Propeller Head Plaza Technical and scientific discussion of amps, cables and other topics.

---

[ Follow Ups ] [ Post Followup ] Thread: [ Display  All  Email ] [ Propeller Head Plaza ]

---

[ Alert Moderator ]

---

RE: This is interesting

Posted by andy_c (A) on August 24, 2007 at 09:19:21

In Reply to: RE: This is interesting posted by morricab on August 24, 2007 at 02:02:05:

It was my understanding that negative feedback was found to not successfully address crossover distortion and that other distortion reducing techniques (like High idle bias) were required. Is this right or not?

The discussion I linked to has a graph that came from some simulations of distortion vs. feedback for a solid-state class AB output stage. See this post and click on its accompanying graph. The description of the simulation circuit is in the post above it. That simulation includes the effects of frequency compensation, which reduces the feedback as frequency increases, in order to keep the system stable. The signal level chosen is such that the crossover distortion without feedback is maximized. This was done because the idea was to show the effect of feedback on crossover distortion only, not other effects like, say, the approach of clipping.

The zero-feedback case is all the way at the left of the graph. You can see that there is monotonically decreasing distortion with increasing feedback for every harmonic - even the higher-order harmonics. If you read somewhere that feedback does not reduce crossover distortion, then I'd say the person who wrote that doesn't know what they are talking about. Pavel Macura shows an example of the extreme case of an output stage with zero bias and feedback applied here , and in the post that follows. Surprisingly, the feedback still provides some improvement, though nobody in their right mind would actually design a circuit this way. Still, I thought Pavel's illustration was interesting.

You mention in your other posts that amps with gross non-linearity open loop did exhibit distortion behavior like Baxandall's examples. Why do you think that is and what percentage of commercial designs do you think would fall into this category?

For the case where the open-loop distortion is only second-order (the FET case), the "why" is given by Baxandall's mathematics, where he predicts the values of each harmonic as feedback is varied. Unfortunately, the mathematics become intractable with more complex examples. When Baxandall analyzed the BJT case, he had to give up after the third harmonic. To compound the conceptual problem, the effect of feedback on distortion does not lend itself well to an intuitive approach either. When such a situation occurs, engineers must resort to measurement, simulation or both. Neither of these approaches give results that are applicable in the general case, only to the specific circuit examined.

One might be tempted to take the mathematical results for the case we do know, and claim they generally apply. That's one of the traps that Cheever fell into. The results of the linked discussion from diyaudio show that the statement "Feedback always increases high-order harmonics" is just flat wrong. Sometimes it does, sometimes it does not. The results depend on the specific circuit one is considering.

Regarding commercial designs, I suspect that most SS amps for the audiophile market that use feedback are quite competently designed with regard to open-loop distortion. But I'm sure there are many products of the "consumer" variety that are not. But this is just speculation and probably not worth much.

Finally, at the amount of average power needed to listen to music the distortion levels of well designed tube amps are quite low (well under 1%) and of generally low order harmonics.

If you check my original statement, you'll see I was referring to SET amps, not tube amps in general. I don't have anything against tubes at all, but I think the current obsession with SETs in some quarters is just silly.

This post is made possible by the generous support of people like you and our sponsors:

Sonic Craft

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---