balanced output - is it as simple as doubling of the stereo circuit? (using opamp)

[ccaudle]

But there is current flow on the reference/return side. It's the same current as supplied to drive the transducer. Current flows in a loop.

Of course, but I thought that in the context of a paragraph discussing common mode interference currents that it would be obvious that the statement meant no common mode currents were flowing. It would be nonsensical to interpret as no signal current flowing.

In theory, better stereo separation

Yes, but is that useful? The very slight possible crosstalk from common impedance coupling of the left and right channels should be well below the masking threshold, and orders of magnitude less than the acoustic crosstalk when listening to speakers.
In fact many headphone amplifiers have circuits to specifically introduce crosstalk and make headphone listening a little closer to speaker listening, so I would argue that slightly lower crosstalk is of no benefit.

In theory... better electrical damping, less dynamic sound compression, etc.

You will need to explain the theory behind that. Electrical damping is essentially describing the resistance in the circuit, and there is no reason to think that adding an additional amplifier channel into the circuit is going to be lower impedance than just a direct connection to the power supply return.
Dynamic sound compression sounds like a fancy way of saying onset of clipping, which can be addressed by increasing the power supply voltage.

There is absolutely nothing wrong with using bridge drive amplifier circuits for headphones, it works on headphones just as well as it does for speakers. The only problem I have with it is when people start making up nonsense pseudo-technical justifications for it, or somehow try to conflate the very valid reasons for using balanced connections for line level connections between equipment with symmetrical drive of output transducers. Start with first principles and understand the reasons behind the various connection choices and eventually everything should make sense.

 

[k brown]

Another thread seemingly abandoned by the OP?

 

[Newmarket]

Of course, but I thought that in the context of a paragraph discussing common mode interference currents that it would be obvious that the statement meant no common mode currents were flowing. It would be nonsensical to interpret as no signal current flowing.

You could simply explicitly state "common mode" and avoid any misunderstanding with little effort. Whilst I have a good understanding of these things due to my "trade", this is not an explicitly "professional" forum - clue in "DIY" imo - and your statement could easily be misconstrued and so required correction. The distinction between DM and CM signals is not easily understood by many.

 

[kags]

They don't. Like everything else, they only care about the voltage across their two terminals. Balanced of course refers to symmetry of impedance not signal.

Cheers

hi-fi reviews and marketing material promotes “balanced” as a cure-all. It magically increases power, reduces noise and distortion, increases slew rate and damping factor. Any buzz words we missed?

They don't. Like everything else, they only care about the voltage across their two terminals. Balanced of course refers to symmetry of impedance not signal.

Cheers

Ian

Exactly - but somehow hi-fi marketing proclaims driving headphones with separate pairs of wires (“balanced”) to be a cure-app for whatever ails you.

I’ve never seen a hi-fi review or article that’s correct in describing CMRR or symmetrical impedance. Somehow they seem to charge more for their gear than pro audio manufacturers.

 

[moamps]

Yes, but is that useful? The very slight possible crosstalk from common impedance coupling of the left and right channels should be well below the masking threshold, and orders of magnitude less than the acoustic crosstalk when listening to speakers.
In fact many headphone amplifiers have circuits to specifically introduce crosstalk and make headphone listening a little closer to speaker listening, so I would argue that slightly lower crosstalk is of no benefit.

First, I quoted the discussion where only the completely separate wiring of the two headphone drivers was mentioned, and not the introduction of a balanced amplifier.

Masking threshold does not have an absolute value for all people, but depends on the individual, so your argument is completely wrong. The introduction of a system for crossfeed listening with headphones is completely wrong also because this mechanism of mutual influence is different (it can also include delay). In fact, the overall quality of the crossfeed system will be better if the separation in the headphones itself is better. That's why most such systems I've seen have separate headphone driver wiring.

You will need to explain the theory behind that.

Do I really have to explain that? What will happen if I don't explain it?
I would kindly ask you to address me with a little more respect next time.

Electrical damping is essentially describing the resistance in the circuit, and there is no reason to think that adding an additional amplifier channel into the circuit is going to be lower impedance than just a direct connection to the power supply return.
Dynamic sound compression sounds like a fancy way of saying onset of clipping, which can be addressed by increasing the power supply voltage.

