Sioux - 26 Watt OPA552 with Diamond Buffer output

The lateral Mosfets have a negative temperature coefficient once they are conducting sufficient bias current. Thus no source resistors are needed.

The early Hafler DH-200 and DH-220 amps used two sets of these in parallel to achieve ca. 200 Watts of power in class A/B operation. Bias current was 275mA with +/- 65V power rails in the DH-220.
 
This was fun stuff to do when testing
output stages. Just a opamp driving it.

Trying to remember when playing with higher
voltage opamps if I got better results with 552
or just used LM1875 as high voltage opamp.

Unfortunately it would add more transistors
on the keeping it on the simple side.
But if R13/14 are important for bias
and the rails will swing few volts or two under load.
Maybe should be constant current sources for
R15/16 instead set to 10ma. Likely high impedance
could improve THD as well.
 
Not to be negative, but not really.
rather frustrating device to model with.
Unless somebody has something that works
with texas spice / Tina.
The ones that " worked" were difficult and
unstable.

Keep in mind Elliott's opinion highly subjective.
Since Cordells opinion much the opposite.
His Error Correction amplifier with Vertical Mosfets.
Far far exceed Rods simple circuits.
And does so with more favorable behavior of
widely available vertical

Vertical mosfet library one of the best in Tina.
So got over non working unstable lateral models.
Being market availability and price is not great real world.
Unfortunately
 
While I don't doubt this is a simple and very good little amp, seems kinda like cheating putting an op-amp in front of a basic output stage?
Although I have been working on an output stage that would be perfect for this application, but probably overkill.
Have you got any THD and FFT readings?
Using an opamp front end is a fantastic way to incorporate a state of the art voltage amplifier stage with heaps of gain into an amplifier with a minimal parts count. I'm surprised it's not done more often.