i agree with you guy, the lens should be custom made and this wold be very expensive.
about upscaling the triplet to a new focal, it is a good idea but this lens was optimized for paralell rays only. If you play with conjugate distances, (place the source beam at 300cm for instance) you will see the aberrations (spot size) will increase.
I said to bring source light to somewhere 3Meters, thats because the lens is being used that way, the focused point would be the lcd and the source beam would be the projected image (like pjotography objetives).
I am doing some litle changes on the shapes to check the spot sizes, but it is a hard work. It is a powerfull tool but should hard to understood.
about upscaling the triplet to a new focal, it is a good idea but this lens was optimized for paralell rays only. If you play with conjugate distances, (place the source beam at 300cm for instance) you will see the aberrations (spot size) will increase.
I said to bring source light to somewhere 3Meters, thats because the lens is being used that way, the focused point would be the lcd and the source beam would be the projected image (like pjotography objetives).
I am doing some litle changes on the shapes to check the spot sizes, but it is a hard work. It is a powerfull tool but should hard to understood.
I have contacts in China that can make a custom lens. The lens itself is not too expensive but the custom pattern must be CNC milled to make a master for each surface. The equipment will follow the tooling pattern. Several grinding stages must be tooled. Its significant that aspherics can be made for similar cost to sphericals. This used to be much much more expensive. Tooling will be a few thousand so I would rather go with 2 aspherics than cutting patterns for a triplet. Aspherics would be far superior. I need a lens design that will cover future 17" LCDs to make this worth the cost. Tooling will be spread over expected sales of 300 sets. I'm hoping to get an exceptional lens for about $100-150. If i have to go to a design house the cost may sink the deal. I dont expect people to be interested if it gets too pricey. I have to find a good design.
Me2!,
That is totally exellent. Now if we can get together with a few knowledgable guys.
In a few months I will be able to do some CNC molds at school but you would need to get in touch with your lens builders because there are some manufacturing issues here. In fact I could do one now but the machine I am presently working on is not the best choice.
In some cases the molds are finished to very high standards to the point where only a little polishing needs to be done. This type of mold quality requires special CNC machines that are not common and a lot of hand finishing. We would need to know the finish requirements of your manufacturer. In some cases the molds may only have to be good enough to compression mold a glass blank to approximate size so only a little material needs to be ground off.
Aspherical surfaces puts the whole operation out of the capacity of ordinary CNC machines and requires special high precision machines that are specifically designed for lens mold making or lens grinding.
However, it would be possible with spherical surfaces to use common CNC machines and hand finish the final molds. They could then be shipped to your China company. They would have to set up the final grinding and polishing tooling.
We would need to have the specs for the total mold dimensions because the outside of the mold has to interface with the moulding machine. We would also need to put together a pool of funds to cover the material costs for the molds as they would be expensive if made from stainless steel. Or even a lesser material.
Rox and Guy,
Zemax has the ability to do computer optimization according to a prescribed set of parameters. In some ways it is far more advanced than Oslo. Like I said if you guys want to work more on this e-mail me your address and I will send you a CD with some goodies on it. I would like to get a handful of more knowledgable and expert DIY's together to try and design the best lens possible. Then if we can pool a few funds to make the molds and send them to China where they can build the lenses we can have some good lenses.
Hezz
That is totally exellent. Now if we can get together with a few knowledgable guys.
In a few months I will be able to do some CNC molds at school but you would need to get in touch with your lens builders because there are some manufacturing issues here. In fact I could do one now but the machine I am presently working on is not the best choice.
In some cases the molds are finished to very high standards to the point where only a little polishing needs to be done. This type of mold quality requires special CNC machines that are not common and a lot of hand finishing. We would need to know the finish requirements of your manufacturer. In some cases the molds may only have to be good enough to compression mold a glass blank to approximate size so only a little material needs to be ground off.
Aspherical surfaces puts the whole operation out of the capacity of ordinary CNC machines and requires special high precision machines that are specifically designed for lens mold making or lens grinding.
However, it would be possible with spherical surfaces to use common CNC machines and hand finish the final molds. They could then be shipped to your China company. They would have to set up the final grinding and polishing tooling.
We would need to have the specs for the total mold dimensions because the outside of the mold has to interface with the moulding machine. We would also need to put together a pool of funds to cover the material costs for the molds as they would be expensive if made from stainless steel. Or even a lesser material.
