Is all negative feedback created equal?

[jupiterjune]

I have been wondering for some time, and I don't have the time to set up two circuits and do extended listening tests, so I was wondering.....

In Nelson's Balanced Zen Linestage article, there are two sets of performance curves, one where r15 is 430 ohms and the gain is 10dB. The distortion is much better than that measured when r15 is 124 ohms.

I have read here and there that this source degeneration resistor introduces negative feedback, and that is perhaps the reason why the performance is soooo much better with r15 at 430 ohms.

this leads me to wonder...

Is this negative feedback the same as, or basically equivalent to using resistors from output to input in what we often refer to as the X topology?

I think the X feedback reduces 2nd order harmonics -- would the frequency spectrum of harmonic distortion look the same for both cases if the feedback (via X resistors) were adjusted to give the same THD for the two different circuits.

Has anyone done listening tests on these two variations?


Beyond just mental meandering, the source degeneration resistor seems to be the fundamental difference between the original BZLS and the circuit used as a really good unbalanced to balanced converter...

hope this question makes sense.
signing out for now, jenny lewis and the watson twins await

JJ

[Babowana]

Originally posted by jupiterjune


In Nelson's Balanced Zen Linestage article, there are two sets of performance curves, one where r15 is 430 ohms and the gain is 10dB. The distortion is much better than that measured when r15 is 124 ohms.

I have read here and there that this source degeneration resistor introduces negative feedback, and that is perhaps the reason why the performance is soooo much better with r15 at 430 ohms.

this leads me to wonder...

Is this negative feedback the same as, or basically equivalent to using resistors from output to input in what we often refer to as the X topology?

R15 is also doing negative feedback role, sensing output current. So, I'd like to say that it is a kind of current feedback. The bigger R gives more feedback, lowering down both the gain and the distortion.

In contrast, the feedback from the output to the input is voltage feedback, introducing part of the inversed distortion voltage to the input and making the amplified distortion lower. The smaller R gives more feedback, lowering down both the gain and the distortion.

I have never compared these two with respect to the sonic effect. Nevertheless, If I look into them using microscope, I would believe that the current feedback would wok faster than the output voltage feedback. So, I simply prefer the current feedback in this case.

I am writing down a non-fiction . . . ?

[jupiterjune]

Babo-
My assumption was pretty much the same (but with less understanding of how adding R15 reduces distortion). I just assumed that using a large value of r15 would not induce any of the sonic nastiness attributed to "negative feedback" resulting in obedient mosfets, but not maliciously obedient. But our Serbian friend suggested it might be otherwise.......

Hoping to get some sort of feedback on this.

JJ

[Babowana]

Originally posted by jupiterjune


(but with less understanding of how adding R15 reduces distortion).

I'm a simple man, trying to easily give up unnecessary things . . .
I like light weight than heavy burden . . .


The transconductance curve of the gain MOSFET is non-linear shape. So, the wider swing of the Vgs creates higher distortion, riding up and down the non-linear transconductance curve. If I increase the value of R15, I could narrower the Vgs swing, reducing the distortion.

If I increase the volume (i.e. increase the swinging amplitude of Vgs), it would be the same as I would be increasing the distortion.

[jupiterjune]

If I increase the volume (i.e. increase the swinging amplitude of Vgs), it would be the same as I would be increasing the distortion

Well, that makes sense. I would think this would be a much better way to make a circuit more linear than NFB from output to input.

I also wonder about the fact that no-one else has replied.
My first thought is, everyone is thinking:
"Yes, what Babowana said, that sums it up"
or perhaps:
"I have no idea what the ____ he's talking about"
or
"JJ has no idea what the ____ he's talking about"
or
"NFB makes no difference whatsoever, as long as you have 2.5mm wire in your power supply chokes"

come on--anyone? anyone? Bueller? Bueller?

OK, I do have several BOSOZ boards laying around. If I get hold of balanced source, I may have to listen to both circuits myself and see if there is a difference. Oh bother--I wanted to figure this out BEFORE I acquire a CD player with balanced outs--because maybe the balanced outputs aren't needed.

JJ