| woneill |
Hi Aleksey,
I am not sure about this, but in my investigations, it appears that the Xenon HID Projector beam headlamps are only running at 35 Watts. Also they are very blue, which might not be good for colour rendering.
What I do know, is that unless you find a really cool source, they seem to cost more than the standard MH projector bulbs for commercial projectors - I have not seen a complete assembly for less than $350.
I was experimenting with some cheap halogen pod ones because they are small, cheap, and have an exit beam diameter of about 1.5". Oh, and did I mention that they were cheap??? ;)
If you know a source of the Xenon projectors, please let me know 'cos I want some blue eyes for my pet motorbike... :cool:
The big thing for me here is not the specific use of car headlamps themselves, it is the possibility that we can merge two separate beams from different fixtures into one single beam, without losing the quality of focus that comes from a small bulb in a well focussed assembly.
Also, for experimentation purposes, the lower colour temperature of halogens does not matter too much in my case, because my eyes seem to auto-adjust very easily to the colour balance of these bulbs. (My room lighting uses halogens, and my walls are cream. Before I stuck the shower curtain up, I was projecting onto a cream wall without too much discomfort...)
There are, actually, "ultra-white" car headlamp bulbs available, though, and some claim to be brighter than normal - check out the PIAA offerings. But, even if the light output per bulb is actually reduced, if we can feed an array of these things into one single beam, then it might not matter as much.
The main reason I am leaning in this direction is that it seems SO difficult to utilize all the light from the big 400 Watt bulbs without resorting to a massive reflector. I am just guessing, but if we could make use of maybe 80%-90% of the output of a lower powered light source, could we still get a brighter image than most people are getting from the big MH bulbs?
I don't know, and if someone can hit us with cheap, small, powerful HID bulbs that could be focussed with a small precision reflector into a well collimated beam, I would definitely follow that path instead!!!
All good fun!!!
Bill. |
|
|
| woneill |
Hi Xblocker,
You make an interesting point about the dates these MLA layers began to be introduced. I must do more research, and get my microscope out!
In my experience, though, I have been playing with a number of panels of the RGB OHP type, mostly being originally manufactured by Sharp, and having manufacture dates from '92 to '97.
The non-TFT ones in the old monochrome panels are nice simple optical filters: you can hold the panel up in front of an image you are viewing (e.g. a computer screen), and not see any distortion.
For four out of five of the TFT panels I have played with, if I hold them up in front of an image, then I get some serious distortion. The distortion gets worse as the angle increases (panel gets tilted), and seems to be in some part due to the shutters acting as a diffraction grating.
However, even directly holding the screen up with the light passing through perpendicular to the surface, there is significant image disruption. The image is recognisable, and not completely obliterated, but "something" is definitely interfering with the light as it flows through the panel - creating a much more diffuse output.
I had believed that the MLAs were introduced into LCD panel manufacture in order to bring the TFT devices into the mainstream - they have always suffered from having much less shutter area than the STN types (because the transistors on the panel took up a large surface area), and they were originally very directional. They needed (and most still do) a much brighter light source than the STN types to be usable. My understanding was that the MLAs were introduced to panel manufacture to reduce the demands on the light source - funneling more light through the small apeture, and as a side effect of their design, increasing the viewing angle of the light coming out of the front.
I could well be wrong on all this. But, the fact still remains that in many of the panels I have played with, there was and is this image disruption effect - ranging from mild diffusion to full scale ghosting and total loss of focus. (No, it's not the vodka - honest!!!)
If it is not MLAs, then why are the TFT screens not as good optical filters as the STN devices?
These panels themselves seem fully functional in use, and work quite well as a projection panel - except that they are interfering, to some extent, with the operation of the fresnels between the panel and the light source, which would send the light to the objective. Light does get scattered, and while an image (often quite cool) gets projected, it bothers me that there is a potential loss of contrast and brightness to this phenomenon.
Looking for enlightenment,
Bill. |
|
|
| woneill |
Hi Invictum,
I don't know much about the outputs of modern computer power supplies - I am a few years out of date, unfortunately.
For my experimentation, I have been using a car batery charger to drive the bulbs. It probably has too high a voltage, and will shorten bulb life, but the bulbs I'm using are cheap, and easy to replace untill I get things right.
If anybody is playing with this stuff, check out the reflectors, bulbs, and any lensing setups. Quite often car headlamp fixtures have opaque panels inserted to shape the beam so that oncoming motorists don't get blinded...
The same can be said for the bulbs - some are designed to only send light in one direction. If you play with these, make sure you are using your full output capabilities!
Bill. |
|
|
| tech head |
Aleksey,
Rear screen projection screens are usually expensive. I haven't seen any material that is specifically made for that. And all the info that I have seen on them shows that the entire back of the screen is a frensal lens focused to a certain distance to increase brightness and the front is covered in micro lenses to increase viewing angle. If you can find one, you might try the screen from a rear projection t.v.
Undream,
The other day one of the bulbs in my proxima projector burned out ( don't worry, it has two :-) ) When I pulled the bulb out I found that it is integrated with a reflector. The interesting thing is that the reflector is a "color correction filter" it only reflects about 70% of the light that hits it. The bulb itself has a large coil that is tightly packed. IT IS NOT A POINT SOURCE! which really supprised me. Another thing about the reflector is that it is not smooth. It is a bunch of small rectangle reflectors that make a virtual eliptical reflector. and the bulb is not really powerful. it is 430w 82v.
What I am trying to say is, The cut mirror reflector is probably the way to go. If that is the way the big boys do it we might want to try it too.
Joe |
|
|
| remp |
Woneill
Lack of efficiency in lighting is a real problem not easy to solve for DIY people.
For a start you loose 50 percent light because projection panels work with polarised light. Half the light we supply is wasted.
Modern manuifacturers use polarising recyclers by prisms and wave retarders so can utilise much more light than we can.
Also we use lights that do not have sharp point source so reflector efficiency not so high as manufactured items.
Some of our reflectors are not optimum.
In addition 3 panel manufacturers use polysilicon panels which have greater transmissivity. than projection panels.
3 panel manufacturers use MLA (micro lens array) to direct light parallel through panel which further increases light throughput and increases contrast by reducing stray light.
What can we do. Short of spending a lot of money on polarising recyclers and buying a sheet of MLA and buying a point source bulb and matching reflector, probably nothing.
People do what they can with whats available at good prices and some results are pretty darn good. But I take your point. Without investigation we get to a certain level and then stop.
Maybe we could find out how to use the half of the light that is wasted. Thats a big gain for a start. So we have firstly get it going, then improvements if economic and possible. |
|
|
| eebasist |
ap0the0sis:
I dont have the perfect light source, I was just trying to help you with your suggestion, by all means go ahead and try those, but I don't think you will have much to be happy with afterward.....there is ~3inches of fillament in those to do the 300 or 500, etc watts, the light is distributed too much.
The best light source most likely like in the first set of posts says, is Metal Halide (MH) Ushio actually has some really cool MH assemblies of lower wattage ~150-250 watt that i want to get more info on, I think that they have some serious potential, but back to your bulbs.....you might get a decent effect out of them if you encase them in an old elipsoidal stage light....still not as good as metal halide, but it would work and would focus the light as much as possible |
|
|
| Undream |
| well, i decided I'm going to pickup a glass cutter. I think I've got a big mirror in storage that I can cut up. I might just glue the pieces right on my curved aluminum sheetmetal, I duno yet. |
|
|
| SuperZoboo |
its been a while since i've done any projector work. but we did reclaim our cloth and still hope to build a screen sometime. really, we like our setup as is for the moment and we just enjoy watching movies, playing nintendo 64, and bigscreen soccer games.
funny thing is, ive never watched soccer in my life, but something about watching it 8 foot wide makes it exciting!
anyway, no new info, but i'd still like you guys to submit your results at our site. i think a list of 15 or 20 submissions with good results would be a good collection of information.
