A lot of high-end headphone amps these days sport so-called 'balanced' headphone outputs to drive modified 'balanced' headphones where the left and right channel grounds are run separately in a 4-wire arrangement (i.e. L + Lgnd and R + Rgnd) instead of the single-ended 3-wire (L, R and Lgnd+Rgnd) arrangement used in a conventional TRS headphone plug.
My first contention is that these so-called 'balanced' outputs aren't balanced at all since they don't feed a balanced receiver (either a transformer or electronic) which provides common mode rejection of noise and interference, the very reason for balancing in the first place. What they really are are bridged amplifier outputs, where one amplifier provides one phase of the headphone coil drive voltage and another amplifier provides the opposite phase.
Bridged amplifiers, while offering higher drive voltage and power output into the headphones (often not required) have the disadvantage that each amplifier sees half the total headphone load impedance and so has a harder time driving them, potentially resulting in higher distortion.
My second contention is that all the benefits of so-called balanced outputs can be easily met by adopting a simple 4-wire single-ended approach to driving headphones with the left and right grounds separated and sent to each headphone drive unit via twisted pair wiring. This will eliminate the biggest problem in 3 wire headphone outputs i.e. shared left and right channel ground crosstalk and interference without the need for more complex 'balanced' circuitry.
Am I wrong here? Are balanced headphone outputs just a marketing gimmick?
My first contention is that these so-called 'balanced' outputs aren't balanced at all since they don't feed a balanced receiver (either a transformer or electronic) which provides common mode rejection of noise and interference, the very reason for balancing in the first place. What they really are are bridged amplifier outputs, where one amplifier provides one phase of the headphone coil drive voltage and another amplifier provides the opposite phase.
Bridged amplifiers, while offering higher drive voltage and power output into the headphones (often not required) have the disadvantage that each amplifier sees half the total headphone load impedance and so has a harder time driving them, potentially resulting in higher distortion.
My second contention is that all the benefits of so-called balanced outputs can be easily met by adopting a simple 4-wire single-ended approach to driving headphones with the left and right grounds separated and sent to each headphone drive unit via twisted pair wiring. This will eliminate the biggest problem in 3 wire headphone outputs i.e. shared left and right channel ground crosstalk and interference without the need for more complex 'balanced' circuitry.
Am I wrong here? Are balanced headphone outputs just a marketing gimmick?
Last edited:
@Gopher, my opinions match yours. I have ceased tearing my hair out over this (virtually none left!). I think that there is a distinct lack of technical understanding in the "Head Fi" community that causes them to promote mythological benefits of what they believe to be "balanced". The so-called "balanced" headphone output stages are actually bridged amplifiers. It seems to be isolated to that single community of users, quite different to the world of pro audio and audio engineering.
I disagree with your first contention:
A single headphone speaker has perfect common mode rejection, so why not take advantage of this.
I have also found that a proper balanced circuit can be simpler than a single ended circuit.
I would like to state that it really is a balanced world and single ended design is an simplification of the signal return path design.
Tim
Can you provide evidence for your statement that a single headphone speaker provides perfect common mode rejection? Also, how can a balanced amplifier be simpler if it requires more components?
Are Murray and I both wrong in saying that so-called balanced headphone outputs are nothing more than bridged amplifiers? Perhaps they may sound better than a poorly implemented single-ended amp with insufficient current/power output because they deliver twice the current and power into the load, but that's a design issue, not a consequence of being 'balanced'.
Can you provide evidence for your statement that a single headphone speaker provides perfect common mode rejection? Also, how can a balanced amplifier be simpler if it requires more components?
Are Murray and I both wrong in saying that so-called balanced headphone outputs are nothing more than bridged amplifiers? Perhaps they may sound better than a poorly implemented single-ended amp with insufficient current/power output because they deliver twice the current and power into the load, but that's a design issue, not a consequence of being 'balanced'.
Last edited:
A balanced impedance receiver has the same impedance seen at the Cold input as it has at the Hot input.
Now look at the two wire input of a single driver speaker/headphone.
The two inputs have an identical impedance. They can only be a balanced impedance load.
And just to reinforce that, look at the impedance with reference to "Earth". It is enormous. So enormous your low voltage resistance/impedance measuring instruments read infinity !
Now look at the two wire input of a single driver speaker/headphone.
The two inputs have an identical impedance. They can only be a balanced impedance load.
And just to reinforce that, look at the impedance with reference to "Earth". It is enormous. So enormous your low voltage resistance/impedance measuring instruments read infinity !
Andrew
So are you saying that a headphone driver coil is a perfect balanced load because it has the same impedance at both inputs? So does a nail.
Where does the CMR come from since there's no electrical output from coil, only a mechanical/audio one?
So are you saying that a headphone driver coil is a perfect balanced load because it has the same impedance at both inputs? So does a nail.
Where does the CMR come from since there's no electrical output from coil, only a mechanical/audio one?
Last edited:
A current will only pass through a coil of wire if only there is a voltage difference at its terminals.
Eg I could put 1000000V on one terminal and 1000000V at the other terminal but no current will flow since there is no voltage difference.
I.e. perfect common mode rejection.
A proper balanced circuit is made of 2 identical halves, but each half doesn't need to be so complicated since it doesn't have to manage its dc offset, like an op amp does.
