Hi!!!
I am trying to gather the materials and i ended a with this configuration:
1) Osram HQI TS 400W 5200K 36000Ansi
2) 1 UV filter
3)330mm 17" condenser fresnel
4)17" tft screen
5)330mm 17" field fresnel
6)135mm Long-Throw Projection Lens Triplet with this specs:
Outer Diameter: ~135mm
Back Focal Length: 380mm
Effective Focal Length: 450mm
Largest Image Source: 15 inches diag. (still testing larger sizes)br> Barrel Length: 159mm from135mm triplet
Is all these materials compatible with each other?
Can i use this triplet with these fresnel?
Will i be able to focus the image in virtually every different distance from the screen?
Can I use a 220mm condenser fresnel?Advantages-Dissadvantages.
Can anyone speculate the final ansi capabilities of a diy projector like this?
Thank you in advance.
Theo.
I am trying to gather the materials and i ended a with this configuration:
1) Osram HQI TS 400W 5200K 36000Ansi
2) 1 UV filter
3)330mm 17" condenser fresnel
4)17" tft screen
5)330mm 17" field fresnel
6)135mm Long-Throw Projection Lens Triplet with this specs:
Outer Diameter: ~135mm
Back Focal Length: 380mm
Effective Focal Length: 450mm
Largest Image Source: 15 inches diag. (still testing larger sizes)br> Barrel Length: 159mm from135mm triplet
Is all these materials compatible with each other?
Can i use this triplet with these fresnel?
Will i be able to focus the image in virtually every different distance from the screen?
Can I use a 220mm condenser fresnel?Advantages-Dissadvantages.
Can anyone speculate the final ansi capabilities of a diy projector like this?
Thank you in advance.
Theo.
feedback
if you use a 330 mm fl condensor fresnel, then you should also use a pre-condensor lens to capture more light. With a 220 mm fl condensor fresnel, it is not so important. 330 + pre-condensor costs more but might light the corners better.
Since you are using a 450 mm fl projection lens (highly recommended), then you might want to use a LumenLab 650 mm or 790 mm field fresnel. The 330 mm field fresnels are just too short for this lens.
Find a good spherical reflector. It will increase your light output quite a bit.
You won't be able to throw an image from ANY distance, because you will not have a zoom lens. If you try to get over 20 feet or so, the image will be so big that you won't have enough light to see it well. But you should be able to project up to a 12 foot image with this setup. It does get a bit silly when part of the image is on the ceiling and part is on the floor!
Actually, most DIY projectors don't have a way of adjusting the condensor system to handle big changes in the projection lens position. You could design that in to your projector if you plan on moving it around to different size venues.
if you use a 330 mm fl condensor fresnel, then you should also use a pre-condensor lens to capture more light. With a 220 mm fl condensor fresnel, it is not so important. 330 + pre-condensor costs more but might light the corners better.
Since you are using a 450 mm fl projection lens (highly recommended), then you might want to use a LumenLab 650 mm or 790 mm field fresnel. The 330 mm field fresnels are just too short for this lens.
Find a good spherical reflector. It will increase your light output quite a bit.
You won't be able to throw an image from ANY distance, because you will not have a zoom lens. If you try to get over 20 feet or so, the image will be so big that you won't have enough light to see it well. But you should be able to project up to a 12 foot image with this setup. It does get a bit silly when part of the image is on the ceiling and part is on the floor!
Actually, most DIY projectors don't have a way of adjusting the condensor system to handle big changes in the projection lens position. You could design that in to your projector if you plan on moving it around to different size venues.
About the lens not being compatable with the 330 fresnel because its 450mm EFL... it has a rear f.l. of only 380, shouldnt that be close enought? thats the number that matters to the fresnel, the rear focal lenght... as far as I know...argument begins.... nnnnow!
🙂
Cheers Guy,
Alex
🙂
Cheers Guy,
Alex
where the light goes
I know some people will tell you it is fine if the cone of light has gone past its focal point and is diverging before it hits the projection lens. Try a ray tracing to see if it matters: Most of the light from the edges of the LCD will enter the lens at such an angle that it won't get through the lens. The only way around this is to use a lens that is much wider than the arc image.
If you want all of the light to make it to the screen, then you adjust the fresnels to focus the image of the lamp arc at the center of the lens. And I don't mean the middle of the first surface it hits. If the lens is 6 inches long, then it should be focussed 3 inches into the lens. So the first surface of the projection lens intersects a converging cone.
