howzabout this for a poweramp design?

[mlloyd1]

OK, so I stretched things a bit. This is an internal schematic (simplified of course) from the datasheet for the AD8022 op amp. Add some emitter degeneration to the input devices (maybe even make them JFETs with the appropriate cascoding, of course ), beef up the output stage (those 15ohmies go down a lot; maybe even redesign the topology to make this stage a bit more conventional - I THINK I remember seeing this particular output stage design in Otala's low TIM power amp from way back in the 70's - not that it's bad, mind you. I'd go with a emitter follower triple like Leach's low TIM project) and add a bias "regulator" and you could have a decent power amp. Current sink for the input diff pair tail current, but resistors for the diff amp loads (Oh yeah, definately more than 600uA for the diff sink!).

Just to yank some chains, I'd try these devices:
Input stage: 2SC3381 (or 2SK389 diff+ 2SC3381 cascode)
Folded Cascode + mirror: 2SA1210, 2SC2912
Output triple: 1st - 2SA1210/2SC2912,
2nd - 2SA1112/2SC2592,
Output - MJL1302A/MJL3281A
Why these devices? You'll have to guess!

Of course, you could always try a preamp first to make sure you can get it to work and be stable.

So, who's game? Simulators, start your engines....

mlloyd1

[Jocko Homo]

You could just build it and see. I already know how it works without having to simulate, although I used different devices. And no global feedback.

Jocko

[jam]

Looks like the front end of a Jeff Rowland amplifier.

Jam

[sonnya]

Looks a bit like the one a build some time ago. Maybe you can use some of it or get some idears?

Jfet/bipolar PCB and Schematic (finaly)

But be aware that some of my calculations is way of.

Sonny

[Ren Hoek]

Hey, jocko, man, they know your secret.

Ren