EDIT: I am completely wrong, I read the highly coincidental post about LEDs before me and used a calculator (http://led.linear1.org/lumen.wiz) to find that the with the 12 degree beam angle and the 130,000MCDs that the luminous flux is 4.475 lumens. An array of these would be way too dim. Looks like the MH lamp will be king for quite some time....
EDIT 2: With the revised lumen rating, I calculated that this hypothetical projector might produce a mere ~62 lumens at the screen. Not good at all.
I decided to revisit the LED lighting topic today, so I did some research.
I visited ebay to see what kind of LEDs were available (online retailers were really expensive) and I found 1cm 130,000mcd LEDs. Just search ebay for 130,000mcd. I found out online that mcd stands for millicandela, which I assume means 1/1000 candela. That would mean each LED puts out 130 candela. A quick entry into google calculator tells me this:
130 candela = 130 lumens
130 Lumens! That is quite a bit for an LED, although it is rather large at 1cm. Then I decided to calculate how many of these LEDs I could fit in an array behind my 17" panel:
LCD dimensions are roughly (I cant measure right now) 13"x11" = 143in^2. Another google conversion tells me that:
143 (in^2) = 922.5788 cm^2
Assuming I can fit 1 LED per cm^2, that means I can probably make an array with 922 of these super bright LEDs.
Some dimensional analysis:
130 Lumens/LED * 922 LED/Board = 119,860 Lumens/Board! That is ridiculous, almost 3 times the output of the Ushio T15, which most of us know comes in at "40,000+" lumens.
So how much light makes it to the screen? Since the array fits well behind the LCD, I assume we could use some type of diffuser layer in between the array and LCD (salvaged backlight elements?). Let's assume that half of the light makes it to the LCD, the LCD is 5% transmissive, and losses after there are 40% (I assume it would be less). Calculations:
119,860 lumens * .5 (pre-LCD loss) *.05 (LCD transmittance) * .6 (light that makes it on screen) = 1,797.9 Lumens.
Almost 1800 lumens, that is ridiculous.
Sadly, I suspect that there is something seriously wrong with my calculations. That, or the manufacturer is lying to us about the 130,000mcd brightness.
I know that LED lighting has been discussed many times before, but I couldn't keep this to myself. A side-note: the LEDs are going for $8.59/100 LEDs bidding price. The BIN is $27, but it doesn't look like there are any bids.
If I am right about this, we could seriously improve our projectors' brightness. But I doubt I am right about this, anyway =).
Someone want to prove me wrong?
Nick
Full Version: LED Lighting (with calculations, might it be feasable?)
QUOTE (nickestorga @ Apr 10 2006, 01:42 AM) 
EDIT: I am completely wrong, I read the highly coincidental post about LEDs before me and used a calculator (http://led.linear1.org/lumen.wiz) to find that the with the 12 degree beam angle and the 130,000MCDs that the luminous flux is 4.475 lumens. An array of these would be way too dim. Looks like the MH lamp will be king for quite some time....
EDIT 2: With the revised lumen rating, I calculated that this hypothetical projector might produce a mere ~62 lumens at the screen. Not good at all.
I decided to revisit the LED lighting topic today, so I did some research.
I visited ebay to see what kind of LEDs were available (online retailers were really expensive) and I found 1cm 130,000mcd LEDs. Just search ebay for 130,000mcd. I found out online that mcd stands for millicandela, which I assume means 1/1000 candela. That would mean each LED puts out 130 candela. A quick entry into google calculator tells me this:
130 candela = 130 lumens
130 Lumens! That is quite a bit for an LED, although it is rather large at 1cm. Then I decided to calculate how many of these LEDs I could fit in an array behind my 17" panel:
LCD dimensions are roughly (I cant measure right now) 13"x11" = 143in^2. Another google conversion tells me that:
143 (in^2) = 922.5788 cm^2
Assuming I can fit 1 LED per cm^2, that means I can probably make an array with 922 of these super bright LEDs.
Some dimensional analysis:
130 Lumens/LED * 922 LED/Board = 119,860 Lumens/Board! That is ridiculous, almost 3 times the output of the Ushio T15, which most of us know comes in at "40,000+" lumens.
So how much light makes it to the screen? Since the array fits well behind the LCD, I assume we could use some type of diffuser layer in between the array and LCD (salvaged backlight elements?). Let's assume that half of the light makes it to the LCD, the LCD is 5% transmissive, and losses after there are 40% (I assume it would be less). Calculations:
119,860 lumens * .5 (pre-LCD loss) *.05 (LCD transmittance) * .6 (light that makes it on screen) = 1,797.9 Lumens.
Almost 1800 lumens, that is ridiculous.
Sadly, I suspect that there is something seriously wrong with my calculations. That, or the manufacturer is lying to us about the 130,000mcd brightness.
I know that LED lighting has been discussed many times before, but I couldn't keep this to myself. A side-note: the LEDs are going for $8.59/100 LEDs bidding price. The BIN is $27, but it doesn't look like there are any bids.
If I am right about this, we could seriously improve our projectors' brightness. But I doubt I am right about this, anyway =).
Someone want to prove me wrong?
