well, there is a point that i can´t make you understand, lets try once more;
yes, there is a real arc and a virtual arc. Do you agree that if i we do correct maths, we can forget about the real arc and work if we only had a virtual arc at the real focal?
If you agree with this, then we can follow this way;
the virtual arc is at rear fresnell focal point. (the uper side of the arc, lets say 18mm uper side) needs to pass trhow the "window".
Do you understand what i mean by "window"?
Window=hole is the aperture of the precondenser lens of 100mm (50mm upside).
Thats very important; The virtual arc goes trhow the window WITHOUT REFRACTION. the real arc goes throw the lens BEING REFRACTED, but as i said since we work with the virtual arc we can forget about the real arc and forget about the refraction on the lens BUT WHAT WE CAN NOT FORGET IS THE WINDOW APERTURE.
Why can´t we forget the window aperture? because the virtual arc is so big that NOT ALL PIXELS WILL SEE FULL ARC THROW THE WINDOW. On the image i drow with 3 pixels, there is the green pixel that sees the edge of the arc that intersets with the window edge. next pixel will NOT SEE THE ARC EDGE, and finally the last pixel (at 15" corner) will only see Half arc throw the window.
please tell me what do i need to xplain farther.
yes, there is a real arc and a virtual arc. Do you agree that if i we do correct maths, we can forget about the real arc and work if we only had a virtual arc at the real focal?
If you agree with this, then we can follow this way;
the virtual arc is at rear fresnell focal point. (the uper side of the arc, lets say 18mm uper side) needs to pass trhow the "window".
Do you understand what i mean by "window"?
Window=hole is the aperture of the precondenser lens of 100mm (50mm upside).
Thats very important; The virtual arc goes trhow the window WITHOUT REFRACTION. the real arc goes throw the lens BEING REFRACTED, but as i said since we work with the virtual arc we can forget about the real arc and forget about the refraction on the lens BUT WHAT WE CAN NOT FORGET IS THE WINDOW APERTURE.
Why can´t we forget the window aperture? because the virtual arc is so big that NOT ALL PIXELS WILL SEE FULL ARC THROW THE WINDOW. On the image i drow with 3 pixels, there is the green pixel that sees the edge of the arc that intersets with the window edge. next pixel will NOT SEE THE ARC EDGE, and finally the last pixel (at 15" corner) will only see Half arc throw the window.
please tell me what do i need to xplain farther.
Attachments
I think I know what you are trying to explain. I’m not sure whether I agree with you yet, I’ll have to do some more thinking.
DJ
DJ
yes, thanks, i also need to think about how to solve this issue.
by now i would say that 15"/220 or 15"/330 triangle to determine the lens aperture and placement, is wrong.
we have to ensure full arc lengh (mean full virtual arc lengh) needs to be seen from every pixel of the lcd.
by now i would say that 15"/220 or 15"/330 triangle to determine the lens aperture and placement, is wrong.
we have to ensure full arc lengh (mean full virtual arc lengh) needs to be seen from every pixel of the lcd.
Guys, ive got some news for you, particulary Rox, it is impossible to get a uniform image on a lcd, its cos of the way the light goes through the polarisers on the sides. I got this out of a DMD article in a mag tonight, i had my suspions.........
As for upgrading the lamp to a more powerful one, that should only be the last resort, collecting more light from the source and directing it in the right direction is the key, and acurately.
Trev🙂
As for upgrading the lamp to a more powerful one, that should only be the last resort, collecting more light from the source and directing it in the right direction is the key, and acurately.
Trev🙂
yes i know it is not posible to have a unform image at the lcd, but there are two ways doing makin things work, 1)the best way posible and 2)wrong.
I find a problem on the way teorically ideal precondensor lens placemnt method (there is a great post from guy grotke with a graphical method to determine the lens placement as well as bulb position) but it only works idealy with 0 arc lengh (infinitesimal light source point). If we introduce real arc lengh there, we see that part of arc lengh is lost and never gets to some lcd area.
i would like to check this issue with you guys 😀. Is this a good place for?
I find a problem on the way teorically ideal precondensor lens placemnt method (there is a great post from guy grotke with a graphical method to determine the lens placement as well as bulb position) but it only works idealy with 0 arc lengh (infinitesimal light source point). If we introduce real arc lengh there, we see that part of arc lengh is lost and never gets to some lcd area.
i would like to check this issue with you guys 😀. Is this a good place for?
