DIY Video Projector

Status
Not open for further replies.
CRT lenses

I have a set of adjustable lenses which I got from an old CRT projector, which I presume are roughly the same as yours. They are very good but you can only have a small LCD, which fits on the front of the lens. Unless you can bend the light through your panel or something.

I haven't bothered using fresnels although I do have a 17" one which came from a friend (not me, honest) who bought it thinking they could make a projector from a normal 17" TV. Obviously I have it because it was hopeless for that. But it is useful for taking the light from the light source and turning it into a parallel beam going into the back of the LCD. LCDs like parallel beams of light apparently.

Putting the fresnel in front of the image isn't a good idea, because it is relatively low quality and while that doesn't matter as much for just light, it matters a lot for an image. Notice how in an OHP, there are two fresnels, one which takes the light from the bulb and turns it into a parallel beam, then another fresnel directly stuck to that which takes the parallel beam and converges it through a slide (or an LCD panel) into the high quality glass optics at the top of the OHP arm. Or whatever they're called.

Hope this helps.
 
..OK so does image quality suffer?

So I gather from Scot_lad's reply that in OHP projectors the light does not enter in though a straight line through the LCD panal ( as I suspected). So from what we know of LCD panals, viewing angle problem could lead to a problem where the image is very good dead center, but more dim and washed out at edges. At this point in game, many of us looking at least ~$450 costs for this project, is this qualtily acceptable? (please look at the reviews for panasonics new $1.5K projector, it looks very good for a low cost projector, and has low per hr costs).

I really wish there was more discussion about image quality, and what is reasonable and what is not. For those that already have theres running, please post pics so we know if the OHP + LCD panal combo needs to be improved on quality wise.

Secondly, if we are going to insist on using a large lense to converge the lcd panels parallels light (assuming non parrallel is unacceptable), is glass (could plastic suffice?) lense unreasonable at that size (15" across). And is a there a supplier for these large low cost lenses?

Attached is a little bitmap I made of the OHP + LCD with mspaint (as oppose to the more IDEAL situation I drew last time)
 

Attachments

  • bitmap of ohp lcd problem.png
    bitmap of ohp lcd problem.png
    3.4 KB · Views: 187
Alternate Light Source....

Is this thing on???

Well, I've just finished reading all of the posts in this thread...yes, all 70 pages.... Unbelievable!!!

This is my first post....

You're all a bunch of Bastards - now I'm sucked in also!!!


First a thank you to Vince for the idea of this thread and the willingness to keep trying... :)

To all the others, thanks for sharing and not giving up as well!

Bitch - You make me laugh.... Great posts! :D


I haven't seen anything like this posted so I figured that I'd throw it out there...

Think Pyramids.... :rolleyes:

As I was reading, an old camera lens that I used to use kept popping into my head... Now the idea is not exactly the same but it got my wheels turning... The idea behind this lens is that it uses mirrors in such a way that the spot in the center of the lens is reduced to a point that is so small it's essentially eliminated. The picture in the link doesn't show it very well, but the "spot" in the middle of the lens is about 1" in diameter and there's a mirror behind it...

http://www.tamron.com/35mm/35mm_mf/500.htm

If we were to apply similar thinking to the light source...arranged as follows, one could generate 27,300 lumens using 4 of the LOA units...

http://www.lightsofamerica.com/products/worklights.htm

The advantages would be long life, less heat and most importantly, all the light striking the LCD at the same angle. You would simply size the mirrors so that the finished dimensions of the pyramid base match (or make it a little bigger) the size of the LCD screen. The length of the sides would be determined by the length of the Florex bulb... Make the overall length a little shorter than the bulb length... I am guessing that the steeper the angle the better off you'd be...

Check out the attached drawing...