Submit you projector details here |
|
|
| Invictum |
Will do Zoboo, as soon as I get a projector, and a panel, and a screen, and a... ;)
I'm not so sure I should go with headlights now. For one, I'm pretty sure that one wouldn't be strong enough, and if the power supply can only power one, it's not really worth it. Even if it could power both, I'd need something to combine the lights etc.
I might screw around with it later today anyway though. |
|
|
| *_if_I_know |
Hey,
I have a suggestion.
Maybe good information could be gathered if someone goes to the patent division in a local library and looks up, “projector,” technology. The details should have parts, measurements, scientific data etc., that we could use with the DIY project. It might be a good step forward. If somebody is willing [I don’t have the time now {soon}], that would be great.
Also a link for an LCD list: http://www.commspecial.com/resolutions.htm |
|
|
| mycamel |
| quote: | Originally posted by eebasist
.....you might get a decent effect out of them if you encase them in an old elipsoidal stage light.... |
That's close to what I'm looking into. After seeing this product:
http://www.premier-lighting.com/tipa/tipa.html
I ordered a reflector and lens from an 8" fresnel stage light from Secoa and intend to experiment away. Not having specific data on the reflector, my fear is that the light source of my big honking 400w MH bulb won't be able to fit anywhere near the focal point. From what I've seen so far, stage lighting fixtures use very small high power bulbs which can fit deeply into a reflector. We'll see once I get it.
I tried to pick up a reflector like marklar's. Cooper makes the Portolio line and sells though Home Depot, so I called them with the model - H7684T. The quote was $181.56 plus $25.58 for the trim ring plus shipping. Youch! Well beyond what I wanted to spend on a reflector!
So, tech head and undream may have the practical answer. What sort of adhesive do we use to allow fairly precise tweaking and aiming of little mirrors inside a form which won't care about heat? Someone a while back had suggested using one of those laser pointers for checking focal points, and that might really help in aligning them in this case. |
|
|
| eebasist |
I ran into the same problem with Marklar's projector. It was just too expensive.......so i decided the ellipsoidal stage light would be better.....the thing is, with a quartz halogen bulb it could be very good, they are desined to use a relativly large bulb fillament. But i probably wont go that route yet since i've got two working projectors and that only need light sources....time to modify stuff
BTW guys i work in a tech transfer office and have access to many differrent patent search tools. |
|
|
| prjctr_builder |
wait a minute, the thing in the middle is where the image that is to be projected is placed, right? but the size of this thing seems to be too small to fit a projection panel....are we still going to use a panel.
and, if we use a panel, then to illuminate it's 10.4" we use say 575W metal halide bulb. but if we use a smaller screen, look at this:
10.4 / 49000 = 5.6 / x where the x is the numberof lumens and the first numbers are the diagonals
in this case we will need about 22000 lumenswhich we can easily get from a variety of CHEAP bulbs.
MY POINT IS>>> do we want to use the projection panel and have to use very powerfull and big bulbs? if at the same time we could use a small bulb on a small LCD...
aleksey:) |
|
|
| woneill |
Aleksey,
If my memory serves, the calculation you propose should be based on area and not on the screen/image diagonal.
Unfortunately, even doing it that way, it still only translates to the projected image if it is the same size of the panel...
Whether you use a 10.4" panel, a 5.6" panel, or a 1.8" panel: if they are each used to project a 100" image at the same brightness, then the total light flux pasing THROUGH each panel will be equal. (The original light source energy needed to achieve this throughput is dependent on the efficiency of the panel.)
It is the size of the image, and how thinly spread the light energy is over its area that is the final determinator in the required output brightness of the projector...
The flashy new little devices with thousands of ANSI lumens are intended to project big images in big bright auditoriums. To achieve the same brightness in a 4' image on your bedroom wall, you will probably need only a couple of hundred lumens.
The reason for using a bigger panel is cheapness/availability, and the possibility that in the bigger panels, the LCD shutter area is bigger with respect to the TFT transistor area, than it is in the smaller panels. (The LCD shutters compete for space on the panel surface with the TFT transistors which are mounted directly on the silicon surface.) If this increased area ratio is the case, then the bigger panels would be more efficient. Unfortunately, their size introduces other problems as we all know...
Bill. |
|
|
| CHARTO |
headlights
When i mentioned headlights in my post, i was'nt thinking of using
the headlight bulb, they are just not bright enough
Im just saying use the reflector
on my car i have two fog lamps on the front spoiler,they are about
6"+4", if you were to get somthing like that and just use the
reflector part, adapt it to take one of those high power capsule bulbs, you end up with somthing like whats in an overhead |
|
|
| Invictum |
| Well, I tried out the headlight...at full brightness, the power supply can't handle it. It only stays on for about half of a second. Also, like Charto said, they're just not very bright. Looks like metal halide is still the best bet. |
|
|
| xblocker |
May be of interest! Here's a pic from 10" single panel projector light unit. Projectors name is 'GP Movin 2001". It's a 4-pack xenon unit with 4x35 Watts. It should last up to 6000h and costs about 76 EUR. Xenon lamps dont produce so much heat and can work with smaller, more silent fans. If somebody is interested:
http://www.gp-elektronik.de |
|
|
| mycamel |
| quote: | Originally posted by xblocker
May be of interest! Here's a pic from 10" single panel projector light unit. Projectors name is 'GP Movin 2001". |
Cool! I'd never even considered an ARRAY of smaller lights and reflectors! This would help keep the light rays going straight through the LCD apertures. They sort of look like light headlight reflectors, don't they? I'd guess they must use a large high quality fresnel after the panel to make this work. Anyone understand enough German (or whatever it is) to know if its mentioned on the site? |
|
|
| tomithy83 |
| quote: | Originally posted by mycamel
...I'd guess they must use a large high quality fresnel after the panel to make this work... |
youd need the fresnal in front of (before) the lcd. so that the rays are paralell, thats the point of parabolic reflectors. scratch that as long as all four reflectors are parabolic it might work??????? someone let me know please!!!!!!!!
anyway here's a site for metal halide bulbs
home depot dont have ballests
i dont know how much you guys are paying but i've heard/read prices hier than this
sorry bout the spelling im on a wave key board its confusing im a pecker ;) |
|
|
| Invictum |
Hey, I just noticed this bulb at bulbs.com:
http://www.bulbs.com/products/produ...&inventory=9722
Now I know this thing is expensive, but otherwise it seems perfect. The color temp. is good, it lets out 60K lumens, it's only 175 watts, it lasts 7.5k hours, it's not too big, and best of all, the reflector is built onto it. Is this a great bulb, or am I missing something here? |
|
|
| prjctr_builder |
| quote: | Originally posted by xblocker
May be of interest! Here's a pic from 10" single panel projector light unit. Projectors name is 'GP Movin 2001". It's a 4-pack xenon unit with 4x35 Watts. It should last up to 6000h and costs about 76 EUR. Xenon lamps dont produce so much heat and can work with smaller, more silent fans. If somebody is interested:
http://www.gp-elektronik.de |
dude, where did you get the picture from?
i wanna research this array of lights!!!
if they are good enough, we will have one hell of a lightsource...
aleksey |
|
|
| prjctr_builder |
| quote: | Originally posted by Invictum
Hey, I just noticed this bulb at bulbs.com:
http://www.bulbs.com/products/produ...&inventory=9722
Now I know this thing is expensive, but otherwise it seems perfect. The color temp. is good, it lets out 60K lumens, it's only 175 watts, it lasts 7.5k hours, it's not too big, and best of all, the reflector is built onto it. Is this a great bulb, or am I missing something here? |
the bulb seems ok, but IT REQUIRES BALLAST!!!!
that's like $60 more. turns out we will use a $160 light source?!?!?!
hmmmmmmm, not a good idea!!!
aleksey;) |
|
|
| eebasist |
| I like the bulb.....60K lumens says it all....... |
|
|
| xblocker |
Additional information about 4-pack xenons: You can find it at the above link and then click 'Beleuchtung', which means lighting.