I mean no disrespect there Ace, but I beg to differ on uniform focus(in my expereince even witha 15" LCD its do-able, depends on lens, and im talking enough focus to see screendoor fine in the corners.)..... but uniform brightness well.. thats a dead issue, its un-doable unless you use a very small panel(5" or smaller.)
And a brighter light source simply compounds the issue, it will be even brighter in the centre, with the same reductions att he side.
And a brighter light source simply compounds the issue, it will be even brighter in the centre, with the same reductions att he side.
This stuff has been hashed over several times on the LL forums 🙂
Basically, with a spherical or no-reflector setup, you need spherical abberation to correct uneven light distribution. Anything else, using ideal lenses, ends up being a balance between taking advantage of a greater angle of light from the lamp (higher brightness) and a better light uniformity.
This causes some problems, though I think that you could have an abberating lense and an aspheric? lense to collimate the rays again.
disclaimer: IANAO (I am not an optician)
Basically, with a spherical or no-reflector setup, you need spherical abberation to correct uneven light distribution. Anything else, using ideal lenses, ends up being a balance between taking advantage of a greater angle of light from the lamp (higher brightness) and a better light uniformity.
This causes some problems, though I think that you could have an abberating lense and an aspheric? lense to collimate the rays again.
disclaimer: IANAO (I am not an optician)
well, I just discover this topic (an the other one , with math inside 😀 ) while I'm trying to place correctly my condenser . I got lucky.
Fortunately, I realize that my experimental setup match the theoritical place for the precondensor lens in my conditions. But As highlighted in this topic, to improve the lumen is a multifactorial organisation.
I finally succeed to get 100-110 lumen with a 7'lcd and a 250w metal halide by adding a precondenser lens but also by increasing the diameter of my projection lens (125mm diameter , focal 412mm) and also by decreasing the fresnel condenser focal (by adding 2 lens 220+330 instead of one) then decreasing the distance between light and lcd. Finally due to my need for a projection lens with a 400mm focal, I'll think to move for a 550 focal fresnel lens as collector (i have a 380mm actually). So a lot of tweaking to get just 100-110 lumen.
The last move to improve the image brightness is to project on a gain screen (i have a old one with a gain around 2). I 'm not decide yet but i think I'll go for a dalite videospectra (gain 1.5) or an even higher gain with a highpower screen (gain 2.8). Then I think I'll be done for this project 😎
just to come back to the beginning of the topic, I recently got a good deal on ebay for an ASK 1280 impression projector. This projector is a 1998 model and rated for 450lumens. This is a 13" lcd system similar to our diy ones. The fresnels are 200/320mm focal, there is a condenser lens of 100mm diameter (focal 140mm) and a varifocal 305-330mm focal with an aperture of 54mm.
They get 450lumens with an HMP 575w double ended bulb. This bulb gives around 50000lumen.
As the ballast and bulb are dead, I'll put a 400W MH in this projector. Doing this I' ll loose some lumen since the bulb gives only(!!) 30000lumens.
If the lumen decrease just linearly with the bulb output, I should get around 270lumen.
If I get that I'll be very happy.
But as the arc lengh will be bigger , I'll certainly loose more lumens..
So to conclude after that long story. A professional projector sold in 1998 roughly 20000$ 🙂bigeyes: ) which use the same setup than our DIY projector with a 400w mh bulb will not give more then 200-300 lumens. FAR From 1300 lumen values announced by some people who want to make some money through DIY projector business...
So now two point to solve:
1)Does your math for precondensor lens ,Rox, are good for the industrial setting of the ask impression (in a first quick calculation, I think it is)
2) does a 550mm fresnel and the 135mm beseler triplet will help me to score 300 lumens on this projector. And does the triplet works for a 13" lcd.
let see.
EDIT: I didn't mention it but I have a luxmeter😉
Fortunately, I realize that my experimental setup match the theoritical place for the precondensor lens in my conditions. But As highlighted in this topic, to improve the lumen is a multifactorial organisation.
I finally succeed to get 100-110 lumen with a 7'lcd and a 250w metal halide by adding a precondenser lens but also by increasing the diameter of my projection lens (125mm diameter , focal 412mm) and also by decreasing the fresnel condenser focal (by adding 2 lens 220+330 instead of one) then decreasing the distance between light and lcd. Finally due to my need for a projection lens with a 400mm focal, I'll think to move for a 550 focal fresnel lens as collector (i have a 380mm actually). So a lot of tweaking to get just 100-110 lumen.
