I'm designing a Lock In amp which after processing will have a bandwidth of 10Hz - so the noise floor is very low ~40nV RMS (heavy loading by feedback network obviously) at ADA4895 output, G=10. It is therefore important to know how the distortion varies with Vout below the 2Vpp condition of the datasheet distortion measurements. In particular I expect with my signal processing (which will be differential) the dominant contribution to distortion will be HD3 from around 400kHz, HD2 will depend on matching. Does anyone have any knowledge of this amplifiers small signal distortion.
The other part I'm considering is AD8021 which has a non-rail-to-rail output and is proud of its <2Vpp distortion, ADA4895 has worse 2Vpp than 4Vpp distortion. I'm also considering the archaic THS4022 from the other side.