Myk that is VERY impressive...one of the best looking i have seen (image wise).
GOt a question, what type of heat does the exhust hole put out (you have a fan blowing in, where does the heat come out?) OverHead projectors put out about as much as a Haridryer on low... is this roughly the same?
thanks
Clok
GOt a question, what type of heat does the exhust hole put out (you have a fan blowing in, where does the heat come out?) OverHead projectors put out about as much as a Haridryer on low... is this roughly the same?
thanks
Clok
Hi Xblocker,
You are correct that in itself, a lens would not inrease contrast in a classical optical system. If anything, a lens introduced into the optical path would tend to decrease contrast due to distortions, internal reflections, and scattering etc.
A "contrast lens" would be nice indeed!
Where this system doesn't fit the classical model is that most LCDs are acting not as a passive filter in a larger optical path (as a transparency slide would be), but as an active image source - an image source that is both semi-diffuse, and larger than the primary objective lens.
As you say, contrast is inherently determined by the quality of the light source and is definitely limited by the qualities of the LCD's ability to block out the light. Unfortunately, the OHP method of forcing light through the LCD at an angle which eventually hits the objective lens is definitely NOT a classical optical system - it is a hybrid and a compromise. A compromise which, while it is surprisingly effective, it is not completely effective, creating the scatter you mentioned that reduces contrast.
My point was that the LCD itself causes scattering, and as a result, not all the light from the much-larger image of an OHP panel hits the objective (and some light that does, shouldn't).
A much larger lens introduced in the optical path after the LCD will ensure all the light from the LCD gets through to the objective as a focussed beam, and this increase in overall light, coupled with the fact that the objective is focussing ALL the light from the lens, not just a subset, WILL increase final image contrast. (If you can stop scattering "softening" the black levels, and increase the brightness of the bright levels by capturing all available light, you have actually increased contrast...)
The problem with most optics theories - and scientific theories in general - is that they cover ideal cases. Very few talk about practical scenarios e.g. where the image is both semi-diffuse AND larger than the lens; the classical theories work on the principle that the image is infinitely small compared to lens dimensions...
In the end, I would say, try it.
Your LCD might be less diffuse than mine. I have experience of a Proxima 9100 projector, a Sharp QA 1100 Panel, and an nView Spectra C.
I went the OHP route with the Sharp and a Dukane SunSplash (4000 ANSI Lumens). I then got hold of the DOA Proxima 9100 which is just an OHP with built in panel.
What intrigued me was the fresnel lens AFTER the LCD on the Proxima. Anyway, after much playing, the guts of the proxima were removed, and replaced by the working Spectra C panel. Then the experimentation started, and after similar playing with my Sharp/OHP, I increased the contrast by placing a Staples Page Magnifier and a 6" DCV lens from SurplusShed between the LCD and the main objective. (Without the DCV, the image was too zoomed with the Sunsplash objective.)
Bill.
P.S. In practice, a Fresnel WILL actually increase contrast in another way... It too, like the LCD has a very narrow working angle: scattered light that is too far from the optical path will not be passed very effectively. In effect, it DOES, in this instance, appear to be that optics holy grail - a "Contrast lens".
You are correct that in itself, a lens would not inrease contrast in a classical optical system. If anything, a lens introduced into the optical path would tend to decrease contrast due to distortions, internal reflections, and scattering etc.
A "contrast lens" would be nice indeed!
Where this system doesn't fit the classical model is that most LCDs are acting not as a passive filter in a larger optical path (as a transparency slide would be), but as an active image source - an image source that is both semi-diffuse, and larger than the primary objective lens.
As you say, contrast is inherently determined by the quality of the light source and is definitely limited by the qualities of the LCD's ability to block out the light. Unfortunately, the OHP method of forcing light through the LCD at an angle which eventually hits the objective lens is definitely NOT a classical optical system - it is a hybrid and a compromise. A compromise which, while it is surprisingly effective, it is not completely effective, creating the scatter you mentioned that reduces contrast.
My point was that the LCD itself causes scattering, and as a result, not all the light from the much-larger image of an OHP panel hits the objective (and some light that does, shouldn't).
A much larger lens introduced in the optical path after the LCD will ensure all the light from the LCD gets through to the objective as a focussed beam, and this increase in overall light, coupled with the fact that the objective is focussing ALL the light from the lens, not just a subset, WILL increase final image contrast. (If you can stop scattering "softening" the black levels, and increase the brightness of the bright levels by capturing all available light, you have actually increased contrast...)
The problem with most optics theories - and scientific theories in general - is that they cover ideal cases. Very few talk about practical scenarios e.g. where the image is both semi-diffuse AND larger than the lens; the classical theories work on the principle that the image is infinitely small compared to lens dimensions...
In the end, I would say, try it.
Your LCD might be less diffuse than mine. I have experience of a Proxima 9100 projector, a Sharp QA 1100 Panel, and an nView Spectra C.
I went the OHP route with the Sharp and a Dukane SunSplash (4000 ANSI Lumens). I then got hold of the DOA Proxima 9100 which is just an OHP with built in panel.
