Audio Project Amplifier Speaker Loudspeaker Kit
diyAudio.com diyAudio Forums Archive > Top > The Moving Image > Optics
 
DIY-Projector Plans and theory collection !! - Click HERE for Original Thread
mathias
First I want to say this is a public diy forum, we share what we know and what we have learned, if you donīt want to share what you know, please leave this forum. Many people have leaved this part of the forum, and I will also do that, because here are very little new ideas, what people are doing today is just repeating of what other people have done many years ago. The only difference is that we have more high resolution lcd:s today, but in the same time commercialprojectors have been very cheap with the same resolution and better contrastrate.

I have here collected the most common projectorplans, some are cheap and some are more expensive to build. The difference in light output between the designs is about 40%. There is not such thing like 2-3 times brightness, if you get 2-3 times more lightoutput you have used wrong condensorlenses from the beginning.

Please share your projectorplans here and help me complete this
collection!!!

This projector design is the most common and use the parts from an ohp-projector, works with 250-575W double ended lamps. 65% of the light from the lamp will hit the fresnel.
mathias
This projector design is very cheap but is only for 1-ended lamps, the picture can be little dark in the corners. 80-90% of the light from the lamp will hit the fresnel.
mathias
This projector design is the best, very even light on the surface of the lcd, use one half of the slideprojector condensor and the other from a ohp. 99% of the light from the lamp will hit the fresnel.
mathias
This projector design use the parts from an ohp-projector, works with 250-575W double ended lamps. 65% of the light from the lamp will hit the fresnel.
mathias
This projector design is very cheap but is only for 1-ended lamps, the picture can be little uneven in brightness. 80-90% of the light from the lamp will hit the fresnel.
mathias
This projector design is the best, very even light on the surface of the lcd, use one half of the slideprojector condensor and the other from a ohp, you need two condensors from a ohp. 99% of the light from the lamp will hit the fresnel.
mathias
All platicmaterial reflecting a part of the light back to the source but in a other direction, this can easy been studied with a laserbeam. So use glass instead if you can, they reflecting a smaller part of the light.
The Alchemist
Great idee!!!! nice overview. Maybe we have to find examples of the specific projectors and also post them as examples.

In your pictures you say that the focal length of the objective is the same as the distance to the lcd, but i believe that it has to be just a little bit longer (240 mm objective ---> ~260 mm distance from tft to objective.)

I still wanna try the design of two half parabolic mirrors and a spherical one.



Maybe i will buy a raylight and test it.
mathias
Here is the most advanced projectordesign plans and also most compact, but most expensive. As you can see you must have pretty big mirrors if you want it most compact. You can use smaller mirrors if you move them closer to the lamp / object, but the distance will be bigger = bigger projector. You can use normal glass mirror on the lamp side, it can also be coldmirror to reduce the heat, but you can also have coldmirror build in to the lamp reflector (cheaper and more easy). The mirror on the objectlens side must be a frontsurface mirror otherwise you get ghosting, you can find these in ohp-projectors.
mathias
Mirrors are good, but what size do I need ?

In small lcd designs 5-7" you can not use so much smaller mirrors than the lcd size depending on that the light from the condensors lenses are almost parallel. But in 15-17" you can use much smaller mirrors than the lcd.
mathias
We need a frontsurface mirror on the object side, but what size do I need, they are very expensive, so do the one in my old ohp-projector fit ?

I have done a calulation example so just follow it and you will get the right size.
ace3000_1
quote:
In small lcd designs 5-7" you can not use so much smaller mirrors than the lcd size depending on that the light from the condensors lenses are almost parallel.

Not true, it depends how far you place the mirror from the lcd, just like how the ohp's mirror is so far away wich in turn alows it to be small.
ace3000_1
mathias, i hate to disapoint you but your drawings are wrong with the condensers lol, sorry to have to break the bad news but dual condensers work as the pick below.

Imo i like your mirror drawings, nice work.

Trev
Hezz
Actually,

I think you are both right. The real problem with lens schematics is that they are so oversimplified as to be almost misleading. There are usually many millions of light paths through the lens. And there will rarely if ever be one simple scheme or focal point. The best you can hope for is to get most of the light to a common focal point. The fact is that positive plano lenses or similar lenses collapse the light beam no matter which way they are oriented. So where the light focuses depends on the angle of the light. However there is a light reflection issue which sometimes effects the orientation of the lens because the closer the light ray is to perpendicular to the surface it enters (or perpendicular to the tangent of a curved surface) the less light will be reflected back. So that is why there are often preferred orientations. The higher the angular incidence that the light ray hits the surface more light is reflected up to a maximum amount of reflection that is specific for each substance. This angle is usually around forty five (but not exactly) degrees for many types of glass.

For this reason you will almost always see the first positive surfaces of lenses rather large curved so that the light rays coming from larger angles hit the lens surface at a more desirable angle thus reflecting less light back.


