Okay, I was talking alot about this in the main thread, but I thought to bring it out to its own area so we don't clutter the main flow of thought. I hope that other people out there are currently as obsessed with LED's as a light source as I am. 🙂
(Note: This assumes you've read Dec 15-17th in the main DIY Video Project II page)
So after researching 1157's, I was told about a new sweet LED called a Star-O
I just skipped more studying for finals to research these things, and they're pretty sweet. 🙂
180,000 mcd with a 10 degree beam width... pretty impressive, and only for $16. Just for comparison, the 1157 LED cluster is 78.720 mcd with a 12 degree beam width.
From my calculations, it seems like it would be cheaper to get the 1157's than the Star-O's -- the Star-O red's are about 5-8 times brighter than the white stars... from my calculations a 180,000 mcd star-o with a 10 degree beam width is 4.3 lumens. A 78,720 mcd 1157 with a 12 degree beam width is 2.71 lumens. At $16 for a white star-o and $7 for a white 1157, I get $2.58 per lumen with the 1157's and $3.72 per lumen for the star-o's.
So roughly $2500 to get the cheapest possible 1000 lumen cluster of LED lights (but then you have to focus them all, but I like the fresnel lens cluster sheet to solve that problem)
However, I'm pretty new to all this, so I might be wrong. The formula I used for mcd to lumens is:
( 1 - cos(beam_width/2)) * 2pi * cd = lumens
Not sure if I'm right on all this, but it seems within the right order of magnitude. 🙂
However, if they ever come out with the a new star-o that's brighter, that would be incredibly sweet. 🙂
Anyway, as far as my own progress goes, I just got a nView Z-Series 640x480 32bit color transparent overhead LCD thingy for $60 off a friend. Works fantastic. I still ordered a raw LCD with controller board so that I can build one the DIY way, but for watching movies over Christmas break, I think I'm gonna' do the old OHP thing. Should get me going for < $100 . I'm still anxious to build my smaller, cooler (temp-wise), quieter, and cooler (geek-wise) LED-lit projector. 🙂
--Clint
(Note: This assumes you've read Dec 15-17th in the main DIY Video Project II page)
So after researching 1157's, I was told about a new sweet LED called a Star-O
I just skipped more studying for finals to research these things, and they're pretty sweet. 🙂
180,000 mcd with a 10 degree beam width... pretty impressive, and only for $16. Just for comparison, the 1157 LED cluster is 78.720 mcd with a 12 degree beam width.
From my calculations, it seems like it would be cheaper to get the 1157's than the Star-O's -- the Star-O red's are about 5-8 times brighter than the white stars... from my calculations a 180,000 mcd star-o with a 10 degree beam width is 4.3 lumens. A 78,720 mcd 1157 with a 12 degree beam width is 2.71 lumens. At $16 for a white star-o and $7 for a white 1157, I get $2.58 per lumen with the 1157's and $3.72 per lumen for the star-o's.
So roughly $2500 to get the cheapest possible 1000 lumen cluster of LED lights (but then you have to focus them all, but I like the fresnel lens cluster sheet to solve that problem)
However, I'm pretty new to all this, so I might be wrong. The formula I used for mcd to lumens is:
( 1 - cos(beam_width/2)) * 2pi * cd = lumens
Not sure if I'm right on all this, but it seems within the right order of magnitude. 🙂
However, if they ever come out with the a new star-o that's brighter, that would be incredibly sweet. 🙂
Anyway, as far as my own progress goes, I just got a nView Z-Series 640x480 32bit color transparent overhead LCD thingy for $60 off a friend. Works fantastic. I still ordered a raw LCD with controller board so that I can build one the DIY way, but for watching movies over Christmas break, I think I'm gonna' do the old OHP thing. Should get me going for < $100 . I'm still anxious to build my smaller, cooler (temp-wise), quieter, and cooler (geek-wise) LED-lit projector. 🙂
--Clint
white LEDS are an intweresting alternative to
light bulbs
the angle of the light spreading is a key issue
light bulbs
the angle of the light spreading is a key issue
Oh sw33t!!!!