What intrigued me was the fresnel lens AFTER the LCD on the Proxima. Anyway, after much playing, the guts of the proxima were removed, and replaced by the working Spectra C panel. Then the experimentation started, and after similar playing with my Sharp/OHP, I increased the contrast by placing a Staples Page Magnifier and a 6" DCV lens from SurplusShed between the LCD and the main objective. (Without the DCV, the image was too zoomed with the Sunsplash objective.)
Bill.
P.S. In practice, a Fresnel WILL actually increase contrast in another way... It too, like the LCD has a very narrow working angle: scattered light that is too far from the optical path will not be passed very effectively. In effect, it DOES, in this instance, appear to be that optics holy grail - a "Contrast lens".
Xblocker, Fender4,
You are both there on the LCD vs DMD front.
The monochrome arrays allow more light through because for a given panel size, there are less transistors (I think), and also, the area per pixel is exclusively devoted to one colour - allowing three times the area and light throughput.
The problem with the panel speed is the reason for modern LCD projectors going to three monochrome panels...
Anyone up for building a three panel projector?
Bill.
You are both there on the LCD vs DMD front.
The monochrome arrays allow more light through because for a given panel size, there are less transistors (I think), and also, the area per pixel is exclusively devoted to one colour - allowing three times the area and light throughput.
The problem with the panel speed is the reason for modern LCD projectors going to three monochrome panels...
Anyone up for building a three panel projector?
Bill.
help with LCD!!!
hi, i have a TOSHIBA LTM10C039 LCD that does 800x600, but i can't find a controller that will run it...im desparate because i really cant afford another display AND a controller for my projector.
this screen is out of a Compaq LTE 5200 laptop if that is relevant....
it has a 31 pin connector also, the "trapazoid" patern with 15 pins on top and 16 pins on bottom, slightly offset..
this screen seems to be in great working condition, i can power up the laptop and get a nice picture without any dead spots. i have been toying with a makeshift projector with a convex lense, a shop light, a piece of glass, and some cardboard and im really excited about getting this thing up and running..
thanks for any help..
hi, i have a TOSHIBA LTM10C039 LCD that does 800x600, but i can't find a controller that will run it...im desparate because i really cant afford another display AND a controller for my projector.
this screen is out of a Compaq LTE 5200 laptop if that is relevant....
it has a 31 pin connector also, the "trapazoid" patern with 15 pins on top and 16 pins on bottom, slightly offset..
this screen seems to be in great working condition, i can power up the laptop and get a nice picture without any dead spots. i have been toying with a makeshift projector with a convex lense, a shop light, a piece of glass, and some cardboard and im really excited about getting this thing up and running..
thanks for any help..
Well I turned it on tonight in the shop and omg I think its gonna be perfect. As for the heat issue I still need a 1' square pane of Low-e glass and to find a vent cover for the top air outlet above the light. Then I can bring it in to test on the big screen, the projection lens is going to be changed to so I can put it 17' away and still have a small image.
woneill
very interesting topic of using fresnel panel as a contrast lens, but I have some questions:
-. since fresnel panel basically is oversize lens, if we put object in a distance smaller than fresnel panel focal length, doesn't it produce spreading light, means unfocused object?
-. do you know fresnel panel focal length you are using? Is it the same as common OHP fresnel panel? (it places beetwen lamp and LCD panel)
-. do you know the DCV focal length? and how far it should be placed from the fresnel panel?
-. if you have pictures of projected image, please share with us in this forum.
I apologizes for so much asking, because it's some kind of new idea for me.
very interesting topic of using fresnel panel as a contrast lens, but I have some questions:
-. since fresnel panel basically is oversize lens, if we put object in a distance smaller than fresnel panel focal length, doesn't it produce spreading light, means unfocused object?
-. do you know fresnel panel focal length you are using? Is it the same as common OHP fresnel panel? (it places beetwen lamp and LCD panel)
-. do you know the DCV focal length? and how far it should be placed from the fresnel panel?
-. if you have pictures of projected image, please share with us in this forum.
I apologizes for so much asking, because it's some kind of new idea for me.
Hi Gunawan,
Good questions!
Placing an image source at a shorter distance than the focal length in a classical optical system would not converge the naturally divergent image source rays enough to produce a "real" image - it would produce a "virtual" image - where the naturally divergent rays from a point on the image source would diverge a little less after passing through the lens.
Optically, with the rays diverging at a reduced angle, it would appear to a "viewer" through the lens that the original point source had been effectively moved further away from the viewer. This is how a reading/magnifying glass works - you can literally get closer to something while your eyes can focus the image as if it were further away. The end result is that the original image is magnified.
The break point in a normal magnifying system is the point where the original image is moved further away from the lens to the point where it and reaches the lens's focal point. Here, the image source has effectively been moved away to infinity, and the rays through the lens are converged to be parallel to each other - like sun's rays.
Now, back to the projection system: we generally have an optical system where the light source, after passing through the OHP fresnels and the lcd is "broadly" convergent on the main objective lens. The system at this point is not perfect, with multiple light paths through the fresnel, and there is some additional scattering (and divergence) after passing through the LCD itself. But, most of the light is going where it should.
The beauty of the original OHP design is that the long(ish) distance to the objective from the original fresnel and transparency, the smaller size of the objective compared to the fresnel, and the right-angle bend in the path, means that much of light that is scattered due to imperfections in the optical path (and many of the non-optimal paths from a non-point light source), hits the ceiling and not the screen.
