Hi John, 

I do work on endoscopes at the University of Arizona. These focal shifts are always a pain. There are some techniques that you may be able to exploit. One is the buried surface (a triplet), an optimize for focal shift in between the two wavelenghts you are using. Another technique is a doublet with an air interface or even a triplet with an air interface. Dispersion matching will be necessary here. I used a dichroic based technique to solve my issue, but in my case it was to separate the focal shift for different applications. Here is my publication:

https://www.osapublishing.org/ao/abstract.cfm?utm_medium=Email&utm_source=FeatureIssueAnnouncement&utm_campaign=AOWyantCollegeFocusIssueAnnouncement&utm_term=2020&utm_content=August&uri=ao-59-22-G71

Please feel free to contact me outside the Zemax if you need help with this. It seems like a problem that I may be able to help. 

David Vega, University of Arizona. 

Hi John, this is Kats Ikeda from Zemax. 

I agree with David that this is always a tricky issue and important in many fields in optics. I also agree with David that three lenses with dispersion matching should get us to your goal, depending on your criteria, of course. 

Just to make sure that I understand correctly, the prism sensor can’t be modified, and you need a focal length adapter that has focal planes (1mm apart) for two different wavelength ranges, Visual and NIR, correct?

In addition to the above comments from David, from a “tips and tricks” point of view, I thought of setting up Multi-config and setting some of the WLWT (wavelegnth weight) to zero for your optimization. I'm not sure if you've tried this already, but just in case, let me explain. 

In the right, I set up the Optimization Wizard for default merit function operands, and added EFFL (to monitor the value). Notice that the DMFS (default merit function) skips any wavelengths that are WLWT=0, so we can optimize for two different planes without either wavelength ranges competing with each other within the merit function. Analyses like FFT MTF will work for both, skipping the contribution from zero weight wavelengths.

Let us know how things go, I hope you find a solution soon.

Hello David and Kats,

Thank you both for your knowledge and advice.

David, I’m so excited to see that we are connected through the UoA. I was just admitted to the optical science graduate program. I’m eager to continue the conversation with you about endoscopes, your paper is very interesting. I’ll reach out to you on your UoA email.

Kats, your restatement of my problem is correct. My first approach was to set the NIR wavelength to zero weight and then use AXCL operands to weight the desired shift between visible and NIR. This was too simplistic since it only considered on-axis shift. I like your multi-configuration approach much better.

I had been working in afocal space since the camera is design for “eyepiece” endoscopes, but I converted to focal space which seems to work better. I have been trying the buried surface and air spaced doublet/triplets and made some headway shifting the NIR, but I’m still a long way from the desired image quality.

I will continue to work on this and let you know if I have a breakthrough. Thanks again, John