I'm just starting my build and have been lurking/reading, but I'm still not solid on my theory. Any help is appreciated!
I just bought a Beseler 18" Triplet (man, that thing is HUGE!) and now I need to choose my fresnels. I know I don't want the 320mm field fresnels or anything so short. 18 inches is 457mm (thanks google calc!), so from what I've read I want a fresnel with >457mm FL. I've also read that the 650mm would be a little too long for the 18", so the 550 would be better. What are the downsides to getting the 650? Is it poor focus, or less light transmission? Just curious, I know I'm probably going with the 550 but I'd like to know what the reasons are behind matching the FL's. What happens if it's too long? What happens if it's too short?
The collimator, from what I gather, is less sensitive about FL's. It just has to match from the bulb to the LCD. What advantages/disadvantages are there between the 220 and the 320? Will the 320 give greater distance and theoretically more even light distribution since the incident angle is less? Does it even matter in this setup? I'm not *too* concerned with box size, I have a 14" room and I'm looking for about a 10" throw, so if the box is about 3" that seems fine.
Thanks yet again, I tried searching for this and reading the FAQ's, but I couldn't find a clear explanation of why we choose certain FL's. It'd be great to have this kind of theory in a Wiki somwhere (maybe I've missed it?).
Full Version: Beseler 18" Triplet - need help with fresnel selection
For the 550 vs the 650 collector:
The problem with the longer FL is a matter of optimisation. You can make it work in one of two ways:
1. Unsplit setup only: Move the fresnels further away from the lens (and LCD) thus allowing more space for the secondary optics to focus. Downsides are that this produces a longer projector, and the fresnels must be larger to compensate. This also increases the arc magnification. Upside is that if you do this well, you get a brighter image.
2. Move the lamp further away from the collimator. This works because both fresnels act as a lens to focus the light back at the projection triplet, so increasing the length on one side decreases the length on the other side. Downside is that this is sub-optimal for the fresnels, but the upside is that it's reasonably easy, and works well. More on the other consequences later.
For the 220 cs 330 collimaotr:
Well, the 330 DOES give more even lighting, at the expense of sheer lighting amount. It's not so much the angle of incidence (Though this is pretty much inexricably linked) as it is square law.
As you get further from a radiating source of light, the brightness (lux) dimishes as the square of the distance. Therefore if you are twice as far away from the lamp, you get 1/2 the brightness. If you are 1.414 times the distance, you get 1/2 the brightness. 330 is 1.5 times the distance away, so you can see that you've already lost more than 1/2 of the amount of light that you get at the center of the LCD.
Therefore, in the interests of maximum brightness, it is best to keep the collimator FL as short as is reasonable for the size panel that you are using. I have this worked out in other places, but it turns out that about a 17" panel is as large as you would want to go with a 220mm collimator. For a 10.6" panel, this can be much shorter. (I pinned the topic that I did all the math for this in, it's up there.)
Precondensers change the math, and a 330mm collimator with a good precondenser could also be a good choice, though that's not without it's price either.
Basically, though the collimator choice depends on the panel that you're using and if you want to use a precondenser lens.
The collector lens choice is up to your choice of triplet, and the other design choices that you make.
The problem with the longer FL is a matter of optimisation. You can make it work in one of two ways:
1. Unsplit setup only: Move the fresnels further away from the lens (and LCD) thus allowing more space for the secondary optics to focus. Downsides are that this produces a longer projector, and the fresnels must be larger to compensate. This also increases the arc magnification. Upside is that if you do this well, you get a brighter image.
2. Move the lamp further away from the collimator. This works because both fresnels act as a lens to focus the light back at the projection triplet, so increasing the length on one side decreases the length on the other side. Downside is that this is sub-optimal for the fresnels, but the upside is that it's reasonably easy, and works well. More on the other consequences later.
For the 220 cs 330 collimaotr:
Well, the 330 DOES give more even lighting, at the expense of sheer lighting amount. It's not so much the angle of incidence (Though this is pretty much inexricably linked) as it is square law.
As you get further from a radiating source of light, the brightness (lux) dimishes as the square of the distance. Therefore if you are twice as far away from the lamp, you get 1/2 the brightness. If you are 1.414 times the distance, you get 1/2 the brightness. 330 is 1.5 times the distance away, so you can see that you've already lost more than 1/2 of the amount of light that you get at the center of the LCD.
Therefore, in the interests of maximum brightness, it is best to keep the collimator FL as short as is reasonable for the size panel that you are using. I have this worked out in other places, but it turns out that about a 17" panel is as large as you would want to go with a 220mm collimator. For a 10.6" panel, this can be much shorter. (I pinned the topic that I did all the math for this in, it's up there.)
Precondensers change the math, and a 330mm collimator with a good precondenser could also be a good choice, though that's not without it's price either.
Basically, though the collimator choice depends on the panel that you're using and if you want to use a precondenser lens.
The collector lens choice is up to your choice of triplet, and the other design choices that you make.
I think it's starting to make sense to me, finally. I'm going to restate some of it to make sure I understand.
Ok... the projection lens has a certain focal length (457mm for me), so it will have to be a certain distance from the LCD to be in focus; I can consider that a "fixed" distance. The field fresnel has a certain focal length that directs the light toward the lens. If I have an unsplit setup, I can move the fresnels away from the LCD and lens, so the fresnel to lens distance is greater than the LCD to lens - this matches the longer fresnel FL (650) vs the lens FL (457). Since the light is already angled and getting "smaller" when it passes through the LCD, the fresnels have to be larger to compensate. Why would this setup be brighter? Due to the larger fresnels catching more light?
