After seeing Tians and friends results with that 900 or so LED's backlight blasting out enough lumens to get a picture.. I thought well perhaps we can use our bigger lumen lamps in a similar manner..

I know this is could be more or less diffuse.. so are CFL's and the main thing I guess is enough rays get through..perhaps BEF would help...in any case the LED projector works against the theory perhaps this could...the trick perhaps is to get an even illumination through the lightguide plate from a nearly point source (our MH lamps).

BTW..a tad quiet around here than normal...even elks ....is everyone busy... !
Perhaps everyone is cooking up something..hope so !

and another..

big ball with a hole is not much different than a spherical reflector with a light gate. I assume the idea is that all light will bounce around in the ball until it eventually exits the hole? Well the likely hood of this light exiting at the appropriate angle will be slim. Also any light hitting the light guide will not be collimated by the fresnel because its incident angle will not be correct.

So you dont think either will work....
I wouldnt think so either..yet the LED projector works... with 30degree spread LEDS..with BEF/similar films I think you would get about the same angle...and we have certainly more lumens than the LED's.

Just chucking some ideas out...

The ball I agree pretty much... however the light guide could play more of a part if it and the exit point were designed to capture the high angle ray escapee's by flattening those out by the time it gets to the LCD...more rays at the near the right angle
Would enough light escape the ball to give enough for a result....I dont know.



Anyone else.

is there a fresnel in the light ball version? If not, all light would have to hit the LCD perpendicularly. There are light tunnels that do this for various optical systems, but I'd challenge anyone to build one exact enough as a DIY project.

A challenge, I like challenges.



I’m working on something similar to this at the moment. The guide is only part of the problem, the magnification it introduces is the biggest issue. But I have some ideas. It’s too late at the moment to post pics, perhaps tomorrow.

DJ

edit:
And don’t get your hope up. My ideas may fall in a heap once I test them.

BTW. LEDs will work just not very efficiently. To get a better idea of what is happening with tian’s setup you need to no the fresnel FLs to gauge how much of the 30 is getting to the triplet.

the prize is 100 points, and points mean nothing

On the subject of crazy ideas, may as well throw this out -



Source 1 and 2 could be part of an RGB luxeon array, traced in different colours only to show that the light from one pixel can (in theory!) be focused in one place...

Where is the snag?

Quote Tian first Said:

"we use 220mm focus lenses,67mm dia. ,with this lenses,led to lcd is 12-15cm,it can see now wave on screen,with 25-30 degree leds in a very close array,if you use lower degree led,or longer focus lenses,or lower dia. lenses, or the leds in a incompact array,you must uprise the distance from lcd to led.
when i use 330mm focus,58mm dia. lense,orther is same,i must keep 20cm from led to lcd to make no waves on screen.
thank you,elektron."

Then he said

"LED to fresnel:200mm
fresnel size:more than 9 inches,330mm focal length.

my friend carpow said,his PJ is brightter than me,he use 220 mm fresnel,220 lence,and 150 mm from fresnel to LED,I think he is right."

hmm so many ideas..so little time...and money...

I gotta say...

except for the very first idea at the top of this page...it seems that all the other ideas expressed here with some exceptions of using a light tunnel... are already being discussed over at

www.allinbox.com/allinbox.htm !!

Yes and no I didnt know about this french projector site either.....I just found it looking for LED's

It seems many minds are churning over a workable new light engine for the DIY projector...

heres the ball.....minus the light tunnel (my idea above) interestingly it has some raytraces...DAZZZ perhaps you can see if you agree with the picture results...


http://64.233.179.104/translate_c?hl=en&am.../language_tools


These guys have got the bug...even trying neons !

I urge the people that have time to have a good look around the site..lots to see.

I didn’t translate the page, just looked at the pics. There are two problems that I can see. They’re technically correct but the interpretation can be tricky. First problem is the number of reflections inside the reflector, look at how many rays finally emerge. The second is that not only is the fresnel projecting an image of the arc but also an image of the reflection. With the normal use of a spherical reflector it will basically reflect the rays back to the opposite side of the arc with slight spherical aberration. Because there is a second half reflector in front each successive reflection will eventually emerge from the hole. This all looks good, if you don’t take into account reflection loss, until you look a bit deeper and realize where these spherically altered rays are originating from. Just taking the end of the arc into consideration, instead of rays only coming from the end, they will be coming from many different arc end reflections at different position. So as far as the fresnel is concerned the arc end is no longer a small point but lots of small points. Effectively the whole arc becomes magnified.