Hezz
Aidan Isaac Lev
@hezz. Very informative post :) Now i think i undesrsand y some projectors have dark corners.
mathias
quote:
mathias, i hate to disapoint you but your drawings are wrong with the condensers lol

Like Hezz say we are both right, and my drawings are NOT wrong, if you have parallel light coming from the lamp they work in the way I have drawn. You have a POINT source and then they work in the way you show .:)
ace3000_1
quote:
However there is a light reflection issue which sometimes effects the orientation of the lens because the closer the light ray is to perpendicular to the surface it enters (or perpendicular to the tangent of a curved surface) the less light will be reflected back. So that is why there are often preferred orientations. The higher the angular incidence that the light ray hits the surface more light is reflected up to a maximum amount of reflection that is specific for each substance. This angle is usually around forty five (but not exactly) degrees for many types of glass.

Hezz and thats exactly right, and its why ive got my condensers coated in anti glare, the anti glare have a much sharper light then with the same without the anti glare, the one with the anti glare is somewhat brighter too as more light is going where we want it.

As for the angles thats where reflectors come in, its not just a matter of getting a spherical reflector, the focal is very important to obtain maximum light from the bulbs light transmision pattern, to get the desired angle through the condenser, and to the frensel at your desired focals for maximum efientcy at the coresponding angles required, even the depth of a spherical reflector matters to get the ultimate desired results.

Trev

Trev
ace3000_1
quote:
if you have parallel light coming from the lamp they work in the way I have drawn. You have a POINT source and then they work in the way you show

Im not talking about the paralelle light ones, i am talking about the point source pics otherwise i wouldnt have posted the way 2 condensers work with a point source.

Also hezz is talking about miss matched light, there is no way u will get rays out of a positive power lens in the way you show it, paralelle or non paralelle light, the light will go to the focal point of the lens.

Trev
mathias
Trev, all drawings use the technics from professional design and I donīt understand why you say they donīt work.
Many of the designs is taken from physics books, and I promise you that they donīt have wrong! The ellipticalreflector design is used in all modern slideprojectors with a lamp of the type ELC A1/259 with built in reflector. And the other designs use the normal ohp-design.

And if you say these design are wrong I suggest that you draw a right one.
ace3000_1
Mathias this stuff has been posted in here a few times and argued about before, i dont have time right now to draw up a pic but i will do abit later on as i have some serious work to do and not enough time. Have a read in here. there are manny formulas and they tell u how basic lenses work, although imo not for the beginner.

http://www.edmundoptics.com/TechSup...m?articleid=267

You can always ring them, ive rung them up before, they are more then willing to help and are all optical engineers.

Trev
ace3000_1
quote:
The ellipticalreflector design is used in all modern slideprojectors with a lamp of the type ELC A1/259 with built in reflector. And the other designs use the normal ohp-design.

There is not 1 eliptical reflector in your design, eliptical reflectors have 2 focal points, i only see a wide angle spot reflector, a spherical reflector and a parabolic reflector in your drawings.


Eliptical reflector:
ace3000_1
Here is a little somthing for hezz to help him along his way with his lens design.

http://wwwhip.obspm.fr/gaia/rvs/wor...emla_optics.pdf
mathias
quote:
There is not 1 eliptical reflector in your design, eliptical reflectors have 2 focal points, i only see a wide angle spot reflector, a spherical reflector and a parabolic reflector in your drawings.
As you can see in the picture the reflector I use and show in my slideprojector designs is a ellipsoidal reflector and the light beam out of it is parallel and this type of reflector is very common in searchlight. I really donīt understand why you write things that is not true.
mathias
quote:
There is no way u will get rays out of a positive power lens in the way you show it, paralelle or non paralelle light, the light will go to the focal point of the lens.
You have also wrong in this statement, positive lenses work exactly as I have drawn.
ace3000_1
Mathias for your post above they are double convex lens, they are not plano convex in what im talking about and they are used as condenser's, if u want to argue on the truth, then dont argue with me, my information is from optical engineers who have phd's in all areas of optics, and this information come from the net acociated to their knewlege, for the record eliptical reflectors dont put out paralelle light and all have 2 focal points as i have stated above with the atached image, that image isnt draw by me btw. So now its upto you to find the truth and do the study for yourself, ive done the study ive contacted who i have had to contact and i know the truth, no guesses but answers from optical engineers who have studdied over this complete matter.

Trev
ace3000_1
Mathias, read this page, look at light, its an eliptical reflector and with the coresponding 2 focals, in the diagram that shows the 3lcd's its clear how it works, it is also clear that your angles are incorect from the condenser to the frensel, so maybe an apology may be in order.

http://www.extremetech.com/print_ar...=15893%2C00.asp

Trev
mathias
I have drawn normal positive lensen because itīs more easy to draw. There are not any diffrence between double convex lenses and plano convex more then plano convex have better optical quality.

Eliptical reflectors actully give parallel light, if you move the lamp backward in the reflector, this is how all searchlight works, if you donīt beleave me, try it yourself !

If you got your information from a optical engineer I think you have misunderstand him a lot, I have myself read many books from the University about optics and I am a enginner myself, and I think itīs unfortunately that you say people have wrong when you donīt really know that.
ace3000_1
Mathias, i realy dont know what your problem is, i see you cant admit to being wrong, my proof is in the pages that ive posted and is very clear, so realy if you take a look at the odds, then your the one who realy needs to read and to realise where you have gone wrong. Ive tried to help you mathias to show you where you are wrong but you cant acept the truth, it is now not my problem and i will not discuss this matter further with you.

Trev
mathias
I donīt understand why you have so hard to understand. That page you linked to prove that you get parallel light out of a eliptical reflectors. When they move the lamp backward in the second picture they get the focus on a longer distance from the reflector. What do you think happen when they move it more backward ? Maybe parallel light ?

And I havenīt any problem more than I want people to know the truth !

But thanks for the link, it helps me with my pro-projector.
ace3000_1
mathis, they arent after a paralelle light read the link again, if they wanted paralelle light they would use a parabolic reflector.
Hezz
Guys,
Correct me if I am wrong,

But my understanding is that parabolic reflectors give off the best parallel light if the light source is small and located near the bottom of the reflector. But you would have to contain the front light cone either by filament design or by blocking it otherwise you have two main light paths. This is the problem with most illustrated examples is that they are only emphisizing one aspect of the light projection.

Elliptical reflectors can be misleading because there are many shapes of elliptical. If you place a light source at one of the focal points of an elliptical reflector the reflected light will be directed at the second focal point. However if you place a small light source which is small in relation to the reflector near the bottom of an elliptical reflector it will be somewhat more parallel but not as much as a parabolic. A parabolic curve is defined by a specific mathmatic equation that shows common trends in statistical sampling. The closer the shape of an elliptical reflector is to a parabolic the more like the parabolic it will act if you set the light source up the same way. However it will not create parallel light as well as a perfectly made parabolic.

Now, even with a parabolic you cannot get perfectly parallel light because in the real world you cannot get a perfect point source of light with the front light cone removed. One has to wonder though if you couldn't use a very small lens close to the light source to collumate the front light cone. And having it just the right diameter as to catch all of the light that missed the outer edges of the reflector. Assuming of course you want parallel light. This does mean that the condenser FL and diameter needs to be designed differently for complimenting the parallel light coming from the reflector.

If you are working with the front light cone and treating the incoming light as a point source then dealing with the back and side reflected light becomes the issue. This is where the circular reflector comes in but one has to wonder if it is perfectly aligned if the light will be dispursed by hitting the filament again. This design seems to be the easiest to execute but one need a spherical light element for it to work at it's best.

Since most light element are a piece of wire or a cylindrical piece of ceramic it would seem that in most cases the element is long and not a point. It almost seems like it would be better to lay the element horizontally in relation to the LCD and have a long filament as wide as the LCD and have a half cylindrical reflector and a cylindrical condenser lens because the light comming off from the ends of the light element is less and is further away from the point source. Maybe it should be disgarded.

Hezz
ace3000_1
Hezz i agree with you 100% and what you say is corect, all reflectors are made for a specific purpose, elipticals are made to make a second focal point to guide the most amount of light through the optics, also they cover more of the frontal light being radiated from the bulb to this point and i may add its also why its the most eficient reflector made to date. On your last statement with the arc being horizontal there are eliptical reflectors that alow the bulb to be placed at such angles for the bigger arced bulbs, its true what you say though no way can u get the perfect focal points with a arc thats big, im talking even 4mm is big, 1mm is ideal.

Trev
mathias
As Hezz write there are many shapes of elliptical reflectors. But I think you have missed the most important why to use a elliptical reflector. I hope my drawing clear this up. The biggest reason why they use elliptical reflectors in most searchlight, is that you can use more light from the front light cone, that will get lost in a parabolic reflector. The elliptical reflector is the most effective reflector you can use.
mathias
This is a picture from a professional light disco equipment, The Alchemist posted this before, this is a typical elliptical reflector. If you donīt beleave me that this type of reflectors is very common, go to the nearest hobby shop and look on some different searchlight-lamps, most of them will have a elliptical reflector.
ace3000_1
mathias, i sujest you do abit more reading on this matter. The link that i posted earlier even states the eliptical reflector has 2 focal points and its the way they are used. The second focal guides the light through the optics in an non paralelle manner, also remember light going through a lens as hezz stated gets reflected back to its source, the ultimate in most cases is 45deg of angle at the lens surface. All of your answers are in here, it is now upto you to see where you are wrong.

Trev
ace3000_1
As for the picture above, how do you know its a eliptical reflector? it doesnt say it is does it? it could be a parabolic with a light guide wich in turn in my eyes is what it looks like to me, or an eliptical with a light guide. Its strange how the radious from the back reflector has a bend to join to the front top half, so i would asume the top half is a light guide. No reflector will have any bend marks in it like the one above if it was all a eliptical reflector, so it has to be half a light guide. If this is a eliptical reflector it would have a long focal point, if its a parabolic well the obvious.

Trev
The Alchemist
I must agree with ace. An eliptical reflector has two focal points, that the nice thing of those refelectors because you don't need a condensor lens.

A parabolic reflector is not an eliptical one.

Look at this article, very nice pic's.


Ps. Ace did you looked at the topic on fresnell condesor distance. Do you understant the equations in my post?
mathias
I know that eliptical reflector has two focal points, but if you place the lightsource BEHIND the first focus, the light beams will NOT go to the second focus point, they will be more like a parallel beam, maybe you get a second focus point 200 meters away.

And the above picture is eliptical, isīt not bend in anyway itīs the reflection that make it look like that.

I havenīt seen any picture yet that say I have wrong, all picture show the lightsource in the focuspoint, but that is not what i use.

Why donīt anyone place the lightsource behind the focuspoint and then tellīs me what happens ?

Have anyone actully seen what type of reflector they use in a normal searchlight ?
ace3000_1
quote:
Have anyone actully seen what type of reflector they use in a normal searchlight ?

I have one right here, it also uses a lens to concentrate the beam.
quote:
I know that eliptical reflector has two focal points, but if you place the lightsource BEHIND the first focus, the light beams will NOT go to the second focus point, they will be more like a parallel beam, maybe you get a second focus point 200 meters away.

Yeah thats right and its not what its designed for or how its used.

quote:
And the above picture is eliptical, isīt not bend in anyway itīs the reflection that make it look like that.

Take another look, it has a form mark to change its shape, that is the light guide taking over from the reflector.
Teleteq
Hi Guy's,

I've had a long time away from the forum, and this thread caught my eye.

If i might be so bold to say I agree with ACE, in that the best reflector to get parellel rays would be parobolic is shape.

This is the one of the most important factors in another hobby
of mine.... Astronomy.

The newtonian telescope uses 2 mirrors to reflect the from distant
start to a single point. this point is then magnifed with the use of an eyepiece to gain a image of the observed object.

The light from these start in our own galaxy may have taveled up to 100,000 light years to reach us. This point of light can then be deemed to be at optical infinaty. thus the light rays are parellel

So in our DIY projectors we are trying to get a parellel light from a point of light. It's a telescope in reverse, on a slightly smaller scale !!:D
ace3000_1
Heya teletag, i had a scope for a while and had a look around for a good while in the skies of Australia, i had a tasco galaxsee 675. Not a bad scope, but nothing compared to a newtonian, ive always wanted to build one of these and may do in the future, but time is always the issue with me latley lol

You are dead right about how we can relate a scope's optics to a projectors light engine, its what you say, its a scope in reverse, and its prety much how i started out in learning about light engines in the beginning. Just smaller focals is what we work with and somtimes a few different reflectors.

Trev
Teleteq
Lol

Well I'm working on projector Mk2.
I have been sucsessful in projecting around 2m wide 16:9

But my light source is a 150watt OHP :bawling:
I'm hoping to get a 250+ MH system up and running
but suppliers in the UK are hard to find for the LCD etc.

Funny you should mention about making a telescope.
Thats exactly what I'm doing right now.

I'm hoping to get a robotic scope up and running in around 2 years. Take a look at my site at http://www.teleteq.mine.nu for
my project details (Still under construction :xeye: )

Sorry of the thread hijack :eek:
ace3000_1
Nice:)

I started to make a robotic tracking satelite dish as i got into downloading live images from space from various satelites before i left Aust, its not very common but its around on the net. Ive got satellite tracking programs and a skywatch program with tracking software for a elqualatoral robotic tracking system if your interested. You just need to make a plc card for the comp and get a hold of 2 ham radio antenna rotators for the motorised mount. use this with the software and you can track anything in space lol.

As for the thread jacking yeah we better call it quits, email might be the go, all the best in your projects anyway.

Trev
mathias
I agree that the parobolic reflectors are best for parallel light raws and that is nothing new, BUT there are alot light we canīt use with the parabolic reflector, there for is the elepticalreflector a better choise because we can collect more light with it and yes the rays may not be completly parallel but that is not a problem, the condensorlenses correct this.

And the above picture is eliptical, isīt not bend in anyway itīs the reflection that make it look like that, this is still true if you look at the picture i took of my reflector you can see that you get strange reflections, the blacklines you see are reflections from the metalbox take another look !
mathias
quote:
I have one right here, it also uses a lens to concentrate the beam.

How can you have a lens that concentrate the beam, when u say that the electpicalreflector do the same ? Strange searchlight ?
Mr.Kh
The image in the post quoted below is incorrect optics btw. If you look at the two pairs of rays furthest to the right, their incident angles of reflection do NOT match their reflected angles. THe incident angles look about 20+deg where their reflected angles are more like <5deg.

quote:
Originally posted by mathias
As Hezz write there are many shapes of elliptical reflectors. But I think you have missed the most important why to use a elliptical reflector. I hope my drawing clear this up. The biggest reason why they use elliptical reflectors in most searchlight, is that you can use more light from the front light cone, that will get lost in a parabolic reflector. The elliptical reflector is the most effective reflector you can use.


Err well this post jumped a page, so look one back to see the image I'm talking about. Here's a link to the post: http://www.diyaudio.com/forums/show...8347#post338347
ace3000_1
mathias, ive been working with steel for the most part of my life, i am also a certified engineer in metal, i also have very good eyes, and im telling you its a bend. Buy one of these and prove us that im wrong because i can garentee you that its a bend. That reflector is a cheap press formed steal reflector and ends at the bend, the rest is the light guide to not waste light, nor to contaminate surounding areas, and to stop light comming out of other areas. This is just press formed into a one peice unit.