Rox and Guy,
Zemax has the ability to do computer optimization according to a prescribed set of parameters. In some ways it is far more advanced than Oslo. Like I said if you guys want to work more on this e-mail me your address and I will send you a CD with some goodies on it. I would like to get a handful of more knowledgable and expert DIY's together to try and design the best lens possible. Then if we can pool a few funds to make the molds and send them to China where they can build the lenses we can have some good lenses.
Hezz
I dont know what equipment they use. I dont think they expect someone to make the mold for them? Im not sure how much of the "tooling cost" is an attempt to get money off me up front. Our conversation was that they wanted a drawing to work out an exact price. They wanted specs and money to provide lenses.
guy would polycarbonite be a good lens material to work with? I can special order lenses also so I may be able to get different kinds of lenses, They are checking to see the larges diameter we can get at my lab its 80mm but I may be able to get bigger from another lab.
Me2!,
Perhaps I was jumping the gun and it would be best if they did all tooling themselfs because they are intimately aware of what they need to do with the machines they have. But the reality is that the tooling costs are the main costs. The glass is not that expensive and grinding only becomes really expensive if it takes a long time because you have big lenses or are removing a lot of material.
The best bet is if we can give them a good design and then someone that has a little cash can flip the bill and then make some profit selling a few hundred lenses to us DIY's.
gguertin145,
In regards to polycarb. It is a good optical material but plastic lens systems require more complexity to correct because there are fewer kinds of optical grade plastic with differing dispursion characteristics. There are only probably less than a dozen kinds of optical grade plastic but hundreds of glass formulations. Plastic while easier to work with is much harder to design for. Also plastic is not manufactured specifically for lenses except for a couple of kinds for making eyeglass lenses. Therefore there will never be enough special formulations for optics. Getting optical information for platics is difficult.
Hezz
Perhaps I was jumping the gun and it would be best if they did all tooling themselfs because they are intimately aware of what they need to do with the machines they have. But the reality is that the tooling costs are the main costs. The glass is not that expensive and grinding only becomes really expensive if it takes a long time because you have big lenses or are removing a lot of material.
The best bet is if we can give them a good design and then someone that has a little cash can flip the bill and then make some profit selling a few hundred lenses to us DIY's.
gguertin145,
In regards to polycarb. It is a good optical material but plastic lens systems require more complexity to correct because there are fewer kinds of optical grade plastic with differing dispursion characteristics. There are only probably less than a dozen kinds of optical grade plastic but hundreds of glass formulations. Plastic while easier to work with is much harder to design for. Also plastic is not manufactured specifically for lenses except for a couple of kinds for making eyeglass lenses. Therefore there will never be enough special formulations for optics. Getting optical information for platics is difficult.
Hezz
I have been put in contact with a design house. I guess it comes down to whether i have the cahones to sink my retirement savings over this. I'll find out how much they want.
Yes, they know their equipment and i cant imagine sending them something made with a dremel LOL That would be embarrasing. This is a fair sized company and they know their stuff. I have no doubt they will deliver a quality product.
Yes, they know their equipment and i cant imagine sending them something made with a dremel LOL That would be embarrasing. This is a fair sized company and they know their stuff. I have no doubt they will deliver a quality product.
Trev,
not just any old 5 axis mill will do because the spindles in typical production mills have too much vibration. THe machines have to have special air bearing ways and air spindles with ultra low vibration. At least for aspherics. With spherical molds the mold can be hand finished.
Hezz
not just any old 5 axis mill will do because the spindles in typical production mills have too much vibration. THe machines have to have special air bearing ways and air spindles with ultra low vibration. At least for aspherics. With spherical molds the mold can be hand finished.
Hezz
still on eyeglass blanks
>would polycarbonite be a good lens material to work with?
Polycarbonate is fine, but it still won't get you an optimized triplet. If you want to see how good (or bad) an eyeglass blank-based projection lens will be, get two ordinary CR39 or glass 80 mm diameter +1.0 Diopter lenses. Get them AR coated with the appropriate coating for their material.
Then put them in a tube about 40 mm apart with the convex sides facing out. This will give you an un-optimized symmetric duplet, with a focal length around 510 mm, and a pretty wide field angle. Try it and you will see what I mean when I say: Good enough for video, but not for anything else. I have one of these with acrylic eyeglass lenses. I can see most of the pixels on my XGA LCD, but the corners are not focussed very well.
Lots of people are making projectors for video viewing with LCDs that have actual resolutions like 320 by 244 pixels. They seem happy with their results. An eyeglass duplet would be fine for such a system.
If you really want to see all 1024 by 768, or 1280 by 800 pixels sharply on your screen, then you might want a better lens. Hezz is exactly right: You have to make optimized triplets out of special glass lenses.
