I am hoping to put my lilliput to use this year. I am hoping to build an led solution. I realize that it will be difficult, but I know that in the long run the benefit will be big.
1. No heat = smaller unit = no fan = quiet
2. 20,000 hour life = 10 normal projector bulbs = 2000$ - $3000 savings
The problems to be addressed, as I am a newbie, are what are my realistic goals as far as number of lights, appropriate driver for the decided upon number, and achieving a good cost versus performance ratio for the leds. Also, combining the lights to give a crisp image. Can we start a thread where we collect knowledge on the diy LED progress? I know there are german forums, but i think we need a translator for accurate information. Anyone wanting to help or add to this, please contact me. I am planning on buying a 2.5" lilliput for experimentation, but I think the actual light output from a lilliput 7" will be better, because there is a larger image area to put leds behind. I dont think I can get more than a few behind a 2.5 ", but someone said they got 21 behind a 5".
I am ready to start purchasing leds and driver, but i need some advice as far as buying a driver that will be suitable on both the small number of leds behind the 2.5 inch as well as the larger behind the 7". Ace has suggested that each light needs a guide, and I take this to mean the snap on colimators. Please contribute to this thread, we all know leds are the future of light sources, and an industry insider tells me the projection industry's biggest fear is the bulb companies going under, and they seem to think this isn't too far away. I think we have the best collection of forward thinkers here, at least for this field, and I think that a useable diy led must be a possibility, or even a laser for that matter.
1. No heat = smaller unit = no fan = quiet
2. 20,000 hour life = 10 normal projector bulbs = 2000$ - $3000 savings
The problems to be addressed, as I am a newbie, are what are my realistic goals as far as number of lights, appropriate driver for the decided upon number, and achieving a good cost versus performance ratio for the leds. Also, combining the lights to give a crisp image. Can we start a thread where we collect knowledge on the diy LED progress? I know there are german forums, but i think we need a translator for accurate information. Anyone wanting to help or add to this, please contact me. I am planning on buying a 2.5" lilliput for experimentation, but I think the actual light output from a lilliput 7" will be better, because there is a larger image area to put leds behind. I dont think I can get more than a few behind a 2.5 ", but someone said they got 21 behind a 5".
I am ready to start purchasing leds and driver, but i need some advice as far as buying a driver that will be suitable on both the small number of leds behind the 2.5 inch as well as the larger behind the 7". Ace has suggested that each light needs a guide, and I take this to mean the snap on colimators. Please contribute to this thread, we all know leds are the future of light sources, and an industry insider tells me the projection industry's biggest fear is the bulb companies going under, and they seem to think this isn't too far away. I think we have the best collection of forward thinkers here, at least for this field, and I think that a useable diy led must be a possibility, or even a laser for that matter.
First off, I am putting in some links to LED manufacturers:
http://www.luxeonstar.com
http://www.lumileds.com
please feel free to add more.
http://www.luxeonstar.com
http://www.lumileds.com
please feel free to add more.
To me it seems odd that we would want to have an "array" of lights making a flat field of seperate points of light hitting different points of a fresnel or lcd panel, as that would just make it harder to focus etc.
What is the possibility of making each of the led's aim at the exact same point on the rear end of a condenser lens? If not fixed to the inner concave of of a spherical shape, then how about in any formation with wide fiber optic cable aiming the individual lights at their destination (the input-end of the condenser). Perhaps have all the points of the led's hit the same part of a condensor that is aimed at the Norpro reflector?
What is the possibility of making each of the led's aim at the exact same point on the rear end of a condenser lens? If not fixed to the inner concave of of a spherical shape, then how about in any formation with wide fiber optic cable aiming the individual lights at their destination (the input-end of the condenser). Perhaps have all the points of the led's hit the same part of a condensor that is aimed at the Norpro reflector?
This is the kind of troubleshooting we need in pre-construction.
The first thing I need to know is what LED to try. There are a lot of choices.
I am hoping that you stumbled onto something:
If we had a perfectly (half)spherical reflector, could I not mark off equidistant points on the reflector , drill it and position the leds on the inside surface? Would colimators improve or degrade the image?
The first thing I need to know is what LED to try. There are a lot of choices.
I am hoping that you stumbled onto something:
If we had a perfectly (half)spherical reflector, could I not mark off equidistant points on the reflector , drill it and position the leds on the inside surface? Would colimators improve or degrade the image?
Attachments
If the design goal is to make a projector using a 7 inch Lilliput, then specific information is already known. There are plenty of 7 inch projectors that have been made with excellent results. Most have used a 250 watt bulb producing an intial 25,000 lumens. Do you think a reflector with LEDs mounted in it would be able to put out this amount of light? If so, how many Leds would it take? The amont of LEDs and thier size would dictate the size of the reflector. I am curious as to how big it actually would be. This sounds like a very interesting project.
Mike
Mike
Ace tells me I am going about it wrong. He says i need to make a light guide, which i translate as a tube, that changed the emitted light shape from round to square, then these squares make up a bigger square. Think flashlight with a piece of paper over the end with a square hole cut out. round light = square emission.
