Several threads have refered to adding extra glass sheets to protect the fresnel lenses from damage or dust. A couple of threads have added these after making and using their projectors only to find a degradation of the image brightness. Others see little difference. Yet others claim a vastly improved image after removing extra glass.

I have spent an afternoon researching this topic to learn the truth.

The transmissability of glass or plastic is determined by several factors. The most important are clarity of the basic material, impurities in the material, and reflectivity. Wavelength of the light is also important as different types of glass or plastic will pass different parts of the spectrum unequally.

Clarity and impurities: Essentially glass and plastic are not totally clear. Even our atmosphere is not totally clear. The transmission of light will lessen with the distance it is required to travel through ANY material as the photons fall prey to either the inherent clarity or the impurities in the material and be absorbed. We can see this simply by looking at the EDGE of a glass panel. As the lower energy wavelengths are absorbed by impurities in the glass or the glass itself, the light travelling through the length or width of the glass tends toward the higher energy green and blue. Eventually, even these shorter wavelengths will be absorbed. Therefore the thicker the glass the less light can be transmitted through it.

Standard window glass (whether tempered or not) will absorb approximately .6% per millimeter of the light transmitted through it. Transparent Lexan or Acrylic plastic has about the same absorbtion pattern as window glass. A standard window pane (~3mm, ~1/8") will absorb about 2% of the light passing through it!

Reflectivity: polished surfaces do not let all of the light in...or out! The highly polished surfaces of the glass WILL reflect a percentage of all light that falls on them. In sheet glass, there are always TWO surfaces... one on each side... and each will reflect some of the light that strikes them. Untreated glass surfaces will reflect about 4.5% of the light that strikes perpendicular to the surface... and it gets worse the greater the angle of incidence until 100% is reached approaching an angle of incidence of 90%.

Special optical coatings can be applied to glass (or plastic) to reduce the reflectivity. The most efficient of these will reduce the reflective index to 1.4%... but we generally don't have access that kind of glass at anything like a reasonable price.

Each surface of a standard uncoated window glass (tempered or not) or Lexan pane will reflect approximately 4.5% - 5% of the light striking that surface depending on the angle of incidence. Since each piece of glass has TWO surfaces, approximately 9% - 10% of the light striking the glass WILL be reflected!

Wavelength: Glass is generally better at transmitting high energy light than low energy light... and that is apparent when one looks at the ability of the light to transmit various colors of visible light. The first chart below shows the ability of glass to transmit the solar spectrum. The second chart shows various types of glass and/or coated glass transmission or transparency.

The charts go from Ultra-violet at the left to Infra-red at the right. The higher the graphline (Y axis) the greater the transmission or transparency. When the graphlines intersect the base line at the bottom it means the glass is opaque (non-transmitting) of those wavelengths.



We can see that all glasses are essentially opaque to most of the Ultra-violet light spectrum, especially to the higher energy Ultra-violets that can do the most damage. Unfortunately, it is also obvious that glass is almost as transparent to Infra-red as it is to visible light. Also unfortunately, to control that Infra-red, coatings to filter out Infra-red either impact the visible light spectrum unequally or cause a 50% reduction in the transmission of visible light! This is why cooling our LCD panels is so important!

Looking at the wavelength charts for Lexan and other plastics we find a similar pattern: blocking of Ultra-violet and transparency to visible light spectrum along with Infra-red (that is why plastics and glass have similar usages in greenhouses... they both let in short Infra-red waves and block long Infra-red waves.)

So, exactly how much light is lost at each additional glass or Lexan panel added to our projectors? Brainchild's original design includes four (4) panels: one tempered glass heat shield, the Field Fresnel Lens, the LCD screen, and the Collector Fresnel Lens. Brain estmates that about 80% of the light that falls on the LCD screen is absorbed. This could be because there are actually several layers in the LCD, each with surfaces and transmission issues... as well as two polarizing filters that cut down quite a bit of the light merely be passing only light that is oriented in a specific angle. So let's use that 80% estimate. That gives the LCD screen about a 20% efficiency. I will also assume that the optics that Brain sells are coated to reduce reflection. We will assume a fresh 40,000 lumen light source with a Norpro bowl reflector, delivering 40% of its output to the first glass element.

