Understanding Balanced Audio (video)

[lurkera]

Thanks amir, I was one of the posters that was confused about the balanced and differential terminology used. For a more technical understanding, is there a circuit/signals book that can be pointed to (I'm comfortable with mathematics)?

 

[Deleted member 2944]

The wikipedia page on this is pretty good.
https://en.wikipedia.org/wiki/Balanced_line
https://en.wikipedia.org/wiki/Balanced_line#Balanced_and_differential

Dave.

 

[PeteL]

Semantics? No, these are actual interfacing issues that could exhibit in real-world systems.
Amir seems primarily interested in the alleviating of ground loops aspect of this, but there are other aspects as well.

This video seems aimed at less experienced audio types, and in that sense it's probably fine and dandy. But there is much more to this topic than he hit on. Anyways, it's not a big deal. It's just something I would have expected Amir to elaborate on.
"Balanced audio" is a premise that a couple of audio hardware manufacturers have placed their whole business model on.

Dave.

Sure, I don't fully disagree with your general point. I was just not getting the "Amir makes the same mistake" part, since he specifically said that there is more to it but that he would not adress this part because for him and his use case, grounding issue is the main reason, which is debatable but not fully wrong, in typical hifi setups. Note as well that I don't think this was targeted at audio professional, just the general hifi crowd who wonder if it's worth going balanced. I personally use balanced at home for this reason, ground related issues. Never had issues with my rca that where related to interference pickup. CMNR in cabling has been thought out to go the distance and he does mention it.

 

[Anthony101]

I have been meaning to write an article for a while on what balanced audio means. Alas, have not had time so I thought I do a video on it instead. Hopefully for people who are new at this, it clarifies different things we call "balanced" and benefits it can provide:

Hope you find it useful.

Thank you for explaining this mess about balanced inputs, cables and headphones in simple, easily-understood terms.

 

[Atanasi]

Question, which I believe it was addressed slightly. Are TRS connection / plugs (like form a mixer or pre-amp) balanced?

Line-level 1/4" TRS in mixers, audio interfaces etc. is most commonly balanced mono. For headphones it is the ordinary stereo connector.

 

[Tks]

When I look at pro interfaces, i see no XLR inputs, yet advertise balanced connections, and I proceed to get lost again due to all the barrel connectors being all weird due to some being mono-balanced or not (ugh, headphone amps, so much simpler).

One thing I wanted to ask. Is a downside to balanced (in theory if you could equalize balanced and unbalanced connection free from ground loops) where I've seen some single-ended portions of an amp perform better in terms of noise (or was it distortion) due to the non-doubling of said power that balanced provides?

If we're doubling power, doesn't that mean there is also an increase in the noise or distortion? I understand because you eliminate things like groundloops, you almost always have better noise performance in most cases, but if that was to be accounted for, wouldn't it mean balanced is worse on paper in that one respect even if by a hair?

Lastly, I know it's been said, voltage is doubled. Does that mean that current remains completely equal?

 

[amirm]

When I look at pro interfaces, i see no XLR inputs, yet advertise balanced connections, and I proceed to get lost again due to all the barrel connectors being all weird due to some being mono-balanced or not (ugh, headphone amps, so much simpler).

They use 1/4 TRS ("headphone") jacks or combo versions that take that and XLR. They are balanced either way though.

 

[Raindog123]

Is audio "balanced" analog signals idea the same as differential signals concept in general? (Eg, like good old RS422, or is there something on top)?

 

[amirm]

Is audio "balanced" analog signals idea the same as differential signals concept in general? (Eg, like good old RS422, or is there something on top)?

Yes, they used balanced and differential signaling. The concept is extremely common in electronic interconnects.

 

[phoenixsong]

Question, which I believe it was addressed slightly. Are TRS connection / plugs (like form a mixer or pre-amp) balanced?

I think it depends, the dual TRS monitor outs from my Motu M2 are

 

[Deleted member 2944]

Yeah, it does depend.
A lot of professional equipment TRS outputs only have active signal on the tip and the ring will have (just) an appropriate resistor to ground.......as mentioned earlier in this thread.
That does create a balanced interface, although you might not consider it differential signaling.
That's what's commonly called a pseudo-balanced output.

Dave.

 

[KSTR]

Question, which I believe it was addressed slightly. Are TRS connection / plugs (like form a mixer or pre-amp) balanced?

If it's spec'd as TRS (vs. TS) then it is usually at least impedance-balanced... and even when it's not it still fulfills the main idea: keep the shield out of the signal circuit, so "ground loop" problems -- which is voltage drop along the shield from balancing currents between gear (from potential differences of the audio grounds) -- are mostly eliminated.
But, this core benefit of balanced audio connection can be had with simple TS or RCA outputs as well with proper adapters to balanced cable, all you need is a balanced input at the receiving end.

 

[MasterMech]

Thanks for the great video Amir & taking the time to make it, I learned quite a bit about balanced connections here.