Yeah this project sounds interesting, that is why I am posting here, I am hoping to get the knowledge to give this project a go.
Yeah this project sounds interesting, that is why I am posting here, I am hoping to get the knowledge to give this project a go.
I shall be watching with interest but can't see how you are going to get anywhere near the lumen output of a 250watt MH without using literally hundreds of LED's
For example
http://www.luxeonstar.com/item.php?id=297&link_str=189::194&partno=LXHL-MWJA
Notice these Luxons (18 of them) mounted on a board only put out 450 Lumens @ a cost of $US268 i might ad
Guarenteed i'll be straight into LED's once you clever fellas work away out to make them work (if they are cost effective) and on a par with a metal halide, but un-fortuneatley at the price of LED's at the present time I simply can't afford to buy them to "experiment" with. At that price seems also out of range of the average DIY'er don't it? aint DIY meant to be budget? I realise in the long run many many many years they will pay for themselves but for the initial outlay it would be cheaper to go straight out and buy a commercial projector.
Just my 10 cents worth (bout all i got left too after building my projector lol)
Anyways good luck with it guys i shall be watching eagerly at this thread.
For example
http://www.luxeonstar.com/item.php?id=297&link_str=189::194&partno=LXHL-MWJA
Notice these Luxons (18 of them) mounted on a board only put out 450 Lumens @ a cost of $US268 i might ad
Guarenteed i'll be straight into LED's once you clever fellas work away out to make them work (if they are cost effective) and on a par with a metal halide, but un-fortuneatley at the price of LED's at the present time I simply can't afford to buy them to "experiment" with. At that price seems also out of range of the average DIY'er don't it? aint DIY meant to be budget? I realise in the long run many many many years they will pay for themselves but for the initial outlay it would be cheaper to go straight out and buy a commercial projector.
Just my 10 cents worth (bout all i got left too after building my projector lol)
Anyways good luck with it guys i shall be watching eagerly at this thread.
The brightest Led is the Luxeon V Emitter which is 120 lumens. They don't make it in white, but suppose they did. it would take 210 of these to equall 25,000 lumens. That would make a sqare panel of 15 by 15 LEDs.(225 LEDs) Since the LEDs are 1/2 inch diameter, the panel would be 7 1/2 by 7 1/2 inches. According to Ace, you would make a light tunnel that would start off being 7 1/2 inches square and expanding to the size of the LCD. That's the good news. Now the bad news. at 45.00 dollars a piece it would cost 10,125.oo plus the cost of the driver added on. So suppose you buy them in bulk and get them for 10 bucks a piece. That's still 2250.oo plus the cost of the driver added on. A metal halide lamp plus a ballast is 150.oo. The only way I can see LEDs ever being feasible is for advancement in the technology to produce higher lumens combined will lower prices for the LEDs.
efficiency of LEDs is a MYTH
People assume that LEDs are more efficient than MH, but the idea that LEDs don't make any heat is based on a false comparison: A tiny 60 milliwatt LED does not generate much heat. But a 60 milliwatt MH bulb (if such a thing existed) would generate far LESS heat! MH bulbs are the second most efficient light generating technology existing. (The most efficient is a sodium bulb, but those have awful color.)
So when you make an array of LEDs that put out as many lumens as a 400 Watt MH bulb, you will have to remove more heat. And you will also have to pay for more electricity.
For example, Luxeon Star white emitters can be driven up to 350 milliamps with a forward voltage drop of 3.42 volts. This is a power input of 1.197 Watts. The output will be 18 lumens. So the efficiency is 18/1.197 = 15 lumens per Watt. (This does not even take into account the power dissipated by the series resistor needed by each emitter.)
A Ushio UHI-S400DD MH bulb puts out 33000 lumens / 400 Watts = 82.5 lumens per Watt. That's 5.5 times as efficient as the Luxeon LED!
So where does all that extra power go in the LED? All energy is conserved, so it comes out of the device as heat.
People assume that LEDs are more efficient than MH, but the idea that LEDs don't make any heat is based on a false comparison: A tiny 60 milliwatt LED does not generate much heat. But a 60 milliwatt MH bulb (if such a thing existed) would generate far LESS heat! MH bulbs are the second most efficient light generating technology existing. (The most efficient is a sodium bulb, but those have awful color.)
So when you make an array of LEDs that put out as many lumens as a 400 Watt MH bulb, you will have to remove more heat. And you will also have to pay for more electricity.
For example, Luxeon Star white emitters can be driven up to 350 milliamps with a forward voltage drop of 3.42 volts. This is a power input of 1.197 Watts. The output will be 18 lumens. So the efficiency is 18/1.197 = 15 lumens per Watt. (This does not even take into account the power dissipated by the series resistor needed by each emitter.)
A Ushio UHI-S400DD MH bulb puts out 33000 lumens / 400 Watts = 82.5 lumens per Watt. That's 5.5 times as efficient as the Luxeon LED!