BTW. this is the last single panel projector i know built up to now. GP-elektonik also builts another type, the GP Movin 2002 with a 4-pack of 4x50 Watts.
I don't know much about the exact optical design, but i suppose, if these xenons have parabolic reflectors, they send parallel rays, which are collimated right before they go through the panel, or direct after panel.
xblocker |
|
|
| Undream |
I dont understand how a 175W bulb puts out 60,000 lumens. Thats 342.8 lumens per watt, about 3 times the efficiency of the highest efficiency Metal Halide bulb that I have seen.
Researching...
a ha. here we go.
http://www.lighting.philips.com/nam...yhid.php?id=124
Watts: 175
Bulb: PAR-38
Base: Med.
Ordering Code: MH175/RSP
ANSI Code/Ballast Ref.: M57/E
Description: †G, Clear, 16 degree Beam
M.O.L. (in.): 5 13/16
Rated Avg. Life Hrs.: 7500
Approximate Mean Lumens: N/A
Approximate Initial Lumens: 60,000 MBCP
CRI: 65
CCT (K): 4100
That site has the specs listed wrong. It is NOT lumens. It is MBCP, a candlepower rating. I would be highly, highly suspicious about purchasing this bulb! I can nearly guarantee you that it won't be bright enough. The *only* thing going for it is that it has a reflector built in. The reason why the candlepower rating is so darn high, is that the beam that this bulb puts off is only 16 degrees.
looking, I see that philips makes another 175W Metal Halide PAR 38, that puts out a 65 degree beam, MH175/RFL
Watts: 175
Bulb: PAR-38
Base: Med.
Product Number: 046677-30858-5
Ordering Code: MH175/RFL
ANSI Code/Ballast Ref.: M57/E
Package Quantity: 6
Description: †G, Clear, 65 degree Beam
M.O.L. (in.): 5 13/16
L.C.L. (in.): N/A
Rated Avg. Life Hrs.: 7500
Approximate Mean Lumens: N/A
Approximate Initial Lumens: 10,000 MBCP
CRI: 65
CCT (K): 4100
As you can see, the changing of the beam from 16 degrees to 65 degrees has taken the MBCP from 60,000 to 10,000.
The efficiency of this bulb is more than likely no more than any other bulb, and, it probably actually puts out at the most, 175*100 = 17,500 lumens. The thing that is cool is that it is extremely efficiently used. But, I still have my doubts. |
|
|
| Undream |
For instance, here is another Metal Halide bulb I found on philips' site.
Description: †G, PAR WISO Spot 15 degrees FadeBlock 54,000 MBCP
Rated Avg. Life Hrs.: 10,000
Approximate Mean Lumens: 4800
Approximate Initial Lumens: 6000
It is a 100W PAR38 that puts out a 15 degree beam, and 54,000 MBCP. Its actual rated lumens, however, is only 6,000!!!!!!!!!!!!!!
CRAPPY, I SAY! hehe. |
|
|
| woneill |
Xblocker,
You are a star!!! :cool: :cool: :cool: Awesome!!! You have opened SO many doors with this find!!!
The fact that miltiple light sources can be used like this without needing to simulate a single point source has SIGNIFICANT implications for everybody to think about, because no matter how good the reflectors are, this arrangement can not produce a single perfectly parallel beam without a significant degree of scattering.
As you mention, some fresnel must be used either in front or behind the panel to cause the emerging light from the LCD to converge on the Objective.
But, whether the fresnel is in front or behind the LCD now has important implications regarding whether it is more important to treat the LCD as a an optical filter in a longer optical pipeline, or, treat it as a completely independent light source itself: a light source whose properties are determined by the evenness of it's back illumination, and the panel's efficiency in using that illumination to produce its image (polarised illuminating light perpendicular to the LCD surface, being the optimal arrangement).
The first thing that triggered me into not accepting the optical pipeline theory, was the original working projector produced by Zark - using LOA bulbs to illuminate a small(ish) LCD panel, and feeding into a Rear Projection TV Lens.
http://www.geocities.com/zark5150/projpicnew.html
It should not have worked at all with the diffuse light from the LOAs if the optical pipeline theory was necessary with LCD panels. (An optical pipeline WOULD still work, but it would not be the only option...)
Anyway, whatever, this find will definitely provoke new research, and could propell everybody to the next level! (Nice, small, light, bright and cheap!!!)
Unfortunately, my subsequent research into these Xenon HID bulbs indicates problems for us in using them: Despite their EXCEPTIONAL price, unless their ignitor and possibly their ballast is included, the total package could become expensive...
Each bulb would need its own ignitor module, and almost certainly its own ballast too.
The best prices I have seen for these ballasts in the US is around $100 each. For four bulbs, this gets a little expensive...
However, what is MOST valuable to me from this is that the TFT LCDs do not need to be driven from a light source that emulates a single point of light...
Interesting...
Bill. |
|
|
| woneill |
Undream,
You are definitely correct in calling the 6000 lumen output from the 175 Watt bulb crappy compared to the 400 Watt beasts.
However, the seemingly accepted efficiency of the LCD panels we are using here is around 10%.
If these bulbs are well focussed and all of the 6000 lumens comes out of the front (unlike most of the designs using the 400Watt bulbs), then with just 10% efficiency in the LCD, we would get 600 lumens coming through the panel...
If we do a cross of Zark's LOA design with the light unit found by Xblocker, and have just two of these bulbs shining through the LCD panel, and pointing towards the objective, then who is to say that the results wouldn't be fantastic???
(I would still shine the bulbs directly at the panel, and put a fresnel between the panel and the objective, but then I am a psycho!!! :eek: )
Bill. |
|
|
| Invictum |
Hey, I have a question about this auction:
http://cgi.ebay.com/ws/eBayISAPI.dl...item=2037292345
How in god's name did this auction for 2 Sharp QA-50s get to 255 dollars? I thought that those were grayscale panels. Who would pay that much for those things? |
|
|
| woneill |
Probably because the people bidding are newbies, and the panels are actually advertised as "Projectors..."
Unfortunate...
Ebay should do a comments/advise thing...
Bill. |
|
|
| xblocker |
Refering again to this xenon lights: Such a design can only work, if real parallelbeams come out of the box. So if someone has now the idea to try a similar arrangement with common cold light mirror halogen lamps or other non parallel lightsources,this might not work, 'cause these have a certain flood angle and a single condensor can't them fit together!
Also brightness isn't kickass, the 'Movin' projectors only reach 200-250 ANSIlumen, although a contrast ratio of 500:1. How they do this ? I don't know. Heard something about a second shutter panel, but no facts here..
xblocker |
|
|
| woneill |
Xblocker,
But the beams coming out of these things CAN'T be perfectly parallel - the HID bulbs don't have a point source spark gap, and the the way the parabolic reflectors are arranged, there will be significant light reaching the (1 st) panel directly from all four bulbs.
I like the thing about using two panels, though - that would definitely eliminate most interference from non parallel light...
VERY interesting...
But... More questions:
1) What about the Zark design? Was it ultimately found not to work? (He seemed to use a large CRT lens, and that could have some bearing on this...)
2) How do the CRT projectors produce a viable image when a CRT tube DEFINITELY produces a diffuse image?
I still think the LCD panel should be treated as the light source by the optics that follow...
If you can explain where I am wrong, I would sincerely appreciate it!!! :confused:
Thanks,
Bill.
P.S. Most car Xenon HID bulbs seem to give 3000 lumens from 35 Watts. 4x3000=12000. This means that the efficiency of their system is higher than I would have expected... |
|
|
| CHRLUC |
Hay guys,
I'm still here!