 

[SHENZHENAUDIO]

Thanks Amir! It's greatly helpful!

 

[charleski]

There may be some cases in which it's beneficial to run a single-ended headphone connection over 4 wires (i.e. separate signal and return for each channel, with the return wires tied together at the jack). Headphones will always produce crosstalk simply by means of the sound from one cup transferring through your skull to the other ear, and with open drivers there will be leakage happening through the air as well. I couldn't find any specific measurement of the point at which crosstalk becomes audible on headphones, but it seems that a figure under -40dB will be acceptable. Subjectivists love their vinyl, and you need a very good vinyl system to get crosstalk under -30dB, so this seems reasonable,

The crosstalk produced by using a common return wire can be calculated as follows (as per the article by @solderdude) :
20 * log (Zreturn / Zreturn + Zsignal + Zheadphone))
Put this in Excel and you get the following:

Return             0.1             0.25            0.5             1
Signal             0.1             0.25            0.5             1
        Crosstalk           
Headphone   16    -44.19030029    -36.39087871    -30.62957834    -25.1054501
            32    -50.15711743    -42.27886705    -36.39087871    -30.62957834
            64    -56.15070056    -48.23239412    -42.27886705    -36.39087871
            128   -62.1577605     -54.21926238    -48.23239412    -42.27886705
            300   -69.54821376    -61.59808935    -55.59192983    -49.60013886
            600   -75.56591982    -67.61146006    -61.59808935    -55.59192983

So in most cases a 3-wire cable will do fine, but if you have very low impedance headphones and a long (or very thin) cable then there's a case to be made for switching to a 4-wire. I use a set of low-impedance (17Ω) planars with a 3m cable, so got a cheap 4-wire replacement off Aliexpress for under £30.

 

[VariousArtists]

Thanks @amirm, another great educational video.

I still have 2 (probably stupid) questions:

1) Is it fair to say that the elimination/reduction in ground loop hum is not applicable in a single unit combining DAC & Amp (e.g. RME ADI-2 DAC)?

2) "Headphone cables are always balanced" - does that mean there is reduced chance of EM interference as it affects both wires the same way? I still hear lots of interference from Wifi when I align the headphone cable along the Wifi antenna on my iPad, why is that?

 

[MRC01]

... 1) Is it fair to say that the elimination/reduction in ground loop hum is not applicable in a single unit combining DAC & Amp (e.g. RME ADI-2 DAC)?
...

No, because there could be a ground differential or loop between the source (CD player or whatever) and the DAC/amp. However, an optical connection can help eliminate any hum that might cause.

Also, I've heard unbalanced headphone amps hum when nothing is connected to their inputs. Since unbalanced audio ties signal reference (ground) to frame/chassis ground, any kind of noise or signal on the frame/chassis ground taints the music.

 

[VariousArtists]

No, because there could be a ground differential or loop between the source (CD player or whatever) and the DAC/amp. However, an optical connection can help eliminate any hum that might cause.

Makes sense - how about other digital connections like USB? I thought ground loop is only relevant in analogue signals / amplification.

 

[MRC01]

... If we're doubling power, doesn't that mean there is also an increase in the noise or distortion? I understand because you eliminate things like groundloops, you almost always have better noise performance in most cases, but if that was to be accounted for, wouldn't it mean balanced is worse on paper in that one respect even if by a hair?
Lastly, I know it's been said, voltage is doubled. Does that mean that current remains completely equal?

Is this what you're asking? Imagine a balanced output as 2 amps, identical yet one's output is inverted from the other. So the downstream device sees the difference between their voltages, which since they're inverted, is always twice the voltage of either. So the power delivered to the downstream device increases by 4x. Since we have 2 amps, each must be producing twice the power. An amp producing twice the power may have slightly higher distortion?

By analogy, a balanced output's differential signal is like bridging the L & R channels of a stereo amplifier. When bridged, each channel operates inverse to the other, and sees half the load impedance. So each delivers the same voltage to half the impedance, which means it delivers twice the current, which is twice the power output. As each channel produces twice the power, it's 4x the power output. The speaker sees the voltage difference between them, which is twice the output voltage, twice the current draw, 4x the power. Of course, the amplifier doesn't magically get 4x power, it can really only produce 2x the rated power. It is current limited and since it forces each channel to see half the load impedance, you can only do this with high impedance speakers.

Most amplifiers that can be bridged, have slightly higher distortion in bridged mode. Does this apply equally to balanced/differential audio outputs?

PS: regarding differential signalling and distortion, I think it reduces or eliminates even order harmonics. I don't know what it does for odd-order harmonics - perhaps nothing?

 

[Speedskater]

Question, which I believe it was addressed slightly. Are TRS connection / plugs (like form a mixer or pre-amp) balanced?

In semi-pro audio equipment, 1/4 inch TRS connectors may be used in different ways.

• mono balanced interconnect
• stereo unbalanced interconnect or headphone cable
• mono unbalanced send/return effects cable
• mono speaker level cable (with heavy conductors)