The design is effectively reducing the apeture of the system at the objective, and selecting a subset of the emerging rays that all follow the same(ish) optical path. Very clever - but very wasteful - trading efficiency against quality...
Ok, my plan is to reduce the amount of light hitting the ceiling because of scattering by the LCD. Stuff that was going there anyway due to imperfections in the original point-source light source will still go there. The fresnel is not a magic bullet fixing this. For that, you need a better point light source as everyone has found out.
Thus, to set things up, you start by placing the fresnel an inch or two from the panel. The best way to determine this distance is to light up the panel with a diffuse light source, and use your eye as the objective... (I originally experimented with my laptop and a Staples page magnifier with a focal length of roughly 4 - 6 inches - easier to get a feel of things this way than burning your eyes out...)
Keeping your eye close to a focal length away from the lens (the fresnels have a VERY narrow viewing angle - the thing I was alluding to at the end of my last post), if you start with the lens about an inch away from the panel with the grooves facing it, and move it forward and backwards, while moving your eye forwards and backwards too, you can get a "feel" for the point where you are looking through the lens at the screen without losing much detail due to moire interference or distortion due to the crap lens. (Lens about 2 inches from the panel, eye about 4-6 inches from the lens.)
If, when doing this, your eye is too close to the lens, or too far away, you see only a small portion of the image due to the fresnel only being able to work with a narrow angle of source image beam divergence. What I mean by this is: The fresnel is optimised to take parallel rays of light and converge them to its focal point. It is not that good at taking massively divergent rays, or massively convergent rays and re-shaping them. Internal reflections, and imperfections in the design just absorb or scatter the light so far away from the optical path that, in our case, it serves our purpose as a contrast filter.
Thus, the ideal optical path of an OHP, and the mild divergences introduced by the LCD are nicely compatible with the fresnel's operating range.
If you have "played" with the lens like this, you will quickly get a "feel" for the optimal distance of the fresnel from the screen, and the optimal point where the resulting light path should be "viewed" from. The optimal view point is where, on the OHP system, I placed the divergent DCV lens. This lens would optimally have roughly the opposite power of the fresnel. I.e. if the the fresnel had a focal length of 12 cm, the DCV lens would have a focal length of -12 cm. (Ideally, the focal length of the fresnel would be as long as possible - reducing the effects of distortions, and increasing the optimal distance from the fresnel to the objective. Mine had a fl of about 12 cm, and the DCV had -35cm, resulting in an overall (2x - ish) magnification of the image by the fresnel-DCV combination, and requiring focussing the main objective closer to the panel - I eventually countered this by moving the light source closer to the OHP fresnel.)
Essentially, the DCV lens (a suitable PCV or negative focal length meniscus would probably be better, but I work with what is available), was introduced at the "eye point" to "straighten" out the projected rays at the optimal viewing point of the fresnel - i.e. lengthen the resulting light cone so that it was closer to the original design spec of the OHP objective. Any further away, and the gross limitations of the fresnel would destroy the image... Similarly, with the objective alone at the optimal point, the image was too zoomed when I eventially got it focussed...
The end result in all this was possibly a 20% (subjective - no ANSI measurements here) increase in brightness and quality over the combination of OHP+Panel alone, due to more of the light from the LCD reaching the objective along an optimal light path.
Subsequently moving the light source slightly closer to the LCD panel than standard, causeed the rays at the OHP fresnel to emerge more parallel than normal. They went straight through the LCD panel - rather than being converged. This also seemed to help slightly too, but I'm not sure if it made a real difference overall. (This was possible because the post-LCD fresnel was now performing the task of converging the light rays to the objective.)
Basically, the issue with contrast is this - where there are multiple "paths" through the objective that need focussing (i.e. where you use the objective to "choose" the "best" part of the image to focus on), there will be a reduction in contrast.
Ideally, if the light source is a pure point source, collimated perfectly and passed through a non diffuse image such that there is only one possible light path to "choose from", then you will get maximum contrast. (I couldn't verbalise this point earlier when trying to answer Xblocker.)
Where the light source is not a perfect point source, or distortions and diffusions in the path produce multiple points of focus (or a single smeared focal point) - requiring the user to choose the "best" point of focus - there WILL be a reduction of contrast, because light rays from an alternate projected image slightly in front of, or behind the image "chosen" will smear the chosen image - adding light to the black areas...
What I am describing is a possible way to enhance the OHP's ability to reject the false paths while using more of the available light. I am not proposing, as it may first seem, to change the laws of physics. I am also not claiming to know more than anybody else here - what you guys here (even Aleksey) have contributed to the field is phenominal. You guys are awesome! I am simply trying to hit things from a slightly different angle, and explore how your successes can be optimised even further.
Bill.
P.S. No pictures yet - my gf "lost" my beautiful digital camera (Fuji Finepix 4800) while on a trip to London, and I am still awaiting a replacement... (Don't know yet whether it is to be a replacement gf or replacement camera...)
