Reading back many posts, I'm not the only one who keeps rethinking what I want to try for my first build.

Unless some affordable stuff falls from the sky I will need to go mostly with what I have for now.

Option 1 already considered:
Use everything from my ohp and my 19" wide panel and use software to re-size the picture as was suggested by someone in another thread. This is better than my idea to chop off the right part of the screen.
====================================================================

Now Option2: is what I am looking at now. Who knows there may be option #3, 4 .....

Anyway here it goes:

The encircled areas in red in the focal calc pic are pretty well fixed values. I have the 342mm triplet from the ohp. I have the 19"wide lcd from user magel. The final diag 16x9 screen size of 108" is approximate and depends on the throw distance /placement of the pj. At 7.5' I know it will be in front of us so it will have to be a low floor model since ceiling mounting is not an option. Venting etc... will have to be directed properly.
The light engine doesn't matter right now because I will use anything cheap that I have or can find until I'm sure. We'll live with the wrong colour temp until I find the right bulb etc...

What I am interested in is the Fresnel. I'm still not sure about that aspect of these calculations.

Does this mean that the calculator is suggesting a 402 mm Fresnel?
If that is so how far off can that be?
Who makes and sells such an animal so I can buy one?

Click to view attachment

btw) the existing ohp Fresnel is about 12" square max and is not wide enough for the option I'm looking at right now.

first off I would be sure the field of vision on your triplet is wide enough to see a 19" WS panel.. you might have to powerstrip it down regardless. This would effect your throw if you wanted to maintain a large image (would need to be further away), and therefore the LCD to triplet distance would decrease. anywho, this wouldn't effect your choice of fresnels since the difference would only be around 20mm.. but then of course, it would make sense to use the OHP fressys

if thats not a problem..402mm is not the fresnel FL you need. You need a little longer FL fresnel, but unfortunately, I don't know of any that would accomodate you. the common options are 220, 317, 330, 550, 650 (really 605-610). Ideally I guess you'd want one with FL ~400-450, but I don't know of any such available.

My triplet has a clear field view similar to the LL standard lens. I guess this is not taken into consideration in the focal calculator. The pics always give the impression that the whole image fits within the lens field of view regardless of the size of the source lcd panel.

I can see that the fresnel's fl has to be more than the triplet because in the un-split design it sits further back than the lcd although I'm not sure if I'm aiming the fres at the front of the triplet lens facing the inside of the box, the middle of the triplet or at some other point in order to calculate the real fl of the fresnel that is needed.

The ohp has the perfect fl for the triplet but it was designed to capture an image sitting on a 10.5" square glass plate. On the other hand, the lcd is 16" x 9". Even pulling the fresnel out of the ohp it is max 12" x 12" minus the edge needed to hold it in place. The reason I want a larger fresnel is to see and concentrate the whole lcd image into the cone. I thought that the fresnel took care of that and this is where I am a bit confused. Exactly how much and how quickly does the fresnel concentrate the image.
I know that the triplet is limited to roughly 342mm from the lcd +- some focusing. Is the problem the fact that the fresnel can't reduce the larger 16" by 9" cone enough within that distance (give or take 20 to 30mm so it will fit inside the clear aperture of the triplet? Are there any fresnels that would make the cone converge faster so that it would fit inside the lens or is the rate of convergence constant for all fresnel regardless of fl?

Ideallly I would need a larger fresnel that sees the complete larger lcd panel and has a similar fl to the one on my ohp but can concentrate the larger image down so it will fit inside the 342mm (68mm approx clear aperture) triplet. Am I dreaming or is it an optical reality that my triplet can never project an image larger than roughly a 15" lcd panel regardless of which fresnel I use?

Click to view attachment

Can someone enlighten me please. Unless you only buy standard stock predefined parts I'm sure this is confusing to many new users like myself.

My triplet has a clear field view similar to the LL standard lens. I guess this is not taken into consideration in the focal calculator. The pics always give the impression that the whole image fits within the lens field of view regardless of the size of the source lcd panel.