From the research done above, each NON-COATED panel will reflect ~10% (5% x 2 surfaces) and absorb 2% (3mm thickness x .6%/1mm)... meaning an ~88% efficiency. Each COATED panel will reflect ~3% and absorb 2% (3mm thickness x .6%/1mm) for a 95% efficiency.

Light striking first element (tempered glass) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (tempered glass) = 16,000 L X 88% = ~14,080 Lumen
Light leaving second element (Fresnel Lens) = 14,080 L X 95% = ~13,376 Lumen
Light leaving third element (LCD Screen) = 13,376 X 20% = ~2675 Lumen
Light leaving fourth element (Fresnel Lens) = 2675 X 95% = ~2541 Lumen

The next element is the Triplet lens... which has six coated surfaces (3 lenses X 2 surfaces) and about 25mm of glass which probably gives it an efficiency of about 76% (9% loss from surfaces plus 15 % loss from glass thickness).

Light leaving the fifth element (Triplet Lens) = 2541 X 76% = ~1931 Lumen

This gives us a theoretical Lumen level available of approximately 1931 Lumen. Pretty impressive. (Without the Norpro bowl, it is probably about 40% less than than that)

Now, let's add the extra window glass designed to protect the fresnels from dust or to stiffen them.

Light striking first element (tempered glass) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (tempered glass) = 16,000 L X 88% = ~14,080 Lumen
Light leaving second element (Fresnel Lens) = 14,080 L X 95% = ~13,376 Lumen
Light leaving first added window glass protector/stiffener = 13,376 X 88% = ~11,771 Lumen
Light leaving third element (LCD Screen) = 11,771 X 20% = ~2354 Lumen
Light leaving second added window glass protector/stiffener = 2354 X 88% = ~2072 Lumen
Light leaving fourth element (Fresnel Lens) = 2072 X 95% = ~1968 Lumen
Light leaving the fifth element (Triplet Lens) = 1968 X 76% = ~1496 Lumen

So, with just two additional panes of window glass we have a theoretical output of only approximately 1496 Lumen which is about 77% of the basic model... or about a 22% loss.

Adding two more as some have suggested with the tempered glass separated a distance from the first Fresnel lens:

Light striking first element (tempered glass) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (tempered glass) = 16,000 L X 88% = ~14,080 Lumen
Light leaving the first added window glass protection/stiffener = ~13,024 Lumen
Light leaving second element (Fresnel Lens) = 13,024 L X 95% = ~12,373 Lumen
Light leaving second added window glass protector/stiffener = 12,373 X 88% = ~10,888 Lumen
Light leaving third element (LCD Screen) = 10,888 X 20% = ~2178 Lumen
Light leaving third added window glass protector/stiffener = 2178 X 88% = ~1917 Lumen
Light leaving fourth element (Fresnel Lens) = 1917 X 95% = ~1821 Lumen
Light leaving fourth added window glass protector/stiffener = 1821 X 88% = ~1602 Lumen
Light leaving the fifth element (Triplet Lens) = 1602 X 76% = ~1218 Lumen

The result of stiffening by sandwiching the Fresnelsbetween two pieces of glass is a reduced output of only approximately 1218 Lumen. This is a significant reduction of 37% of what should be possible. Without the Norpro bowl, the loss of light would be striking! The output would be only about 700 Lumen.

Again, Brainchild's original plans and optic design is best (so long as we add the Norpro bowl!).

Recap of results

Original Brainchild design w/o Norpro reflector = 1160 Lumen,
Original Brainchild design w/ Norpro = ~1931 Lumen
Modified design with Norpro and 2 Fresnel stiffeners w/o Norpro = ~890 Lumen,
Modified design with Norpro and 2 Fresnel stiffeners w/ Norpro= ~1496 Lumen
Modified design with Norpro and 4 sandwiched Fresnels w/o Norpro = ~700 Lumen,
Modified design with Norpro and 4 sandwiched Fresnels w/ Norpro = ~1218 Lumen

The only conclusion I can come to is use the Norpro reflector and no extra glass.

