It appears driving headphones in balanced or bridged mode is very popular these days in the high-end community.
In most cases there appear to be few benefits. Each amplifier will effectively drive half the load impedance, ideally requiring a doubling of output current and increasing distortion. Only even-order harmonics will cancel at the load as well.
The benefits would seem to be limited to crosstalk (connectors/cable) and avoiding any ground-related issues caused by the headphone return currents, which could probably be mitigated by proper layout. Of course, the potential 4x power increase and doubled slew rate are there.
Did I miss anything there?
I have a differential source (PCM1794) component but debating whether or not I should convert to single-ended first and use a single ended headphone amplifier stage or keep the differential outputs and drive the headphone in a bridged mode.
Any thoughts?
In most cases there appear to be few benefits. Each amplifier will effectively drive half the load impedance, ideally requiring a doubling of output current and increasing distortion. Only even-order harmonics will cancel at the load as well.
The benefits would seem to be limited to crosstalk (connectors/cable) and avoiding any ground-related issues caused by the headphone return currents, which could probably be mitigated by proper layout. Of course, the potential 4x power increase and doubled slew rate are there.
Did I miss anything there?
I have a differential source (PCM1794) component but debating whether or not I should convert to single-ended first and use a single ended headphone amplifier stage or keep the differential outputs and drive the headphone in a bridged mode.
Any thoughts?
if you have a balanced source, why wouldnt you use a balanced amp? proper amp design will not see an increase in distortion, any decent amp will be able to deal with the lower effective impedance, nor will it see an increase in noise. a balanced amp when viewed in isolation will have slightly higher noise, but if you have to have a balanced to SE conversion take place beforehand, how is it you will see an increased noise level of the system with less components in the signal path of each phase?
bridged on the other hand I wouldnt bother, all cost with bugger all payoff. to do SE properly you have to take most of the same measures to ensure the headphone has a clean return path, the power supply still needs to be bipolar if you are doing bal-se with anything but a transformer, so no real savings there either.
bridged on the other hand I wouldnt bother, all cost with bugger all payoff. to do SE properly you have to take most of the same measures to ensure the headphone has a clean return path, the power supply still needs to be bipolar if you are doing bal-se with anything but a transformer, so no real savings there either.
I do notice a difference. Balanced sounds better when it comes to using the same amp modules (I've tried things from EL84 tubes to Beta22 to a assortment of other designs). Not sure why that is. My guess is the increased slew rate. We might want more slew rate than the simple math indicates.
Last edited:
Yeah who really knows these things. It's hard to be objective in the audio. I once added 10% harmonic distortion (!) to music files and my friends and family couldn't tell the difference in a blind test. Yet when I changed a coupling capacitor in the amp which measures exactly the same as the capacitor it replaced, the difference was obvious. It's a crazy world.
Or not.
Forget about the math. Show me any single ended headphone amp that's slew limiting in the audio band before it starts clipping. These days, in order to pull that off you'd pretty much have to do it intentionally.
The whole "double the slew rate" thing comes from the fact that if you bridge two amplifier channels together, for a given output level, each amplifier channel only swings half the voltage it would if it were driving the load by itself.
So you could just as well say that you "double the slew rate" if you turn your volume control down 6dB. Or simply use a headphone that's 3dB more sensitive.
se
Yeah I've heard these arguments before, and admit to not understanding why bridged and balanced amps sound different (ie. more dynamic and transparent). If you know why, that'll be great.
If you turn down your volume, if hypothetically there is an increase in sound quality, you wouldn't be able to compare it anyway.
Tube amps have lower slew rates and certain ones do indeed sound like they slew. However, for simple sine waves mathematically the slew rate should be enough. I'm not entirely convinced, however.
If you turn down your volume, if hypothetically there is an increase in sound quality, you wouldn't be able to compare it anyway.
Tube amps have lower slew rates and certain ones do indeed sound like they slew. However, for simple sine waves mathematically the slew rate should be enough. I'm not entirely convinced, however.
Thing is, "sounds different" is really a rather meaningless term outside a particular individual. I mean, some peoples' systems "sound better" to them after they've put photographs of themselves in their freezers. So until things can move from "sounds different" to "audibly different," we're not going to get any conclusive answers.
se
se
So if you've not heard any difference yourself, and you don't think any such difference exists, which is a valid opinion, why not just say so? There is no point in questioning the ears of others and telling them that it's just placebo.
I design and build amps, sources, speakers, etc., for my own enjoyment and I know the basic electronics and measurements. And I hear a difference. And I could not understand why, though I have some very rough theories. If others hear what I hear, I'm curious as to why a difference exists.
Please don't try and put words in my mouth.
All I said was that "sounds different" doesn't establish that there's any actual audible difference. That doesn't mean that there is no or can be no actual audible difference. Only that "sounds different" doesn't answer that question.
But until you can establish that there actually is a difference other than simply "sounds different," then you're never going to be able to get at the "why" of it. I mean, if there is no actual audible difference and the "sounds different" is simply the result of humans being human, but you operate from the assumption that there is actually something going on, then you're just going to end up chasing your own tail and not get anywhere.
se
Well, I'm actually designing the DAC as well, so there won't actually be any extra components if I convert to single-ended. I need (want) an active low pass filter anyway, whether it's a diff-to-SE configuration or inverting handling a single phase doesn't really matter.
When I say bridged I mean that the amplifier design would be the equivalent of 2 single ended amps driving the load, even though they are fed from a true differential source. I'm not talking a true fully differential amplifier although I guess I could use OPA1632 or some SuSy configuration to get that with op-amps.
Yeah I didn't want to get into the can of worms that is the subjective sound evaluation
. I'd have to agree with Steve that it's probably not the slew rate. There are headphone amps using current feedback devices like TPA6120 that have ridiculous slew rate by themselves, higher than almost all popular discrete designs.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.