Amazing DIY camera lens fabrication

[Robin Guymer]

Not sure if this has been linked on APUG yet, but this is a very impressive piece of home engineering. A custom built camera lens including the cutting and grinding of the lens glass. I like modifying cameras too but this is way above anything I could contemplate. Enjoy.
On YouTube link (//youtu.be/_GCwlGZNd3I) or search "DIY camera lens" by Mats Wernersson

 

[ransel]

That was simply amazing.

 

[Pioneer]

I love watching these. Thanks for posting.

 

[Kevin Caulfield]

That was pretty impressive. Thanks for posting.

 

[AgX]

Most of that could be done in a good DIY workshop. The milling/grinding of the negative rough curvatures might be done on a milling machine with swivelling rack, if one mounts a precision rotating head onto the milling bed.
BUT how to mill the positive curvatures with a common milling machine? Anyone make me wiser?

I myself would rather start from a lathe. And mount a spherical turning device on its bed.

 

[Nodda Duma]

AgX, that wouldn't work quite as you'd think. Fundamentally, to obtain a truly spherical surface the tool and blank are counter-rotated against each other at different, non-harmonic rotational frequencies while the tool is swept across the surface. A grinding / polishing slurry is kept loaded between them. The motion automatically generstes a positive spherical surface on one, and negative spherical surface on the other. Which goes to which depends on orientation of tool and blank (positive on bottom). This activity could not be replicated on a mill or lathe. Additionally, water is required to control the generation of dust, or the operator risks getting silicosis (a disease caused by silica -- glass -- dust entering the lungs). The required water would quickly ruin a manual lathe or mill typically found in a DIY shop.

In optical surface generation, the desired spherical radius is actually generated during the grinding stage. The polishing stage does not set the radius of curvature. It removes the damaged layer and performs minor adjustment to meet the surface figure and irregularity requirements.

A lens grinding machine and a lens polishing station are purpose-built machines that replicate the required polishing strokes used to grind and polish soherical optics when not performed by hand. Additional machines are available to generate aspheres.

Texereau "How to Make a Telescope" goes into very good detail about the polishing strokes required to generate a proper optical surface. Grinding and polishing by hand for the DIY'er is the easiest way and not difficult. The grinding stage is a good way to burn calories.

 

[AgX]

I do not understand your comment. The maker is seemingly using a kind of mill. (I'm referring to the rough grinding, not the polishing.) He does not use a pre-formed grinding tool for that but some smaller tool moved in correct radius over the rotating glass.) The same way the industry with CNC lens-making machines does.

Furthemore one can grind with coolant on a lathe too. One can machine a precision sphere-making device to set precisely the radius for moving the tool. I agree though that in such basic set-up only the glass is rotating, not the tool too.
Though on that sphere-making-device one could also install a mill-motor. Then both, glass and tool would be rotating.

 

[Nodda Duma]

The mill and lathe were used to remove the pre-existing surface curvatures and generate blanks. He required this step because he obtained the required glass types as surplus doublets from surplus shed. That step got him to a starting point. He could have acquired blanks from a glass supplier and skipped this step but it would have been more expensive.

The generation of correct surface curvature for the design was performed on the grinding station seen at 0:50. Polishing was done on the polisher at 1:25.

The fellow certainly knows what he's doing. Very impressive.

 

[AgX]

I was referring to that scene starting at 0:30 assuming this was the stage the basic curvature is given.
If that only would be in the next stage, where is the pre-formed shaping cup?

Yes, as you say he is using readymade lenses as source of his glass. But that either neccessitates that he knows what type of glass he is buying or he has to meter the glass after acquiring for refraction/dispersion.

 

[Nodda Duma]

I always enjoy your inquiries, AgX.

I was referring to that scene starting at 0:30 assuming this was the stage the basic curvature is given.
If that only would be in the next stage, where is the pre-formed shaping cup?

No! He only generated the lens blanks in this first stage. The very first step is cutting out the proper diameters. The second step is separating the elements (not shown), then grinding to remove wedge and perform most of the grunt work. Not to generate the proper radius.

You don't need or want to use a cup to generate the radius of curvature. It is not accurate. When he blocked up the blank with green wax, then he began generating the surface curvatures. The surfaces were generated on the station shown at 0:50. Note the jar of 80 grit carbo. What he didn't show was measurements with (likely) a spherometer.

Read Texereau

Yes, as you say he is using readymade lenses as source of his glass. But that either neccessitates that he knows what type of glass he is buying or he has to meter the glass after acquiring for refraction/dispersion.

