There has been a lot of talk about using cold mirrors to remove heat from the light path, but nobody has actually described how to implement it using small cold mirrors.
The only way I can think of designing it is to use two condensers. I’ve attached a drawing that might help explain it.
The design uses two condensers with very short focal lengths (20-30mm). The first condenser bends the rays to be parallel allowing all of the light to strike the cold mirror at 45 degrees for maximum IR removal the second condenser converges the rays to a new point source for the Fresnel to focus on. Although it may be possible do with out the second condenser on a smaller LCD, I can’t see a way to do it on larger LCD, as the cost of the cold mirror may become expensive.
I would be interested if anyone else has any ideas or any improvements
DJ
The only way I can think of designing it is to use two condensers. I’ve attached a drawing that might help explain it.
The design uses two condensers with very short focal lengths (20-30mm). The first condenser bends the rays to be parallel allowing all of the light to strike the cold mirror at 45 degrees for maximum IR removal the second condenser converges the rays to a new point source for the Fresnel to focus on. Although it may be possible do with out the second condenser on a smaller LCD, I can’t see a way to do it on larger LCD, as the cost of the cold mirror may become expensive.
I would be interested if anyone else has any ideas or any improvements
DJ
Hello faithblinded
Thanks for your input. Yes you can tile the mirrors but I can’t see how to make it work without the lamp and condenser getting in the way of the light path. You possibly could split the Fresnel and put the mirror in between them but that makes the mirror and the pj allot bigger.
Or maybe I’m missing something?
DJ
Thanks for your input. Yes you can tile the mirrors but I can’t see how to make it work without the lamp and condenser getting in the way of the light path. You possibly could split the Fresnel and put the mirror in between them but that makes the mirror and the pj allot bigger.
Or maybe I’m missing something?
DJ
Attachments
cold mirror usage
I believe that this idea has been implemented before. According to this picture, we only need to use a small cold mirror. Place the cold mirror near the focal point of the condenser and all the IR light will be filter out completely. The Heatsinker will absorb the IR light and the fan will help to cool the Heatsinker faster.
In the picture, I did not have access to a CAD progrom so I was unable to calculate all the lightrays. But placment of the Coldmirror at a focal point is possible.
I believe that this idea has been implemented before. According to this picture, we only need to use a small cold mirror. Place the cold mirror near the focal point of the condenser and all the IR light will be filter out completely. The Heatsinker will absorb the IR light and the fan will help to cool the Heatsinker faster.
In the picture, I did not have access to a CAD progrom so I was unable to calculate all the lightrays. But placment of the Coldmirror at a focal point is possible.
Attachments
Why not use a hot mirror?
It is the same think but it reflects the ir back.
Also if you do not want to warm up the lamp a lot you can use a 45 degrees hot mirror.The light goes through with 45 degrees angle and the ir reflects to the side.I use an 0 degrees hot mirror and my lamp does not seem to get more hot than without the hot mirror.i use this setup more than a year now with no problem.Beleive me even with the 400w hqi-ts the lcd is very cool with this mirror.It also has a uv filter.it is 90mmx90mm square and i bought it from a 3m supplier for 146 euros last year.it can work upto 500 celcious temperature!!!!
Try edmundoptics.com
http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=1492
It is the same think but it reflects the ir back.
Also if you do not want to warm up the lamp a lot you can use a 45 degrees hot mirror.The light goes through with 45 degrees angle and the ir reflects to the side.I use an 0 degrees hot mirror and my lamp does not seem to get more hot than without the hot mirror.i use this setup more than a year now with no problem.Beleive me even with the 400w hqi-ts the lcd is very cool with this mirror.It also has a uv filter.it is 90mmx90mm square and i bought it from a 3m supplier for 146 euros last year.it can work upto 500 celcious temperature!!!!
Try edmundoptics.com
http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=1492
I think for a hot mirror setup, it would be better to tile up the small ones as well - look at the price difference!!!
55x50mm HotMirror at surplusshed
$2.75!!!
55x50mm HotMirror at surplusshed
$2.75!!!
surplus dichroic mirrors
I have some of the cheap hot mirror tiles from surplusshed.com, and I am not impressed. I think they may be worn out or defective. I don't notice the heat being decreased when I test them with a halogen lamp.
But I also have a 4.75" diameter cold mirror I got from them for $5 and it works very well! I can really feel a lot of heat coming through it, and the reflection of visible light is very bright.
I noticed the same problem with putting a cold mirror into the light path between a condensor lens and a 15" LCD. Very difficult with a 220 mm fl fresnel. A bit easier if you change to a 330 mm fl. But you would still need a huge mirror. I agree that you need to make a stronger condensor so you can reflect a focal point. Or another option is to get the light down to that small focal point using an elliptical reflector.
I am still looking for a better hot mirror, since that would be very easy to use.
I have some of the cheap hot mirror tiles from surplusshed.com, and I am not impressed. I think they may be worn out or defective. I don't notice the heat being decreased when I test them with a halogen lamp.
