I have been reading about projector designs for the last week now and have found it very interesting. I have very little technical knowledge but have a firm grasp of higher physics and optics. I am curious if what i describe below and if anyone else has done it.

A home built projector that has a light source of the end of a fibre optic cable, this cable runs to a parabolic mirror and collects the focused light (like a satellite dish) as the source.

It would take the heat away from the unit, and allow for higher amounts of light.

I see several problems with this and i was hoping that you would be able to help me work these through my head.

I see two situations for this
1. inside a class room during the day, the sunlight may be used
2. in the evening artificial light must be used

the problems i can foresee are that should sunlight be used, one some sort of tracking method might be necessary and two images from the outside are not clean white light - it fluctuates and would also be effected by clouds. Is a lens necessary to even out the fluctuations and also dissipate any images? (there is always an image - the principle of the pinhole camera)

Should artificial light be used can one use multiple lights pointing at the mirror so that the light is condensed to a single point at the other end?, one then has the single point of source needed for projectors. I have heard this talk of arc length and i was wondering if this would effect it? also what is the arc length of the sun?

the great concept behind this is that the technology for this is easily available, by using the design of a compound telescope (http://www.grc.nasa.gov/WWW/K-12/Aero2000/studweb/comp.html) and simply replacing the normal location of the eye piece with the end of the fibreoptic then one could - i believe place numerous lights inside and increase the output dramatically. By using this the parts will also be relatively inexpensive as they are already being manufactured for the alien hunters among us.

The last question i have is to do with the lcd - should my understanding of optics be true and this is possible. then the other consideration is weather the lcd will be able to block out the light for "black" projection. I read somewhere that only 1% of the lumans pass through the lcd. however if i could place 1,000,000 lumans of source within the light funnel then that would still be 10,000 - would this bleed through the darks? do i need to source a extremely high contrast ratio screen?

thank you

D

ps i will post a picture later

Fiber optics sound like a neat idea, but all this talk about satellites to track the sun or (am i reading this right!? ) 1,000,000 lumen bulbs is a little crazy. I don't know if a single light bulb that can put out that much light, but whether you went with a bulb, the sun, or some kind of crazy huge LED array, you would probably consume more energy than most third world countries.

4-9% of lumens pass through the lcd, 1% makes it through the whole projector, so what happens to 95% of the 1,000,000ish lumens that hit the lcd? it becomes heat, so much as to probably kill your lcd instantaneously.

Finally, what would you do with 10,000 lumens anyway, unless you had a 800" screen?

I'm not saying don't try if you want to, ArizonaVideo could use a little healthy competition, but I'm thinking a setup like yours would cost about 2-3 times as much as a 1080p commercial projector, and would cost over $50/hour to run.


WAIT! Doesn't fiber optic cable melt at a certain temp!?

EDIT: The sun isn't a metal halide lamp, it is a NUCLEAR FUSION REACTOR, it's arc length is it's ENTIRE length.

haha...

i especially enjoyed your post - a very fun tone to it.
ok i suppose i will have to make that picture to make myself understood. My talk of the sun was an example of using the array i am describing without the need for electric lighting. in other words the cost to run would be very low. It may be that the sun itself would be to powerful and the array could simply be pointed at the sky. in any case it was more a passing thought.

I am interested rather in using the said array to get around the problem of projectors need for single point light sources (though i have since read a clever scheme using 3 sets of fresnal lenses to get two light sources working in tandem.

800" sounds quite exciting though, that would be a small cinema on the cheap. You mentioned the temperature problem, i am not aware of the melting point of fibre optics but as it is glass it must be quite high. If i understand your second point, 95%of the visible light spectrum (or lumans) hitting the lcd turns into heat. That would imply the need for some form of ver good cooling - a problem i will now consider.

i am interested in achieving a very bright projection, that is why i am talking such high numbers. however if the same design might double as a miniature cinema projector all the better.

to see some different design sketches for copmpound telescopes see here
http://www.myastrologybook.com/Telescopes-...atadioptric.htm

D

ps system components
1 typical diy projector
2 fibre optic cable
3 compound telescope (this is the best solution from a size point of you as it has a large collection surface unlike a regular telescope - alternatively a parabolic reflective surface could be home built - but as it is proffessionally manufactured i thought it would be best)

this is an interesting concept... Looking at the diagram, I have a question: wouldn't the light exiting out of the fiber optic cable be like a laser and be a pin-point light source hitting the Fresnel in only one center spot?

