Solved

Modeling a receiver optic system for a Lidar

  • 10 December 2020
  • 3 replies
  • 198 views

Hello,


 


I am trying to reverse engineer a reciever camera of a lidar.


The wavelength that I am using is 940nm  because to validate this system I will use a 940nm laser as the source.


Which preset for the wavelength should be used to setup 940nm?


The object is at a distance of 3 meters and I am trying have a wide field of view (i started with 90°H and 30°V but I did not succeed).


At the moment I am getting very bad results is terms of spot size and PSF, the image is heavily distorted.


Since I am very new to the field, I am not sure what I am doing wrong.


Can anyone be kind enough to point me in the right direction?


Attaching the zeamx file for reference.


PS: I am trying to use lenses from manufacturer's catalogues so that I can build the model easily.


With best regards,


Amit Pandey

icon

Best answer by Allie 10 December 2020, 21:33

View original

3 replies

Userlevel 6
Badge +2

Hi Amit,


I have taken a look at your file, and I don't see anything particularly wrong with this setup. You have utilized a built-in Merit Function with appropriate boundary constraints. You have also set the thicknesses of the surfaces variable. If you are using manufacturer's pre-made lenses, then those are the only variables you can utilize. When you find the optimization is stuck, there are a few steps you can take:



  1. Try to use the Hammer Optimization or Global Search. These global optimization routines will help find alternatives to your current design form based on your Merit Function. Hammer, especially, may help you find a solution that is close to your current setup. 

  2. Add another lens. If you find that there are a lot of aberrations (or the criterion just isn't working) you may need to start performing aberration balancing. This involves adding another lens to counteract the aberrations placed on the system by an earlier lens. This type of repair is typical and is exemplified in our sample files. 

  3. Add an aspheric lens. I see you already have one Even Asphere surface. You can consider replacing one of your Standard Surfaces with another asphere. These can help to reduce aberrations.

  4. Consider a new starting point. We have several wide-angle lens examples available. You can find a few in the folder Zemax\Samples\Sequential\Objectives. You also have a subscription license which means you have access to the Design Templates. This is a collection of hundreds of lens files which you may use in your own project. 


I hope others in the community will comment with their ideas. In the meantime, let me know if you have any questions about the above. 


Best,


Allie

Userlevel 6
Badge +4

Hi Amit,


I also had a look at your design. Since you are using off-the-shelf lenses, all of the curvatures and glass thicknesses are fixed. The only variables are air thicknesses. Air thicknesses are not very strong variables and generally do not provide sufficent degrees of freedom to affect a good design. I think it is very unlikely you will be able to design an acceptable lens using off-the-shelf components. 


One method to approach this is to design an acceptable lens assuming custom elements (meaning the radii and thicknesses are variables), attempt to locate very similar off-the-shelf componets, and then hope to make it work by varying air spaces. (OpticStudio's Stock Lens Matching tool is intended to help with this, but because it does not try lenses reversed as well as as-drawn, it has limited value.)


But I suspect that a good design of such a compound lens cannot be achieved with off-the-shelf components.


Kind regards,


David

Userlevel 4
Badge

Hi Amit, 


I too have taken a look at your lens, and I have comments as well. But first, to set up your wavelength, you can simply set the wavelength manually. It looks like you have already done so, so no need to worry there. 


I agree with what David and Allie have stated, and I'll echo that the curvatures of the lenses are a major contributor to the performance, so please allow yourself the degrees of freedom that surface variables provide.


I noticed that the majority of your lenses are CaF2. While I can understand your choice for choosing low-dispersion glass, my understanding is that this material is notoriously scratch-prone, so I would not suggest it as your leading lens (in the front). Also, since your last lens, although aspheric, is also a low-ish dispersion glass, or crown glass, all four of your glass elements are low-dispersion crown-type glass. This will not allow for the canceling out of aberrations such as spherical aberration, which can be counteracted with a positive lens and a negative lens, and field curvature (Petzval sum) which can only be corrected by adding from flint glass (glass with high dispersion)


Your lens system has two lenses in front of the stop, and two lenses after the stop, but the two lenses in front are negative lens groups (they have negative focal length), and your two lenses after the stop are positive lenses. This causes an imbalance about the stop, where you have negative lenses in front of the stop and positive lenses in the back of the stop. This asymmetry about the stop will not help you to correct the odd aberrations, which are Coma and Distortion. Odd aberration correction is especially important when you want to increase the FOV, like you state since odd aberrations are primarily corrected with the power about the stop. 


My suggestion for a wide-angle system is to have a strong negative lens in the front, and subsequently have a good-performing normal FOV system but tweaked to counteract the aberrations caused by the front lens. This is easier said than done, but you can tap into years of design iteration by looking at retro-focus type lenses as a starting point. https://en.wikipedia.org/wiki/Ang%C3%A9nieux_retrofocus


Once you have a satisfactory performance with a moderate specification, then you can gradually increase the FOV to your target. In doing so, you may have to go back to the suggestions Allie stated, like adding lenses or adding aspheres. 


I hope this helps, and please don(t hesitate to ask for more help here. 


-Kats

Reply