A couple of points:

I would use front silvered mirrors, and use caution when cutting them. The video image already will have enough red in it.... :eek:

I figure that the best way to assemble the pyramid would be to lay the cut glass panels on a flat surface face down next to each other. Then use silicone to join the seams between each panel. After it's dry, when you pick it up you will only have one seam to deal with and that you can easily silicone from inside. Just tape the thing in shape on the outside and goop the joint on the inside then let dry... Viola! - instant pyramid...

I would support the pyramid from the base using L-brackets or something similar either siliconed or epoxied to the back of the mirrors - this would give you a way to mount it....

The glass panel on my drawing is optional but I figured that it would "smear" the light a little bit...don't know if it'd make any difference... Also, it would make a great place to attach the polarizing filters if you decide to use them...

Obviously, one would still need fans - I think drawing the air in around the LCD using little deflecting "wings" then past the bulbs then out would work well. Locate the fans inside the box pushing the air out to reduce noise...

This config should give LOTS of very white light without all of the other heat problems... Yes, the image theoretically will be most dim in the center, but I don't think in practice that it will be noticeable...the ambient light bounce inside the unit will obliterate it.... just like on the camera lens...

Here's the thing... I can't build this right now due to other projects dominating my time (remodeling my house) so I figured that I'd throw this out there hoping someone will run with it...

If you have any questions please don't hesitate to contact me...I'll help if I can....

To all, keep up the great work! What has been accomplished here is nothing short of amazing....

Steve
 

Attachments

  • lightsource.jpg
    lightsource.jpg
    9.5 KB · Views: 726
That is a great suggestion Apogee, it is good to see some new ideas injected into the forum. Seems like the pyramid mirror scheme has heaps of promise.
Has anyone tried mucking about with polarizing filters and half wave retarding sheets etc. to build a 'polarizing recycler' like someone suggested a while ago? That had potential to double the efficiency of whatever light source we use - sounded great to me!
 
Assembly Clarification

Just wanted to clarify one point in my previous post...

When assembling the glass panels, lay them all flat face down next to each other. Then create a silicone "hinge" using about an inch-wide ribbon of silicone over each seam. Don't worry about putting silicone between the edges of the glass (actually it'd be better if you don't).... Just place a ribbon of silicone over the joint on the back of the mirror and smooth it with your finger to create a ribbon... When dry, it will act as a hinge.....

Also, standard clear silicone should be fine... I think it has a working temp of ~650F. They also make the high temp stuff that you can get at auto supply stores...

Hope this helps,

Steve
 
contrapasta: hrm. this is a very intriguing point, one i actually hadnt considered. i suspect the light collimator has a much greater effect on the light uniformity than the viewing angle. notably, the use of fresnels as light gathering devices rather than something like a ellipsoidal reflector probably contributes to most of the uniformity issues.

does anyone know the effects of the viewing angle? is it an all or nothing thing? whats the % angled to transmisiveness curve look like?



the one thing conventional LCD projectors do have over us it their usage of three seperate panels, which go through some crazy light recombiner. this mixes the colors together, making a solid single colored pixel. the panels we're using have individual RGB pixels which would become apparent if you got really close. for all intensive purposes though, you honestly would never know the difference. notably the higher the resolution, the less you'd notice.



screen door for the 15 inches is completely neglidgible. i got right next to the screen once and used the screen door effect to focus the projector, but my face was practically on the screen so i could see it sharply enough. again, not a problem.




my replacement panel is coming in on friday. the good news is it'sb being replaced, in spite of it having been used as a projeciton panel for quite some time. i'm so excited, i was worried they wouldnt replace it.. the bad news is it feels like its been forever and a day since i've been able to use it, and i'm going nuts with anticipation.



i've got a collection of 50 inch wide screens growing in my basement. couple pearlescents, a beaded high gain, a white, some other random types. i'll try and post an overview of the different types when spring break ends.

myren
 
Myren, what happened to your panel that it needed replacement?
Did you break it in having it out of the enclosure? I'm concerned about ESD failures, which would present as "I picked it up and it just stopped working from then on"

I've posted some pix, folks will have to look back to get the URLs
These were using an OHP pushing light through an Nview Z350, which is basically a 1024X768 LCD panel mounted in a box...There's also pix associated with my Ebay ad which I also mentioned here.