When I mentioned these improvements, then the first word was "in theory", which means that always, not only in this case, reducing the resistance between an amplifier and an electromagnetic transducer will lead to the listed improvements. How much the effect will be audible in practice depends on several factors.
It's really inexplicable to me that none of you critics ask if there are any real, scientifically explainable reasons for using independent wires in headphones. When you look at the evolution of different headphones, for example the first AKG 240 had a 600 ohm system, now they are 54 ohm, can't you see that the working conditions have changed? On the other hand, the connecting cable has not changed, three wires of the same thickness, very thin. Because the "professional" cable must be thin, flexible, low mass and must not pull the headphones from the head. Even "professional" headphone amplifiers still have protective resistors on the output so they don't get destroyed if a short circuit occurs. The first Beyer DT100s that I used to solder the TRS connector to more than 30 years ago came with four wires, with separate systems and had 600 ohms as well, and now they are very often 32 or 16 ohms. Therefore, recent low-ohm headphones will probably sound better if the total resistance (the resistance of the protection resistor at the output of the amplifier and the resistance of the connecting cables) between the amplifier and the driver is reduced. And to come to that conclusion you don't need to be a "professional", it's enough to be a HI-FI lover.
It should be understood that the system we call "professional" does not necessarily mean that it is the best, but usually that it meets certain standards. For headphones, the TRS 6.3mm connector is mostly a professional standard, but it is certainly not the best connector, because the TRS connector itself is very bad and unreliable compared to XLR, LOMO and some other types of connectors.

There is absolutely nothing wrong with using bridge drive amplifier circuits for headphones, it works on headphones just as well as it does for speakers.

I will just state that the use of balanced amplifiers can be helpful because it increases the headroom in systems where the supply voltage is limited, and also the total noise and linear and non-linear distortions can be changed.

The only problem I have with it is when people start making up nonsense pseudo-technical justifications for it, or somehow try to conflate the very valid reasons for using balanced connections for line level connections between equipment with symmetrical drive of output transducers. Start with first principles and understand the reasons behind the various connection choices and eventually everything should make sense.

You definitely addressed this to the wrong person.

 

[LuckyNat]

I'd like to understand why we always tie one leg of an instrument (guitar, bass, etc) to ground.

A pickup is just a coil, just like a transformer, it would be a balanced source, why don't we utilize that and tie one leg to ground ?

We don't do that with microphones, why do we do it with instruments ?

Also goes for loudspeakers, headphones, just a coil, balanced connection, but one leg of the amp's output is always tied to ground.

Why is that ?

You might be surprised - some microphones do to that ... and they cause cause pin 1 problems and require ground lift here and there when interference gets in because of it.

 

[LuckyNat]

I don’t know any respected pro audio company making balanced headphones or amps.

Again, this is in regards to 4-wire headphone cabling, not balanced amp circuits.

They practically all do now .. think of a respected pro company and check out their headphone amp products..

 

[kags]

They practically all do now .. think of a respected pro company and check out their headphone amp products..

Balanced audio inputs and preamp outputs, yes. Balanced (4-wire) headphone connections, I'm not so sure of. I'll poke around at NAMM this week to see what Neve, LittleLabs, Prism, Benchmark, Manley, Dangerous, SPL, Grace, and some others are producing. I haven't seen balanced headphone outputs on any of these brands in the past, but maybe they are trying to woo some hi-fi users. Manley and SPL do make high-end consumer stuff, so they might have some units with balanced outputs. I have been paying close attention to headphone technology over the last few years, as I consult with certain manufacturers of pro audio gear, particularly speakers, headphones, and monitor system accessories.

I recently demoed the newest headphone amp from Schiit Audio (their Midgard amp) which does provide a TRS as well as a 4-pin balanced connector. Their designers claim the outputs are audibly indistinguishable from each other, but the 4-wire output allows a unique form of feedback called "Halo." Here, for what it's worth, is how they describe the 4-wire output (from the Schiit Audio: Audio Products Designed and Built in Texas and California):
"Halo is a mixed-mode feedback system that puts the driver into the feedback loop, a schema that some people think may enhance performance at the transducer, and some people think it may not, and our test results are mixed, and headphones don’t exhibit as much impedance variation and back EMF as big floppy speakers so the results may be harder to find. We figured we could spend 6 years testing it and maybe never release a product, or we could run with it, given that most people who’ve heard it think there’s an audible difference between Halo and non-Halo outputs."

I find Schiit's products to be well-made, fairly priced, and have no B.S. in their marketing or claims. Their products are useful and sound excellent, but may not be for everyone. For those who care, they are made in the U.S. (including the metalwork, and PCBs). Hard to beat for the price.