P.P.S. A fresnel from an OHP is probaly suitable to experiment with, but I would probably use only one of the two panels to reduce scattering and increase focal length. Also, I would worry that the number of grooves per inch on the OHP fresnel might be too low - exacerbating moire effects. Try it, if you have one spare, against a laptop - you will quickly find out if it helps!
Good questions!
Placing an image source at a shorter distance than the focal length in a classical optical system would not converge the naturally divergent image source rays enough to produce a "real" image - it would produce a "virtual" image - where the naturally divergent rays from a point on the image source would diverge a little less after passing through the lens.
Optically, with the rays diverging at a reduced angle, it would appear to a "viewer" through the lens that the original point source had been effectively moved further away from the viewer. This is how a reading/magnifying glass works - you can literally get closer to something while your eyes can focus the image as if it were further away. The end result is that the original image is magnified.
The break point in a normal magnifying system is the point where the original image is moved further away from the lens to the point where it and reaches the lens's focal point. Here, the image source has effectively been moved away to infinity, and the rays through the lens are converged to be parallel to each other - like sun's rays.
Now, back to the projection system: we generally have an optical system where the light source, after passing through the OHP fresnels and the lcd is "broadly" convergent on the main objective lens. The system at this point is not perfect, with multiple light paths through the fresnel, and there is some additional scattering (and divergence) after passing through the LCD itself. But, most of the light is going where it should.
The beauty of the original OHP design is that the long(ish) distance to the objective from the original fresnel and transparency, the smaller size of the objective compared to the fresnel, and the right-angle bend in the path, means that much of light that is scattered due to imperfections in the optical path (and many of the non-optimal paths from a non-point light source), hits the ceiling and not the screen.
The design is effectively reducing the apeture of the system at the objective, and selecting a subset of the emerging rays that all follow the same(ish) optical path. Very clever - but very wasteful - trading efficiency against quality...
Ok, my plan is to reduce the amount of light hitting the ceiling because of scattering by the LCD. Stuff that was going there anyway due to imperfections in the original point-source light source will still go there. The fresnel is not a magic bullet fixing this. For that, you need a better point light source as everyone has found out.
Thus, to set things up, you start by placing the fresnel an inch or two from the panel. The best way to determine this distance is to light up the panel with a diffuse light source, and use your eye as the objective... (I originally experimented with my laptop and a Staples page magnifier with a focal length of roughly 4 - 6 inches - easier to get a feel of things this way than burning your eyes out...)
Keeping your eye close to a focal length away from the lens (the fresnels have a VERY narrow viewing angle - the thing I was alluding to at the end of my last post), if you start with the lens about an inch away from the panel with the grooves facing it, and move it forward and backwards, while moving your eye forwards and backwards too, you can get a "feel" for the point where you are looking through the lens at the screen without losing much detail due to moire interference or distortion due to the crap lens. (Lens about 2 inches from the panel, eye about 4-6 inches from the lens.)
If, when doing this, your eye is too close to the lens, or too far away, you see only a small portion of the image due to the fresnel only being able to work with a narrow angle of source image beam divergence. What I mean by this is: The fresnel is optimised to take parallel rays of light and converge them to its focal point. It is not that good at taking massively divergent rays, or massively convergent rays and re-shaping them. Internal reflections, and imperfections in the design just absorb or scatter the light so far away from the optical path that, in our case, it serves our purpose as a contrast filter.
Thus, the ideal optical path of an OHP, and the mild divergences introduced by the LCD are nicely compatible with the fresnel's operating range.
If you have "played" with the lens like this, you will quickly get a "feel" for the optimal distance of the fresnel from the screen, and the optimal point where the resulting light path should be "viewed" from. The optimal view point is where, on the OHP system, I placed the divergent DCV lens. This lens would optimally have roughly the opposite power of the fresnel. I.e. if the the fresnel had a focal length of 12 cm, the DCV lens would have a focal length of -12 cm. (Ideally, the focal length of the fresnel would be as long as possible - reducing the effects of distortions, and increasing the optimal distance from the fresnel to the objective. Mine had a fl of about 12 cm, and the DCV had -35cm, resulting in an overall (2x - ish) magnification of the image by the fresnel-DCV combination, and requiring focussing the main objective closer to the panel - I eventually countered this by moving the light source closer to the OHP fresnel.)
Essentially, the DCV lens (a suitable PCV or negative focal length meniscus would probably be better, but I work with what is available), was introduced at the "eye point" to "straighten" out the projected rays at the optimal viewing point of the fresnel - i.e. lengthen the resulting light cone so that it was closer to the original design spec of the OHP objective. Any further away, and the gross limitations of the fresnel would destroy the image... Similarly, with the objective alone at the optimal point, the image was too zoomed when I eventially got it focussed...
The end result in all this was possibly a 20% (subjective - no ANSI measurements here) increase in brightness and quality over the combination of OHP+Panel alone, due to more of the light from the LCD reaching the objective along an optimal light path.
Subsequently moving the light source slightly closer to the LCD panel than standard, causeed the rays at the OHP fresnel to emerge more parallel than normal. They went straight through the LCD panel - rather than being converged. This also seemed to help slightly too, but I'm not sure if it made a real difference overall. (This was possible because the post-LCD fresnel was now performing the task of converging the light rays to the objective.)