Yes, I know I just used l33t sp34k in the post subject, but in this case I believe it applies. 🙂
I was told to research Star-O's, which are marginally cool as explained in my last post. HOWEVER, the same manufacturer also makes 5 watt LED emitters. I saw that they only appeared to be available in red, orange, and blue, yet I checked the price list on a distributer's page and they're available in white. 🙂 🙂 🙂
Part#: LXHL-LW5C
Color: White
Radiation Pattern: Lambertian (basically straight-out directed beam)
Color Temp: 5500k (is this good/bad? I don't know anything about color temp.)
Typical Flux (lm): (I'm assuming this is lumens, that's what came up on a google search for the abbrev. for lumens)... get this.... 120 lm . w00t!
Price: $43.20
So that means an incredible price of $0.36 per lumen! (as opposed to $2.43 per lumen on the best thing I found before)That means a 1080 lumen LED array of 9 emitters would cost approx. $388.80.... WOW. Keep in mind these things DO need heat sinks I believe, but still. They've got a metal board on the back that the heat sink attaches to. Just heat-compound them to a solid metal plate and then cool that puppy off with heat sinks on the other side... a couple silent computer fans and you've got a SMALL (3*14.5mm = 1.71 inches per side)... so smaller than 2" by 2" 1080 LED projector.... w0w. I'm in shock.
Anyone else as excited about this as I am? I guess I better give you guys the links before you kill me. 🙂
Mark Hannah Surplus Electronics
Spec Sheet for 5 Watt Emitters
Current things I do not know:
The beam spread on the emitters... I couldn't find specs on that for the *white* emitters, although since they're in the same housing as the normal emitters, it looks like it's a viewing angle of 110 degrees, or the included angle (where 90% of the light is) is 150 degrees. This actually works quite well, as we could have the lights very close to our fresnel lens... wait. No fresnel. We could use real optics here! Wow. Nice when you're dealing w/ small sizes. 🙂 A heck of a lense to make a 110-150 degree spread of light go paralell... that's a high power lens.... but we could find them... and if they're smaller, they're cheaper! Wow. The $43 per emitter is still huge for me as a college student, but I might just ask for some of these puppies for Christmas. Wow they're cool.
Keep the wheels of imagination chuggin' and spinnin'. 🙂
--Clint
P.S. Hah! I just read in the .pdf spec sheet in a side advertising note about possible uses... at the VERY bottom of the list of advertised possiblities is "LCD Backlights / Light Guides".... it seems that we're not the only ones considering the project option. 🙂
Yes, I know I just used l33t sp34k in the post subject, but in this case I believe it applies. 🙂
I was told to research Star-O's, which are marginally cool as explained in my last post. HOWEVER, the same manufacturer also makes 5 watt LED emitters. I saw that they only appeared to be available in red, orange, and blue, yet I checked the price list on a distributer's page and they're available in white. 🙂 🙂 🙂
Part#: LXHL-LW5C
Color: White
Radiation Pattern: Lambertian (basically straight-out directed beam)
Color Temp: 5500k (is this good/bad? I don't know anything about color temp.)
Typical Flux (lm): (I'm assuming this is lumens, that's what came up on a google search for the abbrev. for lumens)... get this.... 120 lm . w00t!
Price: $43.20
So that means an incredible price of $0.36 per lumen! (as opposed to $2.43 per lumen on the best thing I found before)That means a 1080 lumen LED array of 9 emitters would cost approx. $388.80.... WOW. Keep in mind these things DO need heat sinks I believe, but still. They've got a metal board on the back that the heat sink attaches to. Just heat-compound them to a solid metal plate and then cool that puppy off with heat sinks on the other side... a couple silent computer fans and you've got a SMALL (3*14.5mm = 1.71 inches per side)... so smaller than 2" by 2" 1080 LED projector.... w0w. I'm in shock.
