I had no problem opening the document.
The zemax demo is crippleware. It wont even let you modify a lens. Uninstalling also made my system continuously reboot. Took me a couple hours to repair from the WinXP CD. Be wary. Oslo demo seems complicated but good. So much cryptic stuff to play with. This has a learning curve that will put many off.
Rox, the website has a bunch of tutorials for the demo program. This is like open source development 😀
I hope Ace and Guy are also following this thread for their valued input.
The zemax demo is crippleware. It wont even let you modify a lens. Uninstalling also made my system continuously reboot. Took me a couple hours to repair from the WinXP CD. Be wary. Oslo demo seems complicated but good. So much cryptic stuff to play with. This has a learning curve that will put many off.
Rox, the website has a bunch of tutorials for the demo program. This is like open source development 😀
I hope Ace and Guy are also following this thread for their valued input.
Rox,
The triplet that I did was just a smaller pre optimized demo file which I scaled to a larger FL. And then optimized with default values. OSLO is powerful software but hard to use and the user manual is poor. You could enter the data from my lens design as a starting point. Any surface geometry or lens spacing can be defined as a variable or a constant and you need to leave nearly everything a variable except the conjugate distances to the image and object plane. Then the software can optimize the lens system. OSLO is much harder to set up and learn to use and has inferior graphic rendering but in the right hands it is powerful software.
In Zemax the object height is entered as a global parameter and the rays traced at user defined intervals out to the object size. I gave up on OSLO as it was too cryptic and hard to use. One thing to bear in mind is that sometimes the software uses the photographic convention where the object is the physical thing outside and the image is the film inside the camera. With us the LCD is the object and the screen is the image. Just be aware as sometimes the orientation is confusing.
gguertin145,
building the lenses from existing lenses is a good approach but be aware that there is no room for deviation without degrading the lens. All of the lenses need to be reground to match the prescription. You can't just change one surface. In fact you can't change anything without reoptimizing the lens. I don't know why your document is not working. Maybe you need a better unzip utility. If you are certain that your company can build this lens to exact spec then have a bit of patience and I will try to design a perfect no compromise lens.
Hezz
The triplet that I did was just a smaller pre optimized demo file which I scaled to a larger FL. And then optimized with default values. OSLO is powerful software but hard to use and the user manual is poor. You could enter the data from my lens design as a starting point. Any surface geometry or lens spacing can be defined as a variable or a constant and you need to leave nearly everything a variable except the conjugate distances to the image and object plane. Then the software can optimize the lens system. OSLO is much harder to set up and learn to use and has inferior graphic rendering but in the right hands it is powerful software.
In Zemax the object height is entered as a global parameter and the rays traced at user defined intervals out to the object size. I gave up on OSLO as it was too cryptic and hard to use. One thing to bear in mind is that sometimes the software uses the photographic convention where the object is the physical thing outside and the image is the film inside the camera. With us the LCD is the object and the screen is the image. Just be aware as sometimes the orientation is confusing.
gguertin145,
building the lenses from existing lenses is a good approach but be aware that there is no room for deviation without degrading the lens. All of the lenses need to be reground to match the prescription. You can't just change one surface. In fact you can't change anything without reoptimizing the lens. I don't know why your document is not working. Maybe you need a better unzip utility. If you are certain that your company can build this lens to exact spec then have a bit of patience and I will try to design a perfect no compromise lens.
Hezz
Me2!,
The Zemax demo is very bad indeed but the Zemax real software is very good indeed. I have access to an engineering computer with the real stuff installed. Not a demo.
Hezz
The Zemax demo is very bad indeed but the Zemax real software is very good indeed. I have access to an engineering computer with the real stuff installed. Not a demo.
Hezz
i have done it as well with a demo triplet on oslo and the shapes are not compatible with yours, thats why i would like to check the spot sizes on both cases to check whats better.
ok problem we were shooting for a 80mm diam not 135? I am not sure actually pretty sure I cant get 135? rox should I try and get these specs with 80 for a starting point?
also the scaling to 450mm is a good choice but in my case, the resulting diameters aren´t as big as yours (171 diameter, mines are 120mm diameter).
in your original triplet, is it optimized already, but if you change one surface, the rest will be recalculated for you? (i mean are the relations still there?)
in your original triplet, is it optimized already, but if you change one surface, the rest will be recalculated for you? (i mean are the relations still there?)
yes querting, thats the problem i see as well. Resizing the largest diameter to 80mm wont work properly (the resulting focal would be 210mm).