Wonderful post.

I am currently having optics made that will give us ~80% of the lamp's lumens. This will allow us to contain the lightsource and cool it directly, mitigating the heating on the LCD. With a high light output such as this we can safely use some IR glass to further eliminate heating. After this we'd want to make the optics faster and clearer after the LCD. There are some very sophisticated techniques available for the field lenses...ridiculously thin lenses, motheye antiglare technology, coatings etc...but getting them at a reasonable cost is an effort. More soon.

Very impressive research. This is a nice argument against too much glass. But shouldn't your multiplier for the LCD be 80% instead of 20% or am I missing something? Now what are we going to do about that Terrestrial radiation? Thanks for all that work as it is very valuable to the forum.

No, the LCD absorbs approximately 80% of the light passing through it. There are at least six reflective surfaces, and at least one glass or plastic of some thickness.

As I understand it, the greatest users of light in the LCD, though, are the polarizing filters. There is one on the back and one on the front... each of which will pass only about than 50% if the light that hits them. One, for example, will only pass light that is vibrating (if light really vibrates) in a horizontal direction, the other will pass only light vibrating in the vertical direction. (actually, I think they are only slightly out of phase with each other, the more out of phase they can be the greater the contrast ratio) The result is a black screen that passes very little light.

The liquid crystal panel between these two polarized filters "twists" the light from the first panel so that some of it can pass through... the more it twists and allows through the brighter the image. Those panels that can do the best twist will have the best contrast ratio... the difference between the darkest and brightest... that the LCDs can produce. Black exists where no twisting is taking place, white is where the three primary colors are twisted the most... gray is a smaller amount of twisting.

The upshot of this is that only about 20% of the light passes through the LCD screen system... if it is all bright white.

Oh, I thought you were going with Brain's estimate of 80% transmission. 20% seems kind of low, perhaps the fact that these surfaces are bonded together actually lessens the number of reflective surfaces or somehow lessens their impact.

Oops, that is an error in the original article of this thread. The actual is the reverse. 80% of the light passing through an LCD screen is absorbed by the various layers. My bad.

The actual sentence should have read:

"Brain estmates that 80% of the light that falls on the LCD screen is absorbed by the screen."

Sorry.

If the 80% effriciency were correct, then we would be looking at an output of 10,176 Lumens... a ridiculously high amount of light that probably exceeds the Lumen output of your average Rock Show big screen stage projector. You'd need sun glasses to watch it at the distances in the average home!

Ok, that makes more sense to me now. Yes I guess 80% transmission is impossible but it certaintly would make our lives easier if it were.

Thanks for catching my error. The article has been edited to reflect the correction.

Swordmaker

Any timeline on these optics?

ASAP is the best I can do for now.

Brain... how about pinning this thread. The subject of extra glass or lexan comes up often in other threads. This could be useful to others and should be easier to find.

Thanks,
Swordmaker

Swordmaker, how about a re-examination of your calculations with lexan instead of glass. I'm still not sold on the idea of lexan for the lamp isolation, but after seeing your transmission figure for lexan I am considering using it to protect the grooved sides of the fresnels. I will deferr to you to post the calculations since this is your baby. Also light diffraction is a concern to me if the fresnels were protected(any opinions on this?)

Deathray, Lexan is more transparent than glass and apparently it has a lower reflectivity coefficient. The Manufacturer's spec sheets claim an approximate 92% transmission ability.

Using that figure, replacing the Tempered Glass with a lexan sheet would give us:

Light striking first element (Lexan) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (Lexan) = 16,000 L X 92% = ~14,720 Lumen
Light leaving second element (Fresnel Lens) = 14,720 L X 95% = ~13,984 Lumen
Light leaving third element (LCD Screen) = 13,984 X 20% = ~2797 Lumen
Light leaving fourth element (Fresnel Lens) = 2797 X 95% = ~2657 Lumen
Light leaving the fifth element (Triplet Lets) = 2657 X 76% = ~2019 Lumen

The original calculation using Tempered Glass yielded ~1931 Lumen... thus just substituting the Lexan for the tempered glass we get an improvement of 4.5% in brightness.