Anyone else as excited about this as I am? I guess I better give you guys the links before you kill me. 🙂
Mark Hannah Surplus Electronics
Spec Sheet for 5 Watt Emitters
Current things I do not know:
The beam spread on the emitters... I couldn't find specs on that for the *white* emitters, although since they're in the same housing as the normal emitters, it looks like it's a viewing angle of 110 degrees, or the included angle (where 90% of the light is) is 150 degrees. This actually works quite well, as we could have the lights very close to our fresnel lens... wait. No fresnel. We could use real optics here! Wow. Nice when you're dealing w/ small sizes. 🙂 A heck of a lense to make a 110-150 degree spread of light go paralell... that's a high power lens.... but we could find them... and if they're smaller, they're cheaper! Wow. The $43 per emitter is still huge for me as a college student, but I might just ask for some of these puppies for Christmas. Wow they're cool.
Keep the wheels of imagination chuggin' and spinnin'. 🙂
--Clint
P.S. Hah! I just read in the .pdf spec sheet in a side advertising note about possible uses... at the VERY bottom of the list of advertised possiblities is "LCD Backlights / Light Guides".... it seems that we're not the only ones considering the project option. 🙂
Hanclinto
There is one other interesting thing about solid state emmiters like LED's.
You can pulse them.
As an example I have an led emmiter assy pointed down my drive with the receiver in the garage and sending signals into the house. If a car drives down the drive we get a buzz.
Thing is the drive is 30 metres long. Unpulsed led only gets to about 10 metres max. Pulsed it easily gets the 30 metre rain or sunshine.
They are cheap led rated at 100ma. Pulsed they get 1 amp for a tenth of a second.
I think you could probably double your apparent light output, possibly even much higher by pulsing them and if you pulse at say 100 pulses per second you will be well above the flicker rate.
The human eye if it sees a light will continue to see it for a period of time, even if the light has long gone. Persistence of vision.
Might be useful
There is one other interesting thing about solid state emmiters like LED's.
You can pulse them.
As an example I have an led emmiter assy pointed down my drive with the receiver in the garage and sending signals into the house. If a car drives down the drive we get a buzz.
Thing is the drive is 30 metres long. Unpulsed led only gets to about 10 metres max. Pulsed it easily gets the 30 metre rain or sunshine.
They are cheap led rated at 100ma. Pulsed they get 1 amp for a tenth of a second.
I think you could probably double your apparent light output, possibly even much higher by pulsing them and if you pulse at say 100 pulses per second you will be well above the flicker rate.
The human eye if it sees a light will continue to see it for a period of time, even if the light has long gone. Persistence of vision.
Might be useful
hrm....
VERY interesting.... do you think that continuously pulsing an LED would overload it's max amperage rating? Or is the startup current included in all of that max-rating stuff?
Very neat stuff to think about. I'm trying to see what kind of lenses I would need right now in order to condense everything here... the incredibly cool thing about having a 2x2 inch light source is that you can use solid-lens optics rather than fresnels. 🙂 That excites me. 🙂
--Clint
VERY interesting.... do you think that continuously pulsing an LED would overload it's max amperage rating? Or is the startup current included in all of that max-rating stuff?
Very neat stuff to think about. I'm trying to see what kind of lenses I would need right now in order to condense everything here... the incredibly cool thing about having a 2x2 inch light source is that you can use solid-lens optics rather than fresnels. 🙂 That excites me. 🙂
--Clint
Its a solid state device. Lots of solid state devices get pulsed every day. I would imagine as long as you stay within the maximum dissapation ratings it would be OK. Probably need to look closely at the data sheet.
Your main problem is going to be combining the output of X number of emmiters into one beam. You might be able to do it by aiming them all at a parabolic reflector, and collecting the output at the focal point.
Also many high output white LED have not much green content in the light. You need green. That has been posted before. Might show in a search Video projector part 2.
green and led
Maybe you could construct a white beam from regular red,green and blue led's and pulse them to very high output. Perhaps a thousand at 10 cents each or less. Thats only a hundred dollars.
