New VAS simulation.
As you see in the image 1 your version is not enough compensated.
There is an overshoot both in the AC Analysis and the oscilloscope square wave.
It is not to bad, but it is not what I want.
I have changed some component values to make a perfect square wave.
You can see my changes in image 2.
And so the result after changes in image 3.
The distortion, THD, is a little bit worse in your version. But not too much at all.
As you see in the image 1 your version is not enough compensated.
There is an overshoot both in the AC Analysis and the oscilloscope square wave.
It is not to bad, but it is not what I want.
I have changed some component values to make a perfect square wave.
You can see my changes in image 2.
And so the result after changes in image 3.
The distortion, THD, is a little bit worse in your version. But not too much at all.
Attachments
You have really used 1k for R12?
For me I have used 150 or 220 to set the bias in output.
R12 as 1k gives a voltage drop more than 5 Volt.
Attachments
You got the point at last. Professional designers follow some standard rules (I don't claim this role though), they take care to keep all transistors within their linear region by applying a suitable biasing on each one, and they try to keep the current of each branch constant using thermal compensation where is needed. The simulator is very helpful, as e.g. calculates all D.C. points but you should know if those are suitable for each one of the transistors you are using. By changing the values of resistors or substituting diodes with resistors to chance, you could cause high THD, instability and thermal issues, etc. although you believe that "it sounds good", as the sound taste is completely subjective. BTW an excellent and freeware simulator is the LTspice, with schematic capture and waveform viewer.