Trev
ace3000_1
quote:
How can you have a lens that concentrate the beam, when u say that the electpicalreflector do the same ? Strange searchlight ?

LOL the eliptical reflector concentrates all light gatherd to a secondary focal point, normal torch lights use parabolic, u can even test this with your finger. Dont go on shape mathias as parabolic reflectors come in manny shapes and sizes depending on their beam spread and are normally not drawn as a half circle either like you have ilistrated in your drawings, they have a slight PART of an elipse, but once again this depends on the aplication and the desiered amount of light to be reflected, also not all parabolic reflecters are used to reflect perfectly parabolic rays, some alpications like a torch have the beam spread slightly out. Buy a magna light if you want to see what a lens does to a torch and ajust the beam angle.

Trev
mathias
quote:
normal torch lights use parabolic
No they donīt, a parabolic reflector is not so deep.
quote:

Dont go on shape mathias as parabolic reflectors come in manny shapes and sizes depending
A parabolic reflector can not have any other scape than parabolic, if itīs deeper than a parabolic, itīs a elipticalreflector. If they use a mix of parabolic and eliptical in a torchlight I donīt know, but itīs alot deeper than a true parabolic. And this deep shape is what I am talking about. This deep parabolic collect more light than a true parabolic-reflector.

Itīs pointless to argue about the picture, I say itīs not bend, and you say itīs bend, let them how watch the picture self decide.

Anyway I have proved with my own picture from a torchlight that the reflector has a shape more like a ellipse.
mathias
quote:
The image in the post quoted below is incorrect optics btw. If you look at the two pairs of rays furthest to the right, their incident angles of reflection do NOT match their reflected angles. THe incident angles look about 20+deg where their reflected angles are more like <5deg.
If that is my drawing your talking about, I know itīs little wrong, itīs not easy to draw this kind of things, the light source should be more backward in the reflector
ace3000_1
mathias put a peice of paper infront of the torches reflector, without its lens on and see what pattern u get on the paper with the paper standing on its end. Its parabolic.

As ive said you can also test this with your finger.

In the picture below these are all parabolic reflectors, the one on the left and center is a 25deg spot light, the one on the far right is a torch reflector, they are all the same shape and are parabolic.

And mathias im not arguing im stateing facts.
ace3000_1
And here we have big brother and little sister, the big brother is a halogen par spot 30deg spot light, ring a bell mathias? these are the same shape on all reflectors.

Trev
Hezz
Guys,

The fact of the matter is that there are differing shapes of both parabolics and elliptical reflectors. Both kinds can be made shallow or deep. The difference is that they follow a different mathmatical formula for how the slope of the curve changes. It is possible to make an elliptical reflector that is so close in shape to a similar parabolic that it would be hard to tell at a casual glance which is which. You would have to have them superimposed on each other to tell the difference.


Hezz
ace3000_1
Hezz ive just tested my torch out, its parabolic for sure, all of the other parabolics that i have and are known parabolics share the same shape, and thats why i have taken the pictures and posted them up, your right about the depth factor and this is normally used in area size and focal lengthes given to a certian the aplication can fit into, it also has alot to do with eficientcy.

Trev
mathias
Trev, you really got me to laugh:) All those reflectors you show are exectly what I am talking about, they have a scape close to a ellipse, but they are called parabolic anyway. If you try to draw a circle over them you will notice what itīs impossible.
ace3000_1
mathias i dont understad you somtimes, these are parabolic reflectors, not eliptical and no way can they work in the same way, sure they will come close, but its ineficient and they arent designed to do so, if they did well then they wouldnt have made a parabolic reflector and a eliptical would they? Its imposible to turn a parabolic reflector into an eliptical. However you can turn a spherical to radiate parabolic light rays, but its not eficient and thats why we have a parabolic reflector with its acociated shape designed to work in an eficient manner.


Trev
ace3000_1
quote:
If you try to draw a circle over them you will notice what itīs impossible.

I know this but look at your drawings, they are half round when they shouldnt be.
mathias
Trev you must agree that the ellipse is closer in my torchreflector picture than the circle even if itīhalf round, thatīs why I called this kind of reflectors elliptical, but as Hezz say itīs very hard to see the diffrence. Anyway itīs good that we are talking about the same type of reflectors, I would maybe called it "deep parabolic reflectors".
ace3000_1
Yeah somthing like that lol, hey i just tried a parabolic with a condenser, no go , you will end up with a dark patch in the middle, i turned the condenser around the oposite direction also and basically you will loose light so a pcx on a parabolic is a no no and its why they arent comonly used at all on parabolics, this is where an eliptical has another advantage cos with an eliptical we can use a condenser of this nature because it has a focal point. In magna lights i think they use a double convex lens so its probally more the go on a parabolic system.

Trev
mathias
I just want to show the difference between a real parabolic reflector and a elictical/parabolic reflector. The reflector to the right is a true and orignal parabolic reflector, to the left the deep parabolic or eliptical reflector or what you want to call it. In my design where I say eliptical I mean the deep parabolicreflector. This is the type of reflector you must use in the slideprojector design.

If you want to build the slideprojector design, which is the absolute best design use condensors from a slideprojector and turn them in the way they are in the slide-pj. You maybe need bigger slideprojectorcondensors but they must have the same scape as the orginal slideprojector condensors.

Also with slideprojector I mean whose how use a ELC A1/259 multimirror bulb 150-250W.
mathias
This is ofcause more right:

The reflector to the left is a true and orignal parabolic reflector, to the right the deep parabolic or eliptical reflector or what you want to call it
mathias
This is the most important thing if you want to have a movie experience with your diy-projector. The difficult thing is that you must not stop the air circulation. Most people just think: okej I just add some black airfilter that stops the light. Yes it does, but it also stop the most of the air. I did some test with my diy-projector and the absolute best way to solve it, is the way I show in the picture. This is how they do it in all stage and discolight equipments. It works extremly good no light leakage at all and it does not affect the cooling. And itīs very very simple to do.
The Alchemist
So how many fan's do you need?

I did some calculations a while ago and this was the result:

Ok let me first say, that this is a aproximation. The real stuff is mutch more complicated.

So we want to calculate how much flow our fan has to give when we use for example a 1000 watt halogene. So lets start.



What we want to calculate is the volume air per second we are gona need to cool the lamp. So we rewrite the equation into this:



We can now fill it in:

Let say that we lose 90% of our halogene light as heat so
Q=0,9 kW. Cp and the density of air we know. Let say that we want the temperature inside to be max. 40 C and the outside temperature is 20 C. So the temperature difference is 20 degrees.

So the flow we need is ~0,05 m^3/s that is ~90 CFM (cubic feet per minute)

This would be correct if there was no resistant to the flow, but we al know that this is not true. The resistant causes a higher pressure which causes a lower flow. (see fig)



So if you want to be safe, you have to at least double the needed CFM.
So in this case 2x90=180 CFM

A normal 60mmx60mm fan http://www.coolerguys.com/ProductDe...x?productID=485 only moves ~18 CFM. :D
A 80mmx80mm fan does ~39 CFM
And the best i found was a 120mmx120mm fan with 78 CFM.

"Oh no problem! I just install three of them!!"
This is not true, 3 fans does not mean automaticly 3x the flow.
This depents if you place them in serie (so side by side) or in a push and pull situation. But to now which system to use can only be measured and not predicted.

So a 1000 watt halogene can be cooled but you are taking a big risk.

For more information:

http://www.comairrotron.com/engineering/airflow.htm

The next graph shows the relation of CFM against kW halogene.
Remember. To be save double the CFM value.




Ps. This is just a calculation in real live other factors as conduction and raddiation also take place, so some spots in the projector can still get a lot hotter.

I got now reaction then and i couldn't try it in reallife so it can be wrong, so be free to make your point
ace3000_1
quote:
A 80mmx80mm fan does ~39 CFM

This depends on rpm, ive seen 80mm fans push out 80 cf/m and 120mm's push out 94cf/m, the cpu fan on my cpu pushes out 54cf/m and thats an 80mm, bear in mind the noise factor though, the higher rpm the noisier the fan will always be.
ace3000_1
Mathias you've got mail. If you didnt receive it let me know.

Trev
mathias
I am not planing to update this thread any more because I donīt have time with this anymore, so print it out before itīs gone.

Donīt forget Watch Movies !

Trev, I have not got you mail.
harvey
So what's the verdict here guys ? :smash: :D
nmd_sb
why are you going to remove this thread? This thread is very informative and those formulas come in handy. So please dont delete.
multiplexor
you know, theres good info in here, i'm tempted to make it a sticky so it doesn't get lost :-/

why do you guys think?
ace3000_1
Heya multi, yeah i think a sticky it could be, it has got a pile of good info in here. There are a few others around way down the track thats got alot of good info, ill dig those out latter for ya and maybe they could be merged into 1 thread working on its topic ect. There is one hell of alot of good info in here, but its just down the track and all mixed up.

Trev:)
mathias
Itīs okey with me to make it a sticky, useful threads disappear very fast right now.

Anyway I have some more useful things, that I will post later.
mathias
We all know that discharge lamps are the best for a good projector design, not much heat, excellent colortemp and long lifetime. But there is one thing that is not so very good with those lamps, the arcsize is alot bigger than the filament from a halogen lamp, a halogen lamp is a pointsource but a discharge lamp is not a real pointsource, this makes things more complicated when we use them in a projectordesign. Discharge lamps can also be a pointsource but you have to pay alot for that function.

Different discharge lamps:
HQI (MH,cdm) = Arcsize 9-12 mm, Cost $.
HTI = Arcsize 4-5 mm, Cost $$.
UHP = Arcsize 1 mm, Cost $$$.