I read this (part I and part II) and have kept up with it for months.
A few thing I have found (correct me if I wrong)
We need a light with low heat and high output....
Talked to a guy at a Lighting store and we have a few problems with what is available... heat, life of the bulb (that is capable of producing the output we need), size, and reliable and consistent color temperature.
I have to agree that to date it seems that the 400 watt mh is the way to go. However it is far from perfect, it has several problems. The first is the mere size, this thing is really big (respectively), second the power it takes to run it (has anyone see their electric bill and have used this thing as a "main TV." I have not but it but am not looking forward to it. The third is the heat (not a BIG problem it seems) heat comes from the watts the higher the wattage the more heat ... 400 is some heat to deal with. I have a 500 watt halogen (it seems the halogens run hotter, but not confirmed) and, as said in an earlier post, it could also be used as slow roaster. The forth is the cost ($100 in just the light seems too much). The last two are important and have not been addressed, the mh bulb is not consistent in color (mostly over time), I believe that this bulb has not been used long enough (months, maybe even a year or so) to see how it will vary over time. The last is the start up time (time to get to maximum brightness). There are bulbs that have long life, are bright (when used in conjunction), are inexpensive, cooler, and have the correct color temp. The problem seems to be the distortion that mulitple bulb (or light sources cause).......
I don’t claim to be a genius in any way, but why could we not build a projector in a projector.... some thing like this.... |
|
|
| xblocker |
woneill,
i don't say you're wrong, i only would like bring some things to a point.
Exact parallel rays of course is only reachable in a mathematical sense, we can only get near as parallel. So the panel shouldn't be hit at the same area of beams of the other bulbs, paraxial rays should only 'touch' themselves at the outer areas of the beam.
CRT projection optics differ from other optics because, (i try to translate from german, grace!) CRTs are self lighting systems with a large luminance area, which indeed is a diffuse lightsource. Since CRT's brightness isn't so good, the optics must have an aperture as large as possible to catch a maximum of 'image' rays. Important: F-ratio of 1.0 is an optical limitation of objectives, the lens's diameter can't overgo focus length!
This is a classical optical system without condensors and should of course work with LCD panels too, if the sizes match together.
Working with a separate lightsource, you have 2 options:
a) backlight a panel with diffuse light and treat the whole thing like a CRT system.
b)collimate light with condensors through the panel, so that direct light goes through panel and objective. But you can design the beam going to the opjective, so that it hasn't to be so large.
As we all know, b) is the setup of most projectors 'cause it's much more effective.
Must light through LCD panel be parallel? Only if you have more then 1 lightsource (4 xenons). If you have only a single lightsource it isn't mandatory. My OHP+panel give good image. This point seems to be a little overestimated.
What Zark's setup concerns, i don't know much about anymore, saw so much pics over the months...sorry! |
|
|
| woneill |
Xblocker,
We are definitely on the same page now! All the statements I have ever made concerning fresnels (and BIG lenses) were due to my following option a). I use the fresnel a couple of inches after the panel to provide the large aperture for a compound objective.
Doing that, the prime quality needed of the light source became evenness of illumination rather than the perfect point emitter that is so hard to reproduce.
Unfortunately, I wasn't able to express myself very effectively - your English is better than mine! ;)
Bill.
P.S. Somebody seems to have made me a prophet. Shouldn't they have alternative classifications such as "nuissance" and "pain in the a$$"... |
|
|
| xblocker |
| quote: | Originally posted by woneill
Unfortunately, I wasn't able to express myself very effectively - your English is better than mine! ;)
[/B] |
Hehe, that was a good one!
xblocker |
|
|
| jpsexton |
Does anyone know what if anything was wrong with Zarks projector?:confused:
This is the way I'm planning to go right now. I've allready got the Lens and the LOA lights, but I'm kinda lookin for a 5" LCD with better resolution.
I always thought the panel resolution was the only downfall to Zarks projector but the last few posts seem to point at somthing else. |
|
|
| BARBUCHA |
Hi, I have a setupo with Sharp panel and OHP. I used it with the original Halogen bulb in the OHP and after it blew up I changed ii and installed a MH 400W bulb(realy big one). And the problem came. Before I had a beautiful image(a bit redish) and now I have totaly dim picture, unable to focus. I think that the problem is in the position of the bulb(refelctor is OK), because too much light pases to the ceiling(before it was only a little).
Thanks
BARBUCHA
PS: Sorry my english. |
|
|
| woneill |
Barbucha,
Lots of light hitting the ceiling indicates that your light source is definitely configured wrong.
1) What type of halogen did you have: was it an open bulb, or did it have a reflector?
2) If it was an open bulb, and the OHP has a built-in reflector, are you physically able to place the centre of the light bubble of the bulb at the point where the halogen filament used to go?
In the ideal OHP configuration, the light source is a single point of light, at he focal point of the lower of two fresnels. The light from the point source is collected by the first fresnel, and shaped by the second fresnel into a single conical beam of light extending upwards from the fresnel, to the OHP objective lens at the top.
If the bulb is in the wrong position such that the point source is not in the centre of the fresnel, then the light cone will also be offset - causing the apex of the cone to miss the objective lens, and be beamed all over your ceiling.
If the bulb is too close, or too far away from the fresnel, then the light cone will be too tall, or too short, or may not ever reach a point... Again, your ceiling gets the benefit.
Finally, if the light source is in the right place with respect to the fresnel, but has a badly adjusted reflector arrangement, or has too large a light emitting area, then you risk producing multiple overlapping light cones - each with its apex at a slightly differenrt position. (There may be only one light cone, but its apex may be smeared vertically, with no one position that is "the right one".)
In this case, you will get an image with poor definition and contrast, because the objective can only focus an image from the apex of a single light cone - the other light cones will overlay the focussed image with multiple unfocussed ones...
You may have one or more of these effects interfering with your projection.
That said, OHPs are a very clever design, and are much more forgiving than my description would indicate. (e.g. the small aperture of the objective with respect to the LCD image, and its distance from it, allows the objective to ignore all light that is not part of a light cone with its apex roughly at the right position. Only if the apex of a light cone is close to the objective position, does most of the light cone's energy flow through it: too high, or too low, and the light flows past it... The reflector arrangement means that instead of hitting the screen and reducing contrast, all light that does not go through the objective hits the ceiling instead... It is all quite clever!!!)
Thus, your task is to ensure that the light source you use, is correctly positioned and suitably reflected such that all its energy is projected as a single cone of light that hits the objective in its centre. In truth, nothing is perfect, and you can only do your best - if you can position the bulb such that you are satisfied with the image, then your job is done.
Bill. |
|
|
| Gunawan W |
Hi guys,
this already wrote several time before, but I would like to write it again, it's about the specification of projector lamp we should use (correct me if I'm wrong):
Color Rendering Index:
The color rendering index (CRI) scaled 1 to 100. CRI is important in a home because it affects how furniture, decorations, and even people look. Incandescents are considered the best at 95 to 100. Newer Tri-Phosphor Fluorescent and Metal halide is rated above 80, while mercury vapor and old fluorescent are 62 or above. Using CRI with 84 or better gives very little color shift loss.
Light temperature:
If we translate white light to temperature, it's around 6000 degrees Kelvin (sun daylight as an example), below that temperature become redish, above become blueish.
We need white light like sun daylight for perfect projector.
Light intensity:
It measured with lumens, it's still in hot discussion, at the begining 10000 lumens seemed adequate, but recently 20000 lumens still not enough to produce bright projected image!
Arc gap:
short arc gap (shorter than 6mm), so the ideal point source light can be achieved, but now we discuss about more than one light source!
