Starting to build my LCD projector (planning phase yet...)
Do I go for a 400W HQI lamp or the extra heat and price tag is not worth it?
Do I go for a 400W HQI lamp or the extra heat and price tag is not worth it?
Well... Since nobody else has had a stab at it...
IMHO 250watt...that said others use 400...hell Ace gets away with 150! Aint that confusing?!?! All comes down to the design your going to use bud and what and where you going to use the projector for i think
Have a read through the forum...look at some of the designs the other fellas have come up with, checkout their results...and then make an educated decision on what you want, then if you are really confused ask the above question...
IMHO 250watt...that said others use 400...hell Ace gets away with 150! Aint that confusing?!?! All comes down to the design your going to use bud and what and where you going to use the projector for i think
Have a read through the forum...look at some of the designs the other fellas have come up with, checkout their results...and then make an educated decision on what you want, then if you are really confused ask the above question...
400 vrs 250 vrs 150?
From reading thousands of these posts, here and there, I think people are getting about the same amount of light out of projector designs using all of these:
The brute-force approach is a big fat 400 Watt bulb with no reflector or condensor.
The more efficient approach is a 250 Watt bulb with a smaller form factor, so you can put a spherical reflector behind it. These tend to be low-tech reflectors made from soup ladles or mixing bowls.
Trev (aka Ace) uses a very high-tech approach: A small 150 Watt double-ended bulb with a dichroic spherical reflector behind it and a real condensor lens in front, all lighting a 7" panel.
The 400 Watt approach is probably simplest to implement, but then you have to be very careful about cooling.
From reading thousands of these posts, here and there, I think people are getting about the same amount of light out of projector designs using all of these:
The brute-force approach is a big fat 400 Watt bulb with no reflector or condensor.
The more efficient approach is a 250 Watt bulb with a smaller form factor, so you can put a spherical reflector behind it. These tend to be low-tech reflectors made from soup ladles or mixing bowls.
Trev (aka Ace) uses a very high-tech approach: A small 150 Watt double-ended bulb with a dichroic spherical reflector behind it and a real condensor lens in front, all lighting a 7" panel.
The 400 Watt approach is probably simplest to implement, but then you have to be very careful about cooling.
parabolic reflector
Hopefully this parabolic reflector might be of use to someone. The size is LARGE, 12 inches+ diameter.
http://scientificsonline.com/product.asp_Q_pn_E_3053875
Hopefully this parabolic reflector might be of use to someone. The size is LARGE, 12 inches+ diameter.
http://scientificsonline.com/product.asp_Q_pn_E_3053875
Edmonds parabolic reflectors
These are very nice looking reflectors. I think we could use them in a projector design, if we got one that is as large as the LCD and then we left out the bottom fresnel. The reflector would send perpendicular rays to the LCD. Then those would be converged on the projector lens by the upper fresnel (which could be either above or below the LCD).
I do wonder if we would see a hot spot in the middle of the image, since that area would get direct light. The edges would only be getting reflected light. Maybe the 10-20% loss by reflection would not be noticeable.
These are very nice looking reflectors. I think we could use them in a projector design, if we got one that is as large as the LCD and then we left out the bottom fresnel. The reflector would send perpendicular rays to the LCD. Then those would be converged on the projector lens by the upper fresnel (which could be either above or below the LCD).
I do wonder if we would see a hot spot in the middle of the image, since that area would get direct light. The edges would only be getting reflected light. Maybe the 10-20% loss by reflection would not be noticeable.
I would say that using a reflector like that is probably not the best idea. Say we just don't use anything but a fresnel right before the LCD. With that setup, you get to use all of the light that comes from the front of the bulb and actually hits the fesnel. If you want to use the light coming from the back of the bulb, you want make sure that it 'appears' to be coming from nearly the same point as the front light. Otherwise you can only focus on one of the light sources. The front or the back. I found that I could GREATLY improve my brightness by using 2 fresnels between the light and LCD. This let me put the LCD much closer to the light and thus it collected a much larger amount of light. Not only that, but I shortened the distance that the light had to travel. One bad thing theoretically about doing this is that the light in the corners has to travel quite a bit furthur than the light that hits the middle. I have not noticed these effects in my setup yet. I currently have a fresnel 5" from my 250W bulb and then another larger one 2" from that one. This increased the light output by A LOT. I have no light measuring tools, but if you just use some math, you can figure out how much more light is getting there now. I did the math a while back for my specific case and the theoretical light increase was something like going from using 20% of the light to using 40%. Then, because I didnt have a reflector yet, I went ahead and polished the inside of half a coke can and put it behind. Low and behold, it seemed to me as the brightness increased by about the same ammount again.
The second inner fresnel is acting just like a condensor lens, effectively lowering the distance between the bulb and the first fresnel. Either one should produce the same effect. One thing I did notice was that at first heat was no concern. Right now I have 6 12v DC fans running on it. They were all necessary in order to keep the temperature of the LCD down. I would be curious to see what my system would look like with a 400W swapped in. I suspect I would need to ramp up the cooling yet again but that the light output would increase quite a bit. I think the brightness that I have right now is great using a 250w. Its a good combination of brightness of the picture and heat generated in my opinion. I think that ultimately, it depends what you want. 400w will be bright and hot but require not much work from you to produce a bright picture. 250 will be less hot but wll require some work to achieve optimal results. 150w is even less hot but requires even better optics to get amazing results. Using a polished coke can for example probably wouldn't cut it.