So where does all that extra power go in the LED? All energy is conserved, so it comes out of the device as heat.
And it gets worse!
LEDs must be individually current-regulated. If you try to run several in parallel, the one with the lowest forward threshold voltage will get all of the current until it burns up, then the next lowest will take all the current until it burns up, etc.
The cheapest way to regulate their current is to put a resistor in series with each LED. Then the LED will drop its forward voltage (ie. about 3.42 volts for Luxeon stars) and the resistor will drop the rest of the voltage. Unfortunately, this means the resistor is also dissipating power. In the Luxeon datasheets, they use a six volt supply, so the resistor drops about 2.58 volts. Then the combined efficiency goes down to 8.57 lumens per Watt!
Of course, you don't have to use a resistor: You can build a current regulator using a power MOSFET and a few passive components that will be 90% efficient. But you would have to build one of these for every Luxeon Star in your LED array.
Finally, you would have to do a terrific job of cooling them, since they stop emitting visible light if they get too hot. Then they just emit pure IR light. So your array would probably have to be water cooled.
LEDs must be individually current-regulated. If you try to run several in parallel, the one with the lowest forward threshold voltage will get all of the current until it burns up, then the next lowest will take all the current until it burns up, etc.
The cheapest way to regulate their current is to put a resistor in series with each LED. Then the LED will drop its forward voltage (ie. about 3.42 volts for Luxeon stars) and the resistor will drop the rest of the voltage. Unfortunately, this means the resistor is also dissipating power. In the Luxeon datasheets, they use a six volt supply, so the resistor drops about 2.58 volts. Then the combined efficiency goes down to 8.57 lumens per Watt!
Of course, you don't have to use a resistor: You can build a current regulator using a power MOSFET and a few passive components that will be 90% efficient. But you would have to build one of these for every Luxeon Star in your LED array.
Finally, you would have to do a terrific job of cooling them, since they stop emitting visible light if they get too hot. Then they just emit pure IR light. So your array would probably have to be water cooled.
Wow, I never thought there were so many nay sayers in a forum where people are building projectors for under $1000 that compare favorably with the extremely high dollar ($10,000+) high-res home theater projectors. My thought is that we have seen a mans results in the german forum. Flateric has built a 21 led unit, and although the results aren't available yet, led tech is set to expand in a big way in the next couple years. Consider this, they are making car lights using led now, and flashlights with as little as 1 led. LED will take over, it's just a function of time, i.e. will it happen before the rollable displays become cost effective. That tech will kill the market for plasma and dlp probably, because there is less manufacturing and assembly involved. For now I agree that a 250w hqi is prolly the best bet as far as price/performance. I am continuing on my quest to build one. I have been burned out for a while as others were, because finding the right parts is a big job. I am hoping that the new year will energize me and help me get this thing completed, something a lot of us are experiencing is burnout. The LED idea is something that could really recharge the forum community, and believe me it can use some recharging, and if you need proof simply look at the stir the hippad/slide projector combo caused. If someone can make it work, it should be able to happen here first!
naysayer or truth-teller?
I don't want to discourage anybody, and I will say this: A good LCD projector can be built using LEDs as the light source.
But it doesn't hurt to look at the physical limitations of the available devices. If you want to build an LED-based projector just to see if it can be done, then go to it!
But if you think using LEDs will solve the heat problems with current MH designs, then you are just beating your head against the wall.
You just have to be patient: As soon as somebody makes high-powered white (or even red, blue, and green) LEDs that put out more than 85 lumens per Watt, then they will replace MH bulbs. And that may be sooner than you think. The very most efficient LEDs now put out 57 lumens per Watt. Just not the high-powered ones people would like to use for this application. If somebody can double the efficiency of the best LEDs now available, that will be the time to make an LED-backlit projector.
I don't want to discourage anybody, and I will say this: A good LCD projector can be built using LEDs as the light source.
But it doesn't hurt to look at the physical limitations of the available devices. If you want to build an LED-based projector just to see if it can be done, then go to it!
But if you think using LEDs will solve the heat problems with current MH designs, then you are just beating your head against the wall.
You just have to be patient: As soon as somebody makes high-powered white (or even red, blue, and green) LEDs that put out more than 85 lumens per Watt, then they will replace MH bulbs. And that may be sooner than you think. The very most efficient LEDs now put out 57 lumens per Watt. Just not the high-powered ones people would like to use for this application. If somebody can double the efficiency of the best LEDs now available, that will be the time to make an LED-backlit projector.
I love the idea of using led's for a light source, but release that after reading the forums that it has alot of limitations. thanks for all the info as i am a noob.
but i did stumble across this site today, i dont understand all the jargon at the moment trying to get up to speed, but have a look and tell me are these guys all hype or are they on to something.
http://www.laminaceramics.com/news/101804.aspx
but i did stumble across this site today, i dont understand all the jargon at the moment trying to get up to speed, but have a look and tell me are these guys all hype or are they on to something.
http://www.laminaceramics.com/news/101804.aspx
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.