Well it seems like it's time to start a permanent thread on reflection. I'll be adding Joe's lightbox to this thread (but it would be better for you to do it Joe since you have the details). Let me start with a simple reflector I tried today with good results. The standard stainless bowl cut so that the lamp can sit at the focus of the half sphere. It is important to postition the lamp properly, the focus of the reflector should be ascertained. Try to avoid any bowl that isn't a true half sphere, meaning doesn't curve 180º. A mirror polish on the interior is important. I will be expanding on this concept soon. Remember you want the focus, NOT the focal surface. I'll include the standard illustration after the photos. The cool thing about this mod is that it's super easy and shouldn't require much modding to your case or lots of time to implement. You WILL get more light through your system with this mod.

And lit up.

Illustrating the focus (center of sphere) and focal surfaces.

What is the best way to ascertain the focus of the reflector?

Thanks for the great info brain

The best way to be sure is to buy a true half sphere, in which case the focus is easily determined. Imagine that your true half sphere is a bowl, and the bowl has a lid on it. If you put a dot at the center of the lid, that would be the focus of the spherical reflector. A half sphere should look like a globe cut in half. The flat spot in the middle shouldn't matter but try to get something that has the least flat spot as possible.

A spherical reflector that is not a true half sphere will capture less light and require a longer enclosure. These observations only apply to lamp arcs placed at the focus, not to other types of reflection.

Here is the EYE lamp with an 8 inch electric range reflector. Now that you say we need half spheres, this is probably not the way to go. Plus, there is a big hole in the back.

Here is what appears to be an almost spherical bowl I found at Ace Hardware for $12 in their kitchen wares. It is their #1024 Bowl with Lip, 6 cup/1.5 Qt/1.42L stainless, made by Krona Norpro. It is probably undersized as the entire glass envelope will not fit inside. So I need to look at the next larger size tomorrow. There is a 0.25 inch flat spot on the bottom of the bowl that would be perfect for placing a single screw for mounting. It is 6.25 inch across at the mouth, not counting the rim, and 3.5 inches tall, so it is not truly spherical. (edit: Well, it certianly looks spherical, it is probably measurement error. I will have to verify with a circle cut out of paper and match the curves.)

Here is the inside view. It looks like the bottom is flat, but it is perfectly spherical but for a small 0.25 inch flat spot, perfect for a screw head.

i have found that a light tent wil not provide any better results for our design.

i made a temp. setup with my fres. and projection lens and used a laser to beam light through.
this is what i found:
at the focal point of the fres. i could angle the laser just about anyware on the first fres. and get it to come out of the projection lens.
there is a bulls eye to the fres. if found that is approx 2" in dia. if i found the focal point i could move the laser up 1" or down 1" any more than that would reflect/ reject the beam. i moved the laser outside of the bulls eye and angled it to see if i could redirect the light to the center, but the fres would reject it. in short, out side of the bulls eye i could get the light to go everywhere except out the projection lens.
so a light tent that canot bring the light back to the focal point ie. the bulb/ reflector
will not help us.
i recommend trying this yourself with a laser because the effects of the rejected light look really cool in the dark. sort of a 3d holigram type of effect.

Tip: If you find the perfect metal bowl and it's missing the mirror finish and won't polish - try a shop that rechromes bumpers for cars, they may be able to apply a chrome coating to the bowl too!

This link only works for industrial jobs but if the perfect hemisphere was located and needed to be chromed, then maybe a group buy through lumenlab?

Sydney, That bowl from Ace looks almost perfect. The lamp doesn't have to sit in the bowl, the arc would be even with the lip of the bowl. So you would only have cut a notch in the rim.

For the mirror finish; if the bowl is stainless it should polish to a mirror finish without too much difficulty.

Are we talking a huge difference in quality (like joe's)? Is this better, worse, or the same as joe's lightbox concept? It looks definitely cheaper and easier to do. However it doesn't look as cool as Joe's

Instead of the lightbox one could place a 'firewall' between the bulb and the tempered glass/fresnel. The firewall could be a flat piece of metal that is the same size as the tempered glass but with a circular hole cut into it to allow the 'point light' source through.

In theory this concept if properly implemented should be better than Joe's due to the large reflector size. A mirror finish, a spherical reflector and proper alignment is everything.

