Hi Nishie-san,
I wanted to input a 125kHz square wave and output a signal amplified 6 times, but the output voltage became like a Sin wave.
I simulated in Tina-TI and it looks that is fairly close. 100V/usec is defined from 10% to 90% of rising edge on a step response, and settling time indicates that it will take addition time to settling at a setpoint to ±0.01% of the full amplitude.
To duplicate the input square wave, you would need higher BW op amp to replicate square wave's rising and falling edge. To completely duplicate the rising and falling edge of a square wave, you would need up to 13th harmonic (odd harmonic) in BW to fully duplicate the full square wave's edge definition. OPA596 may not have the BW remaining to duplicate the square waveform closely, after the GBW product.
If you want to generate sharp square waveform, you will be better off to use comparator (open collect or push-pull output stage) to generate the required square wave.
If you have other questions, please let me know.
Best,
Raymond