The light is of course pretty much uniform and going up straight through the LCD. All the OH projector is, is a bulb with a reflector behind it and a fresnel lens focused on the bulb. My other model OHP is also a bulb with a reflector behind it, a mirror and a fresnel.
That model is "slimmer" because they reduced the straight path length by bouncing the light off the mirror.

One design I thought of might have one fresnel ordinarily arranged, and the reflector replaced with another fresnel at the same focal distance as the first, with a first surface mirror laminated to it. This would reflect the light back to the point source of the bulb, which would "double" the brightness coming out of it, in the working direction.

I'm suspicious of using flourescent lighting for the purpose of illuminating a panel. I believe (correct me if I'm wrong) that all forms of flourescent lighting are not continuous output, like an incandescent is. They actually are "strobing" on and off, but so fast that the persistance of the eye makes it seem like it's continuous (like the PC CRT screen and the TV screen)

The problem I'm suspecting is that, say, if you have a LCD refreshing at 60hz and a light source also pulsing at 60 hz, you may get strange modulations between the two frequencies. Does anyone know what freqency the LOA flourescents run at? It's something to consider.

I'm thinking of buying a Viewsonic VP140 from a private sale. Anyone have any experience with this panel?
 
Just in case I've been misunderstood...

Just wanted to apologize for calling all of you Bast%rds!

Just wanted to clarify this point....

I meant it as a joke simply because with so many other projects going on (my house) I can't believe that I'm actually now considering building a projector.... All of your posts have totally sucked me in to this project!!!! :) Very fun!!!

As far as a strobing effect using the flourescent bulbs, it's never been an issue that I've heard mentioned before. Very good observation though.... I know at work I always run the refresh rates higher than 60hz on the CRT's due to interaction with the lighting so I presume it could have a similar effect...

My line of thinking was that by building the light source as I described, one could utilize just about any size of display. Hence, the use of laptop or flat panel displays should work better because all of the light would be hitting them straight-on not bent as from a fresnel lens.... In theory if the fresnel is big enough, then the light should be even. The problem is, based on what I've read, that almost everyone using OHP's has the LCD basically at the edges of the usable area of the projector. This outside area will have significantly more distortion of the light than in the center... Hence, why I was looking for another way...

Also, the better/larger displays could have the refresh rate turned up so that there would be no chance of interaction with the light source...

I believe the best person to answer this is Zark. He has had one of these in operation using the LOA lamps for quite some time....

Steve
 
Wooooooooooohaaaaaaaaaaaa!

It works! Last night at 2.00 am I watched TV with an upside down picture on a diagonal of about 4 meters. This is insane. No pixels visible. Bright enough. No pixel smearing. Colors not yet brilliant enough (projecting on a beige wall). I have a quite important exam on wednesday and have a load to learn. Will post more extinsively after that.

g!zmo
 
About fluorescent 60Hz strobing

I can't remember the source right now, but I believe that fluorescent lights that run on ELECTRONIC ballast as opposed to magnetic ballasts don't cycle at 60Hz. I'm thinking that the LOA has an electronic ballast...I see what I can find when I get home from school.

Also, would the viewing angle effect be less a problem with a non-point source light like the LOA, since it more of a "backlight"? It seems like the non-parallel rays would effectively be tamed by the polarizers on the LCD anyway.

It's great to hear all of the fresh ideas! Keep up the good work!

-f4
 
My reflector fresnel ides in error...

I've thought about it and the fresnel with a mirror idea wouldnt work as I described - was late last night.

It would reflect back parallel rays the size of the fresnel, like the reflective OHP designs do - not rays converging at a point at the bulb!

Whoops!
 
