I've seen a few people using precondensor lenses in front of their bulb. I'm not sure exactly what its function is but I assume its supposed to optically shrink the lamp arc to get more light into the triplet. But I also heard some debate about the glass actually filtering out a lot of light and ending up in almost no improvement.

So which is it? An improvement? Or a waste of time?

I don't think it "shrinks" the bulb, I think it catches angles of light that the fresnel would otherwise not see. Have said that it would be interesting to see someone use a cylindrical lens that "wrapped" partly around the bulb catching a wide angle of light and directing it to the triplet. I think the only loss (the one you are refering to) is the light lost by passing through another layer of glass. If I decide to scrap the parabola, which now I toy with the idea (just because I want the length for other ideas) I am definitely going to play with condenser lenses. I have three here, probably a few more on teh way.... maybe thats what that 8" diameter short FL would be good for?

The arc appears bigger, not smaller. I used a lens from a big projecton lenses in front of the lamp, and I got brighter projection.

I, too, am experimenting with precondensers, I also have bought a lux meter. Interesting readings so far. A quick test of a 4" plano convex showed better lit corners. Lucky_Me, what condensers and sizes do you have and where are you getting them? What are you leaning to mostly towards?

Mikau check this link, it will explain a bit of condensers and with the addition of a spherical reflector will double your projected light. http://www.cairnweb.com/tech/techmenu_lamp.html
sim

hi sim, it is the first time i noticed you bought a luxmeter, sorry if you have advertised it somewhere else, but I did not read it yet.

Can you tell me if you did any measuremnts to the projector? lumens output? contrast measuremnt?...

thanks

Hi Sim, I get parts where ever, there is a links page on my site, but if you browse the web, browse ebay etc, there's lots to be found. What am I leaning towards? I don't have enought experience using a collimating fresnel to make that decision so I think once I build my optics section I will have to expereiment a little to see if the parabola is worth it or not. The only reason I say that is length, I think I am ok to keep it, but it makes everything a little tight as I am planning on a "vari" focal pj.

cool...it'll be nice to see some new readings....hey sim, when you do start getting some readings, can you post them over here?
http://www.lumenlab.com/forums/index.php?s...0&gopid=106825&

have you checked out irongeckos pj lumen calculator? he has pre-condensers in there...will be interesting to see how your results mesh with the theoretical calcs...

Rox: Probably not extravegant to yours, seeing that your hardcore into lux readings , but would work for me to see any rise or falls from my testings. http://www.multimeterwarehouse.com/FX101.htm
Interesting, with this meter just by crude testing outside the box, I had the sensor hanging up on the triplet while I was crudely testing a condenser. The readings dropped for some reason after placing a lens almost near the arc but looking at the screen it seemed brighter and evenly lit.

Lucky: Same here, ebay mostly. Understood regarding your parabola, looking forward to your results. I am also getting a heat absorbing glass to somehow protect the lens from shattering.

mikyd: Regarding that link, sure will.

cool, that is the exact same meter I have actually...which means we would have some solid basis for comparison if nothing else!

My experience is that precondensers work. They need to be large - 4 to 6 inches in diameter. And the focal length should be about 12 inches. But I believe many different lens combinations can possibly work also. And the lens glass obviously needs to be heat-resistant - or very far away from the bulb and fan cooled.

And if you find that you just can't illuminate your entire LCD, then try a longer focal length rear condenser fresnel. This will allow you to move the bulb and precondenser farther away and allow the light "cone" to spread out more before reaching the fresnel.

You will place the bulb much closer to the precondenser than the focal length of the lens. That is because you don't want to completely straighten out the light rays. You want to create a cone of light that is the same angle as the focal cone of the rear condenser fresnel.

It works!

Interesting.

Sim you say you can doulbe the light output if you use it together with a spherical reflector? Is that only if your not already using a spherical relfector? I find it somewhat odd that people would go through all this antiglare removal hassle if just a precondensor can double the output. So why doesn't everyone have one? I say theres a conspiracy!

So are these things easy to get? Or is that why not everyone has one?

Rox, Supraguy and I just went over the condenser calculations again today and there was no benefit for lamps 25mm diameter and larger. A precondenser is wasted on our current lamps. The story is different for small arc lamps (coming) and small panels (under 10"); with those a precondenser can have benefit.

You plan to stock lamps with smaller arcs? How small? I take it they'll just be for mini projectors.

Hi Brain,

Not to give you a headache but I am curious to know why? Why is the pre-condensor a waste with larger Arc's and perhaps beneficial with smaller Arc's? I'm just curious to know, I haven't experimented at all with that yet. Does it have something to do with triplet size?

well actually is not a waste.

Is just that there is no much improvement in my opinion. You would say ANY improvement is welcome, true, but I have serious difficulties teorically determining that improvement.

If you have any design, please let us know so we can discuss it.

The model we choosed to work with was this;
1)LCD 18" (averdaged between 17 and 19)
2)220mm rear fresnell focal (where we are going to place the lamp, forget the trhow/lamp position dependency by know).
3)24mm arc gap, 25mm tube diameter lamp.

