[Ancipital]

So, how badly should people be squealing over a budget balanced amp/DAC combo? How much does balanced drive and indeed fully balanced chain actually add? How and why?
 
I'm familiar with balanced audio for pro/broadcast application, but not completely clear What is going on with it in this context. There's a bunch of stuff online where people insist po-faced that it doubles the slew rate and "increases control of the diaphragm" giving that now-familiar "night and day" difference.
 
Given that audiophiles are excitable people and will make similar claims about magic file copying programs, little hurdles to keep speaker wires off the floor and $5000 cables, it's always hard to separate things with a science and engineering basis from the "lies to children" hooey just there to jack up prices.
 
So, at a guess, there is some improvement (noise, especially, maybe a bit of THD)? What else? Some adult answers from people who don't believe in fairies would be appreciated 

 

 

[PleasantSounds]

Faster slew rate and extra power are the only tangible benefits of balanced amps that I can think of. Maybe reduced crosstalk due to avoided common ground.
Noise, THD, output impedance are actually worse than SE, as you have two amps per channel and these things add up.
Plus you get some non-linearity between the positive and negative halves (simply due to parts sample variation) and switching noise at the 0 point. 
 
So in theory a very fast and powerful SE amp should be better than balanced. For the time being the search goes on...

 

[kevin gilmore]

For most solid state balanced amplifiers I can think of, there is actually a reduction of almost 6db for 2nd harmonic distortion.
More if the parts are matched. 3rd harmonic typically goes up slightly.
 
Double the slew rate, true
 
Double the final output impedance, also true, but still typically under 1 ohm.
 
For amplifiers like this, you need 4 distortion graphs to handle all the conditions.
SE-SE, SE-BAL, BAL-SE, BAL-BAL
they are definitely all different.
 
noise is tougher because the different connections result in different gain.

 

[upstateguy]

  For most solid state balanced amplifiers I can think of, there is actually a reduction of almost 6db for 2nd harmonic distortion.
More if the parts are matched. 3rd harmonic typically goes up slightly.
 
Double the slew rate, true
 
Double the final output impedance, also true, but still typically under 1 ohm.
 
For amplifiers like this, you need 4 distortion graphs to handle all the conditions.
SE-SE, SE-BAL, BAL-SE, BAL-BAL
they are definitely all different.
 
noise is tougher because the different connections result in different gain.

 
Hi
 
Maybe you can help me.  I don't really understand what you mean by 2nd harmonic distortion in SS amps.  I know what it is in tube amps but that's as far as it goes.
 
You designed the GS-1 amp I have.  As you know, the THD is 0.006.  How much 2nd harmonic distortion do I have and is it in the audible range?
 
Thanks

 

[kevin gilmore]

total harmonic distortion is the sum of all distortion components, mainly 2nd and 3rd harmonic, but typically all the way to the 5th harmonic
 
a lot of distortion graphs you will see show 1khz at roughly 0db (and this is where a lot of sound cards have issues, with calibration)
and then you can clearly see 2khz and 3khz and the rest of the components including noise (where the thd+n comes in)
 
if you use a 10khz test tone, even the 2nd harmonic is probably outside your hearing range
 
I do things a bit differently for testing where I do sweeps looking just at 2nd harmonic and the rate at which it changes with frequency
and the same thing with 3rd harmonic. bipolar and mosfet output amps perform very differently.
 
2nd harmonic in a tube amp is the same thing as a solid state amp.

 

[icebear]

Have you ever connected a ground cable to a speaker or just L+/L-  and R+/R- ?
A headphone is just a small speaker, same thing. There is no ground connection on the transducer.
Balanced in regards to headphones just decribes the separate L- and R- returns opposed to a "common ground" i.e. L-& R- on the same wire or into the same connector plug.
No one would connect speakers with a common return wire...

 

[kevin gilmore]

Such a small part of a much bigger issue
 
A balanced amplifier can run on much lower power supply voltages for the same output power.
This allows for lower voltage output transistors which are typically a lot less capacitance, have a lot more current output
and have higher current gain and more linear operation. Balanced amplifiers also have matched rise and fall slew rates, something
SE amplifiers cannot do. (other than bryston which has a unique output stage)
 
Although it has nothing to do with headphones, for speakers there is pretty much a hard limit of about 250 watts into 8 ohms for SE amplifiers.
Going past those numbers absolutely requires a balanced/bridged output amplifier.