Qoute from this link maybe relevant with current discussion:
Because of the light emitting from the halide has bad uniformity, adds the light is symmetry to the axle of the light reflected from the face of the reflector, it requires a collection system or so called Integrator to uniform the distribution of light, and convert the circular illumination area into a rectangular area as the LCD panel (4:3 or 16:9) as seen in fig.3. The integrator has been used more often now called lens arrays, the first set of lens arrays divide the light reflected from the reflection into several elements. (Rectangular 4:3 or 16:9 distribution), and the arc will focus on the second set of lens arrays, and the second set of lens will add up the position of the first set of lens arrays on the LCD panel to achieve the uniformity. Using the rod glass is another method, using the effect of Kaleidoscope to create array point source, and uniformity achieved.
Anybody can explain this with more simple words?
And this:
There is another special reflector called orthogonal parabolic reflector. It is able to focus the light from long arc lamp into a single point.
What is orthogonal parabolic reflector look-like? and how's to make it?
see you. |
|
|
| woneill |
Gunawan,
The ortho-parabolic reflector was covered in the first thread. It is like a hybrid between an eliptical and parabolic reflector. It can convert a long filament/spark-gap light source into a point source, but I would suggest that it is not going to be efficient in practice, and it would be very hard to make.
(Based on the same premise as the ortho-parabolic, a simple cone would be all that was needed to produce a parallel beam from the same long filament/spark-gap bulb. You can imagine the losses...)
As for the collector/kaleidoscope stuff, it is hard to say!!! A kaleidoscope is an interesting way to enlarge a light source, and we have seen earlier that multiple point sources can be used. But, the details mentioned here can only be guessed at!
Bill. |
|
|
| Invictum |
Heh, a few of the things you guys are talking about just fly right over my head.
Hey prjctr_builder, did your friend decide if he wanted to sell one of his panels yet? |
|
|
| prjctr_builder |
no, my friend decided that he would keep the QA-2500. the reason is that he already had a Proxima 944+, but after he tried out the QA-2500, he said that there's no way he will sell it. he said that944+ is a piece of cr@p, and he will list it on ebay soon. he is sorry, but he can only sell the items on ebay, he hopes to make more money...
thanks
aleksey
P.S. fact of the day... QA-2500 has a contrast ratio of 150:1, nice , huh? |
|
|
| prjctr_builder |
i did some research on the net today and got to www.digital view.com
there i found their serch screen, plugged in all my panel requirements:
TFT Active LCD
XGA
less than 8"
and here's what i found.
a toshiba LCD panel, you can find it under id of LTM06C310
this LCD is only 6.3" and has an XGA resolution?!?!?!
omg, if it is as good as it seems to me, this is THE lcd for us.
i will call them tommorow, i would call now but i think that they are asleep at 2 am...hehe.
all i need to know is the type of controller, contrast, and response..
can't wait
aleksey |
|
|
| Undream |
They don't make a controller board that works with that panel. Plus, they don't sell panels. They only manufacture controller boards for panels.
The *only* panel in their database with XGA resolution that is less than 12" diagonal that they make a controller board for is the Toshiba LTM10C306L. it has a 10.4" diagonal.
This is the controller board that works with it.
http://www.digitalview.com/products...erModel=SV-1024
Anyways, I don't really think you've found anything we havent before. The controller boards are more than likely uber-expensive, and then your task would expand to actually tracking down a specific model bare LCD panel, which, could possibly be a pain in the neck, and also expensive. |
|
|
| daveb |
Yes, indeed, I've looked into this back in April, when I thought stripped-out laptop displays were "it". I've sourced the controller board locally with one of the Digital View Resellers (Bell) and it's about $450 (Cdn). The good news is that it seems to drive many mainstream displays, including the Sharps, which seem to dominate the market (I know they're in many of the Dell Latitudes).
On one hand, $450 is usually more than the entire budget of the DIYer. On the other, if you're willing to go the distance for a high quality DIY setup at (say) 20% of the cost of an off-the-shelf system, you can justify it. (Given enough contemplation, you can rationalize just about anything...)
Still, it's a big bucket of money, and I'd feel better if I've heard a glowing report from a DIYer about its success. But I don't think anyone here is ready for that outlay and gamble. At lest, not yet. Frankly, if the money just "appeared", I'd do it. I'm as passionate (though not as vocal) about achieving XGA resolution in the set-up as Aleksey is. (Oh yeah, there's also the issue of finding the ribbon cable extender to move the controller board out of the way, but that's a bridge to cross later.)
We can always rely on the "Technological Fix" philosophy: Wait around long enough, and "they" will solve the problem before it gets critical. Who know, eventually someone's bound to make the perfect lcd available.
In the mean time, it's back to stone knives and bearskins.
dave. |
|
|
| Blockhead |
Hello to all...
I have been following this forum for about 2 months now and I have finally made up my mind about doing this. But I do have one quick question before I set my plan in action. I have notice that most people have gone the route of PC connected to panel. What are the problems with running s-video or RCA from DVD player to the panel instead of using a pc. My dvd player is not progressive scan so am I looking at having color issues or Macrovision issues. I guess to simply put it...would I be able to use the popular Nview Spectra C connected to my dvd player via S-video or RCA...I really do not want to go the route of PC dvd playing. Can anyone give me suggestions as what would be a good panel for s-video or rca and what it would take for me to use my dvd connected straight to a panel...
Thanks guys in advance...
-BLOCKHEAD |
|
|
| icculus |
| for that kind of money, you may as well buy a kit from earthlcd. i've been contemplating it, but it's a lot of money to me, even if it is worth it. i'm sort of hoping someone else might be the first to take the plunge. BTW, i found some acceptable ribbon cables inside a busted DVD player. i can now move the controller board about10" away from my panel. scavenge and recycle--something is bound to work. |
|
|
| Gunawan W |
thank's woneill,
if ortho-parabolic reflector is hard to build then forget it.
Back to multiple light source, I draw rough sketch below, if it's correct, then it's easy to make. The setup using small lamp, reflector and lenses, and must cover LCD panel size. lens array function is to produce parallel rays. |
|
|
| icculus |
| composite(rca) and s-video are essentially the same. if your not willing to go the pc route, you're never going to realize the full potential of this project. macrovision sucks. it's annoying as hell, and gets even more annoying on a bigger screen. so you need to get a device to remove the signal. that's betwee$50-$100. without progressive scan, you're really only getting half the resolution. with composite/svideo you also need to use very high quality, shielded cables to carry your signal, or you lose brightness and overall image quality. the pc is the way to go. beleive me, i've spent hours pickling around with cables and DVD players and wasting money. the picture quality from the pc is much better--noninterlaced and no macrovision problems. i'll try and post some comparison pics at some point--ive already disconnected my dvd player. i think i'm going to put it on eBay and use the money towards a tv card for the pc. |
|
|
| Invictum |
| icculus, I have a TV card, Leadtek Win TV 2000 XP to be exact, and it isn't exactly that great. Now I haven't tried other TV cards, but this one doesn't even look that good on a 17" monitor. I mean my generic 19" TV looks noticeably better. All I'm saying is see what these TV cards display before getting one. Don't get burned like I did. |
|
|
| Blockhead |
Thank you so much for your quick response. I can go the route of PC dvd but the reason for my hesitation is how do I utilize my surround sound receiver? Do I just connect from my sound card out to my receiver? Currently I'm connected from dvd player via fiber optic to my Receiver. So will I loose sound quality from PC to Receiver...This has been my only reason for not wanting to go PC route. Sorry for my newbiefied questions. And once again thanks for any responses.
-BLOCK |
|
|
| BARBUCHA |
WONEILL
Thank you. I solved it. The bulb wasn't in the center of the frensel. Now the picture is SUPERB. I can watch films in a darkened room which wasn't possible with the original halogen bulb. The MetalHalide bulbs are absoulutely the best source of light. |
|
|
| Axeman |
Blockhead-
All you need for surround sound is a sound card that supports digital out through coax or optical. These can be gotten for ~$50. You will also need a dvd player that supports SPDIF out, like winDVD or powerdvd.