Please I am all ears.

is the reason you have trouble proving it theoretically because its too small or within the error of margin with the given starting points? at what point do the losses outwiegh the gains? say a 13.3 inch lcd(all other variables being the same)? a 15mm arc? or does it have ot be smaller? like 5 mm?

the problem is I can´t determine the improvement. I see it is not very efficient, but cuantifiyng it is where the headache comes.

Yes, smaller the LCD goes, longer rear fresnell goes, shorther the arc is... does much more effective lens system.

the if someone wants to try it, there are the model specs some posts back. (lcd size.. rear frenell.. arc..)

I will post some pics later, have to go now.

Some or alot have theorized that antiglare removal or panel enhancement was either impossible or a hard road to acheive but it was Elken's action and Mark's brainstorming that got us to where we are at. I just see another road here. So, who's up for a ride?
sim

well, I've got a precondenser coming in .... between the two of us (and our light meters ;-) we can at least see the limits! your panel is 15" ? mine is a virtual 13.3", you've got 400 watts? I've got 250..so between the two of us we have some variations to test anyway..

Well, I have over 30 individual lenses here, I have two spherical mirrors, possibly three more on the way and I plan on buying Brains reflector when it comes out. I also have some Opal glass, Sandblasted glass, two prisms and I will be fooling around but I don't think I will do any serious testing until I complete my "Optics Section". And that I am waiting on Minoten for so who knows.

So far, I have in mind a design to build a light box using the Ushio. Once I build a model of it I'll try the three lenses I have currently. Kinda cold right now in the garage, but will work little at a time.

If anything, a smaller arc would be ideal but there are some if not many of us who can't offord to buy a whole seperate gear @ a higher price when we still have these 25mm arcs bulbs and ballasts in our hands that might still do . Let's see what we can squeeze out of them.
sim

A condenser becomes much more valuable if you increase the rear focal length to 320mm, or use small LCD panels. It also can be used with short arc lamps. I'm attaching 2 of Rox's Oslo ray tracings to demonstrate:

In the first you can see that the long arc prevents us from catching very much stray light with the condensing lens. We may be able to catch a few extra rays but we've put another piece of glass in the lightpath. This is based on a 25m lamp diameter, the situation is worse for larger diameter lamps.

Click to view attachment

In the second picture we see the effect of using a short arc lamp. This is based on an arc length of just a few mm. We can grab lots of stray light with our lens now.

Click to view attachment

dam, i was going to post that images brain ;D

now i have nothing to say .

well, the images are not the final ones... but the basic idea is that;

the arcsize does limit the power on the lens (focal) so the full virtual arc is seen throw the lens at rear fresnell's focal.

Click to view attachment

you can "draw" this kind of ideal things that do work in a happy world, but this lens is not posible to manufacture in real world, the OSLO model is good simulation on what should happen in real world with real lenses.

If someone has found a really effective design, please share it here.

Nicely made condensor I saw.

I've somewhat had an idea, maybe Rox and Brain can kick it around and improve it.

Ok, well it isn't pretty but I am really tired. What if the precondenser was shaped to encompass the lamp with mirror coatings on the sides? It wouldn't really be a lens it would be a fibre-optic-lens-prism-mirror-thing.

It could have a Focal Length or no Focal Length. It could be crystal clear, or it could be slighty diffused to even out any hotspots on the exit side. I guess I just realized this would be difficult to manufacture as one piece, but it is an idea that hopefully could lead to a two piece item that would catch the remaining light (that the pro-reflector doesn't) and direct it forward.

It looked better in my imagination, lol.

I feel dense. I understand the pictures Brain and Rox posted, but I don't understand why it works with Larger LCDs or smaller Arcs. I must be too tired. Anyways if you didn't like my grade 2 drawing (I used crayon) what if we had a glass cylinder say 4" diameter and 3" long. Have a dichrotic coating so the IR escapes, but the visible light doesn't. Instead it gets reflected. The we have two 2" wide slits in the cylinder so it can slide over the lamp and touch the pro-reflector. That must help some light make it to the fresnel that would otherwise be lost?

Why not have the entire reflector engulf the Ushio then place the condenser covering the reflector. Wouldn't most of the light reflect and some how hit the condenser?
sim

Yeah, I guess that is the same idea.. it's just that Brain already has made his reflector. I'd be happy to hear an interest in just a dichrotic cylinder to engulf the bulb. Then it would be up to each person to use a condensor or not (and to choose their own FL).

The reflector is another lens. Putting the arc into a reflector would cause the light to do many things, like become collimated, or hit the condenser at some odd angle that the condenser just bends even more and sends off to the side of the box, missing the LCD (think rays here, not fluids). The optics have to be designed to both capture light and evenly illuminate the panel. That pesky light just doesn't work the way it should.

On the plus side, the holy grail is imminent. With the short arc lamps we are getting next we can design a reflector that can grab 80% of the usