What's happening with this non-inverting amp?

[dentaku]

Here' a beginner OpAmp question.
First of all. YES I KNOW the V- connection of the OpAmp really should be connected to -5V and not ground but THAT'S the point of this question.

The inverting amp does as I'd expect and the blue trace stays at zero when the input (green trace) is positive.

The non-inverting amp, on the other hand, is fine up until the input goes beyond about -0.7V then you get those huge spikes where I'd expect it to stay at zero.

What's happening here when the opamp is powered incorrectly this way?

[c4757p]

Phase inversion. You're saturating the input transistors, so they no longer behave as inverting differential amplifiers and begin passing the signal through directly - so in effect, the op amp's inputs swap.

[dentaku]

Phase inversion. You're saturating the input transistors, so they no longer behave as inverting differential amplifiers and begin passing the signal through directly - so in effect, the op amp's inputs swap.

So... looking at the internals of an LM358 I see that the inputs are connected to PNPs so when the input goes to about -0.65V below GND that's when saturation happens?
Hmmm... Isn't 0.6 to 0.7V a typical diode drop too?

[Andy Watson]

Hmmm... Isn't 0.6 to 0.7V a typical diode drop too?

Yes, it is not a coincidence.