The last move to improve the image brightness is to project on a gain screen (i have a old one with a gain around 2). I 'm not decide yet but i think I'll go for a dalite videospectra (gain 1.5) or an even higher gain with a highpower screen (gain 2.8). Then I think I'll be done for this project 😎
just to come back to the beginning of the topic, I recently got a good deal on ebay for an ASK 1280 impression projector. This projector is a 1998 model and rated for 450lumens. This is a 13" lcd system similar to our diy ones. The fresnels are 200/320mm focal, there is a condenser lens of 100mm diameter (focal 140mm) and a varifocal 305-330mm focal with an aperture of 54mm.
They get 450lumens with an HMP 575w double ended bulb. This bulb gives around 50000lumen.
As the ballast and bulb are dead, I'll put a 400W MH in this projector. Doing this I' ll loose some lumen since the bulb gives only(!!) 30000lumens.
If the lumen decrease just linearly with the bulb output, I should get around 270lumen.
If I get that I'll be very happy.
But as the arc lengh will be bigger , I'll certainly loose more lumens..
So to conclude after that long story. A professional projector sold in 1998 roughly 20000$ 🙂bigeyes: ) which use the same setup than our DIY projector with a 400w mh bulb will not give more then 200-300 lumens. FAR From 1300 lumen values announced by some people who want to make some money through DIY projector business...
So now two point to solve:
1)Does your math for precondensor lens ,Rox, are good for the industrial setting of the ask impression (in a first quick calculation, I think it is)
2) does a 550mm fresnel and the 135mm beseler triplet will help me to score 300 lumens on this projector. And does the triplet works for a 13" lcd.
let see.
EDIT: I didn't mention it but I have a luxmeter😉
great, nice to hear that from you.
mm, now that you have a luxmeter as well, I would e interested on knowing your LCD trasmitance. My hami is 5.2% trasmissive.
mm, now that you have a luxmeter as well, I would e interested on knowing your LCD trasmitance. My hami is 5.2% trasmissive.
yes it's 5% for my lilli.
sorry I did not estimate up to 0.1% 😀
ok I can 5 measure and give you the mean and S.D if you like but i think 5% is a good estimate and more precision is useless.
In contrast, i did not yet measure the transmitance for the 13" ask lcd
sorry I did not estimate up to 0.1% 😀
ok I can 5 measure and give you the mean and S.D if you like but i think 5% is a good estimate and more precision is useless.
In contrast, i did not yet measure the transmitance for the 13" ask lcd
mm thanks; yes, i just divided my two numbers and gave me 5,2% do no worry about the precision. ;D. Actually very close to the hami 😀. Some "users" are claiming 20% trasmitancy on their lcd, and they even don´t have a luxmeter 😀.
do you find 110 lumens confortable for what image size?
do you find 110 lumens confortable for what image size?
I have a 1.6m base image. This is comfortable for me since i'm seated at 3-3.5m. This give between 60-75 lux on the image depending on the area.
Well this is sufficient to watch tv during day time (at least when there is no sun in front of the house between 5pm to 10AM).
You have to consider I have a glass beaded screen (gain 2) which improve greatly the brightness.
If you wan to have an idea go There
In the last two page there is picture with or without flash and with some white sheet of paper on the screen to see the difference between gain 1 and 2. This picture has been done with a 10cm diameter projection lens (and not the 12,5cm) then there is 80 lumens in this setting.
EDIT: you might also notice that the top right corner is not as luminous then other part. My condenser was not set to the right place and is a bit small (54mm by54mm). I hope that with your math I'll be able to adjust it 🙂
Well this is sufficient to watch tv during day time (at least when there is no sun in front of the house between 5pm to 10AM).
You have to consider I have a glass beaded screen (gain 2) which improve greatly the brightness.
If you wan to have an idea go There
In the last two page there is picture with or without flash and with some white sheet of paper on the screen to see the difference between gain 1 and 2. This picture has been done with a 10cm diameter projection lens (and not the 12,5cm) then there is 80 lumens in this setting.