>would polycarbonite be a good lens material to work with?
Polycarbonate is fine, but it still won't get you an optimized triplet. If you want to see how good (or bad) an eyeglass blank-based projection lens will be, get two ordinary CR39 or glass 80 mm diameter +1.0 Diopter lenses. Get them AR coated with the appropriate coating for their material.
Then put them in a tube about 40 mm apart with the convex sides facing out. This will give you an un-optimized symmetric duplet, with a focal length around 510 mm, and a pretty wide field angle. Try it and you will see what I mean when I say: Good enough for video, but not for anything else. I have one of these with acrylic eyeglass lenses. I can see most of the pixels on my XGA LCD, but the corners are not focussed very well.
Lots of people are making projectors for video viewing with LCDs that have actual resolutions like 320 by 244 pixels. They seem happy with their results. An eyeglass duplet would be fine for such a system.
If you really want to see all 1024 by 768, or 1280 by 800 pixels sharply on your screen, then you might want a better lens. Hezz is exactly right: You have to make optimized triplets out of special glass lenses.
well it doesnt sound like something I will have available so where would you order such lenses or are they only going to be custom made? why cant you find a lens that is available and work around that, not that we should lose quality but you could make the other 2 lenses based on the first lens or whatever so that we dont need special lenses? Just seems like there has to be a place that makes the lenses we need even if it is for other applications. I am going to try the duplet but will standard glass the cr39 or poly be the best material to work with.
Hezz said:
Ya i gatherd that but it doesnt have to be as acurate as a grinding machine as the lenses will be ground after and pollished 😉.
Trev🙂
Speaking of plastic lenses I know that it is possible to design a high quality lens in plastic because the popular Fujinon CRT lens that some where using a while ago is a plastic lens I believe. IT actually has very good resolution and it will not work for our application but it does seem to show that a good plastic lens can be built.
You may want to check out this thread as I designed a plastic four element lens based on a tessar a year or so ago. It has good resolution though not as good as the glass triplet. It has some other problems though in that it has a lower FOV and some barrel distortion. THe rays traces with an actual jpeg image are cool though.
I wonder with more time if a better plastic design can be made. This was optimized with infinite conjugates on the long throw side also.
Hezz
http://www.diyaudio.com/forums/showthread.php?s=&threadid=28052&perpage=10&highlight=&pagenumber=4
You may want to check out this thread as I designed a plastic four element lens based on a tessar a year or so ago. It has good resolution though not as good as the glass triplet. It has some other problems though in that it has a lower FOV and some barrel distortion. THe rays traces with an actual jpeg image are cool though.
I wonder with more time if a better plastic design can be made. This was optimized with infinite conjugates on the long throw side also.
Hezz
http://www.diyaudio.com/forums/showthread.php?s=&threadid=28052&perpage=10&highlight=&pagenumber=4
I also think a plastic lens would work if we could find the right material for the middle element which I doubt we can do. I asked today and if I can find lens blanks they will generate them now that I have the shapes and focals that I didnt have the otherday but where do I buy that type of lens material. so I can still have them generated and ar coated just need the lenses. Someone looked at the LL triplet today and said it just looked like our standard glass? I will talk to someone higher up tomorrow and see what they say about getting different types.
Instead of building a triplet from scratch would it be easier to replace one or two elements in an existing triplet (or adding a 4th element) to get the desired results?
Like taking the 80mm standard triplet lens and adding another element after to give it a longer throw?
Like taking the 80mm standard triplet lens and adding another element after to give it a longer throw?
replacing elements in triplet
No, that would throw off all the corrections. It would be very easy to get a specific longer or shorter focal length, but you would lose all the advantages of the original triplet in terms of corrected aberrations and distortions.
What you CAN do is to constrain a new triplet design to use a particular flint glass negative center element that you found at a surplus price. Then you have the lens design program see how good it can do with various positive crown glass positive elements for the outer lenses. That way you just have to get some relatively cheap crown glass positive lenses made.
You would have to plug in all of the surplus negative flint lenses you could find, to see which could work the best.
BTW: You MUST use a combination of materials with different dispersions (ie. flint-like glass and crown-like glass), if you want to get low chromatic aberration. Using all the same glass (or plastic) will give you bad rainbow effects close to the edges. This looks horrible!
No, that would throw off all the corrections. It would be very easy to get a specific longer or shorter focal length, but you would lose all the advantages of the original triplet in terms of corrected aberrations and distortions.