But I also have a 4.75" diameter cold mirror I got from them for $5 and it works very well! I can really feel a lot of heat coming through it, and the reflection of visible light is very bright.
I noticed the same problem with putting a cold mirror into the light path between a condensor lens and a 15" LCD. Very difficult with a 220 mm fl fresnel. A bit easier if you change to a 330 mm fl. But you would still need a huge mirror. I agree that you need to make a stronger condensor so you can reflect a focal point. Or another option is to get the light down to that small focal point using an elliptical reflector.
I am still looking for a better hot mirror, since that would be very easy to use.
I tried those cheap hot mirrors from surplushed and aparently didn´t noticed any good result. But then took one of them and did an experiment; Do you know how night vision works? it is IR light lamp on a camera and then the camera receives the IR reflexion from the objets. So I placed the IR mirror on the cameras IR lamp and found it works nice. Much of the IR waves were reflected and lost (I could notice there was image on the camera visor but trust me, most of it was lost. I took it in front of my bathroom mirror and tried some esperiments. It was intersting...
better hot mirrors
I got some much better hot mirror tiles from the DIY Builder Group online store. They cost a bit more than the ones from Surplus Shed, but they work much better at reflecting the heat from a lamp. I could easily feel the difference when I put one of these between a 50 Watt Halgen flood and my finger two inches away.
I made an aluminum "sandwich" with four of these 50 mm by 50 mm hot mirrors held together in the middle. The result is a 100 mm square heat filter that I am using right next to a 250 Watt MH bulb. I have just a thin piece of lexan to support the fresnels, with no heat control glass at all. My fresnels and LCD don't even get warm after hours of running.
I got some much better hot mirror tiles from the DIY Builder Group online store. They cost a bit more than the ones from Surplus Shed, but they work much better at reflecting the heat from a lamp. I could easily feel the difference when I put one of these between a 50 Watt Halgen flood and my finger two inches away.
I made an aluminum "sandwich" with four of these 50 mm by 50 mm hot mirrors held together in the middle. The result is a 100 mm square heat filter that I am using right next to a 250 Watt MH bulb. I have just a thin piece of lexan to support the fresnels, with no heat control glass at all. My fresnels and LCD don't even get warm after hours of running.
hot mirrors
The surplus shed hot mirrors failed the "halogen flood & burnt finger" test completely. They had no cooling effect at all. I think maybe they were removed from equipment after thousands of hours of use. Or else they have a coating that only removes part of the IR spectrum.
I suspect they all buy their triplets from China or India, just like everybody else these days. You can root around with a search engine to find some optics suppliers there, but it won't do you much good. I am sure that setting up a business relationship with one of them is very involved, with large minimum orders, high shipping costs, customs processing, etc.
On the other hand, if you are interested in getting a longer focal length projection lens for a long throw, large format projector, I am experimenting with making them. I have pretty good results so far making a 500 mm fl symmetrical duplet from $18 worth of positive meniscus lenses from Rolyn. I think this type of lens will work fine at focal lengths above 450 mm. I am looking for more local sources for longer lenses with larger diameters.
The surplus shed hot mirrors failed the "halogen flood & burnt finger" test completely. They had no cooling effect at all. I think maybe they were removed from equipment after thousands of hours of use. Or else they have a coating that only removes part of the IR spectrum.
I suspect they all buy their triplets from China or India, just like everybody else these days. You can root around with a search engine to find some optics suppliers there, but it won't do you much good. I am sure that setting up a business relationship with one of them is very involved, with large minimum orders, high shipping costs, customs processing, etc.
On the other hand, if you are interested in getting a longer focal length projection lens for a long throw, large format projector, I am experimenting with making them. I have pretty good results so far making a 500 mm fl symmetrical duplet from $18 worth of positive meniscus lenses from Rolyn. I think this type of lens will work fine at focal lengths above 450 mm. I am looking for more local sources for longer lenses with larger diameters.
ok, thanks. So yoiu make your own doublets?
what is your doublets effective clear aperture?
both lenses are equal?, what is their separation?
Do you notice much aberrations? I guess you use low power lens, because if you did short focal lenses, the focusing would be imposible (the center could be focussed but not the edges, and who knows about the cromatic aberration...)
thanks, i am interested on your advances. please keep me informed.
what is your doublets effective clear aperture?
both lenses are equal?, what is their separation?
Do you notice much aberrations? I guess you use low power lens, because if you did short focal lenses, the focusing would be imposible (the center could be focussed but not the edges, and who knows about the cromatic aberration...)
thanks, i am interested on your advances. please keep me informed.
duplet
Not doublets, as in cemented achromatic. These symmetric duplets consist of a pair of identical positive meniscus lenses, concave sides facing each other. They have no chromatic aberration correction, but the worst case spot sizes for the 500 mm version I have actually built is under 0.1 mm according to OSLO. Since my LCD pixels are 1.8 mm on the screen, I don't think that aberration makes any difference. I can see the LCD's screendoor pattern over the entire screen surface.
