DIY Video Projector

[Myren]

can anyone confirm this hypothesis:

the higher the distance between the focal points in an ellipsoidal reflector, the worse the light distribution.

thats what a lot of graph paper with a lot of drawings and a lot of math has been telling me, but i'm not at all sure of my work.

if this is the case, then ellipsoidal reflector design is a careful balance between capturing the most light and maintaining the most uniformity.

i'm going to start figuring out if theres a way of using fresnels and spherical reflectors to give better light uniformity (how constant the birghtness is) and light efficiency (% of light that makes it to screen).

for the first time ever, i wish i was taking calc four. knowing how to really integrate across three dimensional surfaces would be very useful.

and is human perception of light logarithmic, linear, or otherwise?

myren

[The Xung]

I'm pretty sure human perception of light is logarithmic, that is why we can take quick glances at the sun and walk around in dark woods.

I was getting a dis-uniformity feeling the closer I put the focal points. The closer you put the focal points, the closer it becomes a circle. light emmited from the center of a circle all gets reflected back to the single center. I was getting better uniformity with larger ellipses and moving the lcd closer to the second focal point.

What I found was that that funny polarizer sheet that came with the LCD that Muzzman was talking about gave the image much more uniformity although it was darker. Howwever, this was with a parabolic refelector.

[Scot_lad]

Hearing is naturally logarithmic, I think vision is different because your eyes adjust the aperture (your pupil) to different light levels, letting more or less in to suit the light conditions.

As for CRTs, sorry I never got very far with my post the other day. I have been investigating a lot and have answers to most of the queries brought up here, I'll just have to wait for another 8 hours to post them til I get home.

I'm glad there's a lot of discussion going on about this method, one way or another. One of the main reasons I took this route is because everyone is doing the LCD thing and I wanted to try a totally different route just for contrast really.

Will post later.

[Jompa]

I, like many of you here, is looking for a lcd that could be suitable for this project.

I found this; 0.55" full color display with NTSC & PAL sync input, 537 x 222 resolution for just $29.

What do you think about it, could this be any good?

A link;
http://www.digisys.net/timeline/hacker.html#mini

It says that "Unit requires a clock, Synchronization and Video", what does that mean? Is it difficult for a non-electrician to handle?

Best regards,

Jompa

[frankbatzen]

Studying for exams and setting up a stable mount for the PCB I haven't gotten to the soldering, yet. It probably is cowardness, thinking of it. Might be a couple of days, before I can tell you more. Actually it is 110 connections to solder, so it really might be some days... One thing that may be interesting to you: The connector that accepts your ribbon-cable is either a ZIF (Zero Insertion Force, a connector that has a sort of locking-mechanism like a slider or a lever) or a LIF (Low Insertion Force, a connector that you just stick the ribbon cable into). ZIFs allow for an endurance of the ribbon cable of just 30! cycles, meaning you can insert and pull out the cable 30 times). LIFs only allow for 10!! cycles. The contacts on the cable are very sensitive and might be the reason for your display not working. You should look at the contacts--if you haven't done so already--with a magnifying-glass. Having taken my ribbon cables out of the ZIF-connectors only once, the contacts already look pretty worn out. Maybe this helps in diagnosis.

g!zmo

[Tim]

Jompa


537 x 222 = 119,214 dots
119,214 / 3 (RGB) = 39,738 pixels

compared to 640 x 480 @ 307,200 pixels

It would have about 13% of the resolution of a VGA monitor.
Not very good.

Tim

[lokost]

Tim

Thanks a bunch. I have been told by the good folks at City Glass to look for one way mirror (a.k.a half-silvered mirror).

Bitch

Wow! thanks for fishing out my post. I was wondering if some UFO snatched it.