Basically, the issue with contrast is this - where there are multiple "paths" through the objective that need focussing (i.e. where you use the objective to "choose" the "best" part of the image to focus on), there will be a reduction in contrast.
Ideally, if the light source is a pure point source, collimated perfectly and passed through a non diffuse image such that there is only one possible light path to "choose from", then you will get maximum contrast. (I couldn't verbalise this point earlier when trying to answer Xblocker.)
Where the light source is not a perfect point source, or distortions and diffusions in the path produce multiple points of focus (or a single smeared focal point) - requiring the user to choose the "best" point of focus - there WILL be a reduction of contrast, because light rays from an alternate projected image slightly in front of, or behind the image "chosen" will smear the chosen image - adding light to the black areas...
What I am describing is a possible way to enhance the OHP's ability to reject the false paths while using more of the available light. I am not proposing, as it may first seem, to change the laws of physics. I am also not claiming to know more than anybody else here - what you guys here (even Aleksey) have contributed to the field is phenominal. You guys are awesome! I am simply trying to hit things from a slightly different angle, and explore how your successes can be optimised even further.
Bill.
P.S. No pictures yet - my gf "lost" my beautiful digital camera (Fuji Finepix 4800) while on a trip to London, and I am still awaiting a replacement...
(Don't know yet whether it is to be a replacement gf or replacement camera...) P.P.S. A fresnel from an OHP is probaly suitable to experiment with, but I would probably use only one of the two panels to reduce scattering and increase focal length. Also, I would worry that the number of grooves per inch on the OHP fresnel might be too low - exacerbating moire effects. Try it, if you have one spare, against a laptop - you will quickly find out if it helps!
Woneill,
Difficult choice new GF or new camera.
If I read you right you are suggesting a more parallel light path through the panel than with a normal OHP and a second fresnel after the panel converging down to the projection lens.
Since the second fresnal is in the image path it will have to be of very good quality with considerably more lines per inch and likely to be expensive.
A lot of activity here is to increase the light source power which gives an immediate increase in light and an increase in bulb life compared to usual OHP lights but improving contrast is interesting. Try turning your panel round so the front faces the light. I tried this and found a slight increase in light transmision and a marginal improvement in contrast. Couldnt get at the buttons on the panel so its back the normal way.
Difficult choice new GF or new camera.
If I read you right you are suggesting a more parallel light path through the panel than with a normal OHP and a second fresnel after the panel converging down to the projection lens.
Since the second fresnal is in the image path it will have to be of very good quality with considerably more lines per inch and likely to be expensive.
A lot of activity here is to increase the light source power which gives an immediate increase in light and an increase in bulb life compared to usual OHP lights but improving contrast is interesting. Try turning your panel round so the front faces the light. I tried this and found a slight increase in light transmision and a marginal improvement in contrast. Couldnt get at the buttons on the panel so its back the normal way.
woneill
very well explanation, thank's.
Long ago, someone, I forgot the name, posted the link:
http://www.display-optics.com/pdf/condenser_lenses.PDF
As we already knew the basic principle of passing the light thru LCD panel to get the best result is:
1. the light must be in 90 degrees when hit the panel.
2. the light must be in the same polarizing with the panel.
I think your setup is similar with the diagram in that link and solve the problem #1 above.
It will improve our setup some step ahead, since most of us don't have perfect point source light, perfect objective lens and rather oversized LCD panel.
But, same as Remp, I'm worrying about putting fresnel in front of the LCD panel, should we use very good quality fresnel panel? Or, because it placed so close to the LCD, it won't reduce projected image quality.
I did some test with OHP fresnel panel (12" fl), placed 1/2" in front of the PC monitor, it has moire effect, further away, moire effect gone but it started distorting image, horizontal and vertical lines bended to the center. Is it normal?
You mentioned that you used DCV to get correct size image produced by OHP objective lens, what if I use longer objective lens and eliminate DCV lens to get the same result? Is there other purpose using DCV lens, such as to reduce distance between DCV and objective lens?
Anyway, again I hope this setup will improve our diy projector, too bad I only have 1 fresnel panel, can't do more test....
very well explanation, thank's.
Long ago, someone, I forgot the name, posted the link:
http://www.display-optics.com/pdf/condenser_lenses.PDF
As we already knew the basic principle of passing the light thru LCD panel to get the best result is:
1. the light must be in 90 degrees when hit the panel.
2. the light must be in the same polarizing with the panel.
I think your setup is similar with the diagram in that link and solve the problem #1 above.
It will improve our setup some step ahead, since most of us don't have perfect point source light, perfect objective lens and rather oversized LCD panel.
But, same as Remp, I'm worrying about putting fresnel in front of the LCD panel, should we use very good quality fresnel panel? Or, because it placed so close to the LCD, it won't reduce projected image quality.
I did some test with OHP fresnel panel (12" fl), placed 1/2" in front of the PC monitor, it has moire effect, further away, moire effect gone but it started distorting image, horizontal and vertical lines bended to the center. Is it normal?
You mentioned that you used DCV to get correct size image produced by OHP objective lens, what if I use longer objective lens and eliminate DCV lens to get the same result? Is there other purpose using DCV lens, such as to reduce distance between DCV and objective lens?
