Skip to main content

I am testing TEZI operand to confirm if it really produces specified RMS surface error.
I created a simple toy lenses

1. flat mirror + paraxial surface to converge rays
2. flat mirror + afocal setting

I added surface errors to be 1um rms using TEZI operand and ran the tolerance command.
I set the wavelength to be 0.5 um.
To my understanding surface error is 2λrms. The mirror doubles the error and result wavefront RMS should be 4λrms.

The configuration 2(the afocal setting) seems to return the correct result, whereas configuration 1(paraxial surface) returns a little different values than I expected.

I wonder why this happens. 
I attached the two zos files.

Thank you for your cooperation.
 

 Screenshot of configuration 1 (paraxial surface)




Screenshot of configuration 2 (afocal setting)

 

Hi Uchida,

This is a really good question and I’m glad you’ve raised it.

The short answer is: use the afocal mode for this system.

The longer answer is: there is a subtle distinction between these two cases, and it lies in the definition of ‘wavefront’. The wavefront in an imaging system is the difference between the optical path length of the rays and some reference. In the afocal case the reference is a plane, and in the focal case the reference is a sphere of 100mm (that being the focal length of the paraxial lens). Given that the path length difference is the same in both cases (as it is induced by a spatial variation of the lens surface), the OPD relative to two different references is slightly different.

Now this difference does not matter in any real sense and you can tolerance successfully with either. Back in the day, using a paraxial lens was our recommended approach but we added afocla mode back in Zemax 5 or similar (a long time ago!) and using the afocal mode will give you results that are directly comparable to what you would measure on an interferometer with a plane reference.

  • Mark

Thank you for the explanation, Mark. It is clear now.


Reply