Hope this helps.
Your main problem is going to be combining the output of X number of emmiters into one beam. You might be able to do it by aiming them all at a parabolic reflector, and collecting the output at the focal point.
Also many high output white LED have not much green content in the light. You need green. That has been posted before. Might show in a search Video projector part 2.
green and led
Maybe you could construct a white beam from regular red,green and blue led's and pulse them to very high output. Perhaps a thousand at 10 cents each or less. Thats only a hundred dollars.
Hope this helps.
Remember that when pulsing the light source this might also be interfering with the refresh rate of the displayed image and if they are not exact multiples (with the light being much higher frequency than the image refrash rate) you might get some pretty strange effects going on at the screen. For normal bulbs I think this is taken care of by the inherent slow nature of the on/off characteristics. For LEDs this is almost instant. Try it first.
/UrSv
/UrSv
remp said:
Actually I think that's probably the easiest problem to solve here. This is not my original idea, although I think it's a fantastic way to combine several light sources easily. I realize that you'd want to get these emitter assemblies as close together as possible, or you could use something like this fresnel array to combine it all together.
P.S. Why can't I do IMG tags correctly? That's bugging me.
Attachments
The 5 Watt Luxeon whites are only rated for 1,000 hours at this point. It will be mid-2003 before they have an extended life. For luminous efficiency, using RGB side emitters into a backlight diffuser plate may actually work well. That would allow full color control as well and make for an easy OHP configuration.
Alvaius
Alvaius
RGB Lights....
Alvaius, do you think it would work to do a combiner... like get an old set of beam-splitting optics and run it in reverse? Put in the three colors and get out white? Then use that as your light source? I would think you'd lose too much light with a diffuser... but maybe I'm wrong.
I'm a little confused as to what the difference is between Side-Emitting and normal luxeons. Can you explain the difference to me?
Thanks! Keep up the great work guys.
--Clint
Alvaius, do you think it would work to do a combiner... like get an old set of beam-splitting optics and run it in reverse? Put in the three colors and get out white? Then use that as your light source? I would think you'd lose too much light with a diffuser... but maybe I'm wrong.
I'm a little confused as to what the difference is between Side-Emitting and normal luxeons. Can you explain the difference to me?
Thanks! Keep up the great work guys.
--Clint
The side emitting Luxeons are just that.... as opposed to emitting forward, they have a small cone on them that emits the lide to the sides in a narrow angle. That can be used to edge illuminate a light pipe such as what you would see in an LCD backlight.
... am I into projection? .... more into DIY audio, but have had some experience in professional projection in a past life and a fair amount of experience in optics and video electronics.... enough that I would be happy to go buy a projector. I find it relatively easy to build audio equipment that is better than what I can buy at a reasonable price. I can't say the same for video. However, always willing to help a DIYER.
Alvaius
... am I into projection? .... more into DIY audio, but have had some experience in professional projection in a past life and a fair amount of experience in optics and video electronics.... enough that I would be happy to go buy a projector. I find it relatively easy to build audio equipment that is better than what I can buy at a reasonable price. I can't say the same for video. However, always willing to help a DIYER.
Alvaius
Member
Joined 2002
I wonder when there going to use white led's in aplictions to light small things.
AHHH i can see it now. MY light bulb will look like a LED soon. and my HYDRo bill will go up from all the aleph amps ( and class a ) all on.
AHHH i can see it now. MY light bulb will look like a LED soon. and my HYDRo bill will go up from all the aleph amps ( and class a ) all on.
Wholesale discounts...