These are 70 mm diameter 1000 mm fl lenses (sold as 1.0 Diopter). They have the standard spectacle curve on the concave side (3.5" radius), because they are test lenses for reading glass fittings. The spacing can be used to adjust the effective focal length a bit, without damaging the performance much. I am using one with 100 mm spacing right now.
I agree that this type of lens will only work at long focal lengths: Short focal length lenses are hard to make! That's why they need to have at least 6 surfaces and 2 or 3 different types of glass, just to make all the needed corrections. But a lot of us making large format DIY projectors really need projection lenses with focal lengths over 500 mm. Nobody seems to make triplets that long. I believe the reason for that is because a symmetric duplet is much cheaper and works pretty well above 500 mm fl.
Not doublets, as in cemented achromatic. These symmetric duplets consist of a pair of identical positive meniscus lenses, concave sides facing each other. They have no chromatic aberration correction, but the worst case spot sizes for the 500 mm version I have actually built is under 0.1 mm according to OSLO. Since my LCD pixels are 1.8 mm on the screen, I don't think that aberration makes any difference. I can see the LCD's screendoor pattern over the entire screen surface.
These are 70 mm diameter 1000 mm fl lenses (sold as 1.0 Diopter). They have the standard spectacle curve on the concave side (3.5" radius), because they are test lenses for reading glass fittings. The spacing can be used to adjust the effective focal length a bit, without damaging the performance much. I am using one with 100 mm spacing right now.
I agree that this type of lens will only work at long focal lengths: Short focal length lenses are hard to make! That's why they need to have at least 6 surfaces and 2 or 3 different types of glass, just to make all the needed corrections. But a lot of us making large format DIY projectors really need projection lenses with focal lengths over 500 mm. Nobody seems to make triplets that long. I believe the reason for that is because a symmetric duplet is much cheaper and works pretty well above 500 mm fl.
if you are using 2x1000mm and 100mm separation, your conbined focal should be 526mm. If you use just one PLX lens with lets say 500mm focal, you should have more focusing issues?
I mean, being them meniscus, does help? I know double overhead proyectors are meniscus as well. And about the symetry, do you know if it is better to be symetrical? I am not sure about it. what do you think?
I mean, being them meniscus, does help? I know double overhead proyectors are meniscus as well. And about the symetry, do you know if it is better to be symetrical? I am not sure about it. what do you think?
duplets
Guy,
would you mind explaining at an "optics for dummies" level how you implement the duplet in a projector? maybe this should be a new thread.
Do you need a long FL fresnel? I'm working on a 15" panel based projector and really would love to have it more than ~8' from my 90" screen
Thanks,
Bob
Guy,
would you mind explaining at an "optics for dummies" level how you implement the duplet in a projector? maybe this should be a new thread.
Do you need a long FL fresnel? I'm working on a 15" panel based projector and really would love to have it more than ~8' from my 90" screen
Thanks,
Bob
long throw lenses
Yes, it should be 526 mm, but that's not what I measure. I think the 1000 mm lenses are actually a bit shorter. I am looking for longer lenses. I think symmetric duplets work better with identical +meniscus lenses, but I have not tried modelling other possibilities. You can download OSLO for free and try modelling with different lenses. It took me about 8 hours to figure out enough to get useful info out of it.
You do have to use a longer fl field fresnel with such long projection lenses. I use a LL "790 mm" fresnel (actually more like 767 mm) in my projector. Both fresnels are together about two inches below the LCD. That gets the image of the lamp arc focussed in my projector lens. I get a 109" diagonal image from 13.5 feet, using a 15" LCD. When I get the lenses I want, I should have a 95" image from 14 feet.
Yes, it should be 526 mm, but that's not what I measure. I think the 1000 mm lenses are actually a bit shorter. I am looking for longer lenses. I think symmetric duplets work better with identical +meniscus lenses, but I have not tried modelling other possibilities. You can download OSLO for free and try modelling with different lenses. It took me about 8 hours to figure out enough to get useful info out of it.
You do have to use a longer fl field fresnel with such long projection lenses. I use a LL "790 mm" fresnel (actually more like 767 mm) in my projector. Both fresnels are together about two inches below the LCD. That gets the image of the lamp arc focussed in my projector lens. I get a 109" diagonal image from 13.5 feet, using a 15" LCD. When I get the lenses I want, I should have a 95" image from 14 feet.
fresnel lenses far from ideal behavior
>if you use 220 fresnell, placing the bulb at 300mm focuses the light to 820mm
If fresnels behaved like spherical glass lenses, then this would work. But they don't! They really work quite well if the rays go through the focal point on the smooth side, and are parallel on the rough side. (In either direction.) You can move a bit from those angles, but if you get too far the light gets messed up very badly. I did try some experiments along those lines, and it did not work at all.
>if you use 220 fresnell, placing the bulb at 300mm focuses the light to 820mm
If fresnels behaved like spherical glass lenses, then this would work. But they don't! They really work quite well if the rays go through the focal point on the smooth side, and are parallel on the rough side. (In either direction.) You can move a bit from those angles, but if you get too far the light gets messed up very badly. I did try some experiments along those lines, and it did not work at all.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.