Most people use the HQI-lamps as a lightsource to the DIY-projectors, because they are the cheapest discharge lamps, but
they are not a good pointsource, in a ohp-design with a big 15-17" tft panel you will not notice this problem, because you
donīt have to magnify the picturesource so very much. But the problem gets bigger with highresolution small tft panels that need more magnifying, the result will be a blurred projected picture. A bigger reflector and more distans to the fresnell helps a
bit. But for small tft-panels the best solution are small arc-size lamps or a HQI-lamp in a big Par-reflector design.
mathias
The Par-reflector is the best design for a 1-ended lamp and more effective than the ohp-design, but you have to use a bigger reflector if you use a big arc-size lamp type HQI,MH or cdm-t to get a good result. A very good thing with the Par-reflector is that the arc-size donīt have any blur affect on the projected picture if you use a big reflector as the ohp-design has.
mathias
The ohp-design is the most simple design to build but it has some disadvantage if itīs not build correct. Most people place the lamp in the center of the spherical reflector, but this wrong ! You must missplace it a little bit, otherwise you will shorten the lifttime of the lamp or it will be overheated, if you canīt missplace the lamp a little or you get blurness if you do that, you have to know that the lamps lifetime will maybe only be the half of what it should be.
Gunawan W
FYI,
don't do Big Arc size in big parallel reflector (Mathias post #73 dwg 1), unless you have a perfect reflector to produce parallel light, I tested it before and got uneven light beam.
Mathias,
"Most people place the lamp in the center of the spherical reflector, but this wrong !"
is this for MH lamp too? because it doesn't have filament, it just has a small tube with chemical filled inside. For Halogen lamp, yes , it blew just for 10 hours use, at least it happen to me.
mathias
You will always get more or less uneven light beam from a Par-reflector, if you use it close to the target, that is why you should use condensors in a such design, all Pro-projectors use condensors after the Par-reflector. But a big reflector with a big arc-size is a must if you want any parallel light at all. With "big arc-size" I donīt mean 40mm or so, more like 10-20mm.

I mean both halogen and MH, but I know itīs harder to missplace a MH-lamp, but the thing is that the lamp gets hotter than it should be in a spherical reflector design even a "small tube with chemical" gets hotter. How much this affect the lifetime of a MH-lamp is hard to say but it gets shorter. The thing you can do to avoid this is to use forced cooling on the lamp and many projectors use that.
declined
however: the combination of a hot bulb and cold air might cause the bulb to crack (same with a condenser)
bobharry
Then why not do the same as ace: hot bulb and cool it with warm air.
mathias
Iīm no expert in cooling, but I know that most pro-projector have two fans one that pull the cool air into the case and one that pull the hot air out from the case, the one that pull the hot air out from the case is always placed near the lamp, so the lamp will probably be cooled with hot air like bobharry said.

I did a different spherical reflector design a while ago, I used a dichroic reflector instead of the normal "soup laddel" reflector, in that way I donīt bounce back the heat to the lamp, only the light, notice that I use a normal magnifying glass as a condensor, not heat proof. The design worked very good and I had no problem with the heat. I know that the reflector is a little to small, but it was only a test setup and you can get a bigger dichroic reflector.
ace3000_1
Mathias nice work, im working on somthing at the moment thats very much the same as a eliptical reflector, when i get the parts i need ill post it up, in my tersting on a prototype i saw a masive increase in light both on the cdm-t and the 250w, the increse on the 250w was so much that it was as bright if not brighter then beforcades 400w. The only problem? heat! no probs for the bulb as they are tested and run at 650deg c in the factory, thats the outer shell case, but for the poor old condenser, well they dont like it weather its mounted loose or tight agains any plate lol. its just too much heat for the glass so ill have to find a very high temp grade of glass or make a baracade inbetween the lamp and condenser with temperd glass or ir filter.

Guys your prety much right all of you on the warm air on the bulb, 2 reasons they have the bulbs this way, 1 running a bulb too cold will change its colour temp and reduce its life, it will also reduce the light output, 2 for the obvious, boomba, the best way to run any bulb especially halogen is hot, they are designed to run hot and thats the way they are designed to stay, if you run them too cold you will have unwanted results. The things to mainly keep cool is the condenser and the reflector and this must be done with warm air and not cold, cold air will cause fractures not only in a glass reflector but also a condenser, trust me, ive tried it.

Trev
rcabezas
On the subject of cooling the bulb. What is the best approach for when the projector is turned off? Is it OK to leave it to cool itself down? Or should I keep the fans running for a while. Any ideas?

Raf.
bobharry
I was just about to suggest the barrier between the lamp and condenser then I read that you'd already thought of it. Wouldn't an IR filter work better? Because it should be designed to take a lot of heat, then theres the other obvious advatages too.
ace3000_1
quote:
I was just about to suggest the barrier between the lamp and condenser then I read that you'd already thought of it. Wouldn't an IR filter work better? Because it should be designed to take a lot of heat, then theres the other obvious advatages too.

Yess and no, an ir filter might over heat the bulb thats where i thought in the temperd glass, ir filter also doesnt stop much heat compared to cold mirror that close to the bulb.
ace3000_1
quote:
On the subject of cooling the bulb. What is the best approach for when the projector is turned off? Is it OK to leave it to cool itself down? Or should I keep the fans running for a while. Any ideas?

