Hi,

i am trying to make a dichroic mirror (beam splitter in this context) work in sequential mode.

I worked through two tutorials, ‘How to model a dichroic beam splitter’ and ‘how to model a beam splitter in sequential mode’. The problem is that my mirror is not a cube but a thin (1mm) plate. The plate has a coating on the back. For the reflection pathway, the rays need to go through the plate, then get reflected and again go through the plate and leave it in a 90° angle in regard to the incoming rays. When I try to use coordinate breaks to bend the optical path by 90° in a way that they are getting reflected I get a weird lens design (see below). I have played around with different parameters on the coordinate break and also used the add fold mirror assisant, but its not doing what i want it to do.

This is how the rays should go. The coating is at the back of the plate, so the rays have to pass the plate twice.

This is how the subsequent lens looks like:

It appears in a weird shape, changing the parameters or readding it from the lens library doesnt help.

If I turn the system 90° to the other direction, with the rays going through the plate only once everything is fine:

How can I model the beam splitter correctly, is it possible in sequential mode?

As its apparently not possible to upload zos files (why?) I made a screenshot of the lens data:

Thank you in advance.