 

[Ancipital]

  For most solid state balanced amplifiers I can think of, there is actually a reduction of almost 6db for 2nd harmonic distortion.
More if the parts are matched. 3rd harmonic typically goes up slightly.
 
Double the slew rate, true
 
Double the final output impedance, also true, but still typically under 1 ohm.
 
For amplifiers like this, you need 4 distortion graphs to handle all the conditions.
SE-SE, SE-BAL, BAL-SE, BAL-BAL
they are definitely all different.
 
noise is tougher because the different connections result in different gain.

 
Right, so would useful benefits accrue from plugging an SE source into a suitable and capable balanced amp, that was driving headphones balanced? It sounds like it wouldn't be completely futile.

 

[upstateguy]

  total harmonic distortion is the sum of all distortion components, mainly 2nd and 3rd harmonic, but typically all the way to the 5th harmonic
 
a lot of distortion graphs you will see show 1khz at roughly 0db (and this is where a lot of sound cards have issues, with calibration)
and then you can clearly see 2khz and 3khz and the rest of the components including noise (where the thd+n comes in)
 
if you use a 10khz test tone, even the 2nd harmonic is probably outside your hearing range
 
I do things a bit differently for testing where I do sweeps looking just at 2nd harmonic and the rate at which it changes with frequency
and the same thing with 3rd harmonic. bipolar and mosfet output amps perform very differently.
 
2nd harmonic in a tube amp is the same thing as a solid state amp.

 
Thank you for answering my layman's question but I'm not completely understanding the audible advantages of a balanced headphone amp other than a volume increase. 
 
With your GS-1's single ended THD of 0.006 is there be anything that would be audible in a balanced version of the same amp that has double the slew rate, double the output impedance and because there are two amps, double the noise and distortion that would be present from a single amp?
 
Edit to ask if the GS-1 was balanced would the THD be doubled over the single ended version?

 

[kevin gilmore]

it is absolutely audible with matched output levels, has nothing to do with volume increase. And i'm not the only one that thinks so.
The 2nd harmonic vs 3rd harmonic is definitely different, so you can attribute your favorite adjectives to this.
 
the 2nd harmonic thd component will be at least 6db LESS when you supply a balanced input signal and output a balanced signal.
in this case the total harmonic distortion is LOWER balanced in/balanced out. The reason is that errors in the amplifiers cancel.
don't remember thd+n
 
The gs1 is unbalanced in-->unbalanced out   or balanced in-->balanced out.   in fact it will do balanced in-->unbalanced out, but that is no different than the unbalanced in-->unbalanced out.
 
my current ssdynalobal and ssdynahibal are fully balanced differential amplifiers. unbalanced or balanced in-->unbalanced or balanced output and the lowest thd is when you
supply a balanced input and use the balanced output. followed by unbalanced input with balanced output.
 
none of this is any different from my stax amplifiers which are all unbalanced/balanced in-->balanced out

 

[RRod]

The THD given for the original product in question is 0.001% 20-20k Hz @ 1Vrms. So are you saying that with this -100dB THD, you are pretty sure you could detect the difference made in the relative levels of the 2nd and 3rd harmonic balanced vs non- when listening at music-listening levels?

 

[kevin gilmore]

I never said that it was that low in distortion, and yes I can both hear and measure the differences.
 
its easy these days for anyone with $500 to buy a quantasylum qa401 and some load resistors and do the same thing.
although a SRS or a AP are better.

 

[RRod]

  I never said that it was that low in distortion, and yes I can both hear and measure the differences.
 
its easy these days for anyone with $500 to buy a quantasylum qa401 and some load resistors and do the same thing.
although a SRS or a AP are better.

 
-100dB is for the product given in the OP, and the interest here is why one would expect the THD benefits from the balanced output on such a product to be audible. At such levels, claims of audible improvement from -6dB 2nd order distortion seem incredible indeed.