I agree, it is a nice design, but i see a few problems with it. There's the tracking system that you mentioned, but there's also:
1: to collect enough light to be optimal, you would need a very large telescope (how big i don't know, i'm not prepared to do the math on it!)
2: last time i checked telescopes are expensive!

So do you have a telescope handy? it might be interesting to make a mock up(minus the projector) and point the fiber optic end at the wall, and give us lux readings, then you would be able to calculate how big of a telescope you need for your application.

P.S. I'm still iffy about the melting point of a fiber optic cable, I'll see if Google could shed some light(get it?) on this.

>>> NOTE: The compound mirror concept is for use with artificial lighting, the solar tracking sat dish for the sun :NOTE <<<


I would like you to know i do intend to build this, when will probably be in 3 months from now. I would also like to point out that i have not tested any of my ideas and am in planning stage and am completely open to suggestions.


To answer the last two posts.

brandinothegreat

The solar version is an optional feature - when i build my own it will be utilising a bank of LED (at the moment). To answer you concerns, tracking systems may not be necessary and if it were they are readily available.
here you will see an example of one that is used in conjunction with fibre optics to get natural light into the middle of a house.

http://pesn.com/2005/07/27/9600139_Fiber_O...ng_Sun_Indoors/
http://www.ecogeek.org/content/view/246/

You also brought up the issue of the telescope - i dont think i would buy one so much as just follow its concept. In other words i would buy a mirror and do the rest myself. I did a quick search and found this site selling mirrors (could probably get them a lot cheaper - but you get the idea. There would be some math to work out later for how big a mirror one would need - but that would be determined by the type of light source and how powerful it is.

http://www.galvoptics.fsnet.co.uk/telescope5.htm

jcherokee_84

My understanding (quickly revised on wikipedia) is that a lasers qualities come about in the way its created - if one could create a laser simply by sending light down a piece of fibre optics you wouldnt need the clunky laser machinery. However even a laser over enough distance will arc out - and thats really what you are asking me "At what angle will the light be emitted from the end of the fibreoptic." - aka dispersal angle. I can happily inform you i havent got a clue A quick nosy about online looking at indore fibre optic lighting solutions would lead me to guess it is about 15 degrees from the centre axis 30 total.

some interesting information about this can be found here http://www.del-lighting.com/products.html, they also have prices on there - american but you get the idea.


-----------------------------------------------------------------------------------------------------------------------------------------------

Looking at the project the most expensive piece will be the parabolic mirror for the light condensing tunnel (lct ohoh i got to make my own anacronym) - i believe this will have to be at least 10" in diameter making it an investment (based on the website above) of about 300 pounds. this is out of my budget for this project so i will start looking for one secondhand and other sources. otherwise it is not worth it. once i have one i will do a lux test.

Q; brandinothegreat, you mentioned that at a certain level i will fry the lcd screen - do you have any idea what that limit is?

for artifical lighting you could simple use a cold light source (dunno what it´s called in english), like the use in medical devices. Pretty much the same concept with fibre optics, but an eliptical reflector instead.
They are very expensive though.

An interesting idea... i have looked around online (for a while now) and it is very hard to get info on these systems - must be the medical industries paranoia that if someone new how simple there devices were then there would be no nead for them I did find this product description

Product Overview:

http://www.myneurolab.com/myneurolab/mnl_p...dsubcategory=19

"This illuminator for Fiber Optics includes a filter holder and IR filter to remove the heat from the light entering the light guides. For most biological uses, microscopy or small animal surgery, this is the appropriate choice. Additional filter choices are available in the optional filter kit.

Bright detail lighting without heat, vibration or electrical interference makes fiber optic illumination the best choice for many small animal surgery, electrophysiology, or microscopy applications. Light can be applied spot focused or diffuse with the aid of adjustable lenses on the end of the fiber optic pipe. Several fiber optic options are available, single pipe, dual pipe, ring lights (for stereo microscopes) or micro probes. For demanding applications such as color photography or machine vision applications, a constant color, stable output model is available."

The interesting points here are that an ir filter and other filter set reduce transmitted heat and that adjustable lenses can be placed over the end of the optic cable giving a variance between a diffused and focused projection of light from the tip.

I do not think the machine itself is an answer (again the cost OMG), but we can take from this the concept of filter use and lenses for the cable. It might even be possible to source both of these as parts.

