top 40mm

above 2 pictures are tilted focus box you can see there is 4mm tilted

I also thought that the new fresnels were a tad thinner - this may be where some of the extra light/brightness is coming from - anyone to confirm (hint hint Brain!!!)

Well I tried to test the fresnels last night with the stickgrip method and had no trustworthy results to report. Today I used a window about 25' away and did the test. It looks to me like the 220mm/317mm figures are correct. With the longer fresnel I was able to get decent focus in the 300-310mm range(venitian blind slats make good lines to focus with). However, even though I slightly lost some apparent focus at 317mm, this was the exact point at which the rainbow abberations converged the best. So I am thinking that 220/317mm is correct.

Do me a favor and pull the blinds up and focus on the light coming through the window. Focus on the objects outside the window. This will ensure parallel light rays. If you have already put them in don't worry about it, but I would be interested to see if you get the same thing.

The blinds were fully open so I could see outside clearly.

Ok thanks. Hate that my little experiment caused you all the trouble. Is it possible that mine is 280? I'm like you, the fact that the first fresnal was 220 seems to add validity to this 280mm measure I keep getting. I'll do it once more, eventhough I know I'll get the same thing.

Deathray,
What went wrong with the stickgrip method?
It's probably in the interpretation. - A picture might help- eh
I was able to get it within a couple of millimeters.

I will try again and refine the technique with my new LL fresnels that arrived last week. The new electronic ballast works the treat but the light output on the d400 is a bit low.
I'm also working on a new focus method to increase total light collection.

Exciting times ahaead

Kirk

I would say user error. I was doing it from memory. I tried it with a little maglite with the reflector unscrewed and the flashlight set in the reflector housing like a candle. It made a great point source, but it was too small of a point to be able to see. I tried moving the fresnel until the edges of the projection were flat and I kept ending up with about 220mm for both lenses. I know that I was doing something wrong. I should have stated that I doubt that the method was flawed.

Hi Brainchild,

Just signed up to the site and looking forward to building my own PJ one I finish with all the details.

Question regarding the pivoting fresnel? Could you not just pivot it near the top of the lens, using a weight attached to the bottom to keep it parallel to the screen? This would eliminate any chance of the top of the lens getting too close to the lcd screen. The bottom would basically do all the moving. A simple line level attached to one side could also confirm its position.

Just a brainstorm for the brainchild!



oops! Look like I found the answer to my question on another thread! LOL!!

me bad..

Hey, guys, how do I "pivot" the center of a fresnel?

Some glass door hinges work great!

To see how I did it have a look here.

http://www.lumenlab.com/forums/index.php?s...opic=2394&st=20

Hello everyone

I'm getting nerves with this Project i've been lookinag at lots of (Lumenlab) forum's over the last few days and must say despire all the infomation here I feel confused.

All I want is the best. Is it possible to have a "state of the art" page on the main lumenlab site??? I would like to know the BEST technique for dealing with ALL the problems. I know this is the height of laziness but this isn't why I make this request. I don't want to do anything wrong that is mentioned on the site... but where?? I want to know ( perhaps an annexed version of the plan's is a good idea??)

I especially want to know what the 'tolerences are in respect to max distance between Screen (Wall) and Triplet lens) and angle of Keystoneing (horizontal and vertical)

Sorry if this seems like I am being rude or lazy

P.S perhaps a Aluminium top cover would be good? Alu displaces heat better than steel and much better than wood! Also perhaps put the Exhust fan above the LCD panel??? this would extract heat better as heat rises anyway ( I think) and how about a tublar fan (for even spread across the panel)

Dave McMahon

This would be pretty bad on the LCD since it would be pulling the hot air from the light source across the LCD causing it to get pretty darn hot.

to m0nket. I understand, but I think it to be the most efficent way of getting the hot air out from the box as quickly as possible. There would also be a constant flow of air on the LCD therefore cooling it (almost) directly. The only way to know for sure is to test!! And as we know Brain's plans work, so this is not required anyway, But it may be better. Perhaps a bigger (top mounted) fan that 'Spans the IR glass , fresnel 1 LCD and Fresnel 2 would be better and quieter (the bigger the quieter) but... this would make "light proofing" harder!!