There’s also the heat to think about as well.


DJ

I’ve altered this pic to show what really happens. The LEDs will emit diverging rays not collimated rays. So unless the LED’s angle is very small not a lot will go though the triplet.

DJ

Thats 2 down ...

DAZZZ whats your ideas..you were gonna post some drawings..

Cheers

That site is pretty cool..they have a competitIon for the Best DIY Proj of the year...!

I wonder who would win at LL...

wouldnt be me...dead last

Here’s another modified pic.

I’m sure tian’s projector is producing an image that he's happy with it but how much light is being wasted?
I don’t mean to sound like I’m negative to everything it’s just that most of this stuff I’ve already thought about in various forms.

DJ

I'm too tied now after all these explanations
Later tonight after I watch Xmen3

wasted light ..not efficient...it works not as well as it could

I agree totally...

what are you thinking ?

damn it you posted before me LOL

Ok Xmen3 it is... till then...

No one has commented on the first idea (top of page..start of topic) ....

I can see what DAZZZ is saying about Tians Proj you can see the wasted light...my thought of using lots more lumens to do the same thing can only have 2 results...

Worse or better ..what would it be ?

LCD monitors used to have notoriously poor viewing angles...and laptops still do. And so I'm guessing the light passing through already has a narrower cone than say a piece of plain glass.

You idea may work with powerful enough lamps. Also how do you keep the light from escaping all the way out the other side?...

Escaping light...yes well how is it stopped in a comercial LCD monitor backlight now I wonder...?

Although you cant see it in the drawing the lamps would have a reflector..so if those rays made it all the way to the other end then they would get reflected back hence they came...

Im thinking an optional idea being the lamps positioned opposite ends of the shortest side of the LCD panel.
Each having a tapered light guide into the lightguide plate.

The other open ends of the light guide plate (longest sides)would have a reflector to bounce the rays back for another go.

The best angles for the tapered light guide..the bestest reflector would give the maximum number of rays hitting the LCD reflector/mirror with the best chance of getting through the Panel..

I guess you know you would get something out of the panel..but with BEF perhaps careful design would it be worse or better than the Tians LED proj...that's the question ?

And would the light be even...? That's another question...so thats 2....questions

Could you have a BEF against the LCD mirror as well as the LCD......to add more efficiency.

so if a ray hits the mirror side it is roughly collimated thru the BEF to the mirror behind then reflected at the bestest angle to wards the LCD....

Can you go back thru the BEF like this ?

Answer:
hmm ..just had a look at 3M site I think the answer is NON...

However I did learn that for max increase a wedge lightguide with 2 BEFII films (crossed 90degrees)...

when you see the magnified BEF film ....it sure does look like those prismatic diffusers you see on fluorescent light fittings...but micro..

I started a new topic for this idea.

DJ

I havnt carried out much digging on LED's but I think the smallest angle is 5 degrees...how would that look.....still going to lose a lot of light.

Do they make a collimating lens/reflector for an LED...Ideally an elliptical lens/reflector that has it's F2 at near the trippy..!

I doubt it..I go and have a search.

Ok from the curved LED light source idea...I found what the author was thinking to reduce the divergence of the LED...

On the same theme this time added a fres..

It looks good but...all those losses..each LED having it's own reflector...1000's of them..

Found a patent that claims up to 7.5 degrees collimation achieved using a tapered light guide..
DIy version is it possible....




Highly collimating tapered light guide

http://www.freepatentsonline.com/6648485.pdf

Now something else..has anyone seen this lamp and reflector before...all I know is it is a 150w MH POWERSTAR HQI-R...is it known...?

That POWERSTAR HQI-R 150Watt has a parabolic reflector.
connected by optical fibre !?* cant find to much about it..but interesting nevertheless.

The crazy idea (top of topic) of using high lumen lamps as what in effect is a powerful backlight..

After looking at the patent reading the 3M site.. come to a conclusion maybe that first idea diagram I posted does not represent possibly the best method to try.

More..both 3M and the patents refer to a Tapered Light Guide as giving the most benefit for collimating the light..

