Just wondering, in some experiments I've had the bulb very close to the lcd, but my cooling system kept the lcd at a safe temperature. I was thinking, maybe I could get a collimator with a shorter focal length to capture more of the bulbs light?
Oppinions?
Full Version: shorter collimator?
QUOTE (Mikau @ Sep 26 2005, 12:51 PM)
Just wondering, in some experiments I've had the bulb very close to the lcd, but my cooling system kept the lcd at a safe temperature. I was thinking, maybe I could get a collimator with a shorter focal length to capture more of the bulbs light?
Oppinions?
Oppinions?
I asked the same thing in another thread and got this response from supraguy...
QUOTE
QUOTE(SupraGuy @ Sep 22 2005, 10:59 AM)
Why not to use a shorter lens:
In short, the square law. The amount of light hitting the lens diminishes as the square of the distance from the lamp.
On a small lens, this doesn't amount to much. For a 15" lens, it does.
Take a 220mm lens, 15" across.
The distance from the center of the lamp to the center of the lens is 220mm, but using pythagoras, a^2 + b^2 = c^2, a=220, b=190.5 (7.5") we get c=291mm At a distance of only 311mm, you get only 1/2 of the light intensity that you get at 220mm, so you can see that there's already a signifigant difference.
Now, if we change a to 110mm, then we get a c of 219mm. This means that the corners will have less than 1/4 of the light that the center would have. This is SEVERE dark corners, and would be clearly visible to anyone. (Basically the center of the image would be 4X brighter than the current image center, and the corners would have at best the current center image brightness.) This is too great a gradient.
Conversely doubling the FL of the collimator fresnel would improve lighting evenness greatly, but would result in an image of only 1/4 the current attainable brightness.
Why not to use a shorter lens:
In short, the square law. The amount of light hitting the lens diminishes as the square of the distance from the lamp.
On a small lens, this doesn't amount to much. For a 15" lens, it does.
Take a 220mm lens, 15" across.
The distance from the center of the lamp to the center of the lens is 220mm, but using pythagoras, a^2 + b^2 = c^2, a=220, b=190.5 (7.5") we get c=291mm At a distance of only 311mm, you get only 1/2 of the light intensity that you get at 220mm, so you can see that there's already a signifigant difference.
Now, if we change a to 110mm, then we get a c of 219mm. This means that the corners will have less than 1/4 of the light that the center would have. This is SEVERE dark corners, and would be clearly visible to anyone. (Basically the center of the image would be 4X brighter than the current image center, and the corners would have at best the current center image brightness.) This is too great a gradient.
Conversely doubling the FL of the collimator fresnel would improve lighting evenness greatly, but would result in an image of only 1/4 the current attainable brightness.
essentially from that I think it means it would capture more light but the differential brightness between the corners and the center would be too noticeable....
QUOTE (mikyd1954 @ Sep 26 2005, 06:01 PM)
QUOTE (Mikau @ Sep 26 2005, 12:51 PM)
Just wondering, in some experiments I've had the bulb very close to the lcd, but my cooling system kept the lcd at a safe temperature. I was thinking, maybe I could get a collimator with a shorter focal length to capture more of the bulbs light?
Oppinions?
Oppinions?
I asked the same thing in another thread and got this response from supraguy...
QUOTE
QUOTE(SupraGuy @ Sep 22 2005, 10:59 AM)
Why not to use a shorter lens:
In short, the square law. The amount of light hitting the lens diminishes as the square of the distance from the lamp.
On a small lens, this doesn't amount to much. For a 15" lens, it does.
Take a 220mm lens, 15" across.
The distance from the center of the lamp to the center of the lens is 220mm, but using pythagoras, a^2 + b^2 = c^2, a=220, b=190.5 (7.5") we get c=291mm At a distance of only 311mm, you get only 1/2 of the light intensity that you get at 220mm, so you can see that there's already a signifigant difference.
Now, if we change a to 110mm, then we get a c of 219mm. This means that the corners will have less than 1/4 of the light that the center would have. This is SEVERE dark corners, and would be clearly visible to anyone. (Basically the center of the image would be 4X brighter than the current image center, and the corners would have at best the current center image brightness.) This is too great a gradient.
Conversely doubling the FL of the collimator fresnel would improve lighting evenness greatly, but would result in an image of only 1/4 the current attainable brightness.
Why not to use a shorter lens:
In short, the square law. The amount of light hitting the lens diminishes as the square of the distance from the lamp.
On a small lens, this doesn't amount to much. For a 15" lens, it does.
Take a 220mm lens, 15" across.
The distance from the center of the lamp to the center of the lens is 220mm, but using pythagoras, a^2 + b^2 = c^2, a=220, b=190.5 (7.5") we get c=291mm At a distance of only 311mm, you get only 1/2 of the light intensity that you get at 220mm, so you can see that there's already a signifigant difference.
Now, if we change a to 110mm, then we get a c of 219mm. This means that the corners will have less than 1/4 of the light that the center would have. This is SEVERE dark corners, and would be clearly visible to anyone. (Basically the center of the image would be 4X brighter than the current image center, and the corners would have at best the current center image brightness.) This is too great a gradient.
Conversely doubling the FL of the collimator fresnel would improve lighting evenness greatly, but would result in an image of only 1/4 the current attainable brightness.
essentially from that I think it means it would capture more light but the differential brightness between the corners and the center would be too noticeable....
Yeah but the pro fresnels, for larger lcds also have a 220 mm collimator.
Not much larger. There's only 1" difference in distance from center to edge.
I suppose that a 200mm collimator would yield some improvement. I wouldn't go below 190mm myself.
Kind of like electronic components, however, you have to choose from the list of lenses available to you. Not all values are commonly available. I suspect the same applies to fresnel lenses. 220mm seems to be a common enough size, so it's likely that this is just the most reasonable compromise.
A 200mm collimator fresnel will still capture signifigantly more light. Shorter than 190 would (IMO) not be practical for a 15" LCD at all, I think that the light loss from reflection would be too great, when the light strikes the flat surface of the fresnel at too sharp an angle.
I suppose that a 200mm collimator would yield some improvement. I wouldn't go below 190mm myself.
Kind of like electronic components, however, you have to choose from the list of lenses available to you. Not all values are commonly available. I suspect the same applies to fresnel lenses. 220mm seems to be a common enough size, so it's likely that this is just the most reasonable compromise.
A 200mm collimator fresnel will still capture signifigantly more light. Shorter than 190 would (IMO) not be practical for a 15" LCD at all, I think that the light loss from reflection would be too great, when the light strikes the flat surface of the fresnel at too sharp an angle.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.