BALANCED HEADPHONES? Why?

[Jim Shaw]

Forgive me, but the electrical engineer occasionally leaks out of me and needs an answer to clean it up.

Of what possible benefit are balanced headphones? I fully get the balanced 'thing' and why it's great for high sensitivity / low power / noise-prone signal levels. I've used balanced inputs, outputs, and cables for many decades with microphones and line-level signal transmission. Single-ended cables could never do what balanced do at low signal levels.
Fine. I get that.

But what advantage does balanced provide to headphones? And do those advantages (if any) warrant the additional circuitry necessary to get the balance? Do balanced headphones somehow test better? Sound better? Work better and cost less? I get it that there might be double the differential voltage, but that could be done in single-ended outputs by just upping the circuit to make twice the voltage.
A Corollary: Why don't we have 'balanced' speakers?

Is there something audiophilic here or is it another audiophetish to sell at a premium?

 

[PierreV]

There is that answer that seems (to my uneducated eyes) make some sense

Benefits of balanced headphones interconnection - nihtila.com

In balanced connection both speakers are individually connected to amplifier using two wires for each channel Crosstalk can be improved by tens of dB by using balanced interconnection Common-impedance coupling seriously limits crosstalk in conventional unbalanced headphones Balanced output can...

nihtila.com

Not sure if this is really audible or not, probably in some cases. It doesn't seem completely snakeoilish.

 

[Zensō]

Audio Myth - Balanced Headphone Amplifiers are Better

This Myth Goes Something Like This: "Balanced headphone amplifiers are better." "If balanced line-level connections work well, balanced headphone outputs should also work well." We disagree! Benchmark does not offer voltage-balanced headphone outputs on any of its products. The reason for this...

benchmarkmedia.com

 

[Human Bass]

Forgive me, but the electrical engineer occasionally leaks out of me and needs an answer to clean it up.

Of what possible benefit are balanced headphones? I fully get the balanced 'thing' and why it's great for high sensitivity / low power / noise-prone signal levels. I've used balanced inputs, outputs, and cables for many decades with microphones and line-level signal transmission. Single-ended cables could never do what balanced do at low signal levels.
Fine. I get that.

But what advantage does balanced provide to headphones? And do those advantages (if any) warrant the additional circuitry necessary to get the balance? Do balanced headphones somehow test better? Sound better? Work better and cost less? I get it that there might be double the differential voltage, but that could be done in single-ended outputs by just upping the circuit to make twice the voltage.
A Corollary: Why don't we have 'balanced' speakers?

Is there something audiophilic here or is it another audiophetish to sell at a premium?

Bi-amping speakers can be considered a kind of balanced speaker I guess.

 

[D!sco]

If ground loops are a common problem in your setup, say a USB connection to a busy motherboard, it's supposed to eliminate that issue. This matters more with very sensitive headphones and IEMs. They're also supposed to eliminate crosstalk in the amplifier by using two mono amplifiers, one left and the other right, each with their own separated ground ins/outs. Neither of these should matter in a well designed amplifier, though.

The truth:
More pins=bigger audio peen

 

[Lambda]

I get it that there might be double the differential voltage, but that could be done in single-ended outputs by just upping the circuit to make twice the voltage.

Making the tow opposing Amplifier per channel gives you slightly lower noise or better signal to noise. and some distortion cancels out.

 

[solderdude]

But what advantage does balanced provide to headphones?

All headphone drivers are always balanced. Its the cable and connector that makes the difference.

The only advantage is that there is no common return wire which can mess up stereo imaging when the resistance of the common return wire is relatively high compared to that of the driver.

This can be solved by running 4 wires from the headphone drivers to a 3-pin plug (TRS jack) and only connecting the return wires in the plug and not in the headphone (a 3 wire cable).
Most headphones with dual entry (so a cable going into each cup) can be used on balanced and single ended amplifiers.
There are also headphones with the cable going in one cup. Those can be balanced (4 wire) or SE (4 wire).
The only difference is the plug. In a balanced cable the plug has 4 connections, in a SE the plug has 3 connections but 4 wires.

This is only seen from the headphone side.

Reasons for balanced amps is getting double the output voltage that could be reached with a single amplifier when power supply rail voltages are limited.

 

[Atanasi]

One additional benefit of balanced headphone amps is that that the same circuits double as balanced pre-amps like Topping A90. In comparison, A30 Pro has single-ended amplification and doesn't include balanced pre-outs.

 

[brimble]

Audio Myth - Balanced Headphone Amplifiers are Better

This Myth Goes Something Like This: "Balanced headphone amplifiers are better." "If balanced line-level connections work well, balanced headphone outputs should also work well." We disagree! Benchmark does not offer voltage-balanced headphone outputs on any of its products. The reason for this...

benchmarkmedia.com

I like that article very much. Thank you.

But:

Headphone transducers are balanced devices. They have two wires.

(Most) electrostatic headphone transducers have three wires!

 

[waquo]

I get it that there might be double the differential voltage, but that could be done in single-ended outputs by just upping the circuit to make twice the voltage

Can you do that without changing the performance of the circuit?
My very limited understanding is that when designing an amplifier you have to design the circuit, select the parts and tuning parameters like transistor bias and all that to get the best performance out of the parts you selected.
A balanced amp then uses double the number of parts to get double the voltage (and up to four times the power) out of the same parts with the same tuning.