Dave

I would say go for your own ideas but remember there has been many, many projector builders gone before you and followed Brains basic concept with variations and built very successful projectors.

I am not saying your idea sucks (no pun intended) but you need to read as many of the forum posts as possible. Your idea has been discussed before and has been rejected.

If you have a look at my PLOG I do not have any light spillage what-so-ever and my LCD sits at a constant 33 to 35 degrees C about 96 to 97 degrees F. It may take you a while to go through it as there are 100 posts in it.

http://www.lumenlab.com/forums/index.php?showtopic=2394

I haven't shown anything about stripping the LCD as this has been done by others before me.

My advise read as many posts as possible and be enlightened.

'k will do thanks.

Dave, i'm with you about the "prime techniques" thread.. . i've spent literally days and days (ok i have an ez job... what of it) reading the threads, trying to pick out cool ideas, or what techniques seem to deliver the best results.

as one builder notes this can be hard since "best" is very subjective, (whats best to you is something only you can really decide) so i dont know the "best" way to make such a thread, lol though i like you wish it could be done. It is very arduous to wade through pages after pages of forum threads, and frustrating when looking for an answer and having to hop over off-topics, or theory debates and stuff like that when what you want is DIY info, how-to's etc. Which is what i believe the site is about .

i've found it most helpful to simply look at the project gallery and project logs, note (even bookmark) your favorite ones and look at how they did what they did. If you think their results are excellent, then go for emulating those results and you'll have an excellet result too right?

good luck to you, me, him, her, the guy in the corner, and the rest of us newbie pj builders. and GODSPEED to the LL team at gettin the supplies in and the research and development, heh!

Although I can't seem to find them again, I think I've seen a few vertical designs here that have an adjustable mirror on top (after the projection lens) that can both yaw and pitch.

Does this approach work for keystone correction or is it really necessary to adjust the lcd and fresnel inside the box?

(After sleeping on it, I'm starting to think that the mirror won't do anything but aim the beam, but why aim the beam instead of just aiming the box?)

The keystoning issues are giving me a headache...

So far what I get is that tilting the front fresnel will give you some modicum of control, as long as you can get it parallel to the screen. ~10 degrees seems to be more or less a maximum correction with this.

Tilting the LCD seems to also offer some correction, but I'm concerned about what that does with the focus.

Tilting the projection lens has been suggested as well, but I'm less sure how t hat works...

I'm starting to think about just "living with it" and accepting that there will be some keystoning distortion in the final image, but I don't want to...

SupraGuy: you can always tilt the screen itself, the one being projected ONTO. apparently you can't tell that the screen is tilted when looking straight at it, haven't tried it, so YMMV

You can get a decent amount of keystone with acceptable distortion, just don't give up. Keystone correction is not just a simple perfect fix on DIY PJ. You WILL lose at least a little image quality. You are almost there... tilt fresnel, tilt the LCD to about half that of fresnel, and tilt the triplet. Just watch the image as you tilt the triplet and you will easily see the best position/angle for it. Just make sure you are starting with an image that is focused enough to see pixels... a white or lite colored image is best. Maybe throw an Excel spreadsheet up there with vert and horiz lines.

Hey Supra maybe you should lower your PJ just a few inches that will help some.CS

Tilting Triplet lens will not fix your keystoning.

When you do keystoning over 15° you will not get a even focusing. by tilting the triplet lens will fix the focasing issue up to a extend.

I cannot access your picture. Could you post it again?

Hey all, I haven't started building my PJ yet but it will be placed near my ceiling (9' and 8'). I will be moving it between college dorm and bedroom at home, so I will need to be adjusting the distance from screen (need to refocus and rekeystone). I don't think i understand keystoning however. Will I have to build some sort of special box within a box that has the front fresnel in it to keep it the same distance from the lens? Once again, I haven't even started to build, this is just something I thought might be a problem for me.

That is a great diagram. Now I can see how bad keystoning can be without correction. Thanks.

Hey dudes, got a question. Does keystone correction effect the focus or does it just correct the shape? I am looking for a method to correct the trapazoid WITHOUT effecting the focus whatsoever. Any suggestions?