No need to measure the glasses.
He very, very likely designed using BK-7 and F2...which are very common simple doublet crowns and flints and are all that's needed for a Petzval.

To get the necessary glass types, he did one of two things: 1) Ask if they have lenses of the proper glass types (likely they have the paperwork). 2). Pseudo-randomly purchase lenses and back out the glasstypes after measuring surface curvatures and thicknesses. Then design the Petzval.

He did the former because he didn't separate the doublets before cutting the blanks, which he would have needed to do to to try to back out the glass types.

Getting the necessary glasstypes is not a difficult problem to solve when you know which types are commonly available and use them in the design.

 

[AgX]

I referred to the types of glass as you yourself recently hinted at the problem of designing a lens with off-the-shelf readymade elements, part of which was the limited choice in types of glass.
Of course if one has a design in mind that works with very common types, then it would be easier to reuse surplus/used lens elemenst as source for raw glass.

Aside of the price issue can an amateur get small enough glass pieces directly from the the glassworks if he can't get used elements to work on?

 

[Luckless]

Well, if you're really keen on lens production, then it is not impossible to make and cast your own glass blanks. It is of course a major step up in complexity and science (and risk of bodily harm), but it is an option if you're willing to fund a suitable shop space. Sourcing highly exotic materials in suitable purities may still be a pain of course, but it can open up a wider range of glass types being available for projects. (And if a project fails, well, "just add hammer..." and start over again.)

And if you're only ever making fairly small lenses and with limited runs of glass production, then you actually can get away with some fairly small gear and work spaces. As in, "Small garage/reasonable sized shed" kind of spaces, not "Large multinational steel works with hundreds of employees" kind of setups.

 

[Nodda Duma]

I referred to the types of glass as you yourself recently hinted at the problem of designing a lens with off-the-shelf readymade elements, part of which was the limited choice in types of glass.
Of course if one has a design in mind that works with very common types, then it would be easier to reuse surplus/used lens elemenst as source for raw glass.

Aside of the price issue can an amateur get small enough glass pieces directly from the the glassworks if he can't get used elements to work on?

The difference is that I designed the Cooke using only pre-made catalog lenses. This puts an extreme limitation on glass choice. With his ability to manufacture custom lenses, all he needed to do is find an inexpensive source of the proper glass types in sizes large enough to make his lens. Pretty smart to get them from Surplus Shed. Way cheaper than buying the blanks new.


For your second question, the answer is yes, you could get blanks from a distributor. This is in large part due to the widespread practice of lean manufacturing, where optical shops order blanks on an as-needed basis and do not keep a lot of glass in stock (they sometimes do for rush orders). That includes orders as small as quantity 1. Optical shops with small orders typically deal with a glass distributor. The distributor buys the glass in bulk (melt glass) from Schott, Ohara, etc at much cheaper cost, then cuts and generates blanks for their customers. It is cheaper overall than if you bought a singke blank from the glass factory.

 

[Nodda Duma]

Well, if you're really keen on lens production, then it is not impossible to make and cast your own glass blanks. It is of course a major step up in complexity and science (and risk of bodily harm), but it is an option if you're willing to fund a suitable shop space. Sourcing highly exotic materials in suitable purities may still be a pain of course, but it can open up a wider range of glass types being available for projects. (And if a project fails, well, "just add hammer..." and start over again.)

And if you're only ever making fairly small lenses and with limited runs of glass production, then you actually can get away with some fairly small gear and work spaces. As in, "Small garage/reasonable sized shed" kind of spaces, not "Large multinational steel works with hundreds of employees" kind of setups.

I've helped a colleague do this (cast glass) for a telescope mirror..just so we could say we did it. It is possible for simple soda lime glass to do so at home. It is an incredible pain in the ass, and cost way more than paying $40 for an 8" mirror blank.

However, there is no way whatsoever that a fellow could produce optical quality flints, crowns, etc at home from scratch. That is the equivalent in difficulty of going from taming and feeding a horse to building your own jet aircraft. The process control, learning curve, tooling, and materials are far, far beyond the capabilities of a home lab. You could easily invest millions, if not tens of millions of dollars setting up a precision glass factory to generate the one or two blanks necessary to truly DIY. Then you have two glasstypes that require full characterization so that you can then begin modeling the optical design with them.

No, what that fellow did was manly enough. . Heck, even being able to do the optical design and the anodization bath himself deserves kudos.