This stainless steel bowl is 10.75 inches diameter and 4 inches deep, with a 4.5 inch diameter flat bottom. Ace Hardware item 107, sku 61787 The receipt says it is 8 Qt. $6. It looks spherical, but I haven’t verified it to be an exact sphere. The flat bottom takes away a lot of the spherical wall. This looks similar to Brain's.

This stainless steel bowl is a heavy gage mixing bowl that is 9.25 inches diameter, fully a half sphere (no flat bottom) that transitions into a 1 inch long cylinder at the rim. Krona Norpro brand Ace Hardware sku 631037. The receipt says 4.5 Qt mixing bowl with lip $30. This one is truly spherical and quite polished. Does the 1 inch long cylinder around the rim cause any concern as to light output? Just a little of the glass envelop at the center of the lamp will extend beyond the rim.

Same spherical bowl shown with a grip ring for a 3 to 8 inch Krona bowl so it doesn’t roll around on the spherical bottom when you are mixing. I think this would help stabilize the mounting of the bowl if one screw was used. Item #1022 5 inch grip ring, Ace Hardware sku 631022, $3. There is a larger ring for about $9.

Note the one inch transistion to a cylindrical shape from the half sphere shape.

I would cut the rim off for a couple of reasons. It makes the enclosure longer. It could interfere with the reflection. The smaller reflector you got did not seem too small to me,

I need a cost benefit analysis.

Choices so far:

$6 10.75 inch diameter flat bottom bowl, only one notch needed for the lamp base

$12 6.25 inch diameter full half sphere bowl that needs a notch for the top of the lamp to stick out of, the glass envelope of the lamp will extend beyond the bowl at the top and bottom.

$30 9.25 inch diameter full half sphere bowl needs no notch for the lamp top and only a notch for the lamp base.

Which bowl makes sense to buy? How much better are the more expensive choices?

Would a stainless steel collandar (with holes) be good for air flow at a sacrifice of some light?

Any chance you can send me one of those 1.5 quart bowls? We don't have an Ace around here..

I took the smaller bowl back to Ace, so it looks like I should take the larger bowls back and get the smaller. I like the base ring for stability for $3.

Do you think the lamp will overheat being confined on one side?

At one time, Brain, you said that reflecting the light back to the arc might be bad for the lamp. Do you think it is ok now?

Thanks for your help.

Brain,

Send me your address and I'll get a bowl and a grip ring sent to you tommorrow, my treat.

Thanks for your help.

Much appreciated Sydney. Address in the mail. If you need anything just let me know.

Look how bright are the inner MDF walls, and how wide is the point light. Each pixel from the LCD panel is going to apear at several places on the screen. This should lead to reduced sharpness and reduced ANSI contrast ratio (it might even have a perceivable effect on absolute black levels). Rorshach's idea together with black paint inside the box should really make a big positive impact.

I'm going to get one of those bowls and implement it in my design when I start next week, this keeps getting better and better. With what I know now I would have paid $100 just to get your plans and get in this forum.

Awesome work brainchild

Im with you Nore. Awesome work Brainchild

Yes I originally said that it would be a bad idea especially for a small lamp but I believe the larger lamps can take it, especially because we are using relatively inefficient reflectors. Of course only time will tell and I hope to get enough time on this lamp to know conclusively before it becomes a 'staple' in our projectors. Worst case scenario is the lamp will burn out a year earlier (2008) or it could rupture BUT we are already running these lamps cooler due to forced cooling, which they are not designed for. It's a chance I'm willing to take because well...it's fun.

Explain how each pixel will appear at several places on the screen. Off axis light is simply rejected by the fresnel.

