Balanced vs unbalanced 4 wire connection

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi

The fashion in 'high-end' headphone amps these days is for so-called 'balanced' connection of headphones. IMHO this is a misnomer since what's really being done is to power the headphones with separate amps for each driver's +VE and -VE terminals in a bridge circuit.

The main advantages claimed for this arrangement are that it increases slew rate and power output compared to the conventional single-ended connection approach and that the lack of a shared ground between the two drivers reduces crosstalk, thereby improving the sound stage.

While I don't have any real issues with these claims (other than the balanced/bridged arrangement halves the impedance of the headphone driver as seen by the +VE and -VE amps), could the same benefits not also be achieved by simply using a 4-wire connection with conventional single-ended amps, with two wires for each driver's +ve and -ve/gnd, assuming that the amp has sufficient slew rate and can deliver sufficient current to drive the phones adequately?
 
Last edited:
This is very interesting. Balanced mode with no shared ground. Does it mean there's even no shield which is grounded? Then what's the point of transmitting balanced signal? Twice peak-peak swing?

Yes, that's it really. Bridging increases slew rate since each amp only has to supply half the signal and allows the power delivered to the headphones to be doubled. Apart from that it seems like marketing BS to me so they can sell amps more expensively. In my opinion, the main benefit of so-called balanced headphone amps is that there's no shared ground line between the Left and Right channels.

That said, anyone want to tell me then why a simple 4 wire connection using single ended amps shouldn't work equally well, assuming as I said previously, that the single ended amps have sufficient slew rate and output current for the impedance of the headphones?

The main thing to realise is that it's BRIDGING not BALANCED connection. Balanced would imply that the headphones have common-mode rejection, which they don't.
 
Last edited:
the terms "balanced headphone", "balanced headphone cable" may annoy some but in fact each driver wired with its own pair of wires and contacts is a pretty good "balanced impedance system"
each driver is a very high common mode rejection "differential receiver" - the "common mode impedance" is the very small balanced parasitic capacitance of each cable pair to your body, headphone/driver/earcups, each other
a dynamic driver only responds to the differential V at its voice coil terminals
calculating the cable effects of pF parasitic C, then taking just the unbalanced portion, trying to drive tens to hundreds of Ohms largely resistive drive at audible frequency - you need to be up on exponential notation

there is a technical advantage of separating the returns from reduced common path impedance coupling compared to 3 wire TRS common shaft being used for both channels

the term "balanced" could be less descriptive on the amp end if you want to describe bridged output amps with no actual common mode rejection of their own
 
Last edited:
the terms "balanced headphone", "balanced headphone cable" may annoy some but in fact each driver wired with its own pair of wires and contacts is a pretty good "balanced impedance system"
each driver is a very high common mode rejection "differential receiver" - the "common mode impedance" is the very small balanced parasitic capacitance of each cable pair to your body, headphone/driver/earcups, each other
a dynamic driver only responds to the differential V at its voice coil terminals

I don't think this is correct. The driver coil isn't a transformer type balanced receiver since there's no galvanically isolated secondary and it clearly isn't an op-amp or transistor type balanced receiver so there can be no CMRR. Just having a parasitic capacitance doesn't constitute a balanced or differential receiver.

there is a technical advantage of separating the returns from reduced common path impedance coupling compared to 3 wire TRS common shaft being used for both channels. And I think this is the real reason why so-called 'balanced' headphone amps may sound better than single-ended amps. My original point was couldn't this be more simply and more cheaply achieved by using a 4 wire connection with conventional single-ended amps, where the L and R ground/return wires to the headphone drivers are kept separate up to the point in the amp where they have to be connected (star gnd)?

the term "balanced" could be less descriptive on the amp end if you want to describe bridged output amps with no actual common mode rejection of their own. Balanced headphone amps ARE bridged amps. They are not the same as the line level balanced output configuration used in true balanced systems. This means that each half of the bridge (i.e. each +VE and -VE amp) sees half the impedance of the headphone coil, which doubles the current delivery requirement for the amp compared to single-ended.
 
Last edited:
I think there are a few sources of confusion - "balanced" in EE, Audio has multiple context dependent meanings, some involving Common Mode vs Differential

while you are sticking to physical implementation details it will be hard to see principles

in your 1st instance it can even be argued that a dynamic headphone driver is "a transformer with an isolated secondary"
because a dynamic driver is usefully modeled as a electro-mechanical-acoustic transformer - the sound radiated from the diaphragm has no strong relation to the local Voltage Common to the headphone, both its voice coil terminals (the "transformer primary"), you, the amplifier...
you could use a DAP in your hand or pocket and a walk across a carpet in a dry room could charge you, the amp and headphones to >10 kV without changing the sound
 
Last edited:
JCX

Well it's practical physical implementations that matter isn't it, not esoteric reasoning?