But hey, that's just what I see by modelling projection lenses in OSLO. Maybe somebody else has a "faith-based" opinion that is just as valid! 😀
I know some people will tell you it is fine if the cone of light has gone past its focal point and is diverging before it hits the projection lens. Try a ray tracing to see if it matters: Most of the light from the edges of the LCD will enter the lens at such an angle that it won't get through the lens. The only way around this is to use a lens that is much wider than the arc image.
If you want all of the light to make it to the screen, then you adjust the fresnels to focus the image of the lamp arc at the center of the lens. And I don't mean the middle of the first surface it hits. If the lens is 6 inches long, then it should be focussed 3 inches into the lens. So the first surface of the projection lens intersects a converging cone.
But hey, that's just what I see by modelling projection lenses in OSLO. Maybe somebody else has a "faith-based" opinion that is just as valid! 😀
a tiny bit of math
That "380 mm Back Focal Length" is a bit misleading. It implies that the space between the LCD and the lens will be 380 mm. But that would only be true if your screen was an infinite distance away! To calculate the real-world distance:
1/EFL = 1/LCD-to-lens + 1/lens-to-screen
Let's try this for a 10 foot throw distance:
1/450 = 1/LCD-to-lens + 1/(10 feet * 12 inches/foot *25.4 mm/inch)
So LCD-to-lens distance = 528 mm (referring to the center of the lens)
The difference between EFL and BFL is 450 - 380 = 80 mm, so we can calculate the distance between the LCD and the back of the lens = 528 - 80 = 448 mm
If you want to fit a 330 mm fresnel in there to focus the arc image into the projection lens, it would have to go more than 100 mm past the LCD! Way too far!
That "380 mm Back Focal Length" is a bit misleading. It implies that the space between the LCD and the lens will be 380 mm. But that would only be true if your screen was an infinite distance away! To calculate the real-world distance:
1/EFL = 1/LCD-to-lens + 1/lens-to-screen
Let's try this for a 10 foot throw distance:
1/450 = 1/LCD-to-lens + 1/(10 feet * 12 inches/foot *25.4 mm/inch)
So LCD-to-lens distance = 528 mm (referring to the center of the lens)
The difference between EFL and BFL is 450 - 380 = 80 mm, so we can calculate the distance between the LCD and the back of the lens = 528 - 80 = 448 mm
If you want to fit a 330 mm fresnel in there to focus the arc image into the projection lens, it would have to go more than 100 mm past the LCD! Way too far!
So you are suggesting a LumenLab 650 mm or 790 mm field fresnel. Isn't their focal lenth way to big?
Is there a good setup for a 17" tft that i can use and have excelent result?
What would be the expected luminence with a good reflector?
Is there a very good field fresnel with ~450mm fl?
450 mm fl fresnel
I don't know where you can buy a good 450 mm fl fresnel, but that would be too short as well, even with a split fresnel design.
A few posts back in this thread, I calculated an LCD to projector lens distance of 528 mm. If you use a 650 mm fresnel, then it would be perfect if it was (650 - 528) = 122 mm before the LCD. Then the condensor fresnel could be spaced at exactly its focal length from the lamp arc.
But you could also put the fresnels closer to the LCD and then adjust the lamp to condensor distance a bit to get the arc image focussed into the projector lens.
If you need a large space between the fresnels and the LCD, you have to use fresnels that are wide enough. The light cone converges as it goes from the fresnels to the LCD. You can draw a full-scale picture to decide how large to cut your fresnels, or you can do the math with similar triangles.
I don't know where you can buy a good 450 mm fl fresnel, but that would be too short as well, even with a split fresnel design.
A few posts back in this thread, I calculated an LCD to projector lens distance of 528 mm. If you use a 650 mm fresnel, then it would be perfect if it was (650 - 528) = 122 mm before the LCD. Then the condensor fresnel could be spaced at exactly its focal length from the lamp arc.
But you could also put the fresnels closer to the LCD and then adjust the lamp to condensor distance a bit to get the arc image focussed into the projector lens.
If you need a large space between the fresnels and the LCD, you have to use fresnels that are wide enough. The light cone converges as it goes from the fresnels to the LCD. You can draw a full-scale picture to decide how large to cut your fresnels, or you can do the math with similar triangles.
