My name is Josh Borts. I am currently working for a non-profit organization called GeekCorps Mali. We provide technology solutions to rural Mali in an effort to better disseminate important health and economic information. One such project is investigating the use of satellite televisions. The hope is that these systems will be used by a large number of individuals and as such a large screen is desirable. Purchasing a television with an adequate screen size is outside the budget of our operation and would not be suitable for the largely nomadic users. As such, we are investigating the use of projectors and have become interested in the cost savings that LED based systems can offer.
As you are undoubtedly aware, typical halogen based projectors burn through bulbs quickly and require costly replacements. We cannot burden these poor communities with such a technology. As well, the heat produced by the bulbs, combined with the hot and dusty working environment in Mali makes these projectors lousy solutions. Recently, I read of new projectors that will be built using LEDs. While these commercial versions are not yet available and are expected to be priced out of our range, we are interested in the basic concept. I have read about others attempting DIY LED projectors but have not found any successful examples. This combination should lower both up-front and future costs allowing us to complete our project goals without heavily taxing the Malian community.
Does anyone have any suggestions?
As you are undoubtedly aware, typical halogen based projectors burn through bulbs quickly and require costly replacements. We cannot burden these poor communities with such a technology. As well, the heat produced by the bulbs, combined with the hot and dusty working environment in Mali makes these projectors lousy solutions. Recently, I read of new projectors that will be built using LEDs. While these commercial versions are not yet available and are expected to be priced out of our range, we are interested in the basic concept. I have read about others attempting DIY LED projectors but have not found any successful examples. This combination should lower both up-front and future costs allowing us to complete our project goals without heavily taxing the Malian community.
Does anyone have any suggestions?
Hi Josh,
And welcome to the Forum🙂
About the bulbs - most projectors here use Metal Halide bulbs, very power and heat efficient, and cost about 30-40 $ (more or less...), plus will last 10,000 to 20,000 hours!!! A 100w MH bulb vs a halogen bulb wil be atleast 5 times brighter (right? no first hand experiencs, just reading...)
Cheers,
Alex
And welcome to the Forum🙂
About the bulbs - most projectors here use Metal Halide bulbs, very power and heat efficient, and cost about 30-40 $ (more or less...), plus will last 10,000 to 20,000 hours!!! A 100w MH bulb vs a halogen bulb wil be atleast 5 times brighter (right? no first hand experiencs, just reading...)
Cheers,
Alex
It "MAY" be possible to build a LED projector out of the newest LED's (20,000mcd narrow viewing angle or the 15,000mcd with a wider) but because they only put out ~2-3 lumens, you would have to throw the whole 3200 behind a 15" lcd giving you a good 6-9,000 lumens to work with before dealing with light loss. A 17" would accommodate more but we still don't know how well LED light projects compared to standard bulbs. They're basically tweaked blue LED's so the color reproduction will most likely be poor compared to say a metal halide bulb. Also 3,000+ LED's WILL put out quite a bit of heat, draw a bit of current (60+ amps), and just the initial cost will be up around $8-900USD after the LED's, resistors if used, etc
I don't think they're worthwhile until they hit at least 5 lumens each for a standard LED and it still depends on how much light we can use per LED. If there's high losses through the lcd panel, then it may just be a lost cause
I don't think they're worthwhile until they hit at least 5 lumens each for a standard LED and it still depends on how much light we can use per LED. If there's high losses through the lcd panel, then it may just be a lost cause
There have been several long threads on this board, including one fellow who built his own, very small, LED-lit projector with results IIRC.
Try searching with any keywords you can think. I am afraid it may take a while to dig up the proper threads.
Try searching with any keywords you can think. I am afraid it may take a while to dig up the proper threads.
Thank you for the help. If what you say about Metal Halide bulbs is true then that would certainly be a good choice. Can these be used as replacement bulbs in commercial systems?
As for the LED system, I have looked through this site and was unable to find any finnished product. In any case, I have been in discussion with LumiLeds about their products. They have built a prototype mini projector using three of there Luxeon LEDs that outputs around 15 lm and should be capable of being raised to 24 lm. However, Metal Halides do seem like a better/simpler choice.
As for the LED system, I have looked through this site and was unable to find any finnished product. In any case, I have been in discussion with LumiLeds about their products. They have built a prototype mini projector using three of there Luxeon LEDs that outputs around 15 lm and should be capable of being raised to 24 lm. However, Metal Halides do seem like a better/simpler choice.
