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

[ImInTIgaToRFERY]

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



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.

 

[k brown]

Balanced headphone out?

 

[ruffrecords]

If you have a balanced output (voltage) from your DAC then, yes, you can drive two identical circuits. The output will be balanced although headphones do not require it. What you will get, though, is twice the voltage swing, so four times the potential output power.

Cheers

Ian

 

[LuckyNat]

Have a quick read of this to check balanced is what you think it is:

https://www.edn.com/the-g-word-how-to-get-your-audio-off-the-ground/
Really, balanced is about the difference between two references. That could be a +2v swing version of the signal and a -2v voltage swing of the signal (as you are proposing) or it could be +4v swing against a steady 0v reference. It doesn't make a difference. (I'm probably using wrong terminology - I'm using +2v to denote one phase and -2v to denote the opposite phase ).

So, can you make a better amp with 4v voltage swing instead and simply wire the transducers to they dont share ground and each have their own 0v reference from the amp?

I thought the idea was that interference eliminated and that the transducers are floating and also both sides of the tranducer see the same impedence.

I'm not experienced by the way ...

 

[ccaudle]

I thought the idea was that interference eliminated and that the transducers are floating

Headphones are always floating, so the term "balanced" as commonly used for headphone amps isn't really using it in the strictly proper sense in the way that line level "balanced interface" is used.
This is really what is called a "bridging amplifier," and quite a few amplifiers made for speakers can do the same thing, usually to get high power with a particular power supply voltage, at the expense of using two amplifier channels. It was very common in car radios to get higher power than you could get from a single channel with just a +12V supply (and may still be common in automotive, I'm not so active in that area these days).

To the original question, yes you can do that. Will require a dual pot per channel, so 4-gang pot for typical stereo use. Make sure you connect the pots the correct way so that the wipers of the two pots move toward each other (lower resistance between the two wipers) as you turn counterclockwise. The connection point between the two pots does not have to be grounded, you can have just the connection between pins 2 and 3 from the input connector since you are AC coupling the amp channels. If you do that make sure that there is some other connection which keeps the common mode voltage between the DAC and the amp low if they are in separate chassis (typically the shield connection of the cable should do that if one or both devices has a two wire power input; with three-wire power input the safety earth connection and the cable shield will both work to keep the low frequency common mode voltage low).

 

[Newmarket]

Headphones are always floating,

Are they ? Not ime unless we have a different understanding of the terminology around this.

 

[k brown]

Of course headphones can Float -
https://www.stereophile.com/content/jecklin-float-electrostatic-headphones

 

[ccaudle]

Are they ? Not ime unless we have a different understanding of the terminology

You are right, that was not a correct choice of phrasing on my part. Typically one side of a headphone driver is tied to the circuit return, not floating.

What I meant to convey was that coil headphone drivers are intrinsically floating, they are not like e.g. an input on a bipolar device which has to be tied into the power supply to provide bias current. You do not have to tie one side of the headphone to circuit ground, or to any other power supply node, and unless you have a somewhat exotic driver like biased electrostatics, the headphones have no power supply, so there is no need to deal with inter-chassis currents, which is the primary advantage of a balanced interface between different pieces of powered equipment.

Without a path for common mode currents to flow, headphones do not have the big disadvantage that single ended line interfaces between chassis have, which is that the cable shield is carrying common mode currents (and thus having a common mode voltage impressed across it), and is also the signal reference. The reference side of the headphones should have no current flow, so no voltage drop, so no interference signal.

 

[kags]

Balanced headphones refer to headphones where the left and right drivers do not share a common ground. Right?

They use a 4-conductor cable and require an amp with isolated channels.

In practice (at least in my experience) good amps and headphones don’t sound better when using balanced drive.

Is this what the OP is speaking about?

 

[k brown]

Balanced headphones refer to headphones where the left and right drivers do not share a common ground. Right?

They use a 4-conductor cable and require an amp with isolated channels.

In practice (at least in my experience) good amps and headphones don’t sound better when using balanced drive.

Is this what the OP is speaking about?

If so, what is the supposed advantage of that?

 

[Newmarket]

You are right, that was not a correct choice of phrasing on my part. Typically one side of a headphone driver is tied to the circuit return, not floating.