This is a design with the setup i suggested when i originally posted:
17" tft diy projector
Guy Grotke what is your opinion about this?
If i want to have my fresnels separated then i cant use the confic youre suggesting to your post right?(I want keystone...)
Also lumenlab in their store havent got a 650 mm fresnel.
Is there an optics company that can create a 17" fresnel with a fl~528mm?
17" tft diy projector
Guy Grotke what is your opinion about this?
If i want to have my fresnels separated then i cant use the confic youre suggesting to your post right?(I want keystone...)
Also lumenlab in their store havent got a 650 mm fresnel.
Is there an optics company that can create a 17" fresnel with a fl~528mm?
yes this is possible
In fact, it is likely to work better in a split fresnel design.
The designer of the 17" projector used a 220 mm fl condensor fresnel and a 330 mm fl field fresnel. But he placed the lamp arc closer to the condensor fresnel that the optimal 220 mm. By moving the lamp closer, it makes the rays coming out of the condensor fresnel diverge. In this design, they immediately hit the field fresnel and are converged to the point 68.2 cm away. Sending divergent rays to the field fresnel stretches its point of focus.
The problem with this, is that fresnels don't behave very well when you use them this way. They work best when you put the object (ie. the lamp) at the focal distance on the smooth side, and the light on the grooved side is parallel. The grooves in the fresnel are designed to collect rays from a particular distance. If you move the lamp too far, then you will get less light and more distortion.
If the design you referenced works well, then your design with the 330 mm field fresnel 15-25 mm past the LCD should work even better. The field fresnel focal distance will not have to be stretched as much, so the lamp will not be so close to the condensor fresnel.
One thing I would suggest is that either projector will work better if the field fresnel focusses the arc image right at the triplet's center point. (el centro optico at 51.63 cm from the LCD)
If you need to do optical keystone correction, then you must put the field fresnel past the LCD. So a 528 mm fresnel would be too long. You would need something more like 500 mm. You need to put the field fresnel more than the usual distance from the LCD, because you need some room to tilt it. If you tilt one side too close to the LCD, then you will see the fresnel's rings in the image. But I don't know of any source for a 500 mm fl fresnel. I have seen some reference to a 450 mm fl fresnel, but I don't remember where.
In fact, it is likely to work better in a split fresnel design.
The designer of the 17" projector used a 220 mm fl condensor fresnel and a 330 mm fl field fresnel. But he placed the lamp arc closer to the condensor fresnel that the optimal 220 mm. By moving the lamp closer, it makes the rays coming out of the condensor fresnel diverge. In this design, they immediately hit the field fresnel and are converged to the point 68.2 cm away. Sending divergent rays to the field fresnel stretches its point of focus.
The problem with this, is that fresnels don't behave very well when you use them this way. They work best when you put the object (ie. the lamp) at the focal distance on the smooth side, and the light on the grooved side is parallel. The grooves in the fresnel are designed to collect rays from a particular distance. If you move the lamp too far, then you will get less light and more distortion.
If the design you referenced works well, then your design with the 330 mm field fresnel 15-25 mm past the LCD should work even better. The field fresnel focal distance will not have to be stretched as much, so the lamp will not be so close to the condensor fresnel.
One thing I would suggest is that either projector will work better if the field fresnel focusses the arc image right at the triplet's center point. (el centro optico at 51.63 cm from the LCD)
If you need to do optical keystone correction, then you must put the field fresnel past the LCD. So a 528 mm fresnel would be too long. You would need something more like 500 mm. You need to put the field fresnel more than the usual distance from the LCD, because you need some room to tilt it. If you tilt one side too close to the LCD, then you will see the fresnel's rings in the image. But I don't know of any source for a 500 mm fl fresnel. I have seen some reference to a 450 mm fl fresnel, but I don't remember where.
hello guy;
here you have done a sumary of most important trick of diyproyection. The optimal light cones (bulb-tft-projectlens) where the two cones are exactly the correct size, meaning all the light goes to the projection lens.
now, my only doubt is; when we play with the bulb position to make the perfect light cones, (forget the precondenser lens, and reflector) everybody says that the focal point of the fresnell is the best place. Thats true, but i think moving farther or closer to the rear fresnell, is no problem for for the fresnell operation. Even i could say that moving the bulb back (farther, imagine we have 220mm rear fresnell and no field fresenll, so we need to use the light at 330mm moreless) then the results are even better for the lens.
