Class B w/o crossover distortion (1975)

[Wavebourn]

Originally Posted by Eva

Warning: Performing speaker THD measurements may result in permanent changes in your point of view about amplifier THD

Not necessary. It may cause a curiosity, why hearing ignores distortions caused by mechanical things, but don't forgive distortions caused by electronics. That will lead to Tube forum for answers why.

[AKSA]

Wahab,

You are right. Q1/Q2 are indeed a diff pair, though a series connection. Voltage feedback is the mode of operation, and pnp driver current is indeed limited by the RC boostrap load resistors. And the current variation through the RC is the weak point of this design.

In truth, it is a power buffer, not an amplifier, and presented here as a curiosity, just another way of doing a power output stage.

Rod's article is good, it is consistent with most of my past designs. The road to a commercial amplifier is (and should be) littered with many different theoretical amplifiers, whose pros and cons must be carefully assessed. I am more than happy to throw up some of these ideas, but if you (and others) prefer I'm happy to remain silent.

Please remember that I make no claim to excellent design here, I am merely presenting an idea.

Hugh

[Eva]

Music is currently being recorded by feding the output of a microphone to an ADC and then to the input of a digital data storage device.

The data used to analyze the THD from a loudspeaker or an amplifier is obtained exactly in the same way.

Waveforms are just information, and that's what we hear. Speaker THD and amplifier THD must be the same thing.

[Wavebourn]

Originally Posted by Eva

Speaker THD and amplifier THD must be the same thing.

For computer?
Sure.
For brain?
No.

[AKSA]

Now now you two, just agree to disagree........

[hahfran]

Originally Posted by Eva

Music is currently being recorded by feding the output of a microphone to an ADC and then to the input of a digital data storage device.

The data used to analyze the THD from a loudspeaker or an amplifier is obtained exactly in the same way.

Waveforms are just information, and that's what we hear. Speaker THD and amplifier THD must be the same thing.

No. A well known example is the strike sound of a church bell. When the tongue hits the bell we hear this particular strike sound but objectively it does not exist. As a matter of fact we hear a lot that objectively does not exist and we don't hear a lot details in sound fields which objectively exist.
Your assumption that human audition is in any way linear ( i.e. a linear system) such that it linearly maps an objective sound field onto a subjective impression is entirely wrong. So briefly all that matters is how it sounds and that is not a matter of thd if thd is around some %.

[wahab]

Originally Posted by AKSA

Wahab,

You are right. Q1/Q2 are indeed a diff pair, though a series connection. Voltage feedback is the mode of operation, and pnp driver current is indeed limited by the RC boostrap load resistors. And the current variation through the RC is the weak point of this design.

In truth, it is a power buffer, not an amplifier, and presented here as a curiosity, just another way of doing a power output stage.

Rod's article is good, it is consistent with most of my past designs. The road to a commercial amplifier is (and should be) littered with many different theoretical amplifiers, whose pros and cons must be carefully assessed. I am more than happy to throw up some of these ideas, but if you (and others) prefer I'm happy to remain silent.

Please remember that I make no claim to excellent design here, I am merely presenting an idea.

Hugh

no pun intended, hugh...
just wanted to highlight the fact that numerous topologies
where tested during the 70 s, and that few working ones did
emerge from the then inventive times...
there s 4 schematics mainly, with the three last being evolutions from
the first one..
first, a simple vas + output stage..
second, the same with an added input stage,giving
the cfb i quoted..
three, the latter with yet another device to make a differential
at the input, this is the lin that was and still the most used...
four is symetrisation of this differential, to go to a symetrical
differential...
almost all circuits are variations of these four ones, and it happens
that nothing better has been found since despite the claims
of numerous designers...



regards,

wahab

[AKSA]

Wahab,

The fact this is a road well trod, and one so universally below par (how many commercial amps sound good?), is strong reason to continue the quest.......

I hope you are not discouraged by the conventional, which is all around us. It should be motivation to do better, to come up with something completely different. Certainly that is what drives me, despite having no more than second year undergraduate math.

Hugh

[hahfran]

I found another even older design idea that is intended to avoid switching and crossover distortion. However I am unfamliar with PSpice the free version. I need a bunch of transistor models and a method to load these, if found, in the model library.

[AKSA]

Hahfran,

Simulations won't tell you how the amp will sound, but are useful development tools. I suggest LTSpice, which is quite good and has readily available models, some on this forum. You can insert them into the model library using a text editor like Notepad; PSpice models work just fine.

Would you be happy to share your latest find?

Hugh