the type of MH bulbs used around here wouldn't fit in a newer lcd projector since they're so compact. The only way would be to have the bulb externally and try to direct all the light in to where the stock bulb shines
yeah, I don't think there are any actual finished LED projects yet and even if you used a couple Luxeon's, you'd still have a really dull image. Think about how most commercial projectors are over 1,000 ansi lumens. If you're only starting with say 3x60lumen LED's, thats still only 180 to work with from the beginning, and probably around 50, at best, ansi. If I had $400 to blow right now, I'd try projecting my 10.5" lcd with 1500 led's and compare it to my 250w MH but until something like that happens, it'd be best to stick with whats known to work
yeah, I don't think there are any actual finished LED projects yet and even if you used a couple Luxeon's, you'd still have a really dull image. Think about how most commercial projectors are over 1,000 ansi lumens. If you're only starting with say 3x60lumen LED's, thats still only 180 to work with from the beginning, and probably around 50, at best, ansi. If I had $400 to blow right now, I'd try projecting my 10.5" lcd with 1500 led's and compare it to my 250w MH but until something like that happens, it'd be best to stick with whats known to work
I don't see LEDs being a viable projection source quite yet. With the larger screens we use, it is more of a possibility, since the required power density is WAY off for use with the tiny elements in a commercial projector. However, if OLED screens get cheaper and brighter, we have a screen that actually emits light on its own. Obviously the brightness would be far beyond what people would normally want to look at, so I don't know if we'll ever be able to use a OLED monitor for this purpose. Another possibility is somehow creating a sheet of mass-produced miniature LEDs, which could be used as a very bright diffuse source. Again, probably not something that will ever be produced en masse, and cheaply.
One more possibility, probably the most expensive of all unless mass-produced: make the screen itself active. Each element would be independent except for power and ground planes. The element would sense infrared and translate to an RGB spectrum depending on the wavelength of the IR. So you project an invisible beam, no great intensity required, onto the projection surface and the surface itself lights up with the image. The advantage of this is that every element on the screen is identical, so you could tile as many sheets together as needed. Also, there is no massively complex control scheme needed to power and modulate the RGB LEDs. Granted, this is nowhere near DIY possibilities. But it's one way displays could go.
One more possibility, probably the most expensive of all unless mass-produced: make the screen itself active. Each element would be independent except for power and ground planes. The element would sense infrared and translate to an RGB spectrum depending on the wavelength of the IR. So you project an invisible beam, no great intensity required, onto the projection surface and the surface itself lights up with the image. The advantage of this is that every element on the screen is identical, so you could tile as many sheets together as needed. Also, there is no massively complex control scheme needed to power and modulate the RGB LEDs. Granted, this is nowhere near DIY possibilities. But it's one way displays could go.
Sometimes it takes creative search terms
you have to have been here to read the thread to remember the keywords to search for.
Witness a successful LED lit projector.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46489&highlight=
dang, no screenshots! Where is this cat anyhow?
you have to have been here to read the thread to remember the keywords to search for.
Witness a successful LED lit projector.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=46489&highlight=
dang, no screenshots! Where is this cat anyhow?
He dangled this in front of everyone and then never showed any pictures of the resultant image. Maybe he sells Luxeons? 😉
For your efforts in "disseminating important health and economic information" to rual areas, I would think a standard overhead projector and LCD projection panel designed for this use would be your best bet. These can be purchased very cheaply, <$300 USD for both, and readily available, even over seas.
I think a DIY projector is a little overkill for your needs.
I think a DIY projector is a little overkill for your needs.
Thank you Inkog for your advice. While I agree with your assessment that a DIY projector may be overkill, this is what has been asked of me. I am to evaluate consumer projector models and am trying to present a cheaper (DIY) version. They would like a high enough resolution to show satelite television, as this will help with the sustanability of any project here. But I do agree, for the overall goal, this is too much.
Definitely go with a metal halide setup.
A 20,000 hour 400 watt metal halide bulb, a 15" consumer LCD monitor stripped of its case, a custom made(out of plywood, MDF, or metal) box, along with split fresnel lenses, a polished steel bowl, a triplet lense, and a fan or two. This is the standard build that Lumenlabs uses. At this stage, their standard kit would probably be the most cost effective choice - expect around US$500. I highly recommend you explore their result gallery, and consider buying their guide + private forums: http://www.lumenlab.com/
Luxeon may be able to offer 20 or 30 lumens, but you need 1000 or so to get a decent image on an average sized projection screen. Metal halide is also actually much more efficient than a white LED is, in lumens per watt.
Using a 15" computer monitor panel gets you more than enough resolution for most things, but you need to add the price of a TV tuner that outputs DVI or VGA, $100-$200. Another benefit: The screendoor effect that can become distracting when your pixels are a quarter inch tall is next to nil.
The size is another thing to consider. A projector built this way is about 1/3 meter by 1/4 meter by 0.6 meter. An average consumer projector is a probably less than a tenth this volume. By virtue of this, the LCD panel can get very dirty before you ever notice it, and is relatively easy to clean. Most commercial projector users seem to take them in for service when they get a tiny dust blob in the tiny image element. Cleaning a comparatively huge LCD panel, even if it's somewhat delicate, requires maybe a sponge.
How big do you need the image to be?
What resolution is the satellite TV you mentioned?
A 20,000 hour 400 watt metal halide bulb, a 15" consumer LCD monitor stripped of its case, a custom made(out of plywood, MDF, or metal) box, along with split fresnel lenses, a polished steel bowl, a triplet lense, and a fan or two. This is the standard build that Lumenlabs uses. At this stage, their standard kit would probably be the most cost effective choice - expect around US$500. I highly recommend you explore their result gallery, and consider buying their guide + private forums: http://www.lumenlab.com/
Luxeon may be able to offer 20 or 30 lumens, but you need 1000 or so to get a decent image on an average sized projection screen. Metal halide is also actually much more efficient than a white LED is, in lumens per watt.