Invictum -
Before you give up on your TV tuner make sure you've tried Dscaler. |
|
|
| Blockhead |
You will also need a dvd player that supports SPDIF out, like winDVD or powerdvd.
I'm going to go the route of a panel on Overhead Pro. with my pc that has a dvd player...Why would I need winDVD or powerdvd...over my current dvd software. Guys I'm so sorry for my stupid questions! I'm just trying to get my plan together and since it looks like I ought not go the route of using my current dvd player and concentrate on using one of my PC's I just want to be sure I have it set up correctly.
Thanks for your replies. |
|
|
| eebasist |
Gunawan
would multiple bulbs produce hotspots? |
|
|
| xblocker |
Invictum,
Dscaler is a good choice to tune image quality!I also have read in several HTPC forums, that a Radeon 8500 is actually the best graphics card for use with projectors. As many people may not know, stuttering DVD and TV is an immense problem even with commercial high end projectors, if they are driven through VGA input.
xblocker |
|
|
| multiplexor |
Hey all, i'm not dead or anything....... ;) just very busy recently :(
I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher?
Also, i'm wondering if (some names on this site are technical) it would be a good idea for those who aren't sure what technical words mean, to setup a thread or something where you have an tiny explanation of a technical word?
Name: ortho-parabolic reflector
Description: I wish i knew... :)
This forum went from decently simple, to using very technical terms :) hehe Not a problem for me, i think it's great. But I was curious if people might be interested in a kinda dictionary of terms thread?
EDIT: I noticed i'm a smilies addict.... :D |
|
|
| Invictum |
| I tried DScaler, and messed around with the options. It definately looked better, but a TV still looks better than this thing. I'm not sure if it's interference or something, or if this card just sucks. |
|
|
| xblocker |
Invictum,
there could be several reasons:
1 splitting of antenna signal isn't correct
2. drivers of TV card
3. fine tune of tv stations
4. antenna cable lenght or quality
I drive my WinTV card with Connexant drivers and have good image qualty, in any case better than on regular TV screen.
xblocker |
|
|
| Invictum |
| Hey, can I ask where you downloaded those drivers from? The ones I got from Leadtek sure don't seem to work very well. |
|
|
| eebasist |
dscaler should look awsome if properly tweaked. Actually i heard on avs that the radeon 7500 is best because there are some driver issues with the 7500.
Dscaler looks better than tv because you actually can get 480p and not the 480i that our tv's have |
|
|
| woneill |
BARBUCHA,
Well done!!! :D :D :D
Gunawan,
Your lenses would be needed if the reflectors were spherical. If they are parabolic, then the rays should already be parallel enough... (See the light array found previously by Xblocker!)
If your approach is the same as mine - to treat the LCD as a CRT projector - then the two prime reasons to get the rays going parallel are:
1) Maximise throughput through the LCD;
2) Stress the fresnel as little as possible by using it the way it was intended - to converge roughly parallel beams to a focal point.
This approach should maximise brightness and clarity. Your original suggestions for the placement of the fresnel & primary objective should work optimally.
I would use as long a focal length fresnel as possible, placed as close to the LCD as will avoid moire patterns, and then choose the best objective lens you have with a focal length that gives tha appropriate zoom for your room when combined with the fresnel.
The closer you get the fresnel to the panel, the less the fresnel optics will be stressed: the majority of the light from a given pixel will only be deflected by a few of local points on the fresnel, rather than its whole surface.
In support of this, the more parallel you can get the light beams entering the LCD, the better.
I think that with a more diffuse light source, you may be able to put the fresnel closer without moire being a pain, but I don't think this would give optimal throughput for the LCD, and the emerging beams might not be good for the fresnel quality.
The natural slight diffusion through the LCD panels will be enough for a good fresnel, and will reduce moire.
eebasist,
I think that he bulbs, if suitably arranged, will not produce hotspots. If the LCD is being treated as a giant CRT projector, then you could live with the natural slight diverging beam that any non-perfect parabolic reflector would produce, and place the lights such that the beam patterns overlap enough to give uniform illumination.
For multiple small bulbs, the best pattern is hexagonal - it is the closest way to pack circular objects, and would reduce the intensity differences between the center of the beam, and the gaps between them.
Bill. |
|
|
| woneill |
Gunawan,
Two extra thoughts - related to protecting your LCD:
1) Make sure any bulbs you use are UV filtered;
2) If you are using halogens, would it be worth trying that IR-blocking film posted earlier?
Bill. |
|
|
| eebasist |
| Bill, the problem i see is that the analogy to the multiple bulbs of a crt projector is slightly incorrect. I have a crt projector right now. (well 3 of them). The beam patterns overlap when you manually change where on the crt face the electron gun fires the image. Since he will be placing the bulbs with no real adjustments, i still believe that hotspots will occur. |
|
|
| prjctr_builder |
got my mini website up and running.
if you click on MY projects, you will learn how to make a DIY projection screen for $27.
http://www.geocities.com/predator3do/index.htm
ok, check it out and tell me what you think
aleksey |
|
|
| woneill |
Hi eebasist,
What I was actually thinking of was that the circles of light produced by the beams from the bulbs hitting the back of the LCD panel would overlap slightly - creating a fairly uniform illumination of the back of the panel.
My reference to a CRT projector is to the concept of treating the illuminated LCD panel as a light source - similar to a single tube from a CRT projector (in this case, a very large RGB tube...)
In the same way that a CRT emits diffuse light which is captured by a large lens, and projected onto the screen, the LCD panel will emit a (very) slightly diffuse light when driven by Gunawan's multi-bulb arrangement, which would be captured by a fresnel and fed into a more normal objective to be projected onto a screen.
This arrangement makes much less strenuous demands on the optical point-source quality of the light source/reflector, but, if correctly set-up, with a decent fresnel, gives a high brightness, high quality image.
I did not mean that the bulbs would each project their own portion of the image. :(
Sorry for the confusion - as usual with me: more words, less meaning... :rolleyes:
Bill. |
|
|
| woneill |
Aleksey,
Your screen setup is VERY nice - impressive!!! :cool: :cool: :cool:
Bill. |
|
|
| Gunawan W |
Guys if you want to know more of HTPC (home theater personal computer) setup, please have a look this site, you can find and learn from simple HTPC to sophisticated HTPC.
Woneill,
the purpose using convex lens in front of each lamp array is to produce parallel light rays.If you only use parabolic reflector, the parallel rays only come from the reflector, direct rays from lamp will spread wide in front of the reflector and wasted, unless you cover the front half of the lamp to bounch the rays back to the reflector (same as car headlamp).
All parallel rays will pasthru the second fresnel and hit in the same point. (=fl of the fresnel). If we put objective lens at that point, we will produce sharp projected image (with good obj. lens of course)!
Diffuse light will produce blurry projected image.
Simple test can be done with cheap laser pointer to show it.
Please see attached sketch below:
eebasist,
IMO hotspot will generated by light passthru LCD panel not at 90 degrees to the panel, it depend on how strong the light intensity, LCD type and how much non perpendicular light hit the LCD (placed the spotlight near the side of the LCD panel and spot the LCD in 30 degrees or more, you will see very dimmed LCD screen compared with if you spot in 90 degrees angle).
This link about LCD, maybe good as a reference.
seeyou. |
|
|
| woneill |
Hi Gunawan,
Your diagram is correct as far as it goes - but you stop at the objective, and didn't follow through to the projected image. (I am being lazy myself because I am terrible at drawing diagrams...)
You are absolutely correct that light coming directly from the bulb will be diverging. If you use the MR11/MR16 bulbs with built-in reflectors, then they have transverse filaments, as do some car headlamp bulbs H3 etc. and there will be much light coming from the front of the bulb.
If you use the bulbs with longitudinal filaments, then this is much less of a problem. The question you must ask yourself is whether the light from the filament is more important than the light from the reflector: which plays the biggest part in producing the image?
You could use a diverging fresnel lens within the reflector itself to correct this and produce a single point source, but it gets very messy!
Within limits, though, the simple arrangement should not be a problem. My reasoning is this: If you treat the LCD as a light source (the CRT projector concept), then with a single objective lens, the second of your lens diagrams on your website applies. (Multiple paths of light from the same point on the light source get focussed by the lens to the same point on the image.)
When you put a fresnel panel between the LCD and the main objective, you are not really changing anything from an optical standpoint - the fresnel and primary objective can be treated mathematically as a single compound lens. So again, the second lens diagram on your website applies. The main reason for using the fresnel here is to ensure that as much of the light from the LCD reaches the main objective as possible.
Because of limitations of the way fresnels operate, and because of the small size of the primary objective when compared to the LCD/fresnel, the effect you illustrate on your diagram here is important - wildly diverging rays coming through the LCD will never reach the primary objective, and their contribution to the final image will be lost. This could be due to scattering by the fresnel, or could be that the rays are just too far off the mark.
Luckily, any significant divergence can be limited by the reflector/focussing arrangement on your light source. The LCD will also introduce a little divergence of its own: subsequent research with a microscope on a couple of my panels has indicated that diffraction effects and surface unevenness within each shutter cell are playing a much bigger role than I ever expected. (Despite my prior claims, MLAs are NOT used on the panels I have investigated so far!!! My bad!!! :( There is still a diffusion effect, though... More on this, later...)
Anyway, ideally, due to the nature of LCDs and the limitations of fresnel lenses, the light coming trouogh the LCD should be as parallel as possible, and perpendicular to both the LCD and fresnel.
This will get the maximum amount of light to the primary objective lens without wasting too much, either lost in the LCD, scattered by the fresnel, or plain missing the target...
Diffuse rays from the LCD that DO hit the objective, will not necessarily cause a blurry image, because those rays that make it through the main objective lens SHOULD be focussed by the fresnel/objective combination to the same point on the final image. (Ok, fresnels are not perfect, but if aligned per your original specs, the result is not bad).
This is why I was refering to the CRT projector concept: the diffuse light output from a CRT tube is focussed through a large lens into a sharp image. This is because while the light emitted from the CRT is very diffuse, the image on the CRT which is emitting the light, is sharp. This diffuse emission is why CRTs have a wide viewing angle, and LCDs dont. It is also the reason why you can actually watch a CRT screen - some of the diffuse light from the screen is projected onto the back of your eye into a sharp image.
The fundamental limit on this concept is the quality of the fresnel. The more diffuse the light coming from the LCD, the more work the fresnel has to do in resolving it into a sharp image, and the more its limitations affect output quality.
As you have seen, though, when treated nicely, the fresnel can cause more benefit than harm.
Bill. |
|
|
| prjctr_builder |
DOES ANYONE HAVE ANY IDEAS ON THE LCD PANEL THAT WE MIGHT BE ABLE TO USE IN OUT PROJECTORS???
please let me know...
aleksey |
|
|
| xblocker |
Woneill,
the confusion is still here!
For communication there should be clearity about terms we use.
A lens between an object and projection lens in optics usually is called a 'field lens'.
LCD isn't a diffuse lightsource like a CRT! You can treat it like that, but the results are poor! You said it yourself, viewing angle is not the same as CRT. Even if a diffuse backlight is attached, e.g. laptops. As we know, view is much brighter on the optical axis. I don't say, there is no diffusion of the panel, but we have the fact, that there is a lot more transmission than diffusion. And transmission is the more important part of projection! That's also the reason, why panels work much better with direct light than diffuse light. To much light would be wasted, or you would need an overkill of light.
And, if LCDs were as diffuse as CRTs, why aren't field lenses in CRT projectors? They then should do the same improvement there as in LCD systems. But they don't exist!
Also there is a change of optical system with a field lens. If the FL of fresnel is to short, the objective could run into troubles not to refract properly the 'cone' angles of the lightbeam.
The task for every 'projectorist' remains, to get out most efficient light of these damned panels!
xblocker |
|
|
| fender4 |
Great job with the screen, Aleksey! Thanks for the good description and tips.
-f4 |
|
|
| woneill |
Hi Xblocker,
I will try to be more precise.
CRT projectors don't need a field lens because their objectives are generally large compared to the image they are projecting, and from what I have seen, they use a compound objective with multiple lens elements anyway.
I actually viewed the fresnel field lens as part of such a compound objective (but with a much larger aperture.)
I agree that a LCD is nowhere near as diffuse as a CRT.
My choice to treat it like a CRT is similar to the philosophical choice between front and rear-wheel drive. They both work in their own way, and both have their own limitations and compromises.
For me, the choice of putting a fresnel between the light source and the LCD is equivalent to the rear wheel drive choice: it involves hard work in producing an efficient point source of light with the right beam characteristics. Usually, the beam is not completely uniform, with greater intensity towards the middle of the beam. This beam is "pushed" by the fresnels through the LCD in a conical beam towards the objective. In this arrangement, the further away the light rays are from the center of the fresnel, the more shallow the angle of the light passing through the LCD. The more shallow the angle, the less efficient the LCD becomes - reducing the intensity of the image away from the centre even more. The LCD is slightly diffusive (though, I agree, more transmissive) and some of the light gets scattered. Any light that does not hit the objective is lost to the image. This arrangement works, and is used in commercial OHPs and is used by OHP LCD panels!
The other arrangement, with the fresnel between the LCD and the objective, I liken to front wheel drive: Much less strenuous demands are placed on the light source conforming to a single point source with the right beam profile. Instead, the emphasis is on producing an even illumination of the back of the panel with the beam (or beams) of light having "roughly" parallel characteristics. Some slight divergence allows beams from individual bulbs to be overlapped to reduce hotspots in a multi-bulb arrangement. The light emerges through the other side of the LCD, filtered to favour those rays roughly perpendicular to its surface (hence the reason for a roughly parallel beam). The viewing angle is a few degrees in all directions, so, while it is not vital to get the rays dead straight, a diffuse backlight will not be very efficient. On emerging from the LCD, the roughly parallel rays are "pulled" by a fresnel field lens into a cone of light that beams towards the primary objective (or second lens in a compound objective). This large fresnel captures virtually all the light that emerges roughly perpendicular to the LCD (they are very compatible technologies in this regard). Light that is very non-perpendicular to the LCD is totally lost to scattering within the fresnel. The fresnel has very restricted operating parameters, and must be placed very correctly to reduce horrific distortion, and also limits the flexibility of where the main objective can be placed. In order not to over-stretch the ability of the objective to work within its limits, and to reduce distortion due to the fresnel, the longest focal length possible should be chosen for the fresnel. However, the light cone produced by this arrangement should be similar in characteristics to that produced with the fresnel in the more normal OHP configuration. Thus, any objective that would normally be used for OHP projection should not have any problems here. This arrangement also works, and was used in commercial large panel projectors during the mid 90's.
For me, it is easier to place the fresnel at the optimum position, and worry less about the light source than it is to get the light source perfect, and worry less about the fresnel.
I agree that the task to get the most out of these panels is still in front of us!
Any clearer? :(
Bill. |
|
|
| xblocker |
Woneill,
thanks for further explanation! Field lenses and CRT: Once upon a time...there where CRT projectors on the market, which only used a single color CRT, similar to this '100"TV' thing. The ojectives were large, but nevertheless smaller than CRT size. I had such a monstrum and tried always to increase brightness. Also had the idea with a field fresnel lens. But no improvement of brightness. I came to the conclusion, that only micro lenses attached directly in front of the phosphor layer of the screen could send all the light into one direction. But i never saw a patent or development on this.