Well Luxeon sells these really nice floods as specified here , (note that these are advertised at 100,000+ hours life), and you can buy the two different models like the 18-bulb flood array and the 12-bulb flood array from Future-Active . The cost difference between wholesale and single-bulb is enormous! Almost $100 for the 12-bulb array, and about $130 difference for the larger, 18 bulb array. Great cost jump! How can I get one of those flood arrays for the wholesale price though? Does anyone know what opportunities companies like Luxeon do to help out R&D by aspiring college students? I know one of my friends here built a driver board for a small camera (or is in the process of it... it's more complicated than he thought), but he wasn't able to get a discount on it. (one of those postage-stamp camera thingies). Anyway, does anyone who has experience in the industrial component/manufacturing/assembly world know how companies deal with this sort of thing? I can't afford $260 for one light, but I might be able to swing the cheaper wholesale cost of $120. 🙂
Well Luxeon sells these really nice floods as specified here , (note that these are advertised at 100,000+ hours life), and you can buy the two different models like the 18-bulb flood array and the 12-bulb flood array from Future-Active . The cost difference between wholesale and single-bulb is enormous! Almost $100 for the 12-bulb array, and about $130 difference for the larger, 18 bulb array. Great cost jump! How can I get one of those flood arrays for the wholesale price though? Does anyone know what opportunities companies like Luxeon do to help out R&D by aspiring college students? I know one of my friends here built a driver board for a small camera (or is in the process of it... it's more complicated than he thought), but he wasn't able to get a discount on it. (one of those postage-stamp camera thingies). Anyway, does anyone who has experience in the industrial component/manufacturing/assembly world know how companies deal with this sort of thing? I can't afford $260 for one light, but I might be able to swing the cheaper wholesale cost of $120. 🙂
That cheaper cost is not just wholesale, but volume pricing as well. How do companies help out? Depends what you are doing and what their product is and what they can get out of it. Microcontroller and processor companies will outfit labs because that generates people who are familiar with their product.
If you are doing something truly unique, you may get free samples, etc. if you share your results. If you are just recreating the wheel as 99.9% of the people are (no offence), then there is not a lot of benefit for a company.
If you are doing something truly unique, you may get free samples, etc. if you share your results. If you are just recreating the wheel as 99.9% of the people are (no offence), then there is not a lot of benefit for a company.
alvaius said:
99.9% of the people *here*, or everywhere?
I was thinking that the thing I could emphasize to them would be that this is a fairly new field, and if I can build a really good one-of-a-kind LED powered LCD projector, people might stand up and take notice, and at least take Luxeon seriously when they mean there's a definite future to these things.
Or were you saying I'm reinventing the wheel?
--Clint
I mean't University students in general, not here in DIY, but that is probably true as well, no offence to any, as there is some truly unique ideas out here.
So you want to build a one of a kind LCD projector. Well, do you have a research budget for this? The reality is that unless you have the money for custom optics, etc. you are not going to be even remotely competitive. Do you have optics experience, LCD experience, thermal experience, etc.
Let's start with the basic physics. If you want say 1,000 lumens output, you will need 5,000 - 10,000 lumens of input based on current LCD transmissivity. The average Luxeon die will put out say 25 lumens today. I say die because the 5 watt ones are 4 die (the 1 watts are really a little more than 1 and the 5s a little less). So, you are talking on the order of 200 die at today's efficiency and maybe more. I am assuming Red, Green and Blue LEDS for best color temperature. That actually results in a fairly reasonable cost light source of say $1,000 in volume. However, 200 Luxeon LEDS takes up an awful amount of space. Realistically, if you were serious about doing this, you will need to do it with die on a special carrier..... which is beyond most university project scopes. The reason you would likely need to go die is to get the size down to where you can use reasonable quality collimating optics. You now have around 200 watts of heat you need to get away from those LEDS or that 100,000 hour life will drop considerably, as will the efficiency, and the color temperature will change. That is a lot of heat to get rid of, and you really need to keep those LEDS at a reasonable temperature. So now you have some significant thermal design.
If you were to go to a company like Lumileds with a proposal for a professional quality project, most of the details at least roughed out, and how you were going to pay for it all, they just might support you with samples as you would fall into the 0.01% category and not the 99.99% category.
