The Failure of LED Light Source LCD Projectors

Status
Not open for further replies.
Jump to Latest
W

wazootyman

Member
Joined 2002
e-reinhardts.com
#1
I had a great time reading about everyones progress on their LED sourced LCD projectors and I am in the early phases of designing my own DIY projector. I have done a few calculations based on the LEDs used and I think I know why this whole LED thing won't work (with the current technology anyways).

As an example, Vince's LEDs have a 20 degree veiwing angle and 6.4 candella (6400 mcd) output which is amazing for an LED. One candella is one lumen per steradian. The LEDs veiwing angle of 20 degrees corresponds to about 0.095456 steradians. This means that each LED only has about 0.6109 lumens of total output. If you wanted to make an LCD projector that has about 500 lumens of output, you would need about 820 LEDs assuming that NO light was lost until the image left the entire optics assembly.

This is my first crack at the calculations, so let me know if I have any gross conceptual errors here.

Thanks,
Jason
 
#7
It's just a little impractical

Well, let's see... why wouldn't we do it... hmmmmmmmm...

Ooooo! I know! Hows about at $1.50US+ per LED, the source light portion of you "cheap" DIY projector, would run about $1230US! For that you can practically get a good used or low end new LCD projector. Or maybe if money is no object for you, you may be disuaded by the thought of hand wiring 1640 leads (820 LEDs at 2 conductors each) just to get a light source. Be prepared to do a whole lot of work for not much output!

Also, I have a feeling that if you made a 42 X 32 array of LEDs evenly spaced behind a 10" LCD panel, your image would have 42 X 32 array of hot spots. That is unless you were tricky with a diffuser or something.

Point is, while extremely attractive from a heat and lifetime point of view, LEDs have a long way to come in the Lumens/$ and effort area.

Jason
 
#11
I sort of grouped the LCD opacity into the statement "assuming that NO light was lost until the image left the entire optics assembly" which of course, is impossible. Point is, assuming optimum (100%) transfer of light through the device, an LED light source still is unfeasible. Thank you for the quantification of the LCD transmittance. Good information.

Thanks,
Jason
 
#13
alvaius said:
It is doable, just quite expensive at this point. I think you could do it at a much lower cost, with better efficiency, and with pretty good life using and array of CCFL's.

Alvaius
Click to expand...

I was wondering about that myself Alvaius. 🙂 Something I *REALLY* need to know is how much light CCFL's output in comparison to high-output LED's. Just using a fresnel and a projection TV lens I've gotten viewable pictures from a laptop computer screen.

If we could replace the CCFL with a row of LED's (and it was that much brighter), we would have alot of advantages. Normal CCFL Backlight optics are great for LCD panels because they lose VERY little light, and they even have those wonderful polarizing reflectors in there so that they keep 90% of the light as opposed to losing half of it like a normal LOA or MH bulb would. So if we could either funnel an LOA into backlight optics (hard), or replace the CCFL with a whole row of 10,000 or 20,000 mcd LED's, then that's a viable option.

Check out
http://www.diyaudio.com/forums/showthread.php?s=&threadid=8739&perpage=15&pagenumber=3
where I did more of the calculations. Two posts halfway down the page, one with a diagram and the other with calculations.

Lemme' know what you think.

--Clint
 
Status
Not open for further replies.
Top Bottom