Hi
Another question I calculated the final stage output using your equation and using my new circuit values, R12=1k RL=200 Re5=6.2k and beta=110. I got a Av3 = .748 which is close to the pspice value of .8, do these values seem appropriate?

 

I'm can't really make an intelligent comment about the "appropriateness" of the values.

The results seem consistent with your new design. The value of 'AV3' is highly dependent on the value of β you assume. If you assumed a β value of 250 then Av3=0.87, which is more optimistic than your simulation result.

Presumably even though you found the Darlington was the likely source of the distortion you have decided to stick with that part of the design. I can't think of any obvious reason for the distortion - other than (say) the early sign of positive peak clipping. The latter 'explanation' could be verified by reducing the input source differential voltage magnitude to test whether the distortion reduces with decreasing output swing. Darlingtons have been used extensively in audio applications - though more in the role of final output power stages. They seem to do quite well in that application, otherwise 'audiophiles' would be 'up in arms' over the lack of fidelity .

 

I did back off on the input and it did seem to improve the distortion. I'm around with the gains of the diff amp and darlington to see how that affects things. any other ideas would be helpful.