 

[Luckless]

Well of course you don't just set up some cobbled together glassmaking equipment on your kitchen counter on weekends, but what exactly stops pretty much anyone from spending the time and effort of putting together a properly equipped shop if they have an interest in making glass?

We are talking about technology that is at this point several hundred years old, and originally made with equipment that is laughable by today's standards for precision and accuracy, not to mention the technology for quality control.

"Its scary, hard, and complex!" is a great mindset if you have no interest in actually doing things. It is also the mindset that keeps people from enjoying film photography, so I always find it kind of amusing when film photographers suggest that things to do with regards to equipment fabrication are 'too complex and difficult'.

 

[flavio81]

However, there is no way whatsoever that a fellow could produce optical quality flints, crowns, etc at home from scratch.

Not to mention producing special glass of high refraction index AND low dispersion...

 

[Mats Wernersson]

First of all I apologize for making such a short video lacking all the details you like to know. But I suppose most people are not like us ;o)
I like to fill in some details
I did not buy lenses or achromats just molded blanks from Surplus Shed. The flints are FN11 SH number L14507. They don't have many flints but a lot of crowns. The crowns are L3895 they state index and Abbe number so i concluded they are BK1. The huge flint was $12 and the crowns 80 cents when buing 10.

I started out by drilling out the size of blanks i needed. This was done i my Sherline mill using diamond powder covered bisquit cutters. I bought a whole set of cutter from China for about $10. The i had to mill them to right thicknes. I did that using the same cutters. Water cooling is necessary and for health reasons as pointed out by a previous comment in this thread. I made a water tank of PVS sheets and used a cheap fountain pump for circulation. That way I could use a standard mill.
Next step is curve generation. Same bisquit cutters again. The method is well known. Below is an illustration from "Modern Optical Engineering". It works for concave as well as convex surfaces. Let the glass and tool rotate in opposite directions to minimize the risk of chipped edges. It is not a big deal if the tool hangs over the edge. The important thing is that it is large enough. So you don't need a tool of precise size but you need to KNOW the size. It is the cutting diameter that is important. I made an easy touch down on a piece of glass and measured the diameter of the mark.

Alignment is important to get the desired radius. But how precise you like to be is really a matter of how much time you like to spend on grinding. But note that it is more difficult to arrive at the right thickness if you are far from the target when you start grinding. I fitted a laser pointer to the milling head and made a scale on the wall two meters away. That helped setting the right angle. Centering is more difficult. I deliberately started milling a little bit off center (away from the center, not passing over it). The i inspected the untouched center through a microscope having a measuring scale. Then i knew how much to move the mill in X-axis. I arrived with a fairly good curve. The radius was checked with a spherometer. A spherometer is easy to make using a dial gauge and three steel balls (you can find examples on the net).
You might say that I now have my lens blanks. A little bit oversized pieces of glass with a fairly good shape.
Next is grinding. The grinding makes the shape spherical with the needed precision and correct radius. The tool was i some cases another lens blank made for the purpose. I also used tools made of cast iron the was turned to right shape. I got frustrated over those iron tools as the tended to scatch the glass when coming to the finest grades of grit. I don't know really why. Any advice is welcome.
The grinding went through finer grades until the surface was silky smooth (ending with white aluminum oxide of 3 microns). I stareted with 80 grit as can be seen on the video but found out it was overkill and started with 160 or even 220 for the following surfaces.
Polishing was done with Cerium Oxide and I used Burgundy as lap pitch.
Amateur telescope makers is a good source of experience and information. But it is difficult do find good information on making small lenses. It may sound difficult or even impossible but remember that Newton and Galileo did it by hand a long time ago. By the way, thank you Isaac for the advice of using Burgundy as pitch ;o)

Glass supply is a big issue. Although I found suitable glass for this project I would really like to have a wider choice of glasses. Anyone having a link to a good supplier? And do you have any price examples?

 

[Nodda Duma]

Hi Mats,

That was an awesome project. I shared your video with my fellow optical and opto-mechanical engineers at work and they were impressed as well. Thank you also for filling in the details explanation..I was typing between meetings trying to explain based on educated guess. I'm assuming you have a background in the industry?

Are you familiar with Hank Karow's book, "Fabrication Methods for Precision Optics"? It would be a useful reference if you're not.

You were right to use another glass blank instead of cast iron for fine grinding. In fact you can use the crown and flint as tool and blank to grind the matching inner radius of curvature (I'm guessing you did). The cast iron tooling does pretty much what you discovered...I think it's really only good for rough grinding. Of course, you have to very carefully clean up the work area when shifting to a finer grit. The iron (and brass) tools I've seen are etched like a pitch lap.

