Hey guys,
This is my first post so take it easy on me. I ahve been lurking the forums for a few days now and am going to try and build a DIY projector as soon as I get my house built.
My question is this, I have noticed alot of people using metal flashing and dish shaped light reflector behind the bulbs. Has anyone tried Mirrors as a reflecting device. It seems to me that if I want uniform light over a square LCD panel, then I would rather use a rectagular reflector made out of trapezoidal panels. Kind of like those Holgen flood light you can get at the hardware store.
Also, Would it be of any benefit to put mirrors on all the sides of the light chamber?
Just another new guy Question.
JAFO
Tom
p.s. if anyone picks up on JAFO I will be impressed
Full Version: Light Chamber Configuration
Jusnt Another F*****g Observer 
Ideally our light source should be a "point" so just sticking mirrors everywhere would result in a less-than-optimal solution and could actually make the image worse. The rear fresnel wants to have all light coming from a "point". Check this. This rest of the wiki should help get you up-to-speed as well. 
Also note that our lights get really hot and any old mirror not meant to take the heat may very well break due to thermal stresses.
Also note that our lights get really hot and any old mirror not meant to take the heat may very well break due to thermal stresses.
QUOTE (Durachko @ Sep 1 2006, 04:07 PM) 
Ideally our light source should be a "point" so just sticking mirrors everywhere would result in a less-than-optimal solution and could actually make the image worse. The rear fresnel wants to have all light coming from a "point". Check this. This rest of the wiki should help get you up-to-speed as well.
Also note that our lights get really hot and any old mirror not meant to take the heat may very well break due to thermal stresses.
Also note that our lights get really hot and any old mirror not meant to take the heat may very well break due to thermal stresses.
So if I understad correctly, th point of the first fresnel is to gather all the lite from that single point and correct it to the width of the screen. So what is you jus removed that fresnel and configured a reflector that would place light evenly on the led screen?
Tom
Damn, I thought JAFO would last a bit longer than the first post. But then again I am dealing with fellow movie dorks. I'll have to be a little more obscure next time
QUOTE (Tscott @ Sep 1 2006, 02:19 PM)
So if I understad correctly, th point of the first fresnel is to gather all the lite from that single point and correct it to the width of the screen. So what is you jus removed that fresnel and configured a reflector that would place light evenly on the led screen?
Tom
Damn, I thought JAFO would last a bit longer than the first post. But then again I am dealing with fellow movie dorks. I'll have to be a little more obscure next time
Tom
Damn, I thought JAFO would last a bit longer than the first post. But then again I am dealing with fellow movie dorks. I'll have to be a little more obscure next time
Pretty much... The light from the bulb will spread to the size of the screen on it's own, the problem is it's "going" in the wrong direction. The first fresnel bends, or refracts (I think I'm using that word right) the light so that it is "collimated" (or made parallel) the second fresnel refracts that light again so that it's heading toward the triplet (or projection lens). It's not that the fresnels get the light to the screen, they get the light to the screen so that it's going in the right direction.
If you can get a mirror/reflector that will collimate the light without using lenses you'd be in business, the problem is that (as far as I know) nobody has been able to do this effectively yet. A parabolic mirror will collimate light if it's coming from it's focal point, in theory. In practice the light from the mirror has lots of hotspots, and it prettymuch unusable.
TScott: Welcome to the fun. I am jamming on this whole DIY projector thing as well being both an AV hobbyist and a physics addict.
I think the main issues are brightness, consistency and direction of the light.
1) Light leaving the panel should be as bright as possible AND
2) Be an even "coat" of light to avoid hotspots AND
3) Be moving in as much the same direction as possible, such that one pixel of the LCD panel is illuminated in one spot on your screen (enhancing focus, contrast, color mix, etc.)
However this can be achieved is a "good" solution. One solution is to have a point source lamp set at the focus of a a fresnel lens, with all other light absorbed by a black interior. Light is collimnated (made more parallel) through the panel by refracting it through the fresnel material, and it will only refract light coming from the focal point of the fresnel.
A better solution is to not lose the light that is emitted out the back of the lamp by using a spherical reflector with the lamp at the center of the circle formed by the reflector surface. This solution so far has been the most effective solution for the folks on the forum from what I have read.
That said, there may be more solutions out there, such as a super bright LED panel with polarizers (similar to what was stripped off in the first place but much brighter), or some kind of laser image. Some folks are also trying to work with multiple bulbs that illuminate portions of the panel or all the panel depending on some diffusion technologies.
BUT, I think any solution will still have to meet the 1, 2 and 3 requirements above. If I am incorrect, I am sure others will chime in.
I think the main issues are brightness, consistency and direction of the light.
1) Light leaving the panel should be as bright as possible AND
2) Be an even "coat" of light to avoid hotspots AND
3) Be moving in as much the same direction as possible, such that one pixel of the LCD panel is illuminated in one spot on your screen (enhancing focus, contrast, color mix, etc.)
However this can be achieved is a "good" solution. One solution is to have a point source lamp set at the focus of a a fresnel lens, with all other light absorbed by a black interior. Light is collimnated (made more parallel) through the panel by refracting it through the fresnel material, and it will only refract light coming from the focal point of the fresnel.
A better solution is to not lose the light that is emitted out the back of the lamp by using a spherical reflector with the lamp at the center of the circle formed by the reflector surface. This solution so far has been the most effective solution for the folks on the forum from what I have read.
That said, there may be more solutions out there, such as a super bright LED panel with polarizers (similar to what was stripped off in the first place but much brighter), or some kind of laser image. Some folks are also trying to work with multiple bulbs that illuminate portions of the panel or all the panel depending on some diffusion technologies.
BUT, I think any solution will still have to meet the 1, 2 and 3 requirements above. If I am incorrect, I am sure others will chime in.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.