I tried to simulate the self-noise of ADA4898 to verify if it's following the datasheet.
The results are far worst then the results written in the datasheet.
The datasheet states 0.9nV per root Hz at 1KHz, however the simulation gives 254nV per root Hz.
In addition the downloaded model had a syntax error, one line above the .subckt definition and pin assignment. The line with the pins text should be commented.
Does the spice model is accurate enough for designing circuits,
or my whole design will get wrong output results because of a faulty model?
Attached image of output noise density