We have a circuit that is performing pretty well. However, we still have an issue with rise/fall times. We are looking to get sub 1ns rise/fall times to meet the specifications and we are close. I am attaching screen shots of the signals. The first image shows the input and output (input on the left). The second image shows the output only and you can see that the fall time is pretty close to 1ns. The third image shows a close up on the output and you can see that the rise time of the pulse is much greater than the fall time. There is a rounding off of the output rising edge after the signal transitions past the center (reference) point. We were wondering what accounts for this rounding.
I am also attaching an image of the schematic. We have made some changes to the schematic though. The amplifier is powered off +4.8V and -0.850V to get us to the center (reference) voltage of 1.65V similar to the input. Also, it is providing unity gain with a 432-ohm resistor in the feedback path (R274) and the Rg (R120) is removed. All other components are the same as shown in the schematic. We are driving a 50-ohm load.
We reviewed the layout and some work can be done there to clear out the many vias used in the circuit. What other recommendations are suggested to improve the output performance?