The truth is that I've tried to change the PNP frequency transition ,it look better but I can not solve the problem.If you vary the input voltage keeping the frequency constant, does the glitch stay at the same point in time (but at a different voltage), or does it occur at the same voltage level (but at a different point in time)?
Yes,the more the frequency is increased,the bigger the glitch is.However,at low freq (around 10khz),the glitch is too small to see.The signal output is nice at low freq.
I note that there seems to be a small downwards step in the input voltage at the same time.
No,-24V and +24V is standing still.
This circuit should have unity gain (or about 0.95), so why are the input and output signals of opposite polarity?
I've just change the gain to see it clear.It's not about the amplitude,it's about the glitch waveform at the higher and higher freq.
Ah,I got your point,cause this is the inverting operational output,I supplied to the inverting input,and applied gain.Definitely a small step on the yellow trace corresponding to the glitch one the blue trace.
Are you sure that is the correct circuit diagram? The diagram is of a non-inverting circuit with no voltage gain. The ‘scope traces are for an inverting circuit with a gain of 3.