On 10/28/2013 12:50 PM, Ellis Burman wrote:
> Yes, I've seen this in some of Ruper Neve's designs.  I guess it implies
> that the crossover distortion, even though it is almost unmeasurable low,
> is audible.

I think what's going on, based on some tests I ran a few years ago, is 
that the in/out transfer curve is simpler, with fewer kinks, in the 
single-ended Class-A version than in the push-pull Class-AB version. 
There were fewer high-order distortion products on a mid-freeuency sine 
wave test, and fewer (and lower-order) IM products with 19kHz and 
19.5kHz signals mixed 1:1. Result in practice: a more transparent stage.

Peace,
Paul