Nick
EDIT 2: With the revised lumen rating, I calculated that this hypothetical projector might produce a mere ~62 lumens at the screen. Not good at all.
I decided to revisit the LED lighting topic today, so I did some research.
I visited ebay to see what kind of LEDs were available (online retailers were really expensive) and I found 1cm 130,000mcd LEDs. Just search ebay for 130,000mcd. I found out online that mcd stands for millicandela, which I assume means 1/1000 candela. That would mean each LED puts out 130 candela. A quick entry into google calculator tells me this:
130 candela = 130 lumens
130 Lumens! That is quite a bit for an LED, although it is rather large at 1cm. Then I decided to calculate how many of these LEDs I could fit in an array behind my 17" panel:
LCD dimensions are roughly (I cant measure right now) 13"x11" = 143in^2. Another google conversion tells me that:
143 (in^2) = 922.5788 cm^2
Assuming I can fit 1 LED per cm^2, that means I can probably make an array with 922 of these super bright LEDs.
Some dimensional analysis:
130 Lumens/LED * 922 LED/Board = 119,860 Lumens/Board! That is ridiculous, almost 3 times the output of the Ushio T15, which most of us know comes in at "40,000+" lumens.
So how much light makes it to the screen? Since the array fits well behind the LCD, I assume we could use some type of diffuser layer in between the array and LCD (salvaged backlight elements?). Let's assume that half of the light makes it to the LCD, the LCD is 5% transmissive, and losses after there are 40% (I assume it would be less). Calculations:
119,860 lumens * .5 (pre-LCD loss) *.05 (LCD transmittance) * .6 (light that makes it on screen) = 1,797.9 Lumens.
Almost 1800 lumens, that is ridiculous.
Sadly, I suspect that there is something seriously wrong with my calculations. That, or the manufacturer is lying to us about the 130,000mcd brightness.
I know that LED lighting has been discussed many times before, but I couldn't keep this to myself. A side-note: the LEDs are going for $8.59/100 LEDs bidding price. The BIN is $27, but it doesn't look like there are any bids.
If I am right about this, we could seriously improve our projectors' brightness. But I doubt I am right about this, anyway =).
Someone want to prove me wrong?
Nick
Hello,
Nice post, I see you just confirmed my conclusion I posted 31 minutes before you posted yours... Haha
I also think it is very odd both of us posts now topics about almost the same thing, at the same time, at the same locations... 
Feel free to read it at http://www.lumenlab.com/forums/index.php?showtopic=11267.
With best regards,
TJ
Indeed, I was quite surprised about the coincidence.
Sounds interesting! I agree, you'll need some sort of diffuser as you need
to fill in the gaps between the light cones for each LED. And no collimating
fresnel or UV shield is needed. Probably no IR shield either. So that's
quite a bit of losses that can be eliminated.
Something like ground glass or a diffuser from a fluorescent overhead fixture
(cheap too) could be used. The light hitting the LCD should be within several
degrees of perpendicular otherwise the light passing through the field fresnel
won't converge on the PJ lens correctly.
There is a special lens called a TIR, which means Total Internal Reflection,
used with some LEDs to collimate the light. I believe they're something like
a fresnel only with additional steps and facets. I've seen them discussed
on the candlepower forums but haven't looked into them too closely.
The candlepower forums are inhabited by those seeking the ultimate flashlight.
Those guys will spend up to $800.00 for one!! The best use 10-50 watt MH
lamps.
to fill in the gaps between the light cones for each LED. And no collimating
fresnel or UV shield is needed. Probably no IR shield either. So that's
quite a bit of losses that can be eliminated.
Something like ground glass or a diffuser from a fluorescent overhead fixture
(cheap too) could be used. The light hitting the LCD should be within several
degrees of perpendicular otherwise the light passing through the field fresnel
won't converge on the PJ lens correctly.
There is a special lens called a TIR, which means Total Internal Reflection,
used with some LEDs to collimate the light. I believe they're something like
a fresnel only with additional steps and facets. I've seen them discussed
on the candlepower forums but haven't looked into them too closely.
The candlepower forums are inhabited by those seeking the ultimate flashlight.
Those guys will spend up to $800.00 for one!! The best use 10-50 watt MH
lamps.
A flashlight..with a metal halide lamp..that costs $800. Now I've heard everything!
Sorry to disappoint, but 1 candela does not equal 1 lumen.
You can’t directly convert since they measure different things. The most useful explanation I’ve found is that lumens measure light output at the source, while candelas measure the light that falls on a surface. As the area of the surface increases, the number of candelas will decrease even as the number of lumens remains constant.
* from http://led.linear1.org/how-do-i-convert-be...las-and-lumens/
Candela to Lumen Calculator
http://led.linear1.org/lumen.wiz
You can’t directly convert since they measure different things. The most useful explanation I’ve found is that lumens measure light output at the source, while candelas measure the light that falls on a surface. As the area of the surface increases, the number of candelas will decrease even as the number of lumens remains constant.
* from http://led.linear1.org/how-do-i-convert-be...las-and-lumens/
Candela to Lumen Calculator
http://led.linear1.org/lumen.wiz
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.