The fact is, if you look at the circuits of so-called balanced headphone amps on the web, they're not balanced they're bridged. Sure, some may have a balanced/differential receiver at the input but the output is bridged.

It's a clever marketing ploy I'll grant you, but inaccurate.
 
JCX - do you agree that the biggest factor in whether so-called balanced headphone amps offer improved sound over single-ended amps is most likely the necessary separation of the ground return paths for L and R channels with 'balanced' amps, thereby avoiding a shared ground impedance, rather than the use of what are really bridged outputs?
 
from what i know (not much) a '4 wire' single ended headphone system can only work if the amp is fully dual mono. if the amp is a single board/power stereo amp then it has shared internal ground anyway and as such separating the two headphone ground wires is going to do nothing as its already shared at the amplifier. fully dual mono is a completely separated amp for left/right and as such should have separate power/ground for left/right which would give the headphone 4 wire system a separate ground for left/right. you would also need 4 contact jack/socket for the headphone wires which in turn limits the headphones to the dual mono amp.
 
some advantage can be had - the biggest, most variable common path impedance will usually be the TRS ground - usually a "single point contact" cantilever beam touching the round shaft

for 1/4" TRS mated jack/plug this can be ~ 10 milliOhms, for 1/8" TRS much higher

with 4 pin "balanced" connector and headphone wiring you can use pin/sockets that have multiple contacts - easily achieving single digit milliOhms contact resistance
sub milliohm contact R is possible if you use some of the hyperboloid wire cage or louver multi contact sockets

care with internal amplifier return paths can still give big reduction in crosstalk with 4 wire headphone drive even with common power supply, single ended drive

but yes, "dual mono" does simplify things, maximize some advantages of a 4-pin headphone connection
 
Is it advantageous from a cost perspective to have 2 power supplies and 2 single ended op-amps VS 1 power supply and 2 differential op-amps?

Each channel in my design is:
Differential DAC output -> differential op-amps -> BTL headphone connection, all powered by supplies shared between the channels.

It was drastically more expensive to implement 2 of the supplies I used (one for each channel), as I had originally designed it that way. By using 1 set of supplies and a BTL configuration, I saved on board real estate and parts costs.
 
Hello,

If I read Bruno's paper right (link below), one of the biggest plusses of balanced circuits is that the signal path (both left and right) is isolated from the ground connections and any ground loops and hum.

Keeping the ground circuit pretty much completely separate makes sure your ground currents between different part of the circuits and chassis's doesn't influence the signal. No more moving ground points around trying to minimize problems, just design it out.

http://www.hypex.nl/docs/papers/The G Word.pdf

Excellent if rather technical paper. See what you think.

Regards,
Greg
 
Sorry to disagree, but as you give no reasons at all to support your view I feel you haven't read the paragraph quoted below. It seems to state my case quite clearly, try reading it.

"The hidden assumption is that a signal is just one
wire. But as anyone with a voltmeter knows, the
second wire is every bit as important as the first.
Still we seem to think it makes sense to use as the
second wire the central sewage pipe that also carries
waste electrons, supply return currents, shield
currents etc back to the recycling plant. And then
we’re surprised to find rubbish on it.
The supposed solution is called a “star ground”, a
common point where “different grounds” connect."

Greg
 
Member
Joined 2014
Paid Member
Actually it shows that you are missing the point .Read and re-read
But as anyone with a voltmeter knows, the second wire is every bit as important as the first.

When you are passing signals between units you end up with two parts to your signal. The differential mode part that you want, and the common mode part that you don't. A balanced link allows you to deal effectively with that case as you have the same impedance on both legs (send and return) and are measuring between the two. Even with a mythical perfect ground you would still want that balanced link.

Think of current flowing in loops. Send and return. Suddenly you realise it's not that you are not connecting to ground, it's the fact that ground as a concept isn't there.
 
As nobody said left or right and don't know where you are pulling stuff out of instead of reading what was actually said, still disagree.

Of course as you opinion is worth not a dime more than anyone else's, not going to worry about it. Make up whatever you want. Keeping the grounds totally separate from the signal wires does make sense and will work. Professional equipment makers have had that figured out for a while, but commercial equipment makers just don't seem to get it yet.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.