Don´t you think it is worst for the lens getting closer than getting farther with the bulb? getting farther, the light would be even more homogeneus, so no hot spot...
the fresnells are very simple lens, i know it but we just need them to light the lcd, no quality needed here...
i don´t know if i xplained right.
here you have done a sumary of most important trick of diyproyection. The optimal light cones (bulb-tft-projectlens) where the two cones are exactly the correct size, meaning all the light goes to the projection lens.
now, my only doubt is; when we play with the bulb position to make the perfect light cones, (forget the precondenser lens, and reflector) everybody says that the focal point of the fresnell is the best place. Thats true, but i think moving farther or closer to the rear fresnell, is no problem for for the fresnell operation. Even i could say that moving the bulb back (farther, imagine we have 220mm rear fresnell and no field fresenll, so we need to use the light at 330mm moreless) then the results are even better for the lens.
Don´t you think it is worst for the lens getting closer than getting farther with the bulb? getting farther, the light would be even more homogeneus, so no hot spot...
the fresnells are very simple lens, i know it but we just need them to light the lcd, no quality needed here...
i don´t know if i xplained right.
real experiments with a 220 mm fresnel
Before I cut and installed my lumenlab 220 mm fl fresnel, I did some experiments to see the effect of placing the lamp at various distances. When the fresnel was very close to 220 mm from the lamp, I saw an evenly lit rectangle of light coming out of the fresnel: Perfect for lighting an LCD. I could move the lamp a tiny amount to make that rectangle slightly diverge or converge. But when I moved the lamp out to about 250 mm, the light coming from the fresnel changed to a circular spot. I think the rings in the outer part of the lens failed because of the angle of the light.
But maybe that was just a problem with that particular fresnel lens design. If you have a 220 mm fl fresnel, then I suggest you try some experiments with it to see how it acts with the lamp arc so far from the focal length. Then post the results back here.
Before I cut and installed my lumenlab 220 mm fl fresnel, I did some experiments to see the effect of placing the lamp at various distances. When the fresnel was very close to 220 mm from the lamp, I saw an evenly lit rectangle of light coming out of the fresnel: Perfect for lighting an LCD. I could move the lamp a tiny amount to make that rectangle slightly diverge or converge. But when I moved the lamp out to about 250 mm, the light coming from the fresnel changed to a circular spot. I think the rings in the outer part of the lens failed because of the angle of the light.
But maybe that was just a problem with that particular fresnel lens design. If you have a 220 mm fl fresnel, then I suggest you try some experiments with it to see how it acts with the lamp arc so far from the focal length. Then post the results back here.
Guy that effect is quite normal for an aspherical fresnel. Moving the lamp from its optimum position stuffs up the aspherical correction. I’ve actually got a fresnel here that has the wrong aspherical correction applied to it and when a lamp is placed at the FL, a star shape is projected. After a lot of testing I came to the conclusion that somehow the correction meant for a 220mm fresnel had been applied to this 330mm fresnel. When the lamp is placed at 220mm the projected image becomes square but it has been magnified. If you have one of those page magnifiers, they are generally spherical so you should be able to see the effect that no correction has. In other words you won’t be able to get a square projected image.
DJ
aspherical fresnel
Yes, I thought that effect might just be a product of the design of the lumenlab fresnel. That's why I suggested an experiment with his fresnel. I think if it is a simple spherical design, then it might work better with the lamp away from the focal point.
But then that brings up another point: People go to the trouble of designing aspherical fresnels to get more even lighting and less aberration. They should perform better than a spherical fresnel, if used correctly. So then is it a good idea to use a spherical fresnel, just so you can fiddle with the lamp position.
I think maybe the best performance comes with using an aspherical fresnel with the object at its focal length.
Yes, I thought that effect might just be a product of the design of the lumenlab fresnel. That's why I suggested an experiment with his fresnel. I think if it is a simple spherical design, then it might work better with the lamp away from the focal point.
But then that brings up another point: People go to the trouble of designing aspherical fresnels to get more even lighting and less aberration. They should perform better than a spherical fresnel, if used correctly. So then is it a good idea to use a spherical fresnel, just so you can fiddle with the lamp position.
I think maybe the best performance comes with using an aspherical fresnel with the object at its focal length.
- Status
- Not open for further replies.