inverting vs. non-inverting op amp with diodes in FB loop?

[Derringer]

Hi guys,

I've been mucking around with solid state preamp designs, opamp based fuzzes, etc etc

It seems that op amp based fuzzes tend to favor the inverting input as do some "metal" flavored preamps (ampeg vh140 comes to mind), same with the Amptweaker tightmetal pedal, which is basically a preamp in stombox form.


So what's the benefit here of using the inverting input instead of the non-inverting?

thanks

[amptramp]

With a non-inverting input, you get the signal amplified by 1 + Rf/Ri where Rf is the feedback resistance of the diodes in parallel with the gain setting resistor.  With an inverting input, you get a gain of -Rf/Ri.  If the signal clips hard, the inverting clipper stays at one voltage level, that of the diode forward voltage, as long as the diode is being driven into conduction.  With the non-inverting design, the voltage stays at the clipped level plus the input waveform, so even with hard clipping, you get some remaining signal variation.  You can simulate this in an inverting system with a resistor in series with the diodes.

[Derringer]

so is it fair to say then that the inverting design, in general, will give more total distortion than the non-inverting since the input waveform is not added to the output?

[merlinb]

so is it fair to say then that the inverting design, in general, will give more total distortion than the non-inverting since the input waveform is not added to the output?

Most of the time, yes.

Think of the non-inverting case as being the same as the inverting one, except the clean input signal is also 'added' to the clipped output signal. Kind of an inherent clean blend.

However, this also means it is possible to drive a non-inverting clipper all the way to the rail -a different kind of clipping- whereas the inverting clipper can't. This doesn't normally happen with guitar-sized signals though.