Hi, Steven,
I'm not as good as you in analyzing amplifier
. And worse, I cannot run simulation (I wanted to learn about it)
The CCT is from Hitachi amp, recently submitted by ASHOK.
The reason I'm very interested in this crossover distortion (that generates high order distortion) caused by a simple experiment.
I exclude the output stage from feedback loop (taking the FB point from VAS) and the sound differs quite much, the "Fatigue" in long time listening disappears. ClassA power amp (that do not have this crossover distortion), gives much less fatigue in listening.
I think this comes from crossover distortion, that is not included back to input stage. (I have another suspicion about LoudSpeaker back EMF like said by Graham Maynard, but I dont have enough equipment to investigate which one contributes to fatigue listening/high order distortion generator)
You seems have known about this from 20years ago, since your design is non-feedback type. Plus you put EC to it, gives good damping factor and wide frequency response while you dont use non-feedback design (problem with ordinary non-feedback design). Smart and everything right to target.
I'm not as good as you in analyzing amplifier
. And worse, I cannot run simulation (I wanted to learn about it)The CCT is from Hitachi amp, recently submitted by ASHOK.
The reason I'm very interested in this crossover distortion (that generates high order distortion) caused by a simple experiment.
I exclude the output stage from feedback loop (taking the FB point from VAS) and the sound differs quite much, the "Fatigue" in long time listening disappears. ClassA power amp (that do not have this crossover distortion), gives much less fatigue in listening.
I think this comes from crossover distortion, that is not included back to input stage. (I have another suspicion about LoudSpeaker back EMF like said by Graham Maynard, but I dont have enough equipment to investigate which one contributes to fatigue listening/high order distortion generator)
You seems have known about this from 20years ago, since your design is non-feedback type. Plus you put EC to it, gives good damping factor and wide frequency response while you dont use non-feedback design (problem with ordinary non-feedback design). Smart and everything right to target.