How do you know? I would assume it was designed to see an image not much larger than the 12x12 surface when the image is placed at the correct distance for focusing. you might end up getting dim sides using a 19" WS.


I can see that the fresnel's fl has to be more than the triplet because in the un-split design it sits further back than the lcd although I'm not sure if I'm aiming the fres at the front of the triplet lens facing the inside of the box, the middle of the triplet or at some other point in order to calculate the real fl of the fresnel that is needed.


The fresnels are used to focus an image of the lamp's arc in the middle of the triplet. The best way to do this is to fill the triplet with light from the fresnels.. a good rule of thumb is when the light 'cone' from the fresnels hits the triplet, it fills most of the area. If the light converges too soon, the image of the arc will miss the triplet. if it converges too late, there will be lost light which spills around the edges of the triplet.

The reason I want a larger fresnel is to see and concentrate the whole lcd image into the cone. I thought that the fresnel took care of that and this is where I am a bit confused. Exactly how much and how quickly does the fresnel concentrate the image.

as long as the fresnel is as large (well really you want a little larger) dimentionally, it will be able to cover the panel. The fresnel concentrates the light to its focal point. this is why you need the FL to be slightly past the triplet, so the focal point is on the otherside, and the cone will fill the triplet.


Is the problem the fact that the fresnel can't reduce the larger 16" by 9" cone enough within that distance (give or take 20 to 30mm so it will fit inside the clear aperture of the triplet? Are there any fresnels that would make the cone converge faster so that it would fit inside the lens or is the rate of convergence constant for all fresnel regardless of fl?

its not about converging it faster, its about converging it at the right point. The triplet needs to be ~400mm from the LCD in order to focus the projection at ~7.5ft throw. Therefore the fresnel needs to converge the light a little further than this distance. The commonly available fresnels larger than 16x9 will converge it shorter (330ish mm), or further (550-600mm). the FL is the distance from the center of the fresnel to the point where the image will converge. This point needs to be slightly past the triplet.


Ideallly I would need a larger fresnel that sees the complete larger lcd panel and has a similar fl to the one on my ohp but can concentrate the larger image down so it will fit inside the 342mm (68mm approx clear aperture) triplet. Am I dreaming or is it an optical reality that my triplet can never project an image larger than roughly a 15" lcd panel regardless of which fresnel I use?


yes thats exactly what you need. the problem is, I don't know of any that have the same FL. unless someone chimes in with a source for proper fresnels, I think you might be stuck with a smaller panel. I don't know of anyone using such a large panel with all the OHP components.

@tameone

Thanks for taking the time to clearly answer my questions. Of course I'm still digesting what some of that means.

While I've done a couple of tests with the lcd sitting uncomfortably on top of the ohp, everything is still in the ohp so almost all calculations are on paper and still in the realm of theory. This limits my practical experiments because the ohp triplet focus arm is in the way and the triplet maximum distance is limited by the length of the adjustment arm so I can't go beyond a certain point to test the upper limits of the triplet focus. I also don't have a larger fresnel so can't really tell how much light the triplet would see with the right fresnel for the lcd. I'd need to shine a light through a 17" x 10" fresnel to really get a better idea. I don't really know how else I could test the triplet's maximum field of view in real life.

It appears that if I had the right fresnel the numbers might work out but as you said, it doesn't seem like one exists or is readily available.

I reworked some numbers with hypothetical larger or smaller fresnel focals but there is too much effect on either pj to screen distance and/or the screen size being huge for my space and probably way too dim without a small sun to light it up.

I'm probably back to option 1 and it looks pretty workable using the 19" lcd but with software resizing to about 70% of it's width down to the equivalent total square inches of a 14" lcd panel at 11.2"w x 9"h. I'm not saying those are the dimensions of a normal 14" but just the total viewable surface is about the same. The effect might be a bit worse when I look at the aspect ratio which may resemble a little more closely a 5:4 screen after the resize. Since I mainly wanted this to watch dvd movies the resized 16:9 aspect ratio back into a 11.2" x 9" area might produce a smaller picture than that shown (110") in the attached picture at 10 feet.