EDIT: you might also notice that the top right corner is not as luminous then other part. My condenser was not set to the right place and is a bit small (54mm by54mm). I hope that with your math I'll be able to adjust it 🙂
Hi Renan,
Would you please describe the whole light engine of that ASK projector, or may be you could post some pictures here? What kind of reflector they used there? Spherical, or something other, how big it is, etc. Also the placement of bulb, reflector, condensor and fresnel is important, people are debating how to set those parts, it would be worth to see how they did it in a commercial product (should be done well, and we have hardly any chance to invent something better)
Regards
Would you please describe the whole light engine of that ASK projector, or may be you could post some pictures here? What kind of reflector they used there? Spherical, or something other, how big it is, etc. Also the placement of bulb, reflector, condensor and fresnel is important, people are debating how to set those parts, it would be worth to see how they did it in a commercial product (should be done well, and we have hardly any chance to invent something better)
Regards
why not!
You 'll find some picture there
spherical reflector diameter 65mm (same one that you can find on izzotek.com
bulb is hmp 575w double ended (rated around 49000lumens, arc is 9mm) with ballast and starter.
precondensator lens diameter 100mm focal 140mm
You might have an idea on the picture of how far are bulb reflector and precondensator. I do not have the measurement righ now, I'll edit later. I think that the precondensor is 2-3cm from the center of the bulb
from the flat side of the precondensor there is 18 cm to the first fresnel.
first fresnel (F1) have a 200mm focal.
In fact the light box is tilted by a 10-15degre angle and the f1 is also tilted.
F1-lcd 3cm
LCD to F2 2cm .
F2 is straight (not tilted) focal 320mm
then a varifocal projction lens 54mm of usefull diameter and 305-330mm focal.
So finally very classic compared to what we do in our diy projector
I had to measure the transmittance for the lcd, I think his will be bigger than for a 7' lilliput because the contrast for that project is 1:100 (pretty bad) but as by replacing the bulb and ballast by 400w metal halide I'll certainly loose half of the lumens so I hope to get a better contrast!
This is also a good point for measurement in our system: what contrast we get??? nobody have an idea??
I did some quick measurement with my 7' setup and it's close to 1:100
You 'll find some picture there
spherical reflector diameter 65mm (same one that you can find on izzotek.com
bulb is hmp 575w double ended (rated around 49000lumens, arc is 9mm) with ballast and starter.
precondensator lens diameter 100mm focal 140mm
You might have an idea on the picture of how far are bulb reflector and precondensator. I do not have the measurement righ now, I'll edit later. I think that the precondensor is 2-3cm from the center of the bulb
from the flat side of the precondensor there is 18 cm to the first fresnel.
first fresnel (F1) have a 200mm focal.
In fact the light box is tilted by a 10-15degre angle and the f1 is also tilted.
F1-lcd 3cm
LCD to F2 2cm .
F2 is straight (not tilted) focal 320mm
then a varifocal projction lens 54mm of usefull diameter and 305-330mm focal.
So finally very classic compared to what we do in our diy projector
I had to measure the transmittance for the lcd, I think his will be bigger than for a 7' lilliput because the contrast for that project is 1:100 (pretty bad) but as by replacing the bulb and ballast by 400w metal halide I'll certainly loose half of the lumens so I hope to get a better contrast!
This is also a good point for measurement in our system: what contrast we get??? nobody have an idea??
I did some quick measurement with my 7' setup and it's close to 1:100
Mille fois merci, this is very valuable information. First, that we do not make any fundamental mistake with our design, and second, that there is probably no simple, magical way to improve brightness, as engineers of that projector would have done it years ago.
Looks, like that precondensor lens is very thick, so focal lenght should be pretty short. I wonder why it didn't work in my case, I tried a short focal precondensor (about 8cm FL) and there was no way the light would cover the whole LCD. I think due to a big bulb with an outer envelope, it was impossible to set the precondensor close enough to the source. I see there is a bulb without an outer envelope in your projector, so the lens is closer to the source, and it gathers more light.
About the contrast, I think we get very crappy results, because of this poor brightness. Although our LCDs claim values like 400:1, 500:1 or even more, even if it's true, this value is mostly wasted, because we pull the brightness as much as possible (both with LCD controls and at signal source, PC or tuner). Then we loose huge percentage of the tonal space, otherwise most of the picture would be very dark. There is something visible in those dark areas, but due to the low amount of light we don't see it, so we "amplify" the input signal to make those areas visible, and so we "cancel" some portion of the contrast range. Like there is 400 steps of brightness, but we don't use first 200 or 300 or so. This range would be fully utilized, but much more light is needed. This is my opinion. I have no means to measure the contrast, but I think it's very poor, in the range of 100:1 or so. And the brightness of about 100-150lm would be a typical value for our designs. Sometimes I see people quoting numbers like 1000lm, this is crazy.