What you CAN do is to constrain a new triplet design to use a particular flint glass negative center element that you found at a surplus price. Then you have the lens design program see how good it can do with various positive crown glass positive elements for the outer lenses. That way you just have to get some relatively cheap crown glass positive lenses made.
You would have to plug in all of the surplus negative flint lenses you could find, to see which could work the best.
BTW: You MUST use a combination of materials with different dispersions (ie. flint-like glass and crown-like glass), if you want to get low chromatic aberration. Using all the same glass (or plastic) will give you bad rainbow effects close to the edges. This looks horrible!
Ya,
What Guy said is exactly true.
Regarding plastic lenses. The problem with plastic lenses is that although they are easier and cheaper to manufacture there are two problems. The first is that there are few types of plastic with the right optical characteristics and second because there are so many varying formulations of plastic there are no set standards for optical characteristics. With optical glass types the formulations are standardized and for the most part very uniform in the optical characteristics compared to plastic.
Because of the limited optical plastic issues a good performing plastic lens usually needs four to six elements. Plus you would have to have optical measurements on the exact plastics you are using.
Seriously, I think it would be easier to make a glass triplet lens by hand and have the lenses coated at the lab. It would be time consuming but effective.
gguertin145,
You ought to see if your lab has any of the old style of lens grinding machines out in a back room or something. There might be an unused machine that a tech can set up and grind you the lenses. These machines are more likely to be able to do larger size lenses than the newer production CNC lens grinding machines.
Hezz
What Guy said is exactly true.
Regarding plastic lenses. The problem with plastic lenses is that although they are easier and cheaper to manufacture there are two problems. The first is that there are few types of plastic with the right optical characteristics and second because there are so many varying formulations of plastic there are no set standards for optical characteristics. With optical glass types the formulations are standardized and for the most part very uniform in the optical characteristics compared to plastic.
Because of the limited optical plastic issues a good performing plastic lens usually needs four to six elements. Plus you would have to have optical measurements on the exact plastics you are using.
Seriously, I think it would be easier to make a glass triplet lens by hand and have the lenses coated at the lab. It would be time consuming but effective.
gguertin145,
You ought to see if your lab has any of the old style of lens grinding machines out in a back room or something. There might be an unused machine that a tech can set up and grind you the lenses. These machines are more likely to be able to do larger size lenses than the newer production CNC lens grinding machines.
Hezz
I have access to cheap single element overhead projector lenses.
They are 320mm focal length. Using the psone I find the focal needs to be shorter.
Acording to this: http://www.lightshow.cc/explorer/Calculators/Lens_Designer/lens_designer.html
I can make a 290mm duplet by spacing the singlets 287mm in a tube. What do you guys think. Will it work? Should I just use a 287mm pvc tube and glue the singles on either end?
Remember these lenses were designed for the large overhead projector image size. So if I am asking it to focus only 3 x 4 inch image can I assume I will get corner to corner focus? What about chromatic aberations? Or other artifacts?
They are 320mm focal length. Using the psone I find the focal needs to be shorter.
Acording to this: http://www.lightshow.cc/explorer/Calculators/Lens_Designer/lens_designer.html
I can make a 290mm duplet by spacing the singlets 287mm in a tube. What do you guys think. Will it work? Should I just use a 287mm pvc tube and glue the singles on either end?
Remember these lenses were designed for the large overhead projector image size. So if I am asking it to focus only 3 x 4 inch image can I assume I will get corner to corner focus? What about chromatic aberations? Or other artifacts?
way too long!
That duplet would be so long that it would have a very narrow field of view. Light from the corners of a 5" LCD would not get through to the screen, unless you are talking about using 5" diameter lenses.
A good symmetric duplet design should be not much longer than the diameter of the lenses, if you want to have enough field angle to cover a whole LCD. Start with individual lenses that are about twice the focal length that you want for the whole lens.
Or just buy one of these for $7.50 US:
http://www.surplusshed.com/pages/item/l3334.html
That is a 290 mm fl Fujinon copy lens (probably a tessar) that will be much much better than anything you can assemble yourself.
That duplet would be so long that it would have a very narrow field of view. Light from the corners of a 5" LCD would not get through to the screen, unless you are talking about using 5" diameter lenses.
A good symmetric duplet design should be not much longer than the diameter of the lenses, if you want to have enough field angle to cover a whole LCD. Start with individual lenses that are about twice the focal length that you want for the whole lens.
Or just buy one of these for $7.50 US:
http://www.surplusshed.com/pages/item/l3334.html
That is a 290 mm fl Fujinon copy lens (probably a tessar) that will be much much better than anything you can assemble yourself.
- Status
- Not open for further replies.