Why would anyone want class AB amp when class A always sounds better ?

As no one in their right mind would prefer the sound of class-B xover distortion no matter how little.

There aren't any class B amps available in the consumer market, so this is no worries.

The sound of ANY amplifier is better without ANY distortions whatsoever. (Sure some distortions are not AS bad as others.)
So long as very little distortion is NOT attained by using means that create other problems areas EG: like masses of feedback global or local, global especially!

Of course, this being planet Earth, there are no amplifiers with no distortion.


Feedback is problematic in insufficient amounts- it can cause brightness and harshness as it introduces distortions of its own (mostly higher ordered harmonics). But if you are able to apply enough feedback then this phenomena does not occur. That amount seems to be north of 35dB. This also has to be consistent at all frequencies. Most amps have sufficient feedback at 100Hz (which is why they are usually measured at only 100Hz and not higher), but at 1KHz and 10KHz is where amps get into trouble. As a result most of the amps using feedback made in the last 60 years have increasing distortion as frequency goes up, because the amount of feedback is falling off, due to insufficient Gain Bandwidth Product.


This is why solid state in particular has gotten a reputation for being bright and harsh (and it does not matter what class of operation is being used) because the ear is very sensitive to those harmonics! This is literally why tube amps are still around since they can mask the higher ordered harmonics they make.

So, with less than capable speakers and headphones class AB can sound better ?

It can sound better even with great speakers! The sound of any amplifier is always about what distortion it makes, and what distortion it doesn't. You can prevent crossover distortion in any AB circuit, since its a function of push-pull rather than whether its class A or not. The only exception is if the amp is operating class B but I don't know of a single example of a class B amp in high end audio.

I might be off here, but class A amps don’t drive lower impedance speakers very well either. Most class A amps aren’t stable into 2 ohm loads. Krell being one Of the exceptions.

The class of operation has nothing to do with what load impedance the amp can drive. They simply aren't related, plain and simple.

Are class A amplifiers significantly harder to engineer, or they are about the same but only bring some of the well-known things (heat) with them simply because of class?

Also, how does efficiency of class-A amplifiers compare with other classes (ballpark figures)?

They are not harder to engineer. Its arguable that they are a bit easier, although you have to be more careful about heat issues and the power supply has to be able to support the output section being on all the time, something that you don't get with AB amps.

Of course class A is the least efficient of all classes of operation. The advantage is the output devices are often biased at the most linear portion of their curve. So if all goes well, they will make the least distortion.

In solid state amps only a handful and in tube only a handful also they are incredibly difficult to make and unreliable due to heat generation because of the 10 percent efficiency. So most of the makers if not all stopped making them.

This statement is false. We've been making class A amps for over 45 years. Reliability isn't an issue- that's an engineering problem, not something inherent in the class of operation. They are more efficient than 10% too- that's about a country mile off as the saying goes.