Adding TWO extra elements for stiffening and dust protection would get us:

Light striking first element (Lexan) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (Lexan) = 16,000 L X 92% = ~14,720 Lumen
Light leaving second element (Fresnel Lens) = 14,720 L X 95% = ~13,984 Lumen
Light leaving first added Lexan protector/stiffener = 12,865 X 92% = ~12,865 Lumen
Light leaving third element (LCD Screen) = 11,771 X 20% = ~2573 Lumen
Light leaving second added Lexan protector/stiffener = 2573 X 92% = ~2367 Lumen
Light leaving fourth element (Fresnel Lens) = 2367 X 95% = ~2249 Lumen
Light leaving the fifth element (Triplet Lens) = 2249 X 76% = ~1709 Lumen

The original calculation using Ordinary Window Glass and TWO added glass panels yielded ~1496 Lumen... thus just substituting the Lexan for the tempered glass and TWO glass panels we get an improvement of 14% in brightness over the similar set up using glass!

Adding ALL FOUR extra elements for stiffening and dust protection (The Sandwich approach) would get us:

Light striking first element (Lexan) = 40,000 L X 40% = ~16,000 Lumen
Light leaving first element (LExan) = 16,000 L X 92% = ~14,720 Lumen
Light leaving the first added Lexan protection/stiffener 14,720 X .92% = ~13,542Lumen
Light leaving second element (Fresnel Lens) = 13,542 L X 95% = ~12,865 Lumen
Light leaving second added Lexan protector/stiffener = 12,865 X 92% = ~11,836 Lumen
Light leaving third element (LCD Screen) = 11,836 X 20% = ~2367 Lumen
Light leaving third added Lexan protector/stiffener = 2367 X 92% = ~2178 Lumen
Light leaving fourth element (Fresnel Lens) = 2178 X 95% = ~2069 Lumen
Light leaving fourth added Lexan protector/stiffener = 2069 X 92% = ~1903 Lumen
Light leaving the fifth element (Triplet Lens) = 1903 X 76% = ~1446 Lumen

The original calculation using ONE Tempered Glass and FOUR Ordinary Window Glass yielded ~1228 Lumen... thus just substituting the Lexan for the tempered glass AND FOUR stiffeners we get an improvement of 18% in brightness over the similar set up using glass!

Final:

Source: 40,000 Lumen Lamp with Norpro Reflector.

One Lexan Sheet = 2019 Lumen
Tempered Glass = 1931 Lumen
Three Lexan Sheets = 1709 Lumen
Tempered Glass plus Two Plain Glass Panes = 1496 Lumen
Five Lexan Sheets = 1446 Lumen
Tempered Glass plus Four Plain Glass Panes = 1228 Lumen

As long as we're talking just about the field lens then fine, otherwise I'm not sure those chinese are going to be too happy. OTOH, you can't satisfy everyone I guess...


Brainchild, your new approach of using 80% of the lamp's light is a very important thing, not because anyone really needs so much light (in fact those users who plan to use smaller than 7 feet screens, might compromise their vision if they won't wear protective gear, on themselves or on the pj lens), but because when most of the light will be able to get out of the pj, we will be able to use much less powerful fans. quieter operation.

Greetings,

I did a little research at Lowes. They sell acrylic sheets that are replacements for glass from a company called Plaskolite. The products are Optix acrylic sheets and Duraplex high impact acrylic sheets.

Here's the company website:
http://www.plaskolite.com

If you look in their product guides for these products, they list the total light transmittance at 92%. I'm assuming this has to include reflection because I looked at the Duraplex in the store and it appeared to be very clear.

The price I saw was $7.96 for an 18" x 24" sheet of Optix. I don't remember what the Duraplex was, but I think it was more expensive. The advantage of Duraplex would be that it is stiffer than Optix. It wouldn't be nearly as scratch resistant as glass, so we'd have to be careful.