I just made my plan purchase about 5 hours ago and have been reading up in the members forum ever since, I have read discussions of reflectors, light boxes, and pyramids to reduce the amount of "stray" light. I have an idea to present, but first a little background. At my company we have 6 top dollar commercial projectors, I mean these are good. In meeting rooms with over 100 32 watt flourescents burning in the ceiling, these are crystal clear. They are NECs, of course no one but NEC makes a replacement lamp so the price per bulb is $749 (the reason I joined this forum and ended up buying the plans). Point is, the lightbox is built around the bulb in ours and in most every commercial projector I have seen, and those damn bulbs get hot, so if it works for them... They also use a light tunnel, albeit an awkward one as they use cold mirrors and seperate LCD panels to represent R, G, & B, so again, if it works for them... This image illustrates the use of a reflector (light box a bit tough to draw in PSP) and a tunnel:

There is supposed to be an image here, however my damned free web space prevents linking directly to an image so here is the link to the directory with the image... It's the only file there so it shouldn't be hard to find:
pyramid.jpg

I apologize if someone has already posted a design in this area, I haven't read EVERY post yet.
The tunnel can be whatever material, however I am a damn good polisher, so, stainless polished to a mirror finish, possibly, four walls from glass back to rear of projector, L bend at rear to secure it to the enclosure, each wall wrapped over and secured to the next for stability, in the picture with only a reflector, make it a little larger than reflector diameter, then use a square piece of the tunnel wall material sized to wrap the four walls with a hole cut to fit around reflector perimeter in the center and give light no where to go but toward the LCD.
Sorry for the ungodly runon.
Obviously, design is simplified with a square/rectangle light box to mate the tunnel to.
Hopefully I was halfway coherent in that explanation as it's 1:20am and I have been awake for 22 hours.

Any thoughts?

Welcome jammor, looking forward to your designs.

sydneybird what was the ace hardware item number for the small bowl?

is this it?

http://www.acehardware.com/product/index.j...rentPage=family

Sullivan18, no, those are not the bowls. The bowl in question does not have a flat bottom. Well, there is a 0.25 inch flat spot on the bottom.

I didn't see the same bowl offered on the Ace Hardware website, nor have I found it searching Google for Krona Norpro bowls, but I didn't check every store (I stopped after 2 hours of searching). On my sales receipt, it says 631024, S/S 1.3 Qt mixing bowl w/lip. $11.99. (1.5 Qt is clearly stamped on the bowl)

Looks like I should buy out their supply so that I can offer them to the members. They only had three or four, so I'll ask how I can get more.

I thought that this colandar might be a good choice if we thought we needed air flow around the lamp. Norpro Stainless Steel 8-5/8-Inch Colander. Amazon.com $17

I found the spherical bowl on Norpro's site, who only sells wholesale. Anyone know how to buy stuff wholesale?

We are looking for the Krona #1024 Bowl with Lip, 6 cup/1.5 Qt/1.42L stainless steel;

and the #1022 5 inch S/S Grip Ring. The grip ring might be useful to stabilize the bowl when screwing the bowl to the side of the box.

http://www.norpro.com/pages/krona/mixing_bowls_w_rings.html

We will look for Brain's blessing next week as I will send him one to test.

As far as I understand how a condenser frensel should work (but please correct me if I'm missing something), it expects light to hit each portion of it, at a very specific angle, because each portion (prism) of the frensel, bends light at a different degree. Theoretically, if the light comes only from a point light source, which is at the appropriate focal lengh distance from that specific frensel, then all the rays of light, after they've passed through the frensel, will be parallel to each other (this is what we need from the first frensel).

If the frensel is being hit by a light coming from outside of the point light, one of these two options should happen:

1) If the angle of the hitting light will be such, that after going inside the frensel, it will try to go outside of the frensel at a degree lower than the critical degree for absolute mirroring, then the ray outputted from the frensel will not be parallel to those which came from the point light (hence causing "each pixel to appear more than once on the screen", assuming it will also go through the second frensel, like it went through this one. If this won't happen (i.e. the light won't make it through the second frensel) you can also expect the light to bounce back and forth between the frensels, until it either all turned into heat, or finds a path through one of them).

2) If the angle of the hitting light will be such, that after going inside the frensel, it will try to go outside of the frensel at a degree higher than the critical degree for absolute mirroring, then the light will be mirrored back toward the back of the box (potentially causing interference with light coming out of the lamp).


The point is, as you like to say, "you can't cheat the laws of physics". So, by saying that "Off axis light is simply rejected by the fresnel", you somehow dismiss (or maybe repress) the law of conservation of energy, which dictates that unless this off-axis light will be transformed into heat at the frensel (and it won't do that anymore than on-axis light do it), then it will have to chose either path #1, or #2, and both don't look thrilling.