Your only option is a LCD or PC software with correction. Focus is compromised on the corners/edges when doing keystone with lenses

How easy is it to find keystone correction software. Does it come on a disc? Are there any video processors that have this feature?

My nVidia graphics cards have a keystone adjustment feature, but while this gets past the focal issues, it degrades image quality by forcing the image at some point to scale.

The software is included with the drivers.

What does that mean "one point to scale"?

It means that It's the end of a Monday, and I'm too tired to proofread. Sorry.

Re-read that as:

While this gets past the focal issues, it dgrades the image quality. At some points the LCD has to drop pixels in order to scale down the size.

What is the "Field of View" of the Lumenlab Projector lens?

You mean the definition, or the amount?

I haven't done a limit measurement, but I was playing with an LED along the top and bottom of my LCD sled. I managed to confirm that I can focus at a distance of greater than 9" from the centerline of the lens along the plane of the LCD. This implies that the lens can handle an LCD of over 18" diagonal. This was as far as I measured, but I didn't think it was really at the limit of what the lens can do.

The best way to deal with keystone is to do it the same way the theaters do.

1. Align the projector perpendicular to the screen.
I know, now the image is not where you want it.
2. Use lens shift to move the image where you want.

A lot of debate has flown around about tilting the second fresnel for keystone correction. Either by hinging the lens on top or pivoting it on the center axis. Although both work to some extent, there are drawbacks in picture quality. The degradation of the image is cause by the top and bottom of the image moving away from focus in opposite directions. Sounds good in theory, but not practical with the quality of a fresnel.

Both methods also induce the same effect as lens shift. By tilting the fresnel, you are also moving its focal point which causes the image to enter the triplet off center. This is why there are limitations when it comes to tilting the lens. Once you move the focal point too far, not all of the image enters the triplet.

The reason lens shifting is better is that the image enters the triplet in even focus. Shifting the triplet in the x and y axis shifts the location of the projected image. The nature of optics automatically corrects the keystone effect you might think would be caused by the screen and projector being offset.

To maximize the quality of the projected image the ideal path of light would be:
1. Light leaves the light source at an angel equal to the fov of the first fresnel with the center of the light source at the focal point of the lens.
2. The first fresnel collimates the light and passes it through the LCD perpendicular to the image.
3. The second fresnel focuses the image toward the center of the triplet. The triplet need not be at the exact focal length of the fresnel.
4. The triplet focuses the image to the screen.
The light source, lenses, and LCD, should all be in a straight line though center and parallel to each other and the screen plane.

This exactly what I was driving at with the FOV question above. The lenses should be parallel, questions are whether the quality is good enough to do this without distortion and the FOV. Also how the LCD likes light coming though it off perpendicular. Got to be better than being out of focus as with the current bodge technique.

If the FOV is wide enough it will work.

Incoherent,

There is no need to move anything other than the projector lens. Everything else stays where it is. The portion of the drawing from the projection lens to the screen is very close. When the focused light going into the projector lens moves from center, it passes through the lens differently and as a result the projected image moves as well. With the right setup you can shift up/down and left/right 1/2 the screen each way. Even more if you want to pay a ton for the optics.
This definitely is the way to, may require a better projection lens though. I would need the full specs on the ones being used to be able to tell. I am starting a PJ this weekend and will incorporate the lens shift in the design.

Here is some good info:
http://htrgroup.com/?tab=projector-docs§ion=lens-shift

I'll add a drawing showing the effect with just the lenses if I come across one.

The drawing is very "not to scale" so the lens to screen in the drawing should not be taken literally. (pictorially?)
Regarding positioning of the other optics; for the record I disagree. Happy to be proven wrong however.
If you don't offset the light-source the light will not enter the imaging lens, it will hit the front of the box. The two fresnels have an image plane near the centre of the projection lens, not offsetting the lamp will mean the lamp image remains on the optical axis.
I'll ray trace it.

If you move only the projection lens, the fresnels will not direct the image to the projection lens, and you will get no projection at all. Or rather, you will get SOME, but the vast majority of light will be lost.

