Let me ask a question about imput circuit. When the signal produce an unbalance in the J1-J2 current, this, in turn, produce an unbalance in the Q1-Q2 current, right ?
The question is: what kind of circuit form Q1-Q2 ?(some cascode conection?).
Other point: why you take the reference of the current mirror Q6-Q7 from the output? what´s the idea ?
As always I ask any solid reason not going for a completely complementary circuit?
The input FET differential pair is cascoded with the bipolar transistor pair Q17 and Q18. The latter is <B><I>folded cascode</B></I> connected to Q1 and Q2.
These are loaded by a simple current mirror Q7 and Q8
I must refrain myself not giving a <B><i>Gromanesque</B></I> answer to your two footballplayers metaphor. What about two footballplayers kicking the ball back and forward between each other?!
But as I see it you get twice the drive to the outputstage with a complementary inputstage. Also a full complementary outputstage gives me a better gutfeeling.
Don't get the meaning of this word. Pls explain it in a simple words
>>What about two footballplayers kicking the ball back and forward between each other?!
This is a right metaphor for class AB otput stage, not A.
>>But as I see it you get twice the drive to the outputstage with a complementary inputstage
It's a total different stories about input and output complimetary stages. For input, you do not have to achive perfect symmetry for AC collector currents of Q1 and Q8. This currents just get summed at out1 point. As the first level of approximation, Q1 and Q8 are independent amplifiers. You can even substitute either Q1 or Q8 with constant current source and just loose 6db gain. Imbalance between Q1 and Q8 AC currents does not affect on linearity of the amp.
For output stage, you HAVE to get perfectly the same devices.
But they doesn't exist. So you get some errors at output stage.
Another well-known fact is compensation of even harmonics at push-pull stage. This leads to "comb filtering" effect that human hearing do not like.