I'm using Asylum Research MFP 3D in AC mode with a silicon nitride tip. I get a nice resolution and morphology but the height retrace shows that my sample thickness is ten fold larger than published values. Is this an issue with the tip broadening? How can I account for this deviation?