LOA frequency

f4,

You're right, they do have an electronic ballast, but still it runs at some frequency.

Finding out what frequency it is shouldnt be too difficult, if you have an oscilloscope and one of the CdS light dependent resistor cells handy. That, a 10K resistor and a battery should let you see a waveform.

I built a "talk over a light beam" out of one of these for a school science project. I recall it worked well in rooms with incandescent lighting, but with flourescents around (typical school classroom) the thing just saturated with 60hz buzz.

Steve,

I agree it may not be an issue. My Elmo HP OHP uses an arc light, which I'm sure is arcing at some frequency. I dont see any funny stuff going on with my panel - at either 60 or 75 hz. I dont know what the frequency is....
 
hi have built a pretty good projector out of a old computer monitor
a cardboard box a mirror and a fresnel lens the color and picture looks better than any ive seen here so far. however i would like to try lcd after all thats what the big boys use , im very happy to see there is a place where we with this fasination can share ideas !!! so keep it real lets think outside the projector?

were working with light lumens a very hard beast to tackle...
when sound proofing you should think of water proofing if its water tight its air tight its sound proof ahhh?

what im mean is we need to control are light and shadow to much light in the wrong places can result in poor picture as im sure some of you have figured out for those of you useing fresnels try not useing the whole fresnel i dont mean cut it just cover it with black constuction paper with about a 4 inch circle cut to the center of the lens this may cut harsh light depending on your projector design. well any help can be usefull hope to see some good ideas think light, think contol color and clearness the 3 c's
 
thought I would share some links. Apologies if any of these have already been posted:

Good cheap video to vga scaling hardware:
http://www.nexternal.com/avtoolbox

These guys are really interesting. They sell a really versatile flat panel controller, with options for video and/or vga input, and interfacing to a wide variety of TFT panels. I emailed them about one offs last week, but no reply as yet... I think this would be a great way to go if they were reasonably inexpensive. You could find a dead laptop with a decent display 1024x768 tft, and pair it up with the controller.
http://www.digitalview.com/products/controllers/index.php

Finally, I found a application note from Linear tech which included a schematic for a fairly simple adjustable gamma correction circuit. I think it is shown bare bones here, and might need some additional circuitry to handle sync pulses ok.
http://www.linear.com/pdf/an57.pdf
The app note also has a variety of other handy video circuits.

When I find some time I will be building a gamma correction circuit, so when I get it going I will post the schematic here. From simple tests I have done with my setup, I think applying a gamma curve will improve the picture quality enormously.
 
Re: CRT Options

Timtimes said:
The first I heard about this project was the 100 inch projected crt (big woody), which reminded me of a similar project (probably in popular science/electronics mags) when I was in high school (70's).

If you are going to project a CRT, I think you'll do better with a SMALL CRT. They are just BRIGHTER to begin with and I would think working off a 5" tube would make the optics more managable. I see a couple folks who are trying 14" monitors, which probably isn't going to be as bright/easy to deal with. You can buy a 5" TV with tuner built-in for less than $200.

For what it's worth, most of the rear projection TV's use five or seven inch CRT's. I'm curious whether the quality of a 5" CRT projected to about 40-50 inches (big enough for my tastes) wouldn't be just as clear/bright using a cheap fresnel setup as the more complicated LCD stuff most are working on? It'd be easy to invert a five inch tv set, and I can live with the left/right inversion by switching my l/r stereo leads to match.

Enjoy.
hi use a mirror and u dont have to deal with l/r prob:)
 
comprehensive instructions...yet?

I've been following this and have yet to see any step by step instructions. When I get ready to do this project I'll have about $1000 saved up. Should I just save up a little more and get a projector off of UBID or can i build a better one? I see some really good looking stuff on ubid lately.

Are there any comprehensive instructions on how to do this project in its entirity yet with parts list? I've seen people say they were going to post.
 
Status
Not open for further replies.