Anyway, again I hope this setup will improve our diy projector, too bad I only have 1 fresnel panel, can't do more test....
Hi remp,
Yup, you read me right: my end configuration seems to be a more parallel light path, and fresnel after the LCD - though the more parallel light path had more to do with the effects of the fresnel trailing LCD than a design inspiration...
I don't know - maybe the more parallel light path was actually more valuable than the trailing fresnel in increasing contrast - I never did a regression check.
I'll check out my panel for directionality - though I am surprised yours works the "wrong" way around... I WOULD have expected the panel to be directional if a MLA were included - but I would expect it to be on the "back" side, to collect more light and funnel it through to the front - producing a brighter, but more diffused image in the process... Was it definitely the "front" that was facing the light source when you got a brighter image?
You are right about the quality requirements of the fresnel though... But, if the fresnel is only used in magnifying mode, placed fairly close to the LCD, even the Staples page magnifyer seems to bring benefits.
I would definitely like to play with one of those aspheric fresnel-tech lenses, though - the ones with both aspheric curvature AND aspheric grooves... Mmmm...
Truth is, I really want to get my hands on a suitable panel that is high enough res/contrast, but small enough to fit behind the SurplusShed 5 7/8 " PCX. It is still tempting to stick that over the top of the OHP instead of the fresnel and forget about the pixels at the edge... For a widescreen DVD the corners are all blacked out anyway...
The light source power is still something I am interested in... For the Proxima 9100 thing, I cheated and managed to rig-up a new driver for the MH bulb (575W 750h double ended 95V 7A, 7mm gap). The fixture it was in was the reason I bought the 9100 in the first place. The bulbs are available (Osram/Sylvania or Ushio (???)) from $90.
The Dukane sunsplash uses a FXL 50h bulb which is bright enough for casual use, but not long enough lasting. I'm even tempted to experiment using one of the MH ballasts to implement slow startup/shutof in an attempt to extend the FXL life. My mechanical skills extend to using a screwdriver/soldering iron, so building a creature from scratch is out of my arena...
I think there is still room to explore though...
One area I still want to play with here was first mentioned (and quickly shot down) by Aleksey - the idea of using projector beam auto lamps... I have a couple on my motorbike, and though the power is limited (100W with an "off road" bulb), they seem to make close to 100% use of the available light and produce a VERY collimated beam. I was driving behind my friends who were in an open topped car, and their "finger puppets" could very clearly be seen silhouetted on a large road sign 100 yards away. This indicates a VERY collimated beam, and one that might have potential... (Two of these things fed into a beam merging prism might have useful effects, or maybe multiple collimated beams all firing into a large condenser... TBD... They run cool, though...) For anybody who wishes to experiment, look for "projector fog" on either ebay or Yahoo shopping...
Another possibility I was thinking of is a little more experimental: enclosing multiple LOA tubes in a box whose inside surfaces are front-surface mirrors - with only a single small hole through which the light could escape... Would it work as a high power cool running point source? Shoot me down on this one, folks...
Bill.
Yup, you read me right: my end configuration seems to be a more parallel light path, and fresnel after the LCD - though the more parallel light path had more to do with the effects of the fresnel trailing LCD than a design inspiration...
I don't know - maybe the more parallel light path was actually more valuable than the trailing fresnel in increasing contrast - I never did a regression check.
I'll check out my panel for directionality - though I am surprised yours works the "wrong" way around... I WOULD have expected the panel to be directional if a MLA were included - but I would expect it to be on the "back" side, to collect more light and funnel it through to the front - producing a brighter, but more diffused image in the process... Was it definitely the "front" that was facing the light source when you got a brighter image?
You are right about the quality requirements of the fresnel though... But, if the fresnel is only used in magnifying mode, placed fairly close to the LCD, even the Staples page magnifyer seems to bring benefits.
I would definitely like to play with one of those aspheric fresnel-tech lenses, though - the ones with both aspheric curvature AND aspheric grooves... Mmmm...
Truth is, I really want to get my hands on a suitable panel that is high enough res/contrast, but small enough to fit behind the SurplusShed 5 7/8 " PCX. It is still tempting to stick that over the top of the OHP instead of the fresnel and forget about the pixels at the edge... For a widescreen DVD the corners are all blacked out anyway...
The light source power is still something I am interested in... For the Proxima 9100 thing, I cheated and managed to rig-up a new driver for the MH bulb (575W 750h double ended 95V 7A, 7mm gap). The fixture it was in was the reason I bought the 9100 in the first place. The bulbs are available (Osram/Sylvania or Ushio (???)) from $90.
The Dukane sunsplash uses a FXL 50h bulb which is bright enough for casual use, but not long enough lasting. I'm even tempted to experiment using one of the MH ballasts to implement slow startup/shutof in an attempt to extend the FXL life. My mechanical skills extend to using a screwdriver/soldering iron, so building a creature from scratch is out of my arena...
I think there is still room to explore though...