 

[thor.zmt]

So I have the circuit below (for a very simple headphone amp) for 1-channel (either right or left).

View attachment 120854

If I want balanced outputs, is it as simple as having 4 (R± + L ± channels) of the same circuit? Input will be coming from a fully balanced DAC.

Yes, but only if the potentiometer used has perfect tracking. Good luck finding a 4 track unit that comes close.

Also note that for a headphone amp balanced invariably has more noise and and distortion at any set SPL than SE, except for even harmonics which may cancel and boost odd harmonics even more, which is not a good thing, even overall THD is lowered.

Commonly the topology we see instead is exemplified by Bryston (also seen in Lake People / Violectric and SPL).

Use a frontend that is a balanced receiver, on Bryston's case switchable as high input impedance buffer for SE inputs.

Followed by a single ended volume control (and any possible processing for Crossfeed and bass boost etc. ahead of the volume control.

Followed by a single ended headphone amplifier and the "balanced headphone out" created using the same headphone amp inverting.

Technically this is best described as balanced input with bridged output.

Truly differential headphone amps are exceedingly rare.

I designed one and in manufacturing the highest selection grade of ALPS RK-168 4 track volume controls Alps-Alpine offers has a 70% rejection rate (which is passed to the customer with the rejects dumped on eBay etc.).

This scheme is really only viable with digital volume control in the Dac or stepped resistor ladder attenuators (chip or discrete). Even then benefits are questionable.

Thor

 

[thor.zmt]

Good question. Balanced headphones are promoted in consumer hi-fi world where a lot of principles get twisted and never critically examined or listened for…

Indeed, unless the the headphone needs more drive voltage than a SE amp can deliver, balanced amplifiers are invariably worse than SE for transducers.

The origins of all this balanced HP stuff are a result of 3.5mm jacks on very expensive and and very low impedance in ear monitors.

A 3.5mm common pin can easily exceed 100mOhm and I have seen over 0.5 Ohm on well worn sockets. If then the HF impedance is below 4 Ohm because of using two double barreled Knowles Balanced Armature super tweeters (22 Ohm each) all drivers in parallel, we have a bit of a problem.

On a 6.3mm jack with 300 Ohm headphones, less so, a lot less so.

So while "balanced wiring" is mandatory for low impedance transducers, balanced amplification is often counterproductive.

Sadly this subtle point is commonly lost on both consumers and marketing flunkies who are happy with the mantra balanced good, unbalanced bad, kinda like Orwell's Animal Farm (4 legs good, 3 legs bad).

I don’t know any respected pro audio company making balanced headphones or amps.

Bryston, Violectric (aka Lake People), SPL have been doing it for bleedin' ages.

Again, this is in regards to 4-wire headphone cabling, not balanced amp circuits.

4 wire (or even 5-Wire) headphone wiring for low impedance headphones is a great idea.

In 2024 wired headphones, at least in portable / mobile applications are rapidly becoming extinct, so the only valid application disappears.

Thor

 

[ImInTIgaToRFERY]

Sorry for just coming back to the thread. Was away.

And thanks alot for the replies!

 

[ccaudle]

I would kindly ask you to address me with a little more respect next time.

I became very busy at work and lost track of this thread, but I want to come back to clean up a mess I apparently left.
@moamps I sincerly apologize for giving the impression of disrespect. I assure you the impression was only because I was in a hurry and slipped into some local idiomatic expression that I would generally try avoid in favor of more standard language to avoid misunderstandings such as this. I value your contributions here, and if I were not in such a hurry at the time I would have phrased it more like "I don't understand that, for me to follow your argument you would need to expand the reasoning more for me to follow your point."

You definitely addressed this [complaint of pseudo-technical justifications] to the wrong person.

That was not intended to be addressed toward you at all, again a case of rushed response, that would have been better expounded as a discussion of how mainstream consumer magazines often latch onto particular solutions with a fetish-like devotion without full discussion of the technical reasons and trade-offs behind the design. I was curious whether the OP wanted to just try this approach to evaluate, wanted to recreate a particular commercial design, or needed symmetrical drive because of running from low voltage batteries, etc. or was following some kind of pop-culture "balanced is better" slogan without first understanding why you would or would not use that approach.

 

[RCAguy]

You might be surprised - some microphones do to that ... and they cause cause pin 1 problems and require ground lift here and there when interference gets in because of it.

The microphone doesn’t cause a pin1 problem (unless its metal case comes in contact with a grounded metal object). Usually it’s when an XLR’s shell is bonded to its pin1.