I think I will look at this for when I get the new Zoom lens and the large sized fresnels.

Thanks for laying the groundwork, Swordmaker!

Thanks,
Zendance

Thanks for the hard work Swordmaker.

Hey Brain,

Just curious if you have any updates on the new optics?

Scott

Wrote to them a little while ago, we'll see.

A Question on another thread prompted me to add to this thread.

The question basically was "How many light will our projectors put on the screen?"

The answer is as follows:

My research on this thread provides a comparison figure for use when comparing LumenLab projectors to the reported output of commercial projectors. It does NOT provide the answer to the question regulator was asking:



The amount of light available EXITING the LumenLab Triplex is the ~900 to ~2000 depending on the number of glass or plastic panels used. The approximate lumens at the screen is variable based on the size of the screen.

First we have to understand what a Lumen is. By definition a Lumen is a measure of illumination.

Illuminance is a measure of photometric flux per unit area, or visible flux density.  Illuminance is typically expressed in lux (lumens per square meter) or foot-candles (lumens per square foot).



For our purposes we can use either English or Metric. Since most of the measurements on here are English, we will use that.

A Lumen is the amount of light that will illuminate a one square foot surface one foot from the source with one candlepower (Candela).

At the theoretical brightest of our LumenLab projectors, the entire area of the Triplex lens is radiating at ~2000 Lumen. That means a 1 square foot image at 1 foot from the Triplex would be ~2000 Lumen. At a wall 10 feet away from the projector, we would STILL have the same amount of light... but it would be spread over a LOT more acreage.

Therefore, a 10 foot Diagonal screen (4:3 Aspect Ratio) has 48 Square Feet (8 X 6 = 48). Using the definition we find that each square foot of the screen will have approximately 41 Lumen of illumination. If you are using a lot more glass or plastic, the illumination of your screen may be as low as ~19 Lumen per square foot at the screen.

Now neither of those Lumen figures sounds too impressive but consider what brightness you would have if you put your birthday cake (if you are 48 years old) on its side, one foot away from EACH square foot of your screen... pretty bright.

Practically, we will probably not get the theoretical maximum of ~2000... but I would not be surprised to get ~1500-1600 out of the basic design with the NorPro reflector which would give us an at screen illumination level of approximately ~33-37 Lumen .

Dude. Very cool info. Thanks!

Wait a minute, I read somewhere in one of these posts that Brain esimated around 3,000+ lumens out of the finished project. Is this not correct?

I found a better bulb with a reduced jacket that produces 32,000 mean lumens (after warm-up) and over 40,000 initial lumens. I also spoke with a local plastic company about their lexan. They report a 97% pass through on a 1/8" sheet, and a 205 degree temperature rating. Would this not be temperature resistant enough to replace the tempered glass? It only costs $98.00 plus tax for an 4' x 8' sheet.

I believe they make some a little thinner if all you guys want is a groove cover.
So far this is looking good to me, and if it looks good to you, let me know. If I can get rid of the surplus, I will buy a sheet, cut it to size, and sell the extra to defer the cost.
Now, does anyone know where we can get our hands on some 'paint on' optical coating? lol

As Swordmaker said, The amount of light available EXITING the LumenLab Triplex is the ~900 to ~2000 depending on the number of glass or plastic panels used.
The lexan works fine in place of the tempered glass and it has a UV filter.
Also, you can buy it, $4 (12"X10") at home depot in the glass session.
Just place it, atleast 2", from the bulb.

This may be a stupid question, but how do you figure it has a uv filter, or are you speaking of the particular piece from home depot?

lexan has a uv filter built in, there is a certain side that you will want to face towards the bulb, it is the side of the lexan that says face towards sun. Goto home depot and check it out, I promise it will make sense, and its only 4 bucks.

Thanks. That should eliminate the need for a seperate uv filter. I will see if I can find some rating specs to compare it with Brain's uv filter, unless it has already been done. Do you know if anyone has done this yet? I need all the shortcuts I can get.