One area I still want to play with here was first mentioned (and quickly shot down) by Aleksey - the idea of using projector beam auto lamps... I have a couple on my motorbike, and though the power is limited (100W with an "off road" bulb), they seem to make close to 100% use of the available light and produce a VERY collimated beam. I was driving behind my friends who were in an open topped car, and their "finger puppets" could very clearly be seen silhouetted on a large road sign 100 yards away. This indicates a VERY collimated beam, and one that might have potential... (Two of these things fed into a beam merging prism might have useful effects, or maybe multiple collimated beams all firing into a large condenser... TBD... They run cool, though...) For anybody who wishes to experiment, look for "projector fog" on either ebay or Yahoo shopping...
Another possibility I was thinking of is a little more experimental: enclosing multiple LOA tubes in a box whose inside surfaces are front-surface mirrors - with only a single small hole through which the light could escape... Would it work as a high power cool running point source? Shoot me down on this one, folks...
Bill.
Hi Gunawan,
Your issue about distortion was the reason I mentioned keeping the "eye point" close to the lens - roughly somewhere between the lens and its focal point.
If the view point was too far away from the lens, the limitations of fresnel technology caused MASSIVE distortion of the image with the corners either being distorted horribly or simply blotted out depending on which way the grooves were facing.
I got best results with the grooves facing the panel.
The concave lens was used because even with a longer throw objective, it was outside the optimal image "collection point" of my fresnel. I.e. it was far enough away from the fresnel that the gross distortions mentioned above kicked in. The concave lens polaced closer to the fresnel collected the light and "elongated" the "cone" such that the objective could operate within its normal parameters and not get hit by the gross distortions of the fresnel.
Each fresnel is different - with different focal length and line spacing. Some longer focal length ones might be better without the concave step. The one I was playing with definitely needed something to collect the light, and forward it to the objective.
Bill.
Your issue about distortion was the reason I mentioned keeping the "eye point" close to the lens - roughly somewhere between the lens and its focal point.
If the view point was too far away from the lens, the limitations of fresnel technology caused MASSIVE distortion of the image with the corners either being distorted horribly or simply blotted out depending on which way the grooves were facing.
I got best results with the grooves facing the panel.
The concave lens was used because even with a longer throw objective, it was outside the optimal image "collection point" of my fresnel. I.e. it was far enough away from the fresnel that the gross distortions mentioned above kicked in. The concave lens polaced closer to the fresnel collected the light and "elongated" the "cone" such that the objective could operate within its normal parameters and not get hit by the gross distortions of the fresnel.
Each fresnel is different - with different focal length and line spacing. Some longer focal length ones might be better without the concave step. The one I was playing with definitely needed something to collect the light, and forward it to the objective.
Bill.
Woneill:
Let me first congratulate you on the first rate information you are providing to the forum.
Also, i would advise against the slow startup/shutdown of Quartz lamps. I tried this myself using servos later to find out info that would make impossible. Quartz halogen lamps use special chemicals on the filament that are activated by the heat. As these chemicals are expecting the full voltage/amp input and a high amount of heat they will attempt to disperse some of the heat accordingly. If less the the full voltage/amp amount is fed through the lamp there will not be enough heat, thus the chemicals will instead be attacking the filament, leading to early lamp failure.
Thought i'd mention it to save u a few $
I'm currently looking for a 400W MH double ended lamp and ballast for my OHP. My OHP is an Elite Vision 4000 and was designed for LCD use and has a plug outlet on the front of the unit for an LCD panel. It has a condensor lens just above the lamp assembly so i cannot fit a full sized spherical lamp in the space. It currently takes 400W 50hr EVD lamps, which give out 16,000 lumens. A 400W mh would increase this to around 38,000 which should hopefully give a nicer picture.
Does anyone know where i could get such a lamp/ballast?
Cheers,
Let me first congratulate you on the first rate information you are providing to the forum.
Also, i would advise against the slow startup/shutdown of Quartz lamps. I tried this myself using servos later to find out info that would make impossible. Quartz halogen lamps use special chemicals on the filament that are activated by the heat. As these chemicals are expecting the full voltage/amp input and a high amount of heat they will attempt to disperse some of the heat accordingly. If less the the full voltage/amp amount is fed through the lamp there will not be enough heat, thus the chemicals will instead be attacking the filament, leading to early lamp failure.
Thought i'd mention it to save u a few $
I'm currently looking for a 400W MH double ended lamp and ballast for my OHP. My OHP is an Elite Vision 4000 and was designed for LCD use and has a plug outlet on the front of the unit for an LCD panel. It has a condensor lens just above the lamp assembly so i cannot fit a full sized spherical lamp in the space. It currently takes 400W 50hr EVD lamps, which give out 16,000 lumens. A 400W mh would increase this to around 38,000 which should hopefully give a nicer picture.
Does anyone know where i could get such a lamp/ballast?
Cheers,
It's Coming to an End.
To All,
When I started this thread I didn't think anyone would be interested. I thought maybe 10 people in the world would respond. But like Drill Sergent Hartman said in Full-Metal Jacket, "I'm glad to see that you have grown beyond our control."
It has grown into a monster thread that is out of control, but now it must come to an end.
I think there is enough here for anyone to attempt to build their own DIY projector. This site has a search feature that everyone should embrace. Web sites are popping up on the subject, so this should be another source of information. And, let's not forget E-mail. I still recieve on average 2 emails a week with questions, and I have not contributed in months to this thread.
Maybe the devoute can create a web ring, email list or a news group on the topic.
