Christopher Cotton, ASE Optics GM and optical engineer, answers audience questions in the Laser Focus World webcast, Design for Aspheres
Audience Questions & Answers
Q: How do you tolerance aspheres to manufacturers?
A: The standard way to do this is similar to tolerancing any other optical surface. Define departure you’ll allow from surface. Determine fringes of power and irregularity. As we get more into machines making optics and smaller tool polishing you have to define surface slope to avoid ripples/features you’re not expecting in the system. Aspheric defines the difference between the surface you’re getting and the ideal surface. Allowing change in power on asphere is similar to finding it for a spherical element. In addition to regularity, define slope tolerance.
Q: What are problems and sensitivities of aspherical mirrors?
A: The common aspherical mirror is a parabolic surface, found in Newtonian telescopes. It produces perfect focus on axis but as you move off axis you get coma and astigmatism very quickly. You find that you have to spend a lot of time designing them if you want to get the field angle you’re looking for. These are used in common astronomical telescopes.
Q: Are there databases of previously designed aspheres or do you have to start anew?
A: RPO lenses are available through RPO and in catalogs. All those designs are available in Zemax so you can bring them in and adapt them to your system. If you’re just going to use one or two, they’ll be difficult to manufacture.
Q: Do you prefer odd or even polynomial surfaces?
A: I’m not 100% sure of the definition people use. I look at it based on change in sag of surface. Some people look at change in derivative of surface instead. The even asphere is rotationally symmetric. Odd aspheres have cosine or sine—changes with the angle.
Q: What design tools are best for aspheres?
A: Code V, Oslo, Zemax. All are useful and pretty similar. The biggest change is the way that you use each tool. Overall I like Zemax because I treat optical designs like a video game.
Q: How do you know how many surfaces you need to make aspheric?
A: Look at surfaces that do the most work, have the lowest R number. They’re most likely to benefit from having an asphere.
Q: Are aspheres used in common photographic lenses as well?
A: There is a line of photo lenses that uses aspheres designed for digital cameras to get a smaller angle at sensor. Nikon lenses have plastic aspheres in them. Aspheric can be made cheaply in large volumes.
Q: Does the shape of the asphere affect if it can be asphere-ized?
A: Not necessarily. Shorter radii and small diameters are hard to make aspheric. When molding plastics, it depends on how much time you’re willing to spend to get a good surface with your tools.
Q: Is there any rule of thumb about the cost of an aspheric surface?
A: Depends on plastic, glass, crystal. If you’re diamond turning it’s about $100 per surface. If you’ve got one that is polished there’s not an easy guess. If it’s moldable then you can you can say between 12 and 20,000 dollars tooling (but don’t quote me). For plastic, $25000 to make the tools and then make samples. It really depends on the object.
Q: Are these used in precision optical systems?
A: Yes, military IPT and objective. They have tight tolerances and high performance requirements.
Q: What is limit to slope on an aspheric surface?
A: Stay away from inflection points so slope of zero is bad. When you get to high slope, it depends on surface. Stay away from 30 degree slope and don’t go past that. Whenever you make something more difficult, you make it more expensive.
Q: Which asphere mathematical equation do you recommend?
A: Political question. Each person will have a different answer. Zernicke and Forbes are both great approaches. I prefer Zernicke is it’s easier to get to rotational equation.
Have a question we haven't answered? Please ask it in the comments below. We'll answer them in a future blog post.