Hi Onur,

Mixed mode systems are difficult to use when you need to rely on chief or marginal rays for your calculation (really, anything beyond a simple Spot Diagram or Extended Scene analysis).  In general, you should think of mixed mode as being purely non-sequential with a little sprinkling of sequential mode (most sequential analysis won’t be accurate in mixed mode with complex geometry).  

In sequential mode, the SEQ ray tracing algorithm can keep track of the chief ray from surface to surface.  So, paraxial calculations (like the effective focal length) can be calculated for the entire system.  However, when you use a Non-Sequential Component surface, the SEQ ray tracer “hands off” all the rays to the non-sequential NSC ray tracer.  In non-sequential mode, there is no such thing as a chief ray, so when the SEQ ray tracer passes the rays to the NSC ray tracer, the chief ray “disappears”.  What is actually happening is the chief ray & parabasal rays most likely are seeing a slight deviation due to the STEP file (a very non-precise representation of a 3D object) and this deviation is causing the difference.

• In mixed mode, I would use the Single Ray Trace to look for any deviations of the chief ray.  
• In the Non-Sequential Component Editor, you can also try to increase the resolution of the STEP file. 
• Finally, in your CAD environment, you can try to increase the tessellation of the STEP file.