Nevertheless as long as the image is clear and bright enough I have the room to back up the pj to increase the projected size if I want to.

With Dazzla's focal calculator:

Using 14" as the screen size I get about 110" diag at 10 feet which is about what I would hope for in the primary location I might use for the pj.

Click to view attachment

Oh well until I can find the $$$ to get the pro lens kit, this temporary setup will save me chasing after parts for a little while.

I think resizing the image is a good path at this point. Your LCD is large enough that you could scale down the used pixels and still have enough left over to upscale a DVD to a bit higher resolution. the downside is of course, smaller image at the same throw. i.e. 19" will give bigger projection from 10ft than a 14" LCD at 10ft with all else equal.

There are many people on here using OHPs, and it might be worth a long look through the plogs to see if you can find another solution.

Agreed, I think resizing is the least complicated way to go for now. I'm already looking at some other options too. There is a thread that talked about fresnel doubling to shorten the focal length or something like that. I believe that it's a thread where someone is building a PS-one in a computer tower. I have to find it again to confirm but that would mean finding 2 larger size fresnels that add up to the right focal. If I understood correctly on a quick read it looks interesting.

The next thing I've been playing with is mirrors. It's interesting how a small mirror can reflect a much larger source image making it appear smaller but complete. I wonder if there is a way to reflect the light passing through the lcd to the triplet by using a mirror instead of of a fresnel. It would take some tricks of bouncing the image and it would still have to respect the 342mm focal length of the triplet but it intrigues me that all the light going through the panel could be reflected smaller to the triplet.

As an experiment, in a darkened room, I bounced an image from my monitor onto a smaller mirror and then back through a magnifying lens onto a sheet of white paper to the side of the monitor. As I expected the image is much smaller, upside down and reversed but complete. I had excess light bouncing off the mirror onto the paper and wall so this wasn't done under closed controlled conditions but it got my curiosity going. I wonder what would happen if I used my triplet instead of a regular magnifying lens. Could I possible produce a complete image larger than the source and in focus. Now that is just an uncontrolled test to look at possibilities but I wonder, would that work in a box with light behind a lcd panel reflected off of a mirror with no field fresnel and then on to the triplet. The mirror would have to be a perfect size or masked so as not to reflect extra light like my experiment did because light was reflected from the glass that went beyond the image of the lcd screen.

Doesn't the mirror act like a condenser similar to the fresnel in this case?

Are there not even mirrors like side rearview car/truck mirrors that say "objects may appear smaller...." I wonder how that would work?

Or...... What if I used a small field fresnel but only after the smaller reflected light coming off the mirror? I suspect that wouldn't work because it's too far from the lcd but I'd like to see the effect for myself anyway.

If one mirror doesn't condense enough for the triplet then what about using 2 mirrors to bounce the light twice on the field side of the pj before it reaches the triplet? As a side benefit the second bounce would correct any keystoning issues.

These are some of the things going through my mind right now.

Do you know of anyone who has played with mirrors on the field side of the lcd like that?

Sorry for the length of my posts but I want to provide as much detail as possible.

Took some time off from home repairs to watch some CFL playoff (football) and mull over some pj ideas at the same time. In the cold light of day I was going to retract everything I said about mirrors when posting late last night. By my own admission I had to respect the 342mm focal length of the triplet with my mirror tinkering but I hadn't really done that.

Last night:
I held a vanity mirror (not fs) as much as 24" facing the screen on an angle and then projected that through a magnifying glass almost as far away back in the shade beside the lcd monitor and onto a piece of white paper producing a much smaller but fairly bright sharp image. That could never work in a pj with the lens I have.

Even when I tried 2 mirrors I couldn't get the angles right and only saw part of the picture when I tried to respect the total distance from the lcd to lens of about 13.5" to 16.5" (the best focal seems to be closer to 16") (ohp manual said to be 342mm or 13.5").