I was looking at the previous illustration on irradiance. Given a bulb starting with 40,000 lumens, how did you arrive at the initial output at 40%? Is that based roughly on the surface area of the bulb projecting light toward the fresnel including what's reflected from the Norpro? Just curious if it was measured or there is a way to measure it, as a 10% swing on that output, and including or excluding a reflector, significantly changes everything downstream.
Appreciate all the hard work Swordmaker!

Almost exactly. I calculated the area the angles of the corners of the Collimater Lens and LCD screen made when extended back to the point source (not surface area) of the lamp's arc. This showed that only about 19+% of the light produced by the arc would strike the fresnel at the appropriate angles. With the NorPro bowl, that would double (ignoring losses in the reflection etc. for simplcity's sake).

In actual fact, I believe that the probably only about 33% of the lamp's output lumen will eventually strike the first fresnel lens. The rest is radiated to non-lens areas of the light box, and some is absorbed by the NorPro reflector's non-100% reflectivity.

You are right about small increases in the efficiency of the light box significantly affect the output downstream. I used my best estimates.

The whole purpose of this exercise was to show the losses extra glass or plastic added cause to the output. I used "close" estimates for the initial light. For example, I used 40% of the lamp's 40,000 Lumen instead of the rated 38,000 initial lumen and even lower average lumen ratings of the lamps we actually use. I also accepted Brain's unmeasured 80% estimate of light loss after going through the LCD. So long as these are constant for all set-ups, then the results are comparable. The losses through glass and plastic were taken from industry specifications.

Anybody got a light meter?

It would be really cool for someone to test this on the basic design using a light meter on a screen. If we could get a reliable end point number, we might be able to extrapolate backwards. That would allow us to accurately calculate the effects of a design change.

moose made a semi meter if we could use it and compare top a commercial projector to compare.

What if the non-lens areas of the light box were to be eliminated by funnelling the light from the Norpro to the first fresnel lens with the aid of a reflective material. Would this actually prevent most of the 67% loss of the lamp's output lumen ? Just a thought.

Hey Joe and Brain,

In the reflection list you talked about using mirrors to make a tunnel from the reflector/bulb to the front fresnel. I didnt see the conclusion of this test. Joe I think you said it doubled the output in your tests, but I didnt see anything from brain.

Did you guys give up on the light tunnel? If so why?

Thanks,

Scott

it generated to much heat to the lcd, and i really did not go much further as far as an elaberate cooling system.

Yet another Catch 22 ! Seems the heat is a limiting factor in any DIY projector and the noise of fans would be another Catch 22.

Hi to all,

First, hats off to Sword... this is cool woork.

Newbie questions:

1) Will a 17inch monitor (600:1) increase / diminish the lumens hitting the screen ?

2) Is it true that tampered glass has UV filter ? (thus reducing the extra layer)

3) Why do we need UV filtering ?

4) Why don't we need IR filtering (I know we can't "see" it" but... nananananaa I'm still asking)

If you are trying to build a "light tunnel", you should leave a space between the bowl and the tunnel, say 1/2", and the tunnel should have small holes in it. Alot of them. I am going to try this idea out and see what happens.

The light tunnel is what I'm trying to do - my cheapo diagram is at the bottom of the link

http://www.lumenlab.com/forums/index.php?showtopic=2279

Anybody got specs on the light transmission for the new Pro lens? We have the light transmission for the current triplet at 76%. Is the new lens better than this?

The Fresnels will only columnate the light that hits the lens at the approriate angle. Any excess will not be too useful beyond giving a general illumination in the wrong direction.

The best way to harvest the wasted light would be to redesign the light box using an efficient parabolic reflector instead of the fresnel field lens. This would require a complex rectolinear parabolic reflector as large as the LCD screen with some complex corners to prevent hotspots and dim spots.

Looking at the angles that would be produced by the reflecting light rays from a simple mirrored tunnel, I doubt that much light would be added that would be usable. Any that would is entirely accidental. The light needs to hit the Fresnel field lens at an angle that would seem to have come from the focal point... the center of the arc. If it doesn't then it won't pass through the LCD perpendicular to the screen.