Using a 15" computer monitor panel gets you more than enough resolution for most things, but you need to add the price of a TV tuner that outputs DVI or VGA, $100-$200. Another benefit: The screendoor effect that can become distracting when your pixels are a quarter inch tall is next to nil.
The size is another thing to consider. A projector built this way is about 1/3 meter by 1/4 meter by 0.6 meter. An average consumer projector is a probably less than a tenth this volume. By virtue of this, the LCD panel can get very dirty before you ever notice it, and is relatively easy to clean. Most commercial projector users seem to take them in for service when they get a tiny dust blob in the tiny image element. Cleaning a comparatively huge LCD panel, even if it's somewhat delicate, requires maybe a sponge.
How big do you need the image to be?
What resolution is the satellite TV you mentioned?
Well if its just sattelite tv, than I think its best to find a cheapers lower res panel. Maybe even a projection panel. Why? because even though screendoor (not if you get an XGA panel) will become more noticable, or the contrast wont be soo good, but you can save the money for the tv converter box, and may not even need one, seeing how some panels have svideo input.
Just my 2 cents
Cheers,
Alex
Just my 2 cents
Cheers,
Alex
Image quality is not a great concern. While it would be nice to have the highest resolution, non-pixulated image, price is a more immediate priority. This certainly seems to favor low rez projector panels. I do not know the resolution of the satelite system off hand, I will try and find that information.
Would I still need a TV tuner card to change the station?
As for size, any projection greater than 30 inches would be wonderful. And the size of the projector itself does not matter, expect perhaps for bigger is better (easier for non-techies to replace parts when they break)
Would I still need a TV tuner card to change the station?
As for size, any projection greater than 30 inches would be wonderful. And the size of the projector itself does not matter, expect perhaps for bigger is better (easier for non-techies to replace parts when they break)
Size of the projection seems to be anything from about 60" diagonal to about 144" diagonal, depending on how bright you want the image to be. They don't put out a huge amount of light, compared to business projectors, though I think they're competitive with home theater projectors.
Projectors are drastically different from how a TV works in their reaction to ambient light. A cathode ray tube is a different technology that is capable of using a dark grey screen, which absorbs ambient light. Projectors require a white screen, and the darkest black on the image is only going to be as dark as the screen is when the projector is turned off. This means that they're only practical in places with very, very little ambient light. You want all the lights to be off, all the shades to be drawn, etc. Or, you can project at night.
Also, regarding the lamps, using a 250 watt HQI bulb with comparable life specs is also possible, if you can tolerate a slightly darker/smaller screen (It sounds like you can).
15" monitors run $150-$200 new. I'd really need details of your source material to describe how you'd have to hook things up.
-------------------------
I've been reading your website. Very interesting stuff, really makes you sit back and think about what you have. Regarding cooling the pentiums, I'd like to ask you a few questions:
What kind of availability is there for water in the areas that you're talking about? How many processors per site need to be cooled? How close together are they? Is cooling the processors to within, say, 5C of ambient tolerable?
Also, what is the cost of a kilowatt-hour of electricity in the area that you're working?
Projectors are drastically different from how a TV works in their reaction to ambient light. A cathode ray tube is a different technology that is capable of using a dark grey screen, which absorbs ambient light. Projectors require a white screen, and the darkest black on the image is only going to be as dark as the screen is when the projector is turned off. This means that they're only practical in places with very, very little ambient light. You want all the lights to be off, all the shades to be drawn, etc. Or, you can project at night.
Also, regarding the lamps, using a 250 watt HQI bulb with comparable life specs is also possible, if you can tolerate a slightly darker/smaller screen (It sounds like you can).
15" monitors run $150-$200 new. I'd really need details of your source material to describe how you'd have to hook things up.
-------------------------
I've been reading your website. Very interesting stuff, really makes you sit back and think about what you have. Regarding cooling the pentiums, I'd like to ask you a few questions:
What kind of availability is there for water in the areas that you're talking about? How many processors per site need to be cooled? How close together are they? Is cooling the processors to within, say, 5C of ambient tolerable?
Also, what is the cost of a kilowatt-hour of electricity in the area that you're working?
Mali multimedia projector
This is a site that explains a working led based microfiche projector which is hardened for the third world field use.
http://www.designthatmatters.org/k2/archives/cat_current_design.html
Over a 100 students from places like MIT and Cambridge have been working on this since 2002. As the guy who started this thread said they are looking for a multimedia projector design. There are some nice design disscussions there along with other good research. Most is over my head but might be worth checking out by some of the more experienced and knowledgeable diy pj types.
This is a site that explains a working led based microfiche projector which is hardened for the third world field use.
http://www.designthatmatters.org/k2/archives/cat_current_design.html
Over a 100 students from places like MIT and Cambridge have been working on this since 2002. As the guy who started this thread said they are looking for a multimedia projector design. There are some nice design disscussions there along with other good research. Most is over my head but might be worth checking out by some of the more experienced and knowledgeable diy pj types.
- Status
- Not open for further replies.