DESIGNING A 35MM LENS USING ONLY OFF THE SHELF OPTICS

Having spent many years in the optics industry, I had always been curious if it was possible to design a usable 35mm camera lens using only pre-made off the shelf singlet optics. Generally during optical design there are many variables to consider; Glass choice, lens thickness and diameter, air spaces and lens radii. Even then you can also deviate from pure spherical surfaces and introduce aspheres to help correct aberrations. In today's market, powerful software is used to optimize lens designs that would be impossible in the pre-computer hand drawn ray trace days. This article attempts to show the result of that challenge.

Optical Design

The design criteria were kept as simple as possible:

1. Canon EF mount with 44 mm flange focus distance

2. Optimized for an APS-C sensor size with an image height of 13.41 mm

3. Designed as a 6 element double gauss system

4. 50% minimum MTF at full field, maximum aperture

On the right is an example of a basic double gauss design. Three lenses before the aperture and three lenses after. The design has a mirrored appearance about the aperture stop in the middle. A design similar to this one was the starting point for my design since it already incorporates two plano surfaces(radius=infinity).

This was important because the lenses I had to choose from were all either plano-convex or plano-concave.

The optical design program used was WinLens3D Basic from QiOptic which is free to use. It can model complex systems right out of the box but lacked the ability to optimize the design so it had to be done by hand.

I chose optics from Edmund Optics since they had a large selection of optics available and started designing using trial and error and educated guesses, ultimately resulting in the lens design below consisting of four plano-convex and two plano-concave lenses with one cemented doublet (six elements in five groups). The air spaces were the only variable I had full control of, but I could choose different focal length singlets to swap out to see what effect it had.

Eventually the final design specs worked out as follows:

• 117 mm focal length

• 83.67 mm back focus at infinity

• 98.44 mm back focus at 1 m

• 13.41 mm image height

• 6.54 degree object angle (13.08 degree FOV)

• 6.704 degree image angle

• F/4.5 maximum aperture

• <0.3% distortion

• 1.0 m focus distance

If lens 3 was slightly larger in diameter(the plano-concave lens immediately before the aperture stop), the F# could have been set to F/4. Currently it will introduce vignetting in the corners if used at F/4.

Ultimately, the only specification that matters most is: What is the quality of the lens? How did the MTF turn out? As this would indicate how good the pictures would be.

As shown in this MTF vs Image Height, the MTF at F/4.5 (lower black lines) is above 50% out to about 12 mm from the center and falls to about 45% at the corner of the sensor. Given the design constraints, this will do nicely and although it doesn't quite make the original design criteria, the slight dip in the corners will be unnoticeable.

Opto Mechanical Design

With the optical design finalized, it was time to design the mechanics to hold everything in place with the ability to change focus and aperture size. Fortunately, over the entire focus range all elements move together so it was possible to mount all the optics in a single movable barrel. Often a lens design will need to have some of the elements moved separately to maintain high image quality. After quite a few hours of work, the mechanical design was completed.

To limit the number of parts that I would need to fabricate, I opted to bond the lenses in their mounts thereby eliminating the need for 6 screw rings. Also, the lens 4 mount is separate from lens 5/6 mount because that particular airspace(between lens 4 and lens 5) is the most sensitive and needed to be cut at assembly to be within +/- 0.01 mm. With the mechanical design complete and a full set of drawings made, off to the lathe to turn all the parts.

Assembly

At this point it was complete enough to test assemble. Aperture is fixed at F/4.5 but I also made a disk that I could insert to take it to F/8. Focus was controlled by sliding the lens assembly in or out for now until I'm able to complete the machining on the parts.

The lenses were centered and tacked in place with some silicone for now until the parts can be anodized and painted. Below is the partially assembled lens stand alone and mounted on the camera. View how the assembly is achieved in the short video.

Performance

So how did it perform? All things considered, pretty darn good. The only caveat right now is the fact that it is still unfinished and the optical path is shinny bare aluminum. This causes a lot of light to bounce around and wash out the image. So the example images below have had some post processing to help remove that, mostly tonal adjustments. Below are a couple shots with the full field and zoomed in to actual pixels to see the full detail.

Lastly, a comparison of resolution charts. On the left, my lens at F/8, on the right the

Canon 80-200 mm set at 117 mm F/8 The results are quite similar.

Final Thoughths

This brings us to today. Currently I am waiting for the opportunity to complete the machining which requires some CNC work. Once complete, then anodize, paint, laser engrave, and final assembly. I've also had a request from a professional photographer to try out the completed lens but she uses Nikon, so I also need to machine a spacer to accommodate the Nikon F mount 46.5 mm flange focus distance and am excited to see what a professional can do with it. See her work at LightOnLifeStudio.com

This project originally began as a thought experiment, but quickly grew into a fully functional lens capable of producing some decent pictures. Can you buy a better quality lens for less? Absolutely, this lens was not cheap to create, but that wasn't the goal. The goal was creating a usable lens from scratch, which I definitely accomplished. I will post progress as more happens on this project.

If you would like to see more detail explanations, or are interested in following me, please subscribe for updates and I'd love to hear your comments and suggestions on what else you would like to see posted here.