So to increase collimation incorporate the BEF films x2 ...and after reading the patent possibly try putting the BEF before the rays get to the light guide. The reflector could still be a mirror...

Of course calculations for exact Lightguide design..perhaps raytracing can help..but as we are DIY we are not talking in 0.5 degrees tolerance....

It's a pity that these BEF films work at the micro level..however the patent talks about an acrylic sheet...perhaps there are alternatives..

The other thoughts I had were ..what about the lamp would it be better to use a coated or non coated....in other words diffuse or non diffuse as the lamps source....in other words for maximising collimation what would be best..

The Lamps reflector ...tricky...one idea is a parabolic trough..seems to be favoured by the europeans in their DIY projectors...perhaps then a diffuse lamp would be best in combo with the reflector...however..minds open.

On the same subject luxeon backlights incorporate a mixing guide before it goes in the LCD panel light guide...

It's all about maximising the collimation for us..

The patent says you can make it more or less collimating depending what you want to achieve...

Of course With a tapered Light Guide-(or guides) and 2x High lumen lamps we would get the screen to be quite bright no doubt.. but how much could we collimate for our purposes enough or not enough...how could you calculate what the output could be before you start...?
Is it possible to manage a raytrace...

Or is it impossible to get an idea...and just have try and find out..

It shouldnt be to hard to find out..need a mirror..maybe paint stripped ala Elks FS trick..
Try one lamp first..make a parabolic trough reflector....out of Aluminium or perhaps silver steel sheet..try without the BEF ..if looks promising ....get some BEF/DBEF sample.....look for alternative collimating films sheets..

just noticed I missed the IR FIlter on that..as heat would be a factor.

hehehe just when I had got my head around that Tapered Light Collimating Guide..I find another Patent claiming collimation at +/- 3 degrees.....

http://www.freepatentsonline.com/6464365.pdf


edit:
Having a quick look...interesting thing so far is the integrating sphere concept (ball with a hole) is sorta used in this patent for concentrating all the light from a linear diffuse source (fluro) in the use of a tube with a slit..then attached to a parabolic reflector.....they claim that the tube with a slit is 75% efficient.
But we are talking a 1mm slit !! the tube reflector is also diffuse...interesting.

btw
I worry when I read these patents as has been proven I miss things that could well be useful..so if anyone is interested in these subjects and read the patents..and spots something ...post away

Cheers

It seems Ive been talking to myself a lot in this thread....I think I better shut up for a bit and work on my inner monologue skills

OK...I going to post some info from patents and other sources..so mainly for reference something perhaps to get some ideas from....

First up is one idea from a designer the old sphere trick..

http://www.freepatentsonline.com/4765718.pdf

Another sphere idea..

http://www.freepatentsonline.com/4735495.pdf

Now designs a lot of people are familiar with.. double parabolic reflectors or DPR


http://www.freepatentsonline.com/6227682.pdf

The rest are more exotic examples of multiple lamp source reflector engines

Note: Im not showing all the ideas from the patents

http://www.freepatentsonline.com/6231199.pdf

...

...

yes and there are so many more..all the parabolic patented designs are by Kenneth Li

see more of his here...

http://www.freepatentsonline.com/search.pl...&sort=chron

btw kenneth li is the pres at


http://www.wavien.com/

This site is quite an eye opener...show cases a lot of his patents in lay persons way..with some animations..

Heres an idea of whats at the site ....looks like Dazzz's horn of light ..Tapered light pipe with lens.

last but one

reducing the arc size !

I..know I said last one..but this one is more conventional so useful for ref again..

it shows a parabolic which uses lenslets...but it shows the lenslets smaller than the reflector diameter...the rest of the diameter is using a reflector.

http://www.freepatentsonline.com/20020176255.pdf

That’s identical to what I’ve been trying to do. Even the angle problems he mentions are the same.

DAZZZ yep ! it proves you are spot on with your ideas !!!

and the great thing is that most all of the designs are using a form of this same idea..
so u r on the right track...keep going

also check out the arc reducing design...

Heres last one ..sorta promise..at least for tonight..

This shows the DPR in 3D action..

Now I couldn't quite figure what the setup was with these DPR's now from this I can see that all they are is a parabolic cut in halve down the vertical axis and then the two put together on their edges.

I know a lot of this has been Parabolic biased..however...there is so much info ..honestly that's as far as Ive got so far..ellipsoids another day !