Heya Raf, i let mine cool down without the fan, having a fan running on a hot bulb without the lamp being on is infact more dangerous then with the lamp on with the fan at the same time, having a fan run on after the bulb is off may cause it to crack as the glass may contract too quikly, condensers also are viable to the same problems.

One thing ill tell you guys not to do at all ever, never turn on a fan when a lamp and condenser is hot, the lamp will surley explode and you can kiss your condenser goodbye, this is caused from too much of a temperature variance to soon, turbulent cold airflow over a lamp can cause the same problems hence why sucking the hot air is safer. never run a fan on a halogen either as they will surley explode, if it dont, you run a very high risk, a finger print can wipe out a halogen, same with metal halide, but most metal halides are protected by the outer shell, halogen have no protection whats so ever.

Trev
rcabezas
Thanks for the reply, Trev.
I agree with what you said, it makes sense. It's just that I've noticed that commercial projectors keep the fans running for a while after the lamp turns off, and was wondering why.
Anyway, I will do just that, suck the hot air off of the bulbs case and not blow air directly to it.
I ordered fans that ramp-up as temp rises; are they OK for a PJ, or are they a bad idea?

Thanks!

Raf.
loke33
Hey guys,
I'm new to the forum and have been allocating about 5 hours every day to reading the posts on this site, there is alot of information out there.

Anyway, to my question. I looked at mathias' picture of mirror size calculations and I had a basic question. I from what I have gathered one must match the fresnel's (the one between the light source and LCD) focal length with the distance from the light source, or the condensor. With a mirror does one only need to measure this focal distance as the distance from the fresnel to the center of the mirror and then off at a 90 degree angle to the light source. Sorry if this is a newb question or it has been covered in another forum, I havent found it yet and I have done quite some extensive searching. Thanks in advance.
rcabezas
Well, after seeing smoke coming out of my fans when I turned them off with the light; I decided to leave them running for a while after the lamp is turned of.
I'm thinking about programming a small microcontroller to do all that for me.
(turn fans on, lamp on, lcd on, and also off at the right times when a switch is pressed)
rcabezas
loke33:
The distance is still there, the mirrors just "fold" it. BTW, the mirror needs to be at 45deg, not 90, so that the light gets reflected at 90deg.

So, if distance without mirror is X, then distance from light to mirror and mirror to object should add to X also.
loke33
thanks for the reply rcabezas. Your answer was exactly what I was trying to say lol. Sorry i wasnt clear. Thanks again for the help.
SmC252
quote:
Originally posted by ace3000_1


This depends on rpm, ive seen 80mm fans push out 80 cf/m and 120mm's push out 94cf/m, the cpu fan on my cpu pushes out 54cf/m and thats an 80mm, bear in mind the noise factor though, the higher rpm the noisier the fan will always be.

Size and flow do corelate but I agree that some smaller fans will outflow much larger ones; it's all in the speed and blade design.

About the noise, I disagree 100%. The noise can for sure increase (and decrease) with rpm but the angle and shape of the blades is what will truely determine how loud or quiet a fan is. The S-shaped blades are usually the quietest, and the loudest are the flat-looking ones where you see only the edge.

-Steve
Varonis
Helo,

there I draw an idea, for replacing the first fresnel with right shaped reflector:

http://buildart.lv/Faili/paraboliks.gif

-- the reflector are taken from Satelite antenna (the stuff what catches tv signal from satelites, without any tech, but with its special lever who holds the lamp in precision focus point, that the reflected light makes a paralel beam. and i place aluminium foil, inside, to make better reflection, than original white matte coating.

this i have not tested in projector, but i tested it with 500W conventional light bulb :) - the results vere fantastic - we get a light beam, and we can hit the clouds with this light beam:)! - so, i decided that it works - light beam are parallel. so it is the same what first frensnel does. and many comercial projectors dont have the first frensenl so, they have a reflector what makes paralel light beam.

the fresnel are absorbing the light, but good reflector will be absorbing much smaller percent of light, so, we wil get brighter images?

sory for my bad english:)
cromaclear-crt
did you ever try this with a projector ?
vandevoordekoen
Hello, I now this tread is closed for a wiile, but is there anybody who tried de elliptical reflector setup? And where are those elliptical reflector to find? Thanks.
giorgos18
ywh has ellipsoidal reflectors for sale but they will not be suitable for standard MH, need to use the small UHP lamps (ywh also sells them)

EDMUNDOPTICS.com sells ellipsoidal reflectors that may fit 150W CDM-T Philips lamp if you want to use Metal Halide lamp

Additional optics may be required to expand light cone from ellipsoidal type reflector since beam angle usually not very large

I will be testing with ellipsoidal and short arc uhp lamps in DIY projector soon and will post results (if succesful)

Page generated in 0.25606894493103 seconds with 17 queries,
spending 0.02893758 doing MySQL queries and 0.22713137 doing PHP things.

Powered by: Search Engine Indexer and vBulletin
Copyright Đ1999-2009 diyAudio.com