OpAmp and Crossover Distortion

[Havenwood]

Hi, a simple question, what is the order of Iq in OpAmp?

Why they have very low THD with a "relative" low Iq in the oupput stage?
Global NFB fix these error?

Can we apply this and obtain in a Power Amplifiers low THD and use a low Iq?

[wg_ski]

What is critical about output stage bias is not strictly the number of milliamps, but rather the current per unit area. Those output trannies are a lot smaller than the ones in a power amp so a lot less current is required.

And many op amps do have unacceptably high levels of crossover distortion.

[Eva]

In bipolar class AB the transfer characteristics of each half of the output stage are being added near zero current crossing. There is an optimum bias current that provides the best compensation of the lower "gain" of one half at very low currents with the lower "gain" of the other half.

Lower bias results in a dead band and higher distortion. Higher bias results in the so called "gm doubling" effect and higher distortion.

MOSFET class AB is a different story because the low current region where "gain" is lower is much wider, so more bias is required to get some kind of compensation.

[satoru]

OPAMPs like NE5532 has a local feedback loop from the output stage to the second stage to cancel out the crossover distortion. NE5532 has unusual output stage, so this is very important to reduce the overall distortion.

AD797 also has a feedback from the output stage but in this OPAMP, it is something user configurable. A capacitor connected between Pin 8 and Pin 6 (output) will serve the purpose. Without a cap there, AD797's distortion is not that impressive.

A couple (at least I know) of commercial power amplifiers used similar idea to reduce the distortion generated at the output stage.

[nigelwright7557]

Originally Posted by wg_ski

What is critical about output stage bias is not strictly the number of milliamps, but rather the current per unit area. Those output trannies are a lot smaller than the ones in a power amp so a lot less current is required.

And many op amps do have unacceptably high levels of crossover distortion.

The load on an op-amp is not usually inductive like a power amp so it needs much less Iq.

A power amp has a dirty big inductor (speaker) across its output and so has to soak up the back emf from the speaker without distorting.

[Havenwood]

Here, the equivalent schematics
NE5532 and AD797

So, feedback from the output stage fixed (corrige) crossover distorsion?
How set Iq in AD797? by feedback?

[Havenwood]

I have this simulation, where Iq = 26pA, THD@1kHz 0.3%
No X-Over Distorsion (i can see it)

[Mooly]

Originally Posted by Havenwood

I have this simulation, where Iq = 26pA, THD@1kHz 0.3%
No X-Over Distorsion (i can see it)

There's plenty of distortion Build it for real and see.

All your circuit does is try and include a pair of fast high gain transistors as a class b output stage within the feedback loop of an opamp... and you have it feeding a high impedance load.
If you build it and look on a scope the distortion will be clearly visible at the zero crossings.

[janneman]

Originally Posted by Havenwood

I have this simulation, where Iq = 26pA, THD@1kHz 0.3%
No X-Over Distorsion (i can see it)

Of course there is a lot of distortion. The opamp output goes through a dead-band where both output transistors are off. This is the classic class-B xover.
What do you mean by 'I can see it'?

jd

[Havenwood]

Originally Posted by janneman

What do you mean by 'I can see it'?

I can´t see in simulation: