I also have to consider the OPA698.
Regarding the AD8036, I started playing with the Spice model (using LTSpice) and I am a bit surprised.
I simulated a non-inverting amplifier with a gain of 1, just like circuit TPC 4 of the datasheet, with +Vs=5V, -Vs=-5V, VH=1V, VL=-1V. The simulation does not exactly reproduce figure 1: for example when the input voltage rises,
- below VH, the ouput voltage follows the input voltage. So far, so good
- above VH, the ouput voltage is not constant and equal to VH but goes on rising. (1,04V for an input of 2V, 1,08V for an input of 5V).
Is there a problem with the AD8036's spice model ?