This is not to say, no one should post regarding DIY projectors. To the contrary. I'm just saying this thread is getting tired.
Take what you know and build on it.
Always remember to help each other and make it fun. Don't forget to respect each other too. The world is watching.
Best regards,
Vince Di Nenna
To All,
When I started this thread I didn't think anyone would be interested. I thought maybe 10 people in the world would respond. But like Drill Sergent Hartman said in Full-Metal Jacket, "I'm glad to see that you have grown beyond our control."
It has grown into a monster thread that is out of control, but now it must come to an end.
I think there is enough here for anyone to attempt to build their own DIY projector. This site has a search feature that everyone should embrace. Web sites are popping up on the subject, so this should be another source of information. And, let's not forget E-mail. I still recieve on average 2 emails a week with questions, and I have not contributed in months to this thread.
Maybe the devoute can create a web ring, email list or a news group on the topic.
This is not to say, no one should post regarding DIY projectors. To the contrary. I'm just saying this thread is getting tired.
Take what you know and build on it.
Always remember to help each other and make it fun. Don't forget to respect each other too. The world is watching.
Best regards,
Vince Di Nenna
since everybody seems to read this forum and it sounds like this thread my die, thought i would post this here.
Im currently compiling a FAQ and links and such from this thread and other Websites. Just wanted to say im going to mention a few of your "screen" names in the thank you, and some of your information. If any of you DO NOT want your "screen" name ( no e-mail address's (dont worry) or webpages, or even your ideas posted in this FAQ please let me know.
PLEASE DO NOT E_MAIL me for the FAQ.. i will post it when its done..
clockwork@uswest.net
Clok
Im currently compiling a FAQ and links and such from this thread and other Websites. Just wanted to say im going to mention a few of your "screen" names in the thank you, and some of your information. If any of you DO NOT want your "screen" name ( no e-mail address's (dont worry) or webpages, or even your ideas posted in this FAQ please let me know.
PLEASE DO NOT E_MAIL me for the FAQ.. i will post it when its done..
clockwork@uswest.net
Clok
snoogans,
I haven't lost interest. I don't think my ideas travel the same "thread" as others on this...thread, but that has nothing to do with it. If it did, I would have shut it down a long time ago.
The thread is entirely too long and people do not use the search function to find answers, they just keep asking the same questions over and over again. That's fine, if you have unlimited bandwidth and disk space to give away.
People are joking around and going on tangents, posting stupid pictures and then insulting each other.
I was asked by a moderator to consider closing the thread for other reasons. Their suggestion alone is good enough for me and I will respect it.
It's been a nice run, Boy and Girls. Thanks for contributing!
This thread will be closed tomorrow morning EST.
See you,
Vince
I haven't lost interest. I don't think my ideas travel the same "thread" as others on this...thread, but that has nothing to do with it. If it did, I would have shut it down a long time ago.
The thread is entirely too long and people do not use the search function to find answers, they just keep asking the same questions over and over again. That's fine, if you have unlimited bandwidth and disk space to give away.
People are joking around and going on tangents, posting stupid pictures and then insulting each other.
I was asked by a moderator to consider closing the thread for other reasons. Their suggestion alone is good enough for me and I will respect it.
It's been a nice run, Boy and Girls. Thanks for contributing!
This thread will be closed tomorrow morning EST.
See you,
Vince
Thanks TheLaw,
No point wasting more money on bulbs than necessary...
I don't know of any double-ended 400W MH bulbs. Most double-ended stuff seems to be set for weird wattages with hard to find ballasts.
This link seems to cover the spread of weird MH bulbs that I am aware of:
http://www.eurotekmarketing.com/index.html
Bill.
No point wasting more money on bulbs than necessary...
I don't know of any double-ended 400W MH bulbs. Most double-ended stuff seems to be set for weird wattages with hard to find ballasts.
This link seems to cover the spread of weird MH bulbs that I am aware of:
http://www.eurotekmarketing.com/index.html
Bill.
Its funny how easily some people get on power trips. Your closing a thread that you admit you haven't been a part of lately? Its obvious that other people seem to like the thread.
This would be simular to starting a conversation at a party, then trying to close the conversation long after other people have started in on it and continued it after you left. What sense does that make expect that it is a power trip for you, that you have the ability to create and destroy. (I think of it as more a technicality of how the forum was setup)
If it is a money thing, then fine, I can understand if the bandwidth is too much (is DIY Audio complaining?), but not for one person just getting some sought of feeling that it should be shut down.
So what if you left the converstion and on to something else, others are just joining or are still interested. Let the thread die when intrest for the thread dies. Of course you have no power over that that, now do you?
This would be simular to starting a conversation at a party, then trying to close the conversation long after other people have started in on it and continued it after you left. What sense does that make expect that it is a power trip for you, that you have the ability to create and destroy. (I think of it as more a technicality of how the forum was setup)
If it is a money thing, then fine, I can understand if the bandwidth is too much (is DIY Audio complaining?), but not for one person just getting some sought of feeling that it should be shut down.
So what if you left the converstion and on to something else, others are just joining or are still interested. Let the thread die when intrest for the thread dies. Of course you have no power over that that, now do you?
- Status
- Not open for further replies.