what's a good average bias setting

[DF96]

Yes, the voltage between the bases is not the thing to measure even if that is what you have been told to do. Voltage across emitter resistor is the right place.

I want to know more.

You might find it helpful to look at pt. 1 of the article too (Sept 2000), as this gives the basic analysis of the circuit. I was quite chuffed when I found my simple theory explained some measured/simulated(?) results reported by Doug Self in his book. I drafted a follow-up article with more on output stages but I was never quite satisfied enough with it to get it published. I found that, when correctly biassed, a complementary pair can give much lower crossover distortion than a Darlington but the downside is a rather non-linear input resistance so the driver needs to be low impedance (e.g. a follower).

 

[AndrewT]

.....I drafted a follow-up article with more on output stages but I was never quite satisfied enough with it to get it published. I found that, when correctly biassed, a complementary pair can give much lower crossover distortion than a Darlington but the downside is a rather non-linear input resistance so the driver needs to be low impedance (e.g. a follower).

Hi,
I have copies of the three parts.
Thanks for the link.
What have you uncovered to expand on that 10year old paper?
Is it worth us seeing how your ideas have developed in the intervening period?

 

[DF96]

No, not really! Since then I closed down my IT business, edited a book, then went back to university to get a PhD. During this time I got bored with SS and started using valves.

I hope you enjoy the articles. Wireless World was still a useful magazine back then, but not long after that it seemed to lose its way and ended up as a trade rag.