[Steve Eddy]

Quote:

Originally Posted by jcx /img/forum/go_quote.gif the TRS contact R on the common "gnd" can exceed 10 mOhms - a much bigger source of "gnd contamination" crosstalk - not cured by any active gnd scheme discussed here

Quite so.

I did some experimenting with this a while back.

Took some 18 gauge zip cord and in one leg, I patched in a 1/8" TRS plug and jack. Across this I wired in a 1/8" TRS jack, combining the tip and ring on the one side, and the sleeve on the other.

I terminated one end of the zip cord with an 8 ohm resistive load and plugged the other end into the speaker outputs of a little JVC mini system I have.

Then I plugged a pair of cheapie headphones into the jack across the plug/jack combo.

Turned up the volume a bit and sure 'nuff, there was Wynona's Big Brown Beaver just as pretty as you please. Turned the volume up all the way and was able to get some pretty moderate output from the headphones.

Though to be fair I was running about 5-6 watts into the load.

k

 

[Steve Eddy]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif So what is really going on that the ground channel is potentially improving so much?

Perhaps nothing?

The only other issue here beside grounding is crosstalk through the rails/power supply.

With a constant current draw from the supply, as opposed to a variable current draw that's directly related to the signal, there's no way for the signal to cross couple into the amplifiers.

However this issue is already rather well addressed in a B22 with an σ22 regulated power supply.

The σ22 provides a low source impedance and the B22 is armored to the gills with regard to power supply rejection with all its cascoding and cap multipliers.

k

 

[rds]

Long discussion on something that appears to be very simple. Maybe someone can distill this down to a few sentences.

It seems that 3 channel is pretty much a way of implementing a balanced circuit.
"A common and simple method of designing a balanced circuit is to drive both ends of the loop circuit with matched transformer output (and input) coils."
Balanced circuit - Wikipedia, the free encyclopedia

I realize there's no transformers, but I think the idea is to have symmetry between the two 'ends' of the loop.

 

[Steve Eddy]

Quote:

Originally Posted by rds /img/forum/go_quote.gif Long discussion on something that appears to be very simple. Maybe someone can distill this down to a few sentences. It seems that 3 channel is pretty much a way of implementing a balanced circuit.

I guess you could say that. Though given that the raison d'etre of balanced circuits is to reject any common mode noise induced on the line, and since common-mode noise induced into a headphone cable isn't really any problem to begin with, there's not much point in describing it in terms of a balanced circuit.

Since it doesn't do anything to isolate anything from ground, I think a more meaningful description simply be a bridged circuit, the main purpose of which is to maintain a constant supply current draw when used with complimentary push-pull amplifiers operating in class A in order to reduce interchannel crosstalk through the power supply.

k

 

[rds]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif I guess you could say that. Though given that the raison d'etre of balanced circuits is to reject any common mode noise induced on the line, and since common-mode noise induced into a headphone cable isn't really any problem to begin with, there's not much point in describing it in terms of a balanced circuit.

But isn't this the reason people spend large sums of money 'going balanced'?

 

[Steve Eddy]

Quote:

Originally Posted by rds /img/forum/go_quote.gif But isn't this the reason people spend large sums of money 'going balanced'?

I can't recall seeing any balanced headphone amps being sold by promoting any sort of common-mode rejection as far as the headphones go. It's not really considered a problem on the headphone side of the equation.

And as for balanced inputs, where there may be some benefit of common-mode rejection if you're using unusually long runs from source to headphone amp, pretty much every balanced headphone amp I've seen is made up using two bridged mono channels.

There's no differential input so even though it's technically "balanced," it can't provide any common-mode rejection. Any common-mode noise at the inputs would just be amplified and passed along to the output.

Go figure.

k

 

[Beefy]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif Though given that the raison d'etre of balanced circuits is to reject any common mode noise induced on the line

Surely in the case of headphones a key point is to make sure that the headphone drivers don't see each other across ground, reducing stereo crosstalk.

Not to mention doubled slew rate and p-p voltage swing.

 

[dsavitsk]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif pretty much every balanced headphone amp I've seen is made up using two bridged mono channels.

with one exception ...

 

[Steve Eddy]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif Surely in the case of headphones a key point is to make sure that the headphone drivers don't see each other across ground, reducing stereo crosstalk.

As jcx pointed out previously, this problem is due to the use of a three conductor plug. An unbalanced amp with a four conductor plug and the driver "grounds" properly terminated shouldn't have any problems with crosstalk.

Quote:

Not to mention doubled slew rate and p-p voltage swing.

These two claims are contradictory.

In order to double the voltage swing, each of the two amplifiers making up the bridged ("balanced") pair must be outputting the same voltage that they would have if they were just a single, unbridged channel in an unbalanced amp.

But now they have to deliver DOUBLE the current that they would have otherwise.

So tell me how this manages to double the slew rate of either of the amplifiers.

I'm certainly not seeing it.

I think this notion of doubling slew rate came about based on the assumption that output voltage remains the same as it would have been for unbalanced rather than going by the doubling of output swing.

In this scenario, then each amp would only have to output half the voltage it would otherwise. However they both have to deliver the same current as they would have otherwise.

And even this doesn't double the slew rate. The slew rate of each of the amplifiers is what it is. Bridging them doesn't change their inherent slew rate. So the best you could possibly say is that all else being equal (i.e. no doubling of output voltage), the amplifiers wouldn't come as close to slew rate limiting as they would if used in an unbalanced amp.

But since slew rate is related to current, I'm not sure even that's the case.

And in any case, if your headphone amp's coming anywhere near slew rate limiting, you need to take it back to Toys-R-Us and go get a real one.

k

 

[Steve Eddy]

Quote:

Originally Posted by dsavitsk /img/forum/go_quote.gif with one exception ...

For now.

k

 

[Beefy]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif These two claims are contradictory.

You are wrong, at least for the typical balanced amplifiers we get around here. Argue it with Kevin Gilmore if you think you are up to it.

HeadWize - Project: A Pure Class A Dynamic Headphone Amplifier by Kevin Gilmore

Quote:

The balanced bridge output version of the amplifier (figure 2) is for those headphones that can be wired as dual mono (see the addendum for instructions on converting a pair of standard Grado SR-80 headphones into dual mono headphones). It has twice the voltage swing, twice the slew rate and 4 times the output power

He makes the same claim in many different places.

 

[Steve Eddy]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif You are wrong, at least for the typical balanced amplifiers we get around here. Argue it with Kevin Gilmore if you think you are up to it. HeadWize - Project: A Pure Class A Dynamic Headphone Amplifier by Kevin Gilmore He makes the same claim in many different places.

But it's only just a claim. You can make claims all day long.

k

 

[rds]

The slew rate doubling seems straightforward. It's just dV/dt. So if the output voltage you want is a constant, the slew rates would be additive in balanced configuration.
Slew rate has a current dependence, but it also has a temperature, gain, step, etc dependence. So we can argue that more output current will decrease the slew, but equally we can argue that smaller steps (by half) for each op-amp, and presumably lower gain (by half) will increase the slew. That's all very qualitative though. You'd have to analyse a particular op-amp to see if the individual slew is better or worse in balanced configuration.
I think saying it approximately doubles is a reasonable statement.

EDIT I shouldn't say 'if the output voltage you want is constant'. Slew rate is just change in voltage with time and it is additive in balanced configuration.

 

[Beefy]

Quote:

Originally Posted by Koyaan I. Sqatsi /img/forum/go_quote.gif But it's only just a claim. You can make claims all day long.

So you think you are smarter than Kevin Gilmore now. Huh. Good luck with that.

Quote:

Originally Posted by rds /img/forum/go_quote.gif The slew rate doubling seems straightforward. It's just dV/dt. So if the output voltage you want is a constant, the slew rates would be additive in balanced configuration.

Same for voltage swing. If one amp on its own can swing x volts relative to ground, adding a second amp on the other side swinging -x relative to ground gives a total voltage swing across the driver of 2x.

 

[luvdunhill]

Any bridged amplifier can make those claims.

Any amplifier with double the gain of another amplifier can make some of those claims.