Your illustration is perfect and explains blank generation better than I attempted.

Re: centering. Many shops edge grind after the surfaces have been finished in order to minimize wedge (which can be thought of as variation in edge thickness). You can do this on a lathe with a hollow spindle. Align a laser to the lathe mechanical axis, shining through the hollow spindle against a far wall. You have to mechanically center the laser beam (minimize both runout and tilt misalignment) then chuck up the lens on a wax chuck and center until spinning the lens does not cause the laser to scribe an arc by refraction through the lens. The optical axis is now aligned to the mechanical axis of the lathe. Then, spin the lens while you grind the edge down to required diameter. You have eliminated wedge. There are specialized edge grinders that do this in an optical shop, but it can be done on a mini-lathe with a hollow spindle.

Burgundy is ok pitch but a harder Gugolz (like 80) or Acculap pitch will give you better surface contact and thus better accuracy and surface irregularity. I prefer the feel of Gugolz with totanium oxide or red rouge (if the glass is stain resistant). But if you have good luck with Burgundy then use what works.

A glass supplier will depend on your location. I can find out who shops use for, say, New England or the Midwestern U.S. Other than that, check Schott or Ohara website and determine who their distributor is for your region. That is the best way. I would suggest using Ohara equivalent glasses in your designs..they are usually cheaper, just as good quality as Schott, and the distributors always say they have shorter and more reliable lead times.

For Ohara start here:

http://www.oharacorp.com/group.html

If you're in Germany, go to

http://www.ohara-gmbh.com/

-Jason

 

[Luckless]

I've met a number of small artisans who work part time out of a backyard shop kind of setup, much like you can find modern blacksmiths or even many photographers with home studios, who have enjoyed the challenge of producing clear optical quality glass for their projects.

One of the largest practical differences, beyond volume capacity, between such small artisans and a large firm producing optical glass seems to boil down to design specification and precision. A smaller/hobby shop would more practically aim for a general ballpark on the glass's specification when producing an initial blank, and then adjust the design to the optical properties of the resulting glass, rather than designing a lens and adjusting the glass formulation to match.

Of course you don't sit down and decide to clone a Canon L series or some other high end bit of optics in a weekend, and it could take a hobby glass maker years of experimentation to actually make a successful lens of reasonable value and quality, but I can't imagine why that should stand in someone's way if they have an interest in making a hobby of it.

 

[removed account4]

Not sure if this has been linked on APUG yet, but this is a very impressive piece of home engineering. A custom built camera lens including the cutting and grinding of the lens glass. I like modifying cameras too but this is way above anything I could contemplate. Enjoy.
On YouTube link (//youtu.be/_GCwlGZNd3I) or search "DIY camera lens" by Mats Wernersson

thanks for posting this, its nice to see how things are done

 

[Nodda Duma]

Hi Luckless,

Yes, forget cloning a Canon. I'm talking the basic flints and crowns of, say, 100 years ago...It is not as simple as adjusting the design to match the glass properties. It is unbelievably (obviously, since you don't believe me) difficult..

Imagine making Ektachrome in your basement. That is the level of difficulty we are talking about. At least now you can purchase the necessary platinum alloy crucibles, tongs, etc. necessary to create the glass in rather than having to make your own. So I guess that is easier.

Here is a 200mL platinum crucible. It is only $12,300.

http://carefordescientific.com/xrf-...200l-89401-238-platinum-crucible-200-ml-each/

 

[Andrew O'Neill]

Wow.

 

[Luckless]

Well it is very much possible to create fine optical quality glass in a small shop. I know this because people actually DO it. I have helped make several small pieces of glass from scratch in the past, some of which has been made into small lenses, and have a tiny compound loupe in a box somewhere that was all hand made.

Difficult does not mean impossible, and should not mean that an attempt is never made. Constantly telling people "That's too hard, just leave it to the professionals" is not a sensible way to expand the arts or sciences.

 

[Nodda Duma]

Your opinion has been noted.

 

[removed account4]

"That's too hard, just leave it to the professionals"

before george eastman invented democratic photography ( the KODAK box camera ) that is what they said about photography
its also what they used to say about making emulson from scratch, camera building and a whole lot of other things.
it is also similar to the " shrouded in difficulty and mystery " = >>> large format photography ... its just what's said ..
and just because its said doesn't mean its true as you have noted...