What I meant to convey was that coil headphone drivers are intrinsically floating, they are not like e.g. an input on a bipolar device which has to be tied into the power supply to provide bias current. You do not have to tie one side of the headphone to circuit ground, or to any other power supply node, and unless you have a somewhat exotic driver like biased electrostatics, the headphones have no power supply, so there is no need to deal with inter-chassis currents, which is the primary advantage of a balanced interface between different pieces of powered equipment.

Without a path for common mode currents to flow, headphones do not have the big disadvantage that single ended line interfaces between chassis have, which is that the cable shield is carrying common mode currents (and thus having a common mode voltage impressed across it), and is also the signal reference. The reference side of the headphones should have no current flow, so no voltage drop, so no interference signal.

So - to summarise - you are considering passive devices that are not "grounded" other than by the cable. For common dynamic headphones that is correct. Similar to dynamic microphones.
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.

 

[PermO]

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 ?

 

[Tubetec]

Most old style hifi amps can be supplied with a balanced signal on L/R channels , then the speaker output is connected across the two positive L/R terminals .

I rememeber the older car radio ICs ,
TDA 2005 series , they came in a few varieties , same package , the mono version was exactly double the power per channel of the stereo , it was the same chip just internally connected differently .

Two stereo amp modules of this kind in a box ,running off 12v , able to drive loudspeakers or headphones , Ive been meaning to try it for ages ,

In a tube amp you have high voltages on the output transformer primary , the speaker line is grounded on one side so if a fault developes with the insulation a person isnt exposed to a dangerous voltage when making/breaking the speaker connection .

 

[berret]

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 ?

Because most amplifiers have an u balanced input. Making a balanced input for a guitaramp, especially tube amplifiers, costs money...

 

[Hubbub]

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 ?

In a word, cost. To be more correct, historical cost. In the early tube days the only easy way to get any input or output balanced meant an expensive transformer. Use an unbalanced input and the transformer goes away, the build is easier and costs less. So two conductor connectors abounded, both the old screw type for microphones, RCA pin plugs, and the ubiquitous 1/4" phone plug. That's what early guitar connectors used, as the three conductor (microphone) connectors were really bulky in the pre-XLR days. Even XLRs are a bit bulky for a musical instrument.

 

[kags]

If so, what is the supposed advantage of that?

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…

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.

 

[moamps]

If so, what is the supposed advantage of that?

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

.... and require an amp with isolated channels....

Not necessarily.

 

[FunkyTek]

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 ?

guitar amp inputs were connected to ground because the jacks were hard clamped mechanically to the chassis and the steel chassis was used as the ground plane. In order to isolate the input ground to say the reference of the input amp, they would have to have physically isolated the 1/4" phone jack at the input. That kind of jack was not readily available in the 40's-70's. Worse was the problem of no ground pin on the power cord. 3 prong AC outlets were not very abundant until the 80's and 90's. To make it worse, the 2 prong cords and outlets were not indexed, so a plug could be flipped over and plugged back in the same outlet. Which side of the plug was neutral was completely arbitrary. So the amp "ground" was just an arbitrary choice of AC power pole. You had to very carefully touch the mic with your hand on the guitar strings to find out if your amp just happened to be on the same AC polarity as the mic. If not then you got a shock. Better with your finger than your wet lips when trying to sing! I used to touch the strings to the mic and see if I saw a spark.
Only the oldest mics are single ended. They are almost all fully differential. The design and operation of mic diagrams and cartridges are very different than a guitar pickup. The guitar pickup could be wired as a differential circuit but this would require a differential front end on the amp to actually make it effective. Plus passive tone controls require a single ended circuit, in practice.

Amp outputs are tied to a reference or they can (will) float up in voltage wrt to the power supply ground. This can be especially problematic when global feedback is applied in the circuit. It is necessary if their is global feedback but not necessary if there is no feedback loop, but still a potentially dangerous situation.

 

[Voyager10]

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…

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.

Right - I could be persuaded that a 4-wire connection has some advantages - lower crosstalk between L and R is the easiest to quantify.

But 'balanced' - interpreted as meaning 'symmetrically driven' - is a little baffling. Why would a headphone drive unit care about the potentials of its + and - inputs with respect to an arbitrary 'ground'?

 

[ruffrecords]

Right - I could be persuaded that a 4-wire connection has some advantages - lower crosstalk between L and R is the easiest to quantify.

But 'balanced' - interpreted as meaning 'symmetrically driven' - is a little baffling. Why would a headphone drive unit care about the potentials of its + and - inputs with respect to an arbitrary 'ground'?

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