Today a revelation:
I decided to pull the triplet and fs mirror from my ohp. I needed to separate the 2 because the fs mirror reflects the image after it comes out of the triplet and I needed to reverse that by having the mirror reflect the image into the triplet first and then onto the wall or ceiling (see diagram). I use the term image because the light coming after the lcd panel which forms an image helps me differentiate between wasted light and the one that produces a usable picture.

Today, I held the fs mirror in front of the 19" widescreen lcd, just below the lens on an upward angle until the light of the lcd image was at the right angle to go through the triplet. I could see a dim picture on the ceiling. I adjusted the distance to the mirror/lens combination so that it was about 13" or so from the lcd until I could see the word divx in white clearly. Video was too dark to see well for measuring so I used a bright window explorer screen (black lettering on a white background) and a video window from the software divx player with just the word DIVX in white lettering in the centre. I moved that divx image into the four corners of the screen without changing anything else to make sure I could see all four corners projected on the ceiling. It worked fine and projected an image on the ceiling about 4 feet from the lens that looked to be about 4 to 6 times the original 16" x 9" display.
Juggling all these things made it hard to be precise. Obviously the overall image is dim but large white lettering was clearly visible and in focus. I played some video but they were upside down from my vantage point, darker and dim but in bright outdoor scenes I could easily make out the detail in bright areas and I could see that I was viewing the whole lcd screen.

I know that what I did is similar to those "build a large screen tv for $15" plans some people got suckered into buying but my interest is in the mirror effect being used so a smaller lens can see a much larger surface. I have seen other designs use inside mirrors but have any of them eliminated the field fresnel. If no, why not. Is there too much light lost? How much light is lost using a mirror vs a fresnel?

Would this work in a box under real projection conditions.

This is still theory but a big step closer to a real test.

Even if somehow this works, I still dont have a large enough collimator fresnel so I'll have to study that problem also before I make any serious move otherwise the light from the bulb wont be directed properly.

Click to view attachment

In my test the light source is the panel's backlight which is why my image was dim. I never used the optional mirror shown in the pic (would have needed a 3rd hand) but it would have projected on the wall instead of the ceiling and reversed the image and made it right side up. The fs mirror I used is only 6" x 4" and yet putting it right close to the triplet on an angle it allowed the light from the whole 16" by 9" panel to go through the triplet and I saw no keystoning effect. The mirror and triplet should almost be touching and not separated like I show similar to how they are in an ohp but the mirror would be at the bottom of the triplet instead of at the top. I also turned the mirror 90 degrees so the wide side reflected the wide side of the panel.

This topic has changed from which fresnel can I use? To... Can I use a fs mirror instead?

Don't know if I just should have started a new topic since my plans have changed so much the title of my thread is no longer accurate.

Theory only:
Anyway, if my plan to use only a mirror and no collector fresnel on the field side of the pj makes any sense at all (my post just above) then I need to deal with the incoming light and the fact I don't have a large enough collimator fresnel for my 19" wide panel (16" by 9").

Then lo and behold along comes a link to this thread:

Multi bulb and multi fresnel topic

The bottom line is there may be a workable solution here. I have 2 options depending on which type of bulb I would use.

If I use the bulb from my ohp it's 410 watts and looks like this:
Click to view attachment

or if I use a reflector as a blocker:

In that case I could design a 2 fresnel setup like this:
Click to view attachment

If I use an open bulb that radiates on all sides without the built-in reflector then I could do something like this using 3 fresnels side by side.

Click to view attachment

The light blue lines in this last pic are blockers to prevent light from the center going to the side fresnels although I'll have to see if they are needed.

The fresnels that are discussed in that link are the page magnifier type found at Staples/Office Depot.
It discusses mullti bulbs but mirrors placed like my example if properly alligned could split the bulb sending the light to each fresnels focal point and onto the lcd.

I will need a real life test to see if I can eliminate that seam between the fresnels but that too is discussed in that thread. Also, I hope someone comments on whether anyone has ever tried to eliminate the second fresnel.

edit:
btw) Nothing in the pics is scaled or angled properly it's just to illustrate my ideas.
I also realise that I forgot the Lexan and cooling in the diagram so has not to melt the cheap fresnells