Look at the 100"TV ****. If there would be optically a way to improve brightness with a field lens, they could easily give a second lens to the package..
Back to business: The light loss in parallaxial areas of lighting systems is a common fact, if parallel light or not, which is due to optical properties of real lenses. Even with modern MLAs there isn't a 100% uniform illumination!We could be happy if we reach 85% with DIY resources.
I agree, that it would be better to send light parallel through LCD, i'm only a little sceptical to bring a fresnel into the image path. Fresnels are good in lighting path, but in the image path, image can't be better than fresnels quality, although knowing not to focus on it!
Cya
xblocker |
|
|
| prjctr_builder |
:) To all who might be intersted.
my friend has a projection system combo for sale.
Sharp QA-2500 Projection Panel & Dukane Quantum 680 Overhead Projector
these are top of the line products that provide excellent image quality if combined.
The set is for $430. This may seem a lot but it is well worth it. You get an XGA projection system that can take various video inputs.
Please contact me if you need any information....
aleksey:cool: |
|
|
| woneill |
Hi Xblocker,
I agree that putting a fresnel in the image path is a scary prospect, and that they are definitely not up to being the main projection lens.
However, if you use a fresnel in a strictly controlled way within its "magnifier" range, then it doesn't do too much harm to the image.
To be more precise: the light source is placed closer to the fresnel than the fresnel's focal point, and the viewer (eye/objective) is also placed close to the fresnel's focal point on the other side of the fresnel.
As with any magnifier, the closer the object being magnified is to the lens, the less it gets magnified. Also, the less it gets magnified, the less it gets distorted.
For a fresnel, being used to bring the parallel beams of light from the LCD into a cone, the least distortion would result from placing the fresnel against the LCD itself. Unfortunately, then you get moire effects.
If the light emerging from the LCD is slightly non parallel, then moving the fresnel away by about an inch or so, reduces moire (light from a single pixel is then being bent by more than one point on the fresnel). This is just enough to remove moire without destroying image quality. If the light were perfectly parallel, then distance between the fresnel and the LCD would not matter - there would always be moire effects. (This is not intuitive, but it IS true!)
(The fresnel is effectively a set of concentric rings of micro prisms, and the moire effects result from light from a single pixel being scattered as it passes through a boundary between two of these rings.)
If the objective lens is placed at the apex of the cone produced by this arrangement, then it will have all the benefits of a normal OHP projection setup, with only tiny distortions introduced by the fresnel, but with a couple of additional cool benefits:
1) The light source does not need to be emulating a point source (hard to do efficiently with the big MH bulbs);
2) Any distortions introduced by the LCD will not affect image quality as they would in the OHP setup, because you are focussing on, and not through the LCD;
3) The image intensity can be made much more even.
Of course, the fresnel needs to be good, with lots of grooves per inch, and preferably with a longer focal length.
But, it DOES definitely work!!! My prime unit is now that modified DP9100 I keep talking about, and it uses this arrangement. Its internal optics generate an almost parallel beam from a 575 Watt MH bulb (the type used in the Dukane 680, and Elmo 305), they beam that onto the LCD, and use the exact arrangement I have been describing.
The image quality is beautiful, even with the reduced resolution of the Spectra C that was used to replace the original dead electronics (1024x768).
With an 8' wide image, I can focus it so clearly that the boundaries between the individual RGB shutters is CRISP.
This is not just another one of my whacky theories!!! ;)
I can be an obnoxious pain at times, but I am genuinely trying to contribute to a project whose success has knocked my socks off!!!
Bill. |
|
|
| multiplexor |
| quote: | Originally posted by prjctr_builder
:) To all who might be intersted.
my friend has a projection system combo for sale.
Sharp QA-2500 Projection Panel & Dukane Quantum 680 Overhead Projector
these are top of the line products that provide excellent image quality if combined.
The set is for $430. This may seem a lot but it is well worth it. You get an XGA projection system that can take various video inputs.
Please contact me if you need any information....
aleksey:cool: |
hmmm, how do we know it's not yours, and you're trying to sell it?
;-) hehehe just messin around.
again, I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher? |
|
|
| mycamel |
| quote: | Originally posted by multiplexor
again, I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher? |
If a body happens to have a plain old-fashioned light meter, could the lumens of our setups be measured? If we could start doing some measurements and share them, we might be able to make our efforts more scientific and productive rather than subjective. |
|
|
| biteon |
| this is what i got so far. ive been waiting for my low-e glass for ever. but the guy end up hooking it up with 1/8" low-e glass just till my 1/4" comes in. there you'll see my box mirror reflector and maybe alittle of the cooling sys. . if you look at the left and right side of thereflector you'll see 4 rectangle holes, thats for the cooling system.i got more pictures but i only no how to put one at a time. |
|
|
| biteon |
| this is the hole where the light bulb is and also where the hot air comes out of the reflector to the exsost hole. |
|
|
| biteon |
these are the air intake fans. which go into the tubes then the air
shoots through the sides of the reflector then hit off the low-e glass(that i havent put in yet) then through the light bulb hole the out the big rectangle exsost hole above the fans. |
|
|
| biteon |
this is my f*$ked up lcd between the two boards and my controler chip,the box is to go over the chip to keep it cool..
if you got ?'s ill answer them later cause im working right now. but go ahead and post your ?'s.this area of my box is going to soon have 2 fresnels and the low-e glass by the reflector. each layer of fresnel and glass is gonna sepparrate the box of insulation and heat provention reasons. |
|
|
| biteon |
| this is my plan once again from page #7 at this thread. |
|
|
| prjctr_builder |
| quote: | Originally posted by multiplexor
hmmm, how do we know it's not yours, and you're trying to sell it?
;-) hehehe just messin around.
again, I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher? |
NAH!!! I do not wanna sell my system. i am pretty much satisfied with it for the money. And if i did want to sell it, there is no point to hide behind someone else.
BITEON - what type of LCD are you planning to use, and did you calculate all the focal lengths and stuff?
aleksey |
|
|
| biteon |
lcd= 930~x700~ res. contras=? im guessing it sucks but then agian it dose look that bad to me. i think i bought it from star audio form e-bay for $130.00 thats with a remote.
i think focal calculation are not needed when the projector box is almost three feet long. i just end up using two fresnel lenses to bend / consentrate all the light on to the lcd. kinda like when burning stuff with a magnifie lenses, you consentrate all the light to put all the heat and light on one spot. |
|
|
| prjctr_builder |
hehe, 3 feet , that's huge!!!
aren't you gonn a try to make it smaller.
aleksey |
|
|
| biteon |
why it just makes my goal harder to get to.
all i want is a good big pictrue on my wall. plus the porjector is going to be mounted from the clg so it will be out of the way. its going to look good cause im making it, and if the fans are to loud ill just pump up the volume.
also here is a picture of my lcd before i took it apart.
also in the last post i ment "dosen't look that bad to me"( the lcd projection) yes ive already got it to project a pretty good picture with just a sheet metal reflector. no i havent tryed the mirror reflector. |
|
|
| prjctr_builder |
what is the diagonal of the LCD?
al |
|
|
| biteon |
| the spy tube how dose it work?. is it like a optic tube?.optic tubes have some sort of jell in them i think,well i think so because some one who lay fiber told me that fiber optic has gell in it. maybe that gel can help us with the cooling system. prjtr builder whats your cooling system?. |
|
|
| prjctr_builder |
i use am overhead for now, but i plan to use the setup that will look something like urs.
aleksey |
|
|
| biteon |
| did your ohp come with a cooling system?.and if so did it work out for you?. |
|
|
| woneill |
Biteon,
Is that a 400W or 1000W bulb? It's enormous!!!
:eek:
Bill. |
|
|
| prjctr_builder |
| i have a Dukane Quantum overhead, and it does have 2 fans installed. but after about an hour you can see the heat f | | | |