Regards
Looks, like that precondensor lens is very thick, so focal lenght should be pretty short. I wonder why it didn't work in my case, I tried a short focal precondensor (about 8cm FL) and there was no way the light would cover the whole LCD. I think due to a big bulb with an outer envelope, it was impossible to set the precondensor close enough to the source. I see there is a bulb without an outer envelope in your projector, so the lens is closer to the source, and it gathers more light.
About the contrast, I think we get very crappy results, because of this poor brightness. Although our LCDs claim values like 400:1, 500:1 or even more, even if it's true, this value is mostly wasted, because we pull the brightness as much as possible (both with LCD controls and at signal source, PC or tuner). Then we loose huge percentage of the tonal space, otherwise most of the picture would be very dark. There is something visible in those dark areas, but due to the low amount of light we don't see it, so we "amplify" the input signal to make those areas visible, and so we "cancel" some portion of the contrast range. Like there is 400 steps of brightness, but we don't use first 200 or 300 or so. This range would be fully utilized, but much more light is needed. This is my opinion. I have no means to measure the contrast, but I think it's very poor, in the range of 100:1 or so. And the brightness of about 100-150lm would be a typical value for our designs. Sometimes I see people quoting numbers like 1000lm, this is crazy.
Regards
about the precondensor of the ask 1280 it is clearly very curved and I'm surprise that the focal is not lower than the 140mm I measured.
However the lens is flat on the bulb side and on this flat side , the surface is not full. There is a hole and the lens in curved inside to.
I'll post a picture here this evening.
using a precondensor lens of 75mm focal in my lilli setup is ok with a 250w mh but not with a 400w mh since the bulb is too large and my precondensor is not large enough too.
However the lens is flat on the bulb side and on this flat side , the surface is not full. There is a hole and the lens in curved inside to.
I'll post a picture here this evening.
using a precondensor lens of 75mm focal in my lilli setup is ok with a 250w mh but not with a 400w mh since the bulb is too large and my precondensor is not large enough too.
mmm, about the 100:1, how did you meassure it? do you remenber your 2 values?
thanks for the allimbox link, but no idea of french ;D. Nice pics 😀.
thanks for the allimbox link, but no idea of french ;D. Nice pics 😀.
the link is only for the picture (I'm too lazy to put here🙄 )
measurements has been done on ansi picture (black and white) by measuring lux on black and white on 1.6 base image.
with the 10cm diameter projection lens white 51lux black 0.5
i did measure the white with the 12,5vm projection lens (70lux) but not the black yet.
Ok I still have very weak leakage of ligh through the upper fan and the black value might be very dependant on it.
I'll try to put a curtain around the box and see what happens with the black value. Sure if I decrease to 0.2 this will give the real contrast. let's do some more measurements
measurements has been done on ansi picture (black and white) by measuring lux on black and white on 1.6 base image.
with the 10cm diameter projection lens white 51lux black 0.5
i did measure the white with the 12,5vm projection lens (70lux) but not the black yet.
Ok I still have very weak leakage of ligh through the upper fan and the black value might be very dependant on it.
I'll try to put a curtain around the box and see what happens with the black value. Sure if I decrease to 0.2 this will give the real contrast. let's do some more measurements
mmm, i wouldn´t consider those two values as accurate enough for the contrast meassurment;
the white meassurment is fine, but the black meassuremnt is very low. I mean the luxmeter accuracy is not very good i believe. I had this same problem with my luxmeter, black levels are hard to meassure. I thought about a solution though;
if you meassure somewhere closer to the projection lens, both values will be much higher. I took 2000 lux just close to the lens (white). Then the black meassuremnt will be somewhere 20 lux if the contrast is 100:1. The 20 lux meassurement on the luxmeter is much believable in my opinion than the 0.5 one.
I bet you don´t have a 100:1 contrast projector 😀, let me know if you take a chance to meassure close to the lens.
the white meassurment is fine, but the black meassuremnt is very low. I mean the luxmeter accuracy is not very good i believe. I had this same problem with my luxmeter, black levels are hard to meassure. I thought about a solution though;
if you meassure somewhere closer to the projection lens, both values will be much higher. I took 2000 lux just close to the lens (white). Then the black meassuremnt will be somewhere 20 lux if the contrast is 100:1. The 20 lux meassurement on the luxmeter is much believable in my opinion than the 0.5 one.
I bet you don´t have a 100:1 contrast projector 😀, let me know if you take a chance to meassure close to the lens.
- Status
- Not open for further replies.