Gerry de la Sel wrote:
Even if the Rat opamp did clip, it would clip at about 7.7Vp-p. However, this signal is immediately delivered to the diodes, which will clip at about 0.7Vp-p. Therefore, whatever bad sound the opamp might have created, it is immediately removed by the diodes. You never hear the opamp clipping. I don't know any distortion pedals where you can hear opamp clipping; it always covered by diode clipping.
ok, the red-marked point is what is wrong, because there is this 1k resistor between diodes and lm386's output that "isolates" the signal in some degree. You may think that resistor's effect is completely unimportant, so,
so, I made a simulation in Tina to look it better.
I feeded input with a 50mV- 600Hz sine, pretty typical for guitar tone. Tone shaping & gain is the same as in real Rat, but I used tl071 (inaccuracy of simulation).
The escilloscope results are enlightening and interesting.
The
green waveform is taken from the "OUT" point. It is clipped by 1N4148 diodes, at 600mV (each grid is 200mV for green signal) and is almost square.
The
brown waveform is taken from the "VM1" point. It is clipped by tl071 only, at 2V - 2,7Volts (each grid is 2V for brown signal) and is somehow sawtooth.