Let's say you selected a specific transistor and dialed in Class-A bias to where it gives you a certain performance. Could you up the voltage of that circuit without changing performance?

 

[AdamG]

Head tilt! Sore neck!

 

[garbulky]

I have an HD800S with a XLR cable. Schiit has some super cheap balanced amps. My only thing is right now, my XLR's are run balanced to my Emotiva XPA-1 gen 2 speaker amps from my Emotiva DC-1. The RCA cables from the DC-1 run to my unbalanced Emotiva Bas-x A-100. My system is voice controlled so I can turn the speaker amplifier off with my voice.
If I was to go balanced I would have to use some sort of splitter or a passthrough on the headphone amp. I'm not sure what kind of repercussions that would have to the impedance or distortion in the signal.

 

[waquo]

A Corollary: Why don't we have 'balanced' speakers?

We do, don't we? Speakers don't have three pole connectors with a shared ground for two speakers preventing balanced wiring, right?
As far as I can tell, the advantages are purely in terms of amplifier design. The headphones or speakers job is merely allow a balanced connection.

 

[Jim Shaw]

Can you do that without changing the performance of the circuit?
My very limited understanding is that when designing an amplifier you have to design the circuit, select the parts and tuning parameters like transistor bias and all that to get the best performance out of the parts you selected.
A balanced amp then uses double the number of parts to get double the voltage (and up to four times the power) out of the same parts with the same tuning.

Let's say you selected a specific transistor and dialed in Class-A bias to where it gives you a certain performance. Could you up the voltage of that circuit without changing performance?

If your amp is too low output, scrap it and redesign the output circuit rather than doubling the number of components a balanced design needs...?

 

[Jim Shaw]

Audio Myth - Balanced Headphone Amplifiers are Better

This Myth Goes Something Like This: "Balanced headphone amplifiers are better." "If balanced line-level connections work well, balanced headphone outputs should also work well." We disagree! Benchmark does not offer voltage-balanced headphone outputs on any of its products. The reason for this...

benchmarkmedia.com

Thanks for this reference. It tends to strongly parallel my thoughts.
Slogans and hearsay are easy; good performance is more difficult.

But I try to be open to other theories. Just using first principles and physics, I have not always been wrong. (Some stuff that could work, doesn't.)

 

[waquo]

If your amp is too low output, scrap it and redesign the output circuit rather than doubling the number of components a balanced design needs...?

I'm not an electrical engineer, but my impression is that at least some amp designers very carefully select parts and then build around that part.

See Flux labs descriptions of their amps, which seem to ascribe differences between them to the silicon used (MOSFET, J-FET, bipolar): Flux FA-10

Or this description of the design process behind the Schiit Magius, especially the output stage under "Internal Thunderdome Time", with previous mentions of how the Magni Heresy used 8 op-amps in its output stage and how they considered using a scaled up version with 32 of those op-amps for the Magnius): Jason Stoddard on Magnius

 

[Jim Shaw]

I'm not an electrical engineer, but my impression is that at least some amp designers very carefully select parts and then build around that part.

See Flux labs descriptions of their amps, which seem to ascribe differences between them to the silicon used (MOSFET, J-FET, bipolar): Flux FA-10

Or this description of the design process behind the Schiit Magius, especially the output stage under "Internal Thunderdome Time", with previous mentions of how the Magni Heresy used 8 op-amps in its output stage and how they considered using a scaled up version with 32 of those op-amps for the Magnius): Jason Stoddard on Magnius

Well, of course, the starting point of a circuit design is the performance you expect, not the individual components you might have on your bench. And if you are designing a headphone amplifier, the starting performance needed is the desired output current and voltage to drive a wide variety of headphones. The power supply voltage and current are a design subgoal, not just whatever you have lying around.

It is the difference between electrical engineering and tinkering, I suppose. Schiit occasionally takes exception to the rules. They have been known to admit that they sometimes design around an abundant part or parts that they can readily procure. That's become increasingly common in today's supply chains. It is especially likely when designing around available vacuum tubes and DAC chips.

 

[egellings]

Dynamic headphone drivers are inherently balanced. Should be no big deal to get the cable for them right.

 

[AnalogSteph]

Reasons for balanced amps is getting double the output voltage that could be reached with a single amplifier when power supply rail voltages are limited.

This is quite important in electrostatic headphones, which need a voltage swing of hundreds of volts p-p. High-voltage transistors are hard enough to come by as-is, no need to double the requirements on top of that. Even common tube types tend to be limited to plate voltages of around 250 V, maybe 300.

When this is not a major concern, e.g. Topping has shown that it is possible to design internally single-ended amps with exemplary performance by any standard (e.g. A50s). In this case the split between ground returns is merely shifted from after the plug to before the jack, avoiding a further 20-50 mOhms of common return resistance depending (on jack type). The important part is in the cable, which might weigh in around 1 ohm per conductor at 3 m.

 

[solderdude]

Yes, for stats you even need 6 wires per cable (3 wires per side)
For electret 4 wires in total would be sufficient.

For those wondering what the difference is between electret and electrostatic.
An electrostaic driver needs an external bias voltage (which requires the extra wire) where electrets have a permanent (very slowly discharges over decades) static charge on its membrane so does not need that bias voltage applied.