Hi Mads,

I looked over your model. In my opinion, you are trying to do too much with a single model. Here is the workflow I have used with similar design projects:

First, the design is optimized for image quality using sequential mode with a model similar to yours. But the slit is modeled as the object -- not as an aperture in an illumination path. The slit dimensions are modeled by field points in postition space. For example, a four field points could be placed at the corners of the slit and a fifth in the slit center. Or, if the slit is so narrow that the width is insignificant with respect to image quality, the slit could be modeled with three points- at the center and the two ends. The charactersitics of the illumination path is modeled by the selection of aperture type.using 'object cone angle' or 'Object space NA' to model the NA of the slit illumination, and perhaps 'Telecentric object space' as well. The image plane is placed at the location of the exit slit for a monochromator or the sensor for an imaging spectrometer. In this model (as in yours) the spots do not represent the image of the slit -- They represnt the spots formed by rays originating at the various field points. They are a measure of image qualtiy. You can, as I think you mention, optimize for small spots ony across the width of the slit, allowing astigmatism to lenthen the slit image rather than damage the resolution.The strike locations on the spot diagram do tell you about the dimensions of the slit image at the image plane. The slit image dimensions can be controlled by real ray targets at the image plane. They can also measure the slit dimensions as described by chiel ray strikes.

The second step is in Non-sequential mode. OpticStudio has a great tool for converting the design to non-sequential. Now we can model the slit. The slit could be modeled as a Source Rectangle with angular control to produce illumination of the correct NA. Or you can model the illumination path from light source to sensor by modeling the lamp and condenser forming a light body to illuminate a rectangular aperture the size of the slit. With this model, a rectangular detector can be used to visualize the image of the slit on the sensor and to measure the width of the image. A smaller detector can be used to model the exit slit or sensor pixel stripe and measure the power delivered. This delivered power can be graphed with respect to wavelength to measure spectral resolution.

I hope this helps,

Kind regards,

David

Hi Mads,

I think there are two reasons why you didn't see much difference in the RMS vs Wavelength plot when the slit width changes from 0.7 to 10 mm.

1. First because in the RMS vs Wavelenght plot you are using the RMS Spot Radius which is a 'radial' measurement of the spot size that combines the effect from both the X and Y directions. In this case, in the RMS vs Wavelenght plot, under Settings, you should choose the 'Data: Spot Y' if you only want to look at the Spot size variation along Y direction.

2. In the RMS vs Wavelength analysis under Settings, you have the 'Method: ' set to Gaussian Quadrature where you should in fact use Rectangular array. When your system has non-circular aperture, for example the Rectangule aperture on surface 5, Rectangular Array is the preferred sampling method because it can account for the effect of non-circular aperture accurately. 

Once you change these two settings, you should observe a clear RMS Spot Y size difference in the RMS vs Wavelength plot when changing the slit width from 0.7 mm to 10 mm.

That said, I do agree that the approach David described above should work as well. If you want to stay in sequential mode, you can set the slit as the object as David has described, and use one of the Extended Scene analysis tools, for example, the Image Simulation or the Geometric Image Analysis to set a rectangle source image to represent your slit at the object plane and observe the simulated slit image at the Image plane. 

Best regards,

Hui

Thank you very much for your help, both of you. Defining the slit in terms of fields did the trick. I’ll have to look more into the Non-sequential mode, as I have not been acquainted with it yet.

Thank you again

Best Regards

Mads