1: No, it will actually INCREASE the total lumens hitting the screen because the image is larger, but the Lumens per Square Foot would probably be about the same. In other words, you would have a larger, but no brighter image. As the size of the LCD grows, the problem with inherent light dimming at the peripheries will also grow. This is because the corners of the screen are farther away from the point light source than at the center and the result of the distance squared law of optics. Soon that dimming will exceed the ability of the human eye and brain to ignore it.

2: No, tempered glass transmits the same amount and frequencies of light that non-tempered glass of the same composition. Changing the surface tension of the glass pane will not affect the reflectivity or transmissibility of the glass.

That being said, almost all glass and plastics are opaque to the shorter ultraviolet wavelengths that do the most damage.

3: It is my opinion that we don't (see the answer to 2). The high brightness florescents that make up the back light of a normal LCD monitor already produce ultraviolet to cause the florescing. Some of that escapes yet it is not a problem for the LCDs as they are designed. Brain is offering it as a safeguard in case I am wrong and the arc produces sufficient UV to damage the LCD.

4: IR filtering would be GREAT... it would keep the temperature of the LCD down and easier to control. The unfortunate thing is that every IR filter ALSO filters some of the longer wavelengths of visible light and reduces the reds by 50% or more while leaving the shorter, bluer wavelengths essentially at 100% which causes a distinct color distortion problem.

Find the total thickness of the glass in the lens and multiply it by .6% per millimeter. Then count the number of coated lens surfaces (entry and exit) and multiply that number by 1.4%. Add all of this together and subtract from 100% and you should have a good estimate of the transmittibility of the lens.

It is probably about the same as the original triplet.

there are plenty of high quality hot and cold mirrors which are very nearly true color, but they are expensive

www.edmundoptics.com

i have thought about trying to use a cold mirror at the light source (cold mirrors reflect IR pass visible), it is about 200$ for a 4x6". If i was building a 7" projector, i would probably spring for one but the filter is too small for the 15" LCD i would guess.

hot mirrors (reflect visible pass IR) are cheaper but the focal length of the field lens is too short to make their use practical or at least easy.

has anyone tried using an IR filter?

Hey Brain

Theres a company in Cinncinnati Oh ,Named UVT {Ultra Violet Technolgies}
10 Years ago when we started Printing thin plastic film 48 Ga
we had problems with heat warping and distorting the film
they make High Intensity UV lamp Systems for the printing industry
they placed a tempered glass with an IR coating to stop heat
the Surface of the film went from 125 F to 85 F
The called the system Cold Cure Technology
I dont know how it would affect optics performance color saturation,ECT.
but it might be a source worth investagating

Aiel Maiden

About reducing heat... what about this solution ? (pic found on http://www.hommie.net/ , I hope he won't be angry for demonstrating it here )

If you look at his other pictures you'll see him holding a piece of flaming cardboard in front of his condenser lens; I wouldn't exactly call that 'reducing heat'. The high bay fixture and 1000w lamp are certainly not appropriate for this project. The design is not based on sound optical principles and his enclosure is 4' long!

I would like to know the groups opinions of an 8mm diameter point light source at the focal point ?

In an ideal world, what would the optimum point source size be ?

Thanks
John

With a perfect fresnel lens, the point light source would be practically non existent and very bright. The fresnels we use are designed for the real world and have a built in 'slop factor' which allows for a larger than ideal point source. A sub 10mm point source would work very well in our systems.

Brain,
Is there a negative to reflecting greater than 80% of the light back to the bulb arc?
John

Shorter lamp life. How much shorter, no one knows.

I figure the light on the first fresnel is only about 4400 lumen of the 40,000 available. This is based on how manufacturers rate the bulbs and the amount of bulb thats actually being used. There is also as much as a 70% drop over mean in the extreme corners. If you then take out the lens/glass loss and the 80-85% loss across the LCD then I would estimate about 700 lumen leaving the projector .
Thats still pretty good. If you could boost this by 20% or 30% from a norpro or similar then thats bonus.
Has anybody bought the $70 light meter to get some "actual" output numbers? I'm keen to get one just to hack around with.

Cheers,
Kirk