Maybe someday in the future, they will invent a frensel that is capable of turning all off-axis light into heat, but until that day, just painting the inside of that box with black (transforming the far off-axis light into heat before it even has a chance to hit to the frensel), and trying to make sure that the source light is closer to point light as much as possible (reducing closer off-axis light from reaching the frensel), should at least theoretically help a lot.

Interference isn't a major concern until after the object, in which case it will degrade picture quality. You said that a pixel would be projected in 2 places (or more) simultaneously; the only thing that could do that is an optic after the object (lcd), and it isn't an optic that we would ever use in projection unless we were intentionally splitting a beam.

The only problem associated with large point sources is a central hotspot, because more light makes it through the center of the fresnel due to less lensing going on there.

Make a ray tracing for me that illustrates how a pixel can be projected in 2 places.

Something along this,

Light is polarized by the panel, no matter what angle light comes into the panel, if it makes it through the panel it is polarized. If light can get through the panel is coming straight out of the panel, where it is collected. It is impossible to 'duplicate' a pixel by your scheme. The implication of off axis light per your illustration would be uneven lighting.

Brainchild, I'm not quite sure what you're saying here. Let's say we chose a certain pixel on the screen and call it "X". Are you saying that each photon that hit X, no matter where it came from, or at which angle it hit X, it will leave X headed in only one possible direction ?.

The ray path that I sketched above doesn't show only one path after each pixel. The ray which comes from the point light will take a different path. Yet you didn't rule out my sketch. I'm a bit confused here...

I added another ray to the sketch (in green), to show the path that a ray will do from the point light, which will make it hit the exact same pixel as the red one hit, at the LCD. Do you see now ?, they are both taking the same pixel to different places on the screen.

Is 10" wide and 5.5" high to big for a completly half sphericle bowl?

brain,

when using a spherical reflector, according to your pictures, the light returns parallel to a straight line from the lamp to the lcd. aren't we trying to have a point light source the spreads to all edges of the fresnel? from my common sense thinking, which is usually wrong, this would make the center of the projection brighter. though it doesn't seem to do that with all the projectors i've seen that use reflectors. could you explain this?

that's a huge bowl! i'm assuming your not going to try the light box. i was thinking about using a bowl around 5-6" in diameter at the most. should it be bigger?

That could only happen if the panel didn't polarize the light. Off axis light can't even get through the panel. There is an X axis polarizer on the back of the panel, and a Y axis polarizer on the front. Any light that makes it through those is definitely going in only one direction.

So, how does the light from the point source goes through the Y axis polarizer ?. Because you use the collector frensel before the LCD, the light which come out of it, is angled down toward the panel. Shouldn't such angled down light be blocked by the Y axis polarizer ? (or at least most of it).

brainchild,

i was wondering if you have knowledge on the construction of lcd
panels, in the where and why they place certain filters. on the original
rear of the panel, there polarizing filters after the cold-cathode lamps
to make the lcd panel function, on the front viewing face, there are
are dispersion films and antireflection coating, that allow wider view
and less glare.

i do not have such knowledge, but have wondered if there are any
negative consequences in turning the monitor back to front and
projecting through the original front viewing side instead of the
original rear.

if there is a difference, then this might have an impact on the level
of off axis light that is rejected when it reaches the lcd panel, and
what is allowed to pass through it. how tolerant the panel is of
point source vrs non point source light, in terms of smear or lack
of focus. just food for thought.

Some of the light is blocked, and depending on the panel and the tolerance of the polarizer this can cause a brighter spot in the middle of the projection, which is why we are switching to using the collector in front.

Tony, those extra sheets are diffusers and collimators. They are designed to even the light from the CCFs. There is a polarizer on each side of the panel that makes the panel look black, they should never be removed (some people have actually done it...oof). The anti-glare coating actually helps with our projection, as it allows less light to be reflected from the surface of the panel.

This is new.
I saw several warnings you made in the guide against such move. Do you now feel that the benefits of having the collector in front of the panel, outweigh the distortion added by it ?. What made you change your mind ?.

I found a source for quality lenses that could be used for a field lens. I tested this lens for image degradation and found none. This is the lens we sell. A budget fresnel should never be used as a field lens.