Also for the record, the fresnels do NOT focus the light. They direct the light. You can get an extreme tilt on the fresnel, and the light will still collect at the same place, as long as the center of the lens is still in the center. You can experiment for yourself. The light does not collect at the focal point of the lens, perpendicular to the lens surface, it only appears to do so in the case where the lens is perpendicular to the path of the light entering it. If you hold one of the fresnels in the Sun, the point of light (Actually an image of the Sun) will focus on the ground (Frying whatever's there. ) If you tilt the lens, the focus will shift, and you may not be able to get a focused image of the Sun anymore, but the place where the bright spot is will stay in the middle of the lens' shadow

In the case of a projector, we are not using the fresnel lens to focus. It's much too close to the LCD to ever get any kind of focus at any distance. It forms part of a lens which is able to focus on the lamp, but since I don't really want to focus an image of the lamp on my wall, this is somewhat unimportant. (Well, what we actually want to do is focus the lamp into the projection lens, but it's not important at all that it be a precise focus -- impossible anyway with the LCD in place.)

What the fresnel is doing in order to correct for keystone is magnifying a part of the LCD. by doing so, the image on the screen can be brought back to square and true. There are, indeed limits on the abilities of the lens to do this. The limits are brought about because of the focal concerns. When the projector is tilted, the distance to the top and bottom of the screen are different, meaning that only a band across the image can actually be in focus. The rest of the image will blur. Incidentally, if anyone is reading between the lines, there is one more thing that the field fresnel can do: It can zoom your projected image, though it can only make the projection larger. By moving the fresnel further from the LCD and closer to the projection lens, you can attain a larger image at a shorter distance. This may be useful for a smaller panel, (Say the 8" Hami) in a smaller room. The focal length of the fresnel may need to be shorter, ie: a 220mm fl fresnel for an 8" project if this is used for a lot of zoom.

Having the projection lens and LCD parallel to the screen is the "right" way to do it. We're used to thinking of the whole thing as one set of optics. It's actually 2. We're actually dealing with 2 different sets of optics in a LL projector. One set gets the image from the LCD onto a screen, and the other set gets the light from the lamp through the projection lens. Only when both of these systems are working do we get a projection. So if you move the projection lens, keeping it parallel to the LCD and the screen, the projection part of the equation still works. Unfortunately without moving (at the very least) the lamp, the light optics will not direct the light into the projection lens, and you've wasted your time. So the lamp optics (light and fresnels) also need to be realigned in order to get your desired results. In this case, the "ideal" solution would be an unsplit fresnel configuration, with the fresnels perpendicular to the path of the light, directing it to the projection lens placement thrugh the LCD. Note that the fresnels COULD be parallel to the LCD< though this is likely to cause more problems with light spillage due to the face that you can no longer get a proper focus of the lamp image into the projection lens. The following diagram (If anyone is still reading!) will illustrate:

Supraguy, that picture looks pretty, but if the triplet can except light at that kind of angle, then we could easily get a 30 inch lcd into focus with the standard triplet, if we had big enough fresnels.

I think your drawing is going WAY out of the triplets field of view.

Good post.

Correct me if im wrong but wouldnt another problem with that setup be uneven distribution of light through the image (lcd), meaning darker on top and lighter on the bottom or vise versa?

Obviously, that picture is not to scale, but is to represent the concept. The distances are also way off, but that's something that I whipped up in Paint to illustrate a point.

Smalls: Actually, as I diagrammed the illustration, the problems would pretty much remain as they are, with a someswhat brighter spot in the center of the screen, and somewhat dimmer corners. Unless, of course, the LCD's viewable angle caused other problems with the angle that the light has to pass through it to get to the triplet.

If a fresnel cannot focus light, how/why does it have a focal length? And how could I, to quote what you posted, "hold one of the fresnels in the Sun, the point of light (Actually an image of the Sun) will focus on the ground Frying whatever's there." Or even see that "If you tilt the lens, the focus will shift," If I were to try this experiment of yours, would it work best if the lens was a certain distance from the object it was burning? Maybe this distance would be equal to the focal length? Is there a difference between focusing light and directing it to a central point? No.
I agree it is not the ideal way to focus the image to the projector lens but to replace it with a plano convex lens (which is what the fresnel acts as) would be cost prohibitive.