If we just keep the shapes and the thicknesess as well as the lens distancements and just manufacture them 3 to 80mm diameter, then some specs will change; the field angle will decrease, the f-stop will change... (the focal lengh will be 450, I think it will be a good starting point yet)
just make shure that resizing the large diameter lens to 80mm the central thicknesses remain equal (this will give you heavyer lenses for the first and third lens but if you change the thickensses, evarithing needs to be recalculated so forget about it by now).
If we just keep the shapes and the thicknesess as well as the lens distancements and just manufacture them 3 to 80mm diameter, then some specs will change; the field angle will decrease, the f-stop will change... (the focal lengh will be 450, I think it will be a good starting point yet)
just make shure that resizing the large diameter lens to 80mm the central thicknesses remain equal (this will give you heavyer lenses for the first and third lens but if you change the thickensses, evarithing needs to be recalculated so forget about it by now).
I was just estimating the diameters from memory. They may be more in the order of 120 mm.
IF you are using OSLO lite or student version it has some good things about it in that it is not disabled software. Only limited to about 10 - 12 surfaces.
THere is a very useful software called LensVIEW. It is essentially a library of about 3000 patented lens designs that can write each design to ZEMAX or OSLO or other high end lens design software. THe beauty is that there is a fully working demo version that only has about 200 designs. You should try to download it. It is very instructive to open up a lens in the software and then see it's performance when imported into OSLO or ZEMAX or Code V. For some reason there is a bug in the software that prevents it working until you close it and open it a second time. I would recommend trying to find it.
ROX,
If you make a small change you have to reoptimize it again but if it was small and the optimization parameters were the same then it only takes a few seconds to reoptimize.
OSLO and Zemax may use somewhat different optimizing methods. ALso there could be some small difference with our setups. I will post a spot diagram of the lens shortly but right now I have to go to work. Talk to you soon.
Hezz
IF you are using OSLO lite or student version it has some good things about it in that it is not disabled software. Only limited to about 10 - 12 surfaces.
THere is a very useful software called LensVIEW. It is essentially a library of about 3000 patented lens designs that can write each design to ZEMAX or OSLO or other high end lens design software. THe beauty is that there is a fully working demo version that only has about 200 designs. You should try to download it. It is very instructive to open up a lens in the software and then see it's performance when imported into OSLO or ZEMAX or Code V. For some reason there is a bug in the software that prevents it working until you close it and open it a second time. I would recommend trying to find it.
ROX,
If you make a small change you have to reoptimize it again but if it was small and the optimization parameters were the same then it only takes a few seconds to reoptimize.
OSLO and Zemax may use somewhat different optimizing methods. ALso there could be some small difference with our setups. I will post a spot diagram of the lens shortly but right now I have to go to work. Talk to you soon.
Hezz
i don´t know what oslo version is mine but it has limited functions as well. (spotsize calculator, crhomatic aberrations functions.... are all desactivated).
I don´t understand what do you mean by the diameters were closer to 120mm. Yours are 171 / 90 / 135. if you just resize the triplet to 450 focal, then you can´t use smaller diameter lenses there.
quertin needs a triplet with a 80mm largest lens and 450 mm EFL triplet. We have aproblem there right now. (I would recomend just to cut the diameters down to 80mm by now but keeping everithing else as is, the field angle will decrease but we can see if the central part works fine)
I don´t understand what do you mean by the diameters were closer to 120mm. Yours are 171 / 90 / 135. if you just resize the triplet to 450 focal, then you can´t use smaller diameter lenses there.
quertin needs a triplet with a 80mm largest lens and 450 mm EFL triplet. We have aproblem there right now. (I would recomend just to cut the diameters down to 80mm by now but keeping everithing else as is, the field angle will decrease but we can see if the central part works fine)
rox can you email me exactly what you think I should get? which lenses and what they should be? I will bring it in and see if this is even possible before anyone spends any more time on it. Remember I really dont understand this stuff and I dont want to get the wrong lens right off the bat because of my own error not knowing what we were trying to get. I saw a pic in your post at school for somereason it isnt showing now at home? so if you could email that too that would be great its gguertin145@hotmail.com thanks
Rox,
you are correct about the lens sizes. I guess my memory is less good than I thought. As long as the curvatures and spacings and lens thicknesses stay the same the 80 mm diameter should work with reduced FOV.
Here is the spot diagram for my lens but remember that it uses infinite conjugates on the long throw side so it will not actually perform this good in reality. I still have not figured out how to optimize a better or as good performance with the exact conjugates. The long side should be about 12 feet instead of infinite.
I have not worked on any design for a year or more and have forgotten much about the software. Maybe this summer we can come up with some really high end exact purpose designs.
I was using OSLO LT when I was trying it. It was fully functional package except for only 10 -12 surfaces. This use to be the demo free version but is now the student or low priced version. As the demo version has more functions disabled.
