Working with OAPs in sequential mode can be tricky.  Below is an example of what I think you are looking to accomplish.  Here are a few tips.  First, use the parabolic surfaces with their full apertures, then once things are aligned you can go back and add the off-axis apertures (in my example I’ve used two configurations, the first without the off-axis apertures and the second with them included).  Next, utilize chief-ray solves to help with nominal alignment of the on-axis field.  Once your system looks good, you can convert the solves to fixed values which locks the baseline design in place and allows for system perturbations to be investigated.  Lastly, it’s important to understand how the local coordinate system should change from surface-to-surface; in this regard it can be helpful to add display of the local coordinate systems as needed.

Hope this helps…

Wow this is a great design. I inputted it and got a much improved layout - then added a precision aspheric lens for a focus.

I had to stop the setup down to 10mm (I wanted 20mm) but the spot diagram is showing more distortion than I would have expected.

Have I missed a value somewhere?

The devil is in the detail!  Your lens parameters are wrong (no asphere coefficients are present, and in fact the last lens surface isn’t even designated as an asphere).  Simply load the lens model from the built-in Edmund Optics catalog and you will see much better results.  When errors arise, it’s important to first carefully check all the component parameters.

Thank you so much for your response. I think I have fixed it now. I found the above performance with 15mm of aperture - it degrades rather severely with 20mm aperture - even after trying a non achromat past the grating. I had hoped the mirrors would have provided better optical performance?