I'm not talking about moving the projection lens to the other side of the box, come on now. If the diag size of the image is equal to half of the diameter of the projection lens when it enters the lens, (the image is smaller as you get closer to the, if I dare say, focal point of the fresnel) you can move half an image either way before you loose anything. More movement would be possible up and down because the image is wider than it is tall.
Now I know the next thing is "what about all the wasted light if your not using the whole lens?" There is no more wasted light this way than if you use the max image that will fit the lens. It is just concentrated in a smaller area (remember your experiment with the sun?). Is the image the same shape as the lens anyway? No. If the second fresnel did not focus the light to the projector lens the projected image on the screen would be round not the shape of your LCD because the projector lens transfers the light that enters it to the screen. Try removing the second fresnel and see what I mean.
Each fresnel is a lens in itself, not just part of a lens. Placing them together simply creates a compund lens.

"there is one more thing that the field fresnel can do: It can zoom your projected image, though it can only make the projection larger. By moving the fresnel further from the LCD and closer to the projection lens, you can attain a larger image at a shorter distance"
This is because you are moving the focal point of the fresnel in relation to the projector lens. The closer you are to the focal length, the smaller the image is when it enters the projector lens and the smaller it is when exiting. And vice versa.

"So if you move the projection lens, keeping it parallel to the LCD and the screen, the projection part of the equation still works. Unfortunately without moving (at the very least) the lamp, the light optics will not direct the light into the projection lens, and you've wasted your time. So the lamp optics (light and fresnels) also need to be realigned in order to get your desired results." So, why not move the whole projector? Or are you just moving the entire contents from side to side?
The limitation of lens shift is just the simple fact that you have to have the entire image enter the projector lens. This is why having the diag size of the image equal to half the dia of the lens allows for the movement of the projector lens without moving anything else.

"We're used to thinking of the whole thing as one set of optics. It's actually 2." Once the light leaves the lamp, it actually passes through 6 stages that change its characteristics. 1st fresnel, LCD, 2nd fresnel, and the triplet which is 3 lenses stacked with each lens affecting the light differently.

I'll just stop there.

Definitly off scale. Way out of scale. The system would work fine if we had a triplet with a wider viewing angle, but we don't. If the triplet is indeed capable of focusing an 18 inch lcd, then you could off set the projection in this fashion by maybe an inch or two. But we need at least 3 inches to make a substantial shift.

Ok since writting the above I figured it out. A normal 15 inch is 9 by 12 (We all knew that) an 18 inch lcd would be 10.8 by 14.4. This allows a bout 2 inch shift up or down, and a 2.5 inch shift left or right. Less then 25% of the image. You might gain a foot or so if your lucky. (which IS better then nothing.) but I think 18 inches is pushing it a lot.

FrugalTech: I'm getting the impression that you are being deliberately obtuse.

Yes, of course the fresnel can focus light. It's essentally a convex lens, in essence. But it's role in focusing the projection image is so minimal as to be negligable. It's role is in directing the light. You're taking bits and pieces out of context.

If you focus an image of the Sun on the ground with the fresnel, you are dealing with a different set of distances. The Sun is some millions of miles away, and the ground (Or whatever you're focusing that image onto) is approximately the focal length of the lens away. That's one set of circumstances. The LCD is mere millimeters away, many orders of magnitude closer than the Sun. In that circumstance, it is impossible for the lens to ever acheive focus. Another lens has to do that job for you.

The fresnels do a focal job. Both of them, working together do a rudimentary job of focusing an image of the bulb arc into the projection lens. Of course we don't care if it's an accurate image of the bulb arc, because that's not what we want to see.

Okay, now back up and think about what's happening to the light. Your statement about the image size in the projection lens is compeltely misleading. There is no image at the lens. If you were to replace the projection lens with a screen, you would not be able to discern any kind of image at all. That's because there is NO image at that point. The closest thing to an image that you can possibly get is an image of the bulb envelope. And if you have that so tightly focused that it's 1/2 the diameter of the projection lens, you'd be a miracle worker. I did a write-up in "how lenses work" -- it's in the tutorials section. The way to think of it is that the light from each individual pixel hits every part of the projection lens, which directs that light back to a point on your screen.