In anycase the research continues - thankyou for the suggestion

D

I have been looking at mirrors and i think i will have to make my own to keep the cost down - the idea is simple, make my own dobsonian mirror array and use it as a focus for bringing multiple points of light into one point, send that through the fibre optic and out as the single source for the projector.

here is a link i found for making your own mirror.
http://stellafane.org/tm/atm/index.html

and a dobsonian array looks like this
http://www.skyandtelescope.com/howto/diy/3306876.html

D

Well, so long ascost isn't an issue, here's an idea which could just get a lot more efficiency from a MH light source, uwsing fibreoptics.

1. Get a prism, or, more accurately, get several prisms. Place them all around your MH lamp so that you can capture as much light as possible and separate it into it's constituent spectra.

2. Use fibreoptic strands to capture red, green and blue wavelengths.

3. Direct these wavelengths to the corresponding locations on the LCD colour filter.

What this will do is allow the LCD to pass much more of the light that strikes it, since the largest portion of the light that an LCD blocks is due to the colour filters.

Alternately, if that level of precision isn't feasable, you could just capture full spectrum light from all around the lamp. This could be in a spherical shape just a couple of inches from the lamp. This eliminates most of the square law losses from lamp to fresnel, plus captures a much larger percentage of the overall lamp's light.

The major problem... Time and Expense. I'd worked out before how much fibreoptic cable would be required to cover a 1024X768 LCD with one strand per sub-pixel (1024X768X3=2,359,296 strands, an average of 8" in length. The end result is just under 298 miles of fibreoptic strand. Add in the effort to place these at the LCD, and well... That idea doesn't work out.

Wow now that is thinking grand!!!

As you iterated this is basically unfeasible from a cost/time point of you. However i think that the concept of using a prism to divide the light into its separate parts is brilliant. When i was looking up colour filters (i did not know what the term referred to - not yet up on my TFT LCD components) i came across this page.

http://images.google.co.uk/imgres?imgurl=h...8NZ308%26sa%3DN

This led me believe that one can find prisms supplied as thin sheets, in other words the sheet would be many prisms in series and would there for have some mathematical traits - such as depth, quantity etc... I would imagine with this information one could start with the pattern in the colour filter as a baseline and with the properties of the prism sheet work backwards calculating the distance and alignment needed for the prism sheet so that the light is divided into the perfect number of columns to correspond to the colour filter.

Wow i had not thought of that - its a great idea.

Thank you for the suggestion

D

Here is some info on arc lamps for fiber optic illuminators.

http://www.cairnweb.com/tech/tech_lamp1.html

Reply to Marco Polo

I looked at the link you sent to me - i cant claim i understand them. Rather i can understand the data, but i do not understand what the optimal holy grail of light sources for projectors comparative graph would look like and so dont understand fibre optic illuminators bare desirable traits.

Do you think that there are different spectral requirements for using fibre optics in this manner. I have started looking into different types of bulbs available and am now completely confused. i had hoped to use a bank of LED lights - as they are fairly efficient and relatively cheap - however i am open to suggestions. As my original post said i am hoping to get a hi luman rig set up, but using a combination of low intensity, long life and cheap bulbs.

hypothetically what is the luman count of a modern day cinema projector - that would be my goal (why - i believe the question is why not? rhetorical - please dont answer that)

D

Are you planning to send the light thru a light pipe ( http://www.physics.ubc.ca/ssp/research/lightpipe.htm )
or a bundle of fiber optic threads. The pattern of light output would be different.

Wow another interesting idea...

Light Pipes or Fibre optics. I had never heard of a light pipe before and it was very interesting to read about. I cant seem to find any other references to this anywhere else - do you know if it has another name?

It is a point to note though. In my idea i am introducing three components apart form the "standard" DIY projector.

1 multiple light source - solution unknown
2 LCT (light condensing tunnel) - design based off of dobsonian telescope, will make own mirror.
3 Optic cable - solution unknown

additional ideas

4 prism split light to increase efficiency of colour filter - solution known but material unsorced.
5 solar collector - solved - need to experiment.

The components 1 and 3 are why i opend this thread - i do not know enough about these yet and wished to gain insights.

the search continues
Thanks for the suggestion

D

light pipes are also called light integrators or homogenizer

here a thread http://www.lumenlab.com/forums/index.php?s...c=14198&hl=