I actually think bridged amplifiers can be balanced amplifiers.
Sent from my iPhone using Tapatalk
Tim - so why does the pro audio field bother with balanced input receivers? Why not use a coil of wire or any other conductor such as a nail and be done with it?
Other Advantages of a balanced/bridged headphone amp:
- Single supply rails.
- Cancelation of second harmonic distortion.
- it is possible to build a very simple discrete circuit (ie 2 transistors) without output decoupling caps.
Sent from my iPhone using Tapatalk
- Single supply rails.
- Cancelation of second harmonic distortion.
- it is possible to build a very simple discrete circuit (ie 2 transistors) without output decoupling caps.
Sent from my iPhone using Tapatalk
Well they do use a coil of wire, it is called an input transformer.
Sent from my iPhone using Tapatalk
The advantage of separate grounds on left and right transducer is channel separation. The shared ground means the current from one driver will impose an error voltage on the other driver from the I*R drop across the ground impedance.
On my HP-1, I measure about 95 dB channel separation on the 1/4" phone connector (shared ground) and 115 dB on the 4-pin XLR (separate ground).
I don't see any advantage of driving the headphones differentially. All you get out of that, as far as I can figure, is higher THD. In theory, you should get cancellation of the even order harmonics, but in reality most cancellation schemes don't live up to the advertising.
Tom
On my HP-1, I measure about 95 dB channel separation on the 1/4" phone connector (shared ground) and 115 dB on the 4-pin XLR (separate ground).
I don't see any advantage of driving the headphones differentially. All you get out of that, as far as I can figure, is higher THD. In theory, you should get cancellation of the even order harmonics, but in reality most cancellation schemes don't live up to the advertising.
Tom
Thank you Tom, my thoughts exactly. I think Tim's claim about perfect CMRR from a coil of wire is spurious. Yes, there's a coil of wire as input to a transformer but there's also a core and an electrical output from the other coil. Where's the other coil in a headphone driver? Why not just use nails instead of coils? They have perfect CMRR too don't they?
My point is, to benefit from CMRR there has to be an electrical output into succeeding circuitry, not an electro-mechanical output.
My point is, to benefit from CMRR there has to be an electrical output into succeeding circuitry, not an electro-mechanical output.
Last edited:
The main reason for balanced plugs in pro audio is there is less danger of someone unplugging unintentionally. That's why many prefer SpeakON connectors.
A speaker is a transformer too, Electrical power goes in acoustic power comes out.
No difference in terminal voltage equals no current in the coil which mean no magnetic field which means no deflection of the diaphragm which means no sound.
I thought the concept is pretty basic really.
Though I think you are getting a little desperate with your nails.
Sent from my iPhone using Tapatalk
No difference in terminal voltage equals no current in the coil which mean no magnetic field which means no deflection of the diaphragm which means no sound.
I thought the concept is pretty basic really.
Though I think you are getting a little desperate with your nails.
Sent from my iPhone using Tapatalk
Say you were to use a single ended design and use a 4 wire setup. How would you isolate the two grounds from crosstalk?
Isn't that the point of the 4-wire setup? With 4 wires there is no common wire between the 2 channels. Do you consider crosstalk on the ground will be a problem in a well implemented amplifier circuit? It's more likely to come from power supply modulation which can be mitigated by having separate supply rails and/or improving PSRR.
We are conflating balanced impedance interfaces with differentially driven interfaces. They are not the same. Balanced impedance signal interfaces offer common-mode noise rejection, especially for power mains ground loop noise manifested between two wall powered active components. Headphones and loudspeakers don't manifest power mains ground loop noise since they are passive components floating with repsect to the power mains.
Differentially drive circuitry is not necessary for common-mode noise rejection. Single-ended drive can benefit equally from use of a balanced impedance interface for common-mode rejection. However,, differential drive can offer a number of other benefits. Such as:
1) Twice the slew-rate of an equivalent single-ended circuit.
2) Reduced signal induced power supply modulation, which includes injecting less signal related noise onto the circut ground.
3) Differential symmetry circuits tend to cancel even order distortion products the same as complementary symmetry circuits do.
4) A differential bridge could deliver twice the voltage in to a load from a given supply rail voltage, producing up to four times the power of a single-ended circuit.
Probably, the greatest knock on differential circuits is that they can require twice the number of parts.
Differentially drive circuitry is not necessary for common-mode noise rejection. Single-ended drive can benefit equally from use of a balanced impedance interface for common-mode rejection. However,, differential drive can offer a number of other benefits. Such as:
1) Twice the slew-rate of an equivalent single-ended circuit.
2) Reduced signal induced power supply modulation, which includes injecting less signal related noise onto the circut ground.
3) Differential symmetry circuits tend to cancel even order distortion products the same as complementary symmetry circuits do.
4) A differential bridge could deliver twice the voltage in to a load from a given supply rail voltage, producing up to four times the power of a single-ended circuit.
Probably, the greatest knock on differential circuits is that they can require twice the number of parts.
Last edited:
Hi Ken,
I agree with you although I wasn't aware of #1 about the slew rate. Is it also true for differential tube amps with output capacitors holding one capacitor per phase?
- Status
- Not open for further replies.