I will see if I have the self installing file somewhere on an archived disk and if I do I can send it to you.
If you want to e-mail me your address I will send you a disk with some goodies on it.
Hezz
you are correct about the lens sizes. I guess my memory is less good than I thought. As long as the curvatures and spacings and lens thicknesses stay the same the 80 mm diameter should work with reduced FOV.
Here is the spot diagram for my lens but remember that it uses infinite conjugates on the long throw side so it will not actually perform this good in reality. I still have not figured out how to optimize a better or as good performance with the exact conjugates. The long side should be about 12 feet instead of infinite.
I have not worked on any design for a year or more and have forgotten much about the software. Maybe this summer we can come up with some really high end exact purpose designs.
I was using OSLO LT when I was trying it. It was fully functional package except for only 10 -12 surfaces. This use to be the demo free version but is now the student or low priced version. As the demo version has more functions disabled.
I will see if I have the self installing file somewhere on an archived disk and if I do I can send it to you.
If you want to e-mail me your address I will send you a disk with some goodies on it.
Hezz
Attachments
Rox,
I have the Oslo LT on an old archived disk. It is about 6.5 MB in size. If you can take at attachment that size I can e-mail it to you.
I think this version has spot diagrams if I am not mistaken.
Hezz
I have the Oslo LT on an old archived disk. It is about 6.5 MB in size. If you can take at attachment that size I can e-mail it to you.
I think this version has spot diagrams if I am not mistaken.
Hezz
sorry it was a confusion there, i can as well draw spot diagrams but it is not what i am triyng to do.
what i would like to chek is error functions (how spot size changes with diferent shapes) are all desactivated on my estudent oslo edu version 6.3.0.
the solution would be to test every shapes manualy and check the spot sizes but it is a hard work.
i see what you mean by conjugate distances. I have to check this on my oslo, all my tests rays are coming from infinite as well.
Hezz did you changed anyparameter on the original lens before/after scaling? (my oslo triplet model is too much diferent...)
what i would like to chek is error functions (how spot size changes with diferent shapes) are all desactivated on my estudent oslo edu version 6.3.0.
the solution would be to test every shapes manualy and check the spot sizes but it is a hard work.
i see what you mean by conjugate distances. I have to check this on my oslo, all my tests rays are coming from infinite as well.
Hezz did you changed anyparameter on the original lens before/after scaling? (my oslo triplet model is too much diferent...)
Rox,
let me know if you want to try this Oslo LT version that I have. I have a few questions myself about the spot diagrams. I'm wondering what the original size of the spot was before the photons travel through the lens. It is easy to see the relative changes at different angles and how they scatter the different wavelengths. But what is not so obvious to me is the relative change from the original unfocused spot. Or I will call it the spot object for lack of a better term.
Also, I made some minor changes as I believe in a attempt to make the lens edges thicker because I though that I might have to eventually make the lens by hand and if the lens edges are too thin it is easy to break them when grinding lenses by hand. So it is possible my lenses are a little thicker and wider and have slightly different geometry.
Gguertin145,
I will send you the Oslo LT in an attachment soon.
Hezz
let me know if you want to try this Oslo LT version that I have. I have a few questions myself about the spot diagrams. I'm wondering what the original size of the spot was before the photons travel through the lens. It is easy to see the relative changes at different angles and how they scatter the different wavelengths. But what is not so obvious to me is the relative change from the original unfocused spot. Or I will call it the spot object for lack of a better term.
Also, I made some minor changes as I believe in a attempt to make the lens edges thicker because I though that I might have to eventually make the lens by hand and if the lens edges are too thin it is easy to break them when grinding lenses by hand. So it is possible my lenses are a little thicker and wider and have slightly different geometry.
Gguertin145,
I will send you the Oslo LT in an attachment soon.
Hezz
gguertin145,
I sent e-mail with attachment but it wouldn't go through. Most likely it was intercepted as a virus since the attachment was an executable file.
Hezz
I sent e-mail with attachment but it wouldn't go through. Most likely it was intercepted as a virus since the attachment was an executable file.
Hezz
when you check the spot size, the incoming spot is infinite (paralell rays) so when we work out how to use conjugate distances, we need to determine object spot size as well i guess. Im yet on it...
gguertin145,
Ok I'll try that. My CAD computer and internet computer are at different locations and CAD computer is not internet connected for data security reasons. So I can't do it right now but I will do it soon.
Rox,
I read a chapter in an optical textbook about the spot diagrams but that was a year of more ago and I was thinking that the spot was a small patch of light that kind of represents a small bundle of rays. It is possible to optimize the lens system by treating light as a ray or a wave in the software. But the ray method should work best for what we are doing. There are some instances were the wave method is prefered but I am not sure under what circumstances this is best.
