Hello again,
You are right about that second diode and the delay to get an op amp out of saturation and i am very happy to see that you understand this phenomenon. It also has something to do with slew rate because if the diode did not clamp the negative output the output of the op amp would have to go very negative if there was a reasonable negative supply and to get the output to go positive again the op amp would have to slew all the way up to whatever positive value was required for the given input. So if the slew rate was 1v/us and the output went down to -10v and had to swing up to +5v, that would be an extra 15us delay on top of the delay caused by saturation of the internal circuitry. IF the output only goes down to -0.5v, then the extra time would only be 5.5us a much shorter time delay.
I am happy to see that you understand this concept because it is often not thought about until the circuit is breadboarded and then the question comes up as to why the op amp is acting so slow.
