OPA454 shows clipping and odd distortion

[Audioguru]

I think a darlington pair (TIP120) is slower (its delay causes phase shift above 1MHz) than a Sziklai pair. The phase shift causes the opamp to oscillate when negative feedback is added without frequency compensation.
Hee, hee. I typed "phase shift" but the "f' wasn't printed. Does it smell like roses?

 

[D.A.(Tony)Stewart]

I also used the CFP in 1975 but never knew it had a name or a patent.

Best bet is to use 3 NPN and 1 PNP with thermal and ESR matching diodes to Re * hFE values

This then becomes a Darlington-Sziklai quad or simply I call it Quasimoto power

.. Quasi mo power

500 Watt peak. excellent linearity, bandwidth and stability,, crossover distortion starts > 500Wpk in 1R load.

The benefit is lower cost, high linearity and matched NPN driver pair.

The down side is heat loss is high and efficiency is low.

 

[E-design]

Why will using a CFP make it more prone to oscillation by the way? What is the op-amp now 'seeing' that makes it want to oscillate?

CFP output stages suffer from high-frequency instability due to more local feedback in the CFP loop. Here the loop gain will be a function of beta, ft( of output transistors ) and Miller capacitance of the output transistors.

Due to the Miller capacitance, the CFP configuration has inherent turn-off problems resulting in high-frequency distortion and possible shoot-through problems. This will be more of a problem in high-power applications.