It seems that I read that the ray method is preferred for wide band visible spectrum situations.
Hezz
Ok I'll try that. My CAD computer and internet computer are at different locations and CAD computer is not internet connected for data security reasons. So I can't do it right now but I will do it soon.
Rox,
I read a chapter in an optical textbook about the spot diagrams but that was a year of more ago and I was thinking that the spot was a small patch of light that kind of represents a small bundle of rays. It is possible to optimize the lens system by treating light as a ray or a wave in the software. But the ray method should work best for what we are doing. There are some instances were the wave method is prefered but I am not sure under what circumstances this is best.
It seems that I read that the ray method is preferred for wide band visible spectrum situations.
Hezz
you are right, the conjugate setup does show diferent spot sizes to my non conjugate optimized triplet.
I have done some test on it and found some interesting points. The object has no spot size, it is infinitesimal (light coming from 0 radius point). But i can define any size object (lcd) and see how works it throw the lens.
Anyway, your custom triplet model at oslo does show very bad results (it could be something done wrong by me anyway). the spot sizes are huge on my tests.
Could you tell me what beams are entering your triplet to give you those spots you posted? ( angle and aperture of the beam)
I have done some test on it and found some interesting points. The object has no spot size, it is infinitesimal (light coming from 0 radius point). But i can define any size object (lcd) and see how works it throw the lens.
Anyway, your custom triplet model at oslo does show very bad results (it could be something done wrong by me anyway). the spot sizes are huge on my tests.
Could you tell me what beams are entering your triplet to give you those spots you posted? ( angle and aperture of the beam)
just found this thread
I don't think you can make an optimized triplet from eyeglass blanks. For one thing, they all have an 89.5 mm radius concave surface on the back side, (facing the eye), that is optimal for linking to human vision. That takes away three of your six adjustable surfaces.
The other problem is the glass types: The positive lenses have to be a crown-like glass, and the negative lens has to be flint-like. The two types of glass have different dispersion (different refractive index for different colors). I think eyeglass lenses are all crown-like. Lots of the flint-like glasses are made from materials that are soft or easily damaged by water, so they don't make durable eyeglasses.
I downloaded OSLO EDU 6.3.0 some while ago. It can do everything you need to analyze a lens, and you can enter a triplet or tessar design, but it does not do automatic optimization. That means you have to try lots of little steps to see if the performance gets better or worse. It can do a whole bunch of different analyses, (most of which I don't understand!). So I have just been looking at the average spot size, since that sums up all the different aberrations and distortions.
I rescaled their Cooke triplet example up to 125 mm aperature and 450 mm focal length, with a 17" object. That gives you a spec for a triplet that would probably work okay for projection. But to build it you would need to start with the proper types of glass blanks, and then grind and polish all three lenses. You will not find these exact lenses all ready made. And the optimization is so exact that changing any of the curves, spacing, or materials throws it way off.
If you can find a copy of Telescope Optics by Rutten & van Venrooij, it has a chapter on optimizing a triplet that is useful. They discuss what each lens surface is correcting, but of course they all interact.
I don't think you can make an optimized triplet from eyeglass blanks. For one thing, they all have an 89.5 mm radius concave surface on the back side, (facing the eye), that is optimal for linking to human vision. That takes away three of your six adjustable surfaces.
The other problem is the glass types: The positive lenses have to be a crown-like glass, and the negative lens has to be flint-like. The two types of glass have different dispersion (different refractive index for different colors). I think eyeglass lenses are all crown-like. Lots of the flint-like glasses are made from materials that are soft or easily damaged by water, so they don't make durable eyeglasses.
I downloaded OSLO EDU 6.3.0 some while ago. It can do everything you need to analyze a lens, and you can enter a triplet or tessar design, but it does not do automatic optimization. That means you have to try lots of little steps to see if the performance gets better or worse. It can do a whole bunch of different analyses, (most of which I don't understand!). So I have just been looking at the average spot size, since that sums up all the different aberrations and distortions.
I rescaled their Cooke triplet example up to 125 mm aperature and 450 mm focal length, with a 17" object. That gives you a spec for a triplet that would probably work okay for projection. But to build it you would need to start with the proper types of glass blanks, and then grind and polish all three lenses. You will not find these exact lenses all ready made. And the optimization is so exact that changing any of the curves, spacing, or materials throws it way off.
If you can find a copy of Telescope Optics by Rutten & van Venrooij, it has a chapter on optimizing a triplet that is useful. They discuss what each lens surface is correcting, but of course they all interact.
- Status
- Not open for further replies.