Of course you would be competing with companies like In-Focus, Sony, Sanyo, as well as companies targeting the high end..... all of whome have multi-million dollar research and development budgets. You could always align with one of these companies for a research project.
Sorry if I have burst any bubbles, but I am just stating reality.
Alvaius
So you want to build a one of a kind LCD projector. Well, do you have a research budget for this? The reality is that unless you have the money for custom optics, etc. you are not going to be even remotely competitive. Do you have optics experience, LCD experience, thermal experience, etc.
Let's start with the basic physics. If you want say 1,000 lumens output, you will need 5,000 - 10,000 lumens of input based on current LCD transmissivity. The average Luxeon die will put out say 25 lumens today. I say die because the 5 watt ones are 4 die (the 1 watts are really a little more than 1 and the 5s a little less). So, you are talking on the order of 200 die at today's efficiency and maybe more. I am assuming Red, Green and Blue LEDS for best color temperature. That actually results in a fairly reasonable cost light source of say $1,000 in volume. However, 200 Luxeon LEDS takes up an awful amount of space. Realistically, if you were serious about doing this, you will need to do it with die on a special carrier..... which is beyond most university project scopes. The reason you would likely need to go die is to get the size down to where you can use reasonable quality collimating optics. You now have around 200 watts of heat you need to get away from those LEDS or that 100,000 hour life will drop considerably, as will the efficiency, and the color temperature will change. That is a lot of heat to get rid of, and you really need to keep those LEDS at a reasonable temperature. So now you have some significant thermal design.
If you were to go to a company like Lumileds with a proposal for a professional quality project, most of the details at least roughed out, and how you were going to pay for it all, they just might support you with samples as you would fall into the 0.01% category and not the 99.99% category.
Of course you would be competing with companies like In-Focus, Sony, Sanyo, as well as companies targeting the high end..... all of whome have multi-million dollar research and development budgets. You could always align with one of these companies for a research project.
Sorry if I have burst any bubbles, but I am just stating reality.
Alvaius
nono, this is good. 🙂 Bubbles need to be poked to see if they're of any value or not. 🙂
Hurm... if the 5 watt emitters ever get any good amount of lifespan, I think that may be the best option. 12,000 lumens means 100 of the 120 lumen emitters, 14.5 mm on a side, so 10x10 means 14.5 cm or a 5.7 inch square... wow. That JUST fits for my 5.7 inch LCD panel. That would rock. 🙂 But anyway, yeah. 100 of the emitters giving off 10,000 lumens for a 1000 effective lumen projector (assuming 90% loss in LCD) in a 6 inch square, give about 6 inches for the various optics in the box, and you could have an INCREDIBLY small presentation machine. 6x6x10 or so.... here's a sample drawing of what I'm thinking, and how it would work. Something this small would be great for presenters on the go (plus if it was really quiet as this looks like it could be, it would be a big selling point too). The fan would be a quiet computer case fan or something, nothing huge, just enough to move air past the heat sink blades.
Whaddya' think?
--Clint
Hurm... if the 5 watt emitters ever get any good amount of lifespan, I think that may be the best option. 12,000 lumens means 100 of the 120 lumen emitters, 14.5 mm on a side, so 10x10 means 14.5 cm or a 5.7 inch square... wow. That JUST fits for my 5.7 inch LCD panel. That would rock. 🙂 But anyway, yeah. 100 of the emitters giving off 10,000 lumens for a 1000 effective lumen projector (assuming 90% loss in LCD) in a 6 inch square, give about 6 inches for the various optics in the box, and you could have an INCREDIBLY small presentation machine. 6x6x10 or so.... here's a sample drawing of what I'm thinking, and how it would work. Something this small would be great for presenters on the go (plus if it was really quiet as this looks like it could be, it would be a big selling point too). The fan would be a quiet computer case fan or something, nothing huge, just enough to move air past the heat sink blades.
Whaddya' think?
--Clint
Attachments
- Status
- Not open for further replies.