Okay, you're only talking about moving the projection lens "a little" -- Small movement equates to small results. It'll HELP, but a 20mm movement of the projection lens only equates to a handful on inches at the screen, and may result in signifigant loss of lumens. By moving the light, I think you could get as much as 80mm movement of the projection lens, and not lose any light. That's 4 times the amount of adjustment, but it would need to be done at build time. (80mm is an estimate. Using an LED, I was able to focus light through the projection lens and get a coherent image focused on my wall from that height above the corners of my LCD. It's possible that there is more adjustment possible.)

Your 6 stage alteration of the light is more or less accurate, but there are only 2 important SYSTEMS at work.

Take a standard LCD and a triplet. Place the triplet at an approriate distance from the LCD, and it WILL project the image of the LCD to a screen. It'll be very dim, but it'll work. The LCD and projection lens are one SYSTEM. That's what gets the image onto the screen. All of the image enters the lens, because the regular backlight scatters the image to all directions. Therefore the lens can "see" the image. All light that enters the lens from the correct source will be redirected to the correct destination on the screen. There is NO "image" at the projection lens, at least no in terms of what's projected.

The light, and both fresnels act as another system. This is what produces the light to make the image brighter, and could even be wholesale replaced by a different system, as long as it got the same end result -- that is, the light going through the LCD to the projection lens. We're using two differnet fresnel lenses, though a single one of shorter focal length could do the trick, in theory. So could an elliptical reflector, if it were large enough. The fresnels take effectively no role in focusing the image. Their purpose is to direct the light. (I could use the term focus here, but "direct" is more accurate in terms of the result.) If the fresnels focused the image, you should be able to see an image of the LCD projected at some distance from the field fresnel WITH NO PROJECTION LENS AT ALL. I will guarantee you that this will never happen at any distance. A lens can only focus an image if both the source and destination are further away than the focal length of the lens. What a lens CAN do with an image closer than it's focal length is magnify it from the perspective of another lens which can focus on the image. If you look at the LCD through the fresnel, you'll note that it does not invert, where the projection does. You can get the same effect with a magnifying glass, and for the same reasons. Close to the image, it bends the light, but cannot focus. The image is right side around, but appears larger. Hold the lens further away, and far from the object, and it inverts. In order for this to happen, both your eye and what's on the other side of the lens needs to be farther than the focal length of the lens.

And I highly recommend that you do try my experiment. You'll find that my statements accurately predict what happens in the "real world"

Edit: A small note about the field of view for the standard triplet: My projection screen is on a wall, approximately 3' off the floor, and about 4'6" tall. As I'm typing this, I can see a light band along the base of the wall, wich goes the entire length of the screen. This light band is caused because where the air goes past the LCD, it allows some light to get past, which hits the top of the projector. The lens is projecting it onto the wall. This represents AT LEAST an offset of 4" from the corners of the LCD, though since it's not an image, and I haven't focused it, I'm not willing to say for certain that it can be that far out. So why not use a larger LCD? Because if I did, my screen would be that much bigger, or my projector would need to be that much closer to the wall. A bigger screen wouldn't be good (It' s this size because that's what I have room for) and shorter throw would be very annoying.

SupraGuy and Incoherent, your set-ups will work but will be limited by the fov and the diameter of the triplet and the size of the arc. Supraguy’s is the most efficient in terms of light loss by the fresnel. TheFrugalTech’s suggestion of moving the triplet only won’t work with the current optics that are available, and also the large arc length that we use and the fresnel ratios. If we could find a triplet that was very large, something like >200mm diameter and wasn’t to long, it could be possible to only move the triplet.
TheFrugalTech I think you might be trying to use film and commercial projector lens shift techniques on diy projection. The two are not quite the same. Lens shift in commercial projectors is more easily done because the objective lens to LCD ratio is allot smaller than diy projectors. In other words commercial objective lenses are very large compered to the LCD. Another advantage the commercial set-up has is that the light travelling through the optics is relatively parallel.

DJ