Hope that helps on your decision!
The second inner fresnel is acting just like a condensor lens, effectively lowering the distance between the bulb and the first fresnel. Either one should produce the same effect. One thing I did notice was that at first heat was no concern. Right now I have 6 12v DC fans running on it. They were all necessary in order to keep the temperature of the LCD down. I would be curious to see what my system would look like with a 400W swapped in. I suspect I would need to ramp up the cooling yet again but that the light output would increase quite a bit. I think the brightness that I have right now is great using a 250w. Its a good combination of brightness of the picture and heat generated in my opinion. I think that ultimately, it depends what you want. 400w will be bright and hot but require not much work from you to produce a bright picture. 250 will be less hot but wll require some work to achieve optimal results. 150w is even less hot but requires even better optics to get amazing results. Using a polished coke can for example probably wouldn't cut it.
Hope that helps on your decision!
what we want
What we want is to capture as much light as possible from the lamp, but to get it all going through the LCD perpendicular to the panel with uniform brightness. Any light that is not going in that direction, or not even striking the LCD panel, is wasted. You can work backwards from that to any number of designs.
For example, a parabolic reflector captures all the light that hits it and then sends it in parallel to the LCD panel. That's good. But almost all of the light that does not hit the reflector is wasted, because it will hit the LCD at such an angle that it will not end up at the projection lens. So the efficiency of such a reflector depends on how deep it is: A deep-dish parabola intersects more of the light than a shallow-dish. Independant of the depth, the diameter of the dish has to be at least as large as the diagonal edges of the LCD panel or some parts of the panel will not be lit. Make it much bigger, and you are wasting light beyond the edges of the panel.
Another example is a spherical reflector and condensor design. The condensor lens is placed very close to the lamp arc so it captures a lot of the light. (The distance does not really matter: Putting it close makes it intersect more of the light emitted by the lamp.) Adding a spherical reflector captures most of the light from the back of the lamp, and sends it back through the center of the lamp to the condensor. So the condensor "sees" a lamp arc that is near 100% brighter. The light from the condensor lens is sent to the lower fresnel and bent to the parallel beam we want. You can make a condensor lens from glass, acrylic, PVC, etc., and it can be aspherical, PCX, PCV, fresnel, or positive meniscus. Any positive lens of any sort could work, and you can put a small lens close to the lamp or a larger lens (ie. a fresnel) further away. But since it will be getting a lot of energy from the lamp, I think glass is the only practical material.
A third example is an elliptical reflector. (See drawing below.) In this design, the lamp arc is placed at one focal point of the ellipse, and the reflected light is focussed at the other focal point. The reflector size can be set so the diverging cone of light beyond that second focal point will fill a lower fresnel. The efficiency can also be increased by adding a large spherical reflector, placed so most of the light that misses the elliptical reflector does hit the spherical reflector and gets sent back to the lamp. When it strikes the elliptical reflector, it is going in a useful direction. (The spherical reflector has a hole in the center at the second focal point of the ellipse.) This is the most efficient design I have seen so far. It captures almost all of the light, but will lose 10 to 20% because of the reflector losses.
What we want is to capture as much light as possible from the lamp, but to get it all going through the LCD perpendicular to the panel with uniform brightness. Any light that is not going in that direction, or not even striking the LCD panel, is wasted. You can work backwards from that to any number of designs.
For example, a parabolic reflector captures all the light that hits it and then sends it in parallel to the LCD panel. That's good. But almost all of the light that does not hit the reflector is wasted, because it will hit the LCD at such an angle that it will not end up at the projection lens. So the efficiency of such a reflector depends on how deep it is: A deep-dish parabola intersects more of the light than a shallow-dish. Independant of the depth, the diameter of the dish has to be at least as large as the diagonal edges of the LCD panel or some parts of the panel will not be lit. Make it much bigger, and you are wasting light beyond the edges of the panel.
Another example is a spherical reflector and condensor design. The condensor lens is placed very close to the lamp arc so it captures a lot of the light. (The distance does not really matter: Putting it close makes it intersect more of the light emitted by the lamp.) Adding a spherical reflector captures most of the light from the back of the lamp, and sends it back through the center of the lamp to the condensor. So the condensor "sees" a lamp arc that is near 100% brighter. The light from the condensor lens is sent to the lower fresnel and bent to the parallel beam we want. You can make a condensor lens from glass, acrylic, PVC, etc., and it can be aspherical, PCX, PCV, fresnel, or positive meniscus. Any positive lens of any sort could work, and you can put a small lens close to the lamp or a larger lens (ie. a fresnel) further away. But since it will be getting a lot of energy from the lamp, I think glass is the only practical material.
A third example is an elliptical reflector. (See drawing below.) In this design, the lamp arc is placed at one focal point of the ellipse, and the reflected light is focussed at the other focal point. The reflector size can be set so the diverging cone of light beyond that second focal point will fill a lower fresnel. The efficiency can also be increased by adding a large spherical reflector, placed so most of the light that misses the elliptical reflector does hit the spherical reflector and gets sent back to the lamp. When it strikes the elliptical reflector, it is going in a useful direction. (The spherical reflector has a hole in the center at the second focal point of the ellipse.) This is the most efficient design I have seen so far. It captures almost all of the light, but will lose 10 to 20% because of the reflector losses.
Attachments
- Status
- Not open for further replies.