Myren

I came across this while investigating the possibility of reducing the arc size of the cheap metal halide lamps.

http://www.fvastro.org/presentations/ray_tracing.htm


As for a wide glass bead screen, I saw your post on the other forum, these guys have a couple for real cheap:


http://www.etsusa.org/templates/stat...b?parm=AVEquip

[jjasniew]

One of the IT folks said they had this old proxima panel, so I took it home to compare. It was a proxima "ovation" 640X480, with a vga, s-vid and composite input.

It seemed to do better than my infocus panelbook only as far as not having the "venetian blinds" artifacts on moving objects, that I could see. The image was brighter, but at the cost of larger pixel size apparently. I couldnt get a good contrast out of it.

The achellis heal of this unit was that the horizontal size was out of scale with the verticle. Frazer's head looked suspiciously wide and nothing like my ordinary TV. I perused the manual and controls and found NO way to correct this.

In contrast, the infocus panelbook displayed in the correct proportions.

How bummed I'd be if I spent $50 - $100 for one of these on ebay - and then found out that WYSIWYG. I guess it might be fun putting a tire iron though it.

Interesting, the specs in the manual said 50 ms response. I didnt see how this panel was "too slow" - as I said, motion looked fine. Maybe this response time actually filters out the venetian blinds artifacts, I dunno.

Correct me if I'm wrong about that horizontal scaling. I'm pretty sure I didnt see a 'width' control anywhere -

[fender4]

I guess I'm back to the LCD search. I received my 5.6" Sharp LQ6NC01 today, and moving the driver board to keep it from obstructing light transmission is IMPOSSIBLE. Well, I'm sure it is possible, but it would involve extending 100's of tiny pins that run around three sides of the panel.

So, alank, it looks like I won't need your lens after all. It will be some time before I can work on this is again (board exams in three months).

This leads me to my request from the board members:

Which small LCD's (like the mobile panels) can be adapted for this project? I know that Vince and zark have done this (I think zark used the PartsExpress #205-013). I think some have been successful with the Sony PS1 LCD's. Please post or email me if you have any suggestions!

nadyleets@hotmail.com

Thanks,
f4

[MuzzMan]

fender4:
Wow, sorry to hear about the sharp panel, sounds difficult. It does sound like it is not a TFT panel (TFT panels have more or less a multiplexor scheme on the glass itself with one driver transistor per pixel, so they require less connections between the driver board and the panel).
My 5" toshiba TFT display from ps1 MadCatz screen addon was quite easy to dissassemble. Connections were only along one side of the panel, and the driver PCB could be folded back out of the way. Which is a good thing, because I believe it would be near impossible to work on the connections if that were needed. In some ways a ps1 display would be a good panel to use, because ALL of them would be recent TFT screens. Be sure to pick one that already has a composite video input (external AV connectors). Mine didn't, and the panel only supported RGB. The drawback to these screens is the resolution - only about 384x240 (real res). Now I am starting to get a really decent picture out of my setup, I want more res than that and I am thinking of ways to cheaply obtain a decent VGA or higher res panel and drive it from a PC.

BREAKTHROUGH!
I have come to the realisation that brightness and even contrast are not the real problem my setup has. I have an ASUS graphics card in my machine with video in and out, and so I hooked that up to the projector in order to have more flexibility in adjusting brightness/contrast/black level etc. I found that I could get a terrific picture in dark night tv scenes by cranking up the brightness control a bit, but I had to turn it back down for brighter outdoor tv scenes. What I really needed was an adjustment for GAMMA, ie. non-linear control of the brightness curve. What I need is blacks to stay black, but colours just a bit off black to get brighter. At the same time bright near white colours have to be unchanged. Most graphics cards provide this gamma curve adjustment now, but I found that when I adjusted it on mine, it applied to everything except the video input graphics. ie. Windows background goes up and down in brightness, but the video in a window stays exactly the same. The controls the ASUS driver provides for video are only black level, brightness and contrast, NO gamma...
The ideal would be to build a standalone circuit to control this, but in the mean time I have to look for better software for the asus card that might help. Anyone have any ideas?