I think there's a typo... twice as far away should be 1/(2^2) = 1/4 the brightness, right? I know the square law, and yet seeing the actual numbers makes me realize how big a drop that is. So theoretically, from the square law alone a 330 would have half the brightness of a 220. However, a precondensor would "capture" more of the light to compensate for the longer distance to the 330, and would have the benefit of more light and more even lighting.
I'm starting to reconsider my "do everything the best possible the first time around" approach. I think I might take it a little easier. This means a 220/650 setup, split or unsplit depending on what's easiest. With a vertical/folded setup, keystoning shouldn't be much of a problem. I can use this setup to get everything working and learn how to optimize things. Then later I can convert it to a precondensor/330/650 unsplit setup to optimize light. Since this will be used in a small dark room, I think I'll be fine with lighting at first.
Thanks so much for the info, I also read through your lens thread and it really does clear up a lot of preconceptions I had from reading other information. I can't wait to start my build.
QUOTE (SupraGuy @ Feb 28 2008, 02:06 PM) 
For the 550 vs the 650 collector:
The problem with the longer FL is a matter of optimisation. You can make it work in one of two ways:
1. Unsplit setup only: Move the fresnels further away from the lens (and LCD) thus allowing more space for the secondary optics to focus. Downsides are that this produces a longer projector, and the fresnels must be larger to compensate. This also increases the arc magnification. Upside is that if you do this well, you get a brighter image.
The problem with the longer FL is a matter of optimisation. You can make it work in one of two ways:
1. Unsplit setup only: Move the fresnels further away from the lens (and LCD) thus allowing more space for the secondary optics to focus. Downsides are that this produces a longer projector, and the fresnels must be larger to compensate. This also increases the arc magnification. Upside is that if you do this well, you get a brighter image.
Ok... the projection lens has a certain focal length (457mm for me), so it will have to be a certain distance from the LCD to be in focus; I can consider that a "fixed" distance. The field fresnel has a certain focal length that directs the light toward the lens. If I have an unsplit setup, I can move the fresnels away from the LCD and lens, so the fresnel to lens distance is greater than the LCD to lens - this matches the longer fresnel FL (650) vs the lens FL (457). Since the light is already angled and getting "smaller" when it passes through the LCD, the fresnels have to be larger to compensate. Why would this setup be brighter? Due to the larger fresnels catching more light?
QUOTE
2. Move the lamp further away from the collimator. This works because both fresnels act as a lens to focus the light back at the projection triplet, so increasing the length on one side decreases the length on the other side. Downside is that this is sub-optimal for the fresnels, but the upside is that it's reasonably easy, and works well. More on the other consequences later.
For the 220 cs 330 collimaotr:
Well, the 330 DOES give more even lighting, at the expense of sheer lighting amount. It's not so much the angle of incidence (Though this is pretty much inexricably linked) as it is square law.
As you get further from a radiating source of light, the brightness (lux) dimishes as the square of the distance. Therefore if you are twice as far away from the lamp, you get 1/2 the brightness. If you are 1.414 times the distance, you get 1/2 the brightness. 330 is 1.5 times the distance away, so you can see that you've already lost more than 1/2 of the amount of light that you get at the center of the LCD.
For the 220 cs 330 collimaotr:
Well, the 330 DOES give more even lighting, at the expense of sheer lighting amount. It's not so much the angle of incidence (Though this is pretty much inexricably linked) as it is square law.
As you get further from a radiating source of light, the brightness (lux) dimishes as the square of the distance. Therefore if you are twice as far away from the lamp, you get 1/2 the brightness. If you are 1.414 times the distance, you get 1/2 the brightness. 330 is 1.5 times the distance away, so you can see that you've already lost more than 1/2 of the amount of light that you get at the center of the LCD.
I think there's a typo... twice as far away should be 1/(2^2) = 1/4 the brightness, right? I know the square law, and yet seeing the actual numbers makes me realize how big a drop that is. So theoretically, from the square law alone a 330 would have half the brightness of a 220. However, a precondensor would "capture" more of the light to compensate for the longer distance to the 330, and would have the benefit of more light and more even lighting.
QUOTE
Therefore, in the interests of maximum brightness, it is best to keep the collimator FL as short as is reasonable for the size panel that you are using. I have this worked out in other places, but it turns out that about a 17" panel is as large as you would want to go with a 220mm collimator. For a 10.6" panel, this can be much shorter. (I pinned the topic that I did all the math for this in, it's up there.)
Precondensers change the math, and a 330mm collimator with a good precondenser could also be a good choice, though that's not without it's price either.
Basically, though the collimator choice depends on the panel that you're using and if you want to use a precondenser lens.
The collector lens choice is up to your choice of triplet, and the other design choices that you make.
Precondensers change the math, and a 330mm collimator with a good precondenser could also be a good choice, though that's not without it's price either.
Basically, though the collimator choice depends on the panel that you're using and if you want to use a precondenser lens.
The collector lens choice is up to your choice of triplet, and the other design choices that you make.
I'm starting to reconsider my "do everything the best possible the first time around" approach. I think I might take it a little easier. This means a 220/650 setup, split or unsplit depending on what's easiest. With a vertical/folded setup, keystoning shouldn't be much of a problem. I can use this setup to get everything working and learn how to optimize things. Then later I can convert it to a precondensor/330/650 unsplit setup to optimize light. Since this will be used in a small dark room, I think I'll be fine with lighting at first.
Thanks so much for the info, I also read through your lens thread and it really does clear up a lot of preconceptions I had from reading other information. I can't wait to start my build.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.