual representation of the path taken by the audio signal as well as an understanding of how the signal is “handled” by the different components it passes through (ie a working knowledge of the function of each electronic “part” along the signal path).

This seems an excellent explanation but I simply don’t have the basic knowledge required to understand the terms used to allow me to visualize what happens to an electronic signal (be it analog or digital in it’s “representation”) as it goes through the various stages of it’s generation into audible sound waves.
I’m sure the information you provide here may make things much clearer for many with a reasonable understanding of electronic signal generation.

3 Likes

Do you mean each mono amp is single-ended or that running both…one to each stereo side/ear cup is still not balanced.
If so, then I’m more confused than ever, as it would seem that each channel would have an independent ground (and independent everything else).
I don’t see how two completely separate amplifiers would not produce the same result as a balanced amp. And if that is in fact not the case, I’m obviously not interested (or focused) enough at this point in my life to try to understand the difference without a teacher and the required gear (for a demonstration of the practical, working differences) right in from of me!

I moved these posts to a more appropriate thread (one I created earlier in the week, prompted by these posts) …

If you want to learn more, start by reading the first post in this thread. It may not help, but it’s not really a big deal anyway, but to respond to your last post properly:

In the case of my Woo WA234 MK2 MONO … the internals of the amplifier are single-ended. The inputs and outputs have balanced options.

I don’t know what amplifier you’re using, so I cannot say how it is operating.

What I can say is that no, just because something is operating in dual-mono, with separate grounds for each channel, with no communications or connections between each mono channel, does not mean it is balanced.

“Balanced” refers to the transmission line. To be balanced, per the above (first) post, both the signal and ground/return lines must have the same impedance. That’s all “balanced” means.

So, if you implemented the amplifier such that the signal line (the part being amplified) was at 2 Ω, and the ground/return line was at 1 Ω, which is not unheard of , then it would not be balanced.

In a dual-mono amp, that isn’t specifically balanced:

What you would lose (or rather what you wouldn’t gain) is the noise rejection of the balanced transmission line. Which you probably don’t care about anyway, since short cable runs in domestic environments aren’t generally subject to high levels of (E)MI.

You’d gain the better separation and lower crosstalk (effectively none), and the increase in resolution that tends to accompany it.

The term “balanced” has been badly abused in the world of headphone amplifiers. Most of the time it is used when what is meant is that it is both balanced and differential. Again, this has nothing to do with being dual-mono or not and, frankly, a proper dual-mono implementation is more important than being balanced or differential when it comes to home use.

2 Likes

Also, just be clear (and probably redundant) …

A dual-mono amp certainly can be balanced. But it is balanced because it is balanced, not because it is dual-mono.

Similarly a dual-mono amp can be differential. And where it is, it is differential because it is differential, not because it is dual-mono.

—-

As a point of interest, most differential amplifiers are also balanced. This is because differential amplifiers are usually implemented as two amplifiers each operating in opposite phase, which means the + and - signal lines should innately have the same impedance.

2 Likes

Things are becoming much more clear here Ian. Thanks for taking the time to go into such detail in an attempt to clear up some (as you say) often misused and confusing terminology.

I’d still likely need a demo/lesson or two to be truly educated on the topic. But with what’s available online and your clarification of terms and functionality in these recent posts, it will be much easier to gain a complete understanding of different designs and the logic behind choosing one or the other in any given situation/application.

Here’s a link to a couple of photos (and description) of the the amp in use.
The original LD MKVII (NO +) had much more power but some units had a persistent buzz with lower impedance headphones.
It seems a simple design but the soud is clean and pleasing. I realize it’s low cost impedes better design and performance in some areas but it sounds very good with HD 650s.

I’ve heard far more expensive amps but they were tube amps so I have no idea what’s wanting in comparison to high cost SS amps. I imagine bass reproduction would be more intense with more available power @ 300ohms. And I doubt it would drive planar cans all that well

http://www.littledot.net/forum/viewtopic.php?f=9&t=822&sid=ef03d8b2769665de09a4629b57fa7049

2 Likes

That’s balanced, differential and dual-mono. That’s as much as you can ask for.

Assuming competent design/implementation (and I am):

  • Dual-mono is invariably better than not.

  • Balanced for input can help with ground loops, and reduces susceptibility to (E)MI on cable runs but is otherwise not necessarily interesting ( i.e. it’s benefit is situational, but it won’t make anything worse - it just may not make anything better).

  • Balanced headphone output (i.e. 4-wire, separate grounds for each channel) is almost always better.

  • Differential amplification will always have double the noise and distortion of non-differential, but will also offer double the voltage swing and double the available drive current for the same fundamental amplifier topology.

Additionally it can be less expensive (overall) to hit one’s power-targets with dual, inverted, single-ended amplifiers in a differential configuration than in implementing a single-ended amplifier of the same raw power output.

And, in general, simple designs tend to be better. The exceptions tend to get extremely complicated.

3 Likes

Thanks again for the information and your time.

David (at Little Dot) said several years ago that the amp was “retired” simply because it became too expensive to produce/offer for the amount indicated in the link.
LD then began focusing on producing higher priced balanced (and less costly SE) tube amps. Little Dot has always seemed a pretty decent company without serious QC issues.
I believe “Sword Yang” was/is the designer (perhaps owner?) and that David Zhezhe left the company awhile ago.
I’m not sure if David was just the main distributor or was involved with other aspects of Little Dot.

1 Like

Apologies for reviving this thread.

I wonder if his first bullet is referring to – what we know today as – Bifrost 2? Timing-wise, makes a lot of sense. There seems to be a lot of hype for this DAC in 2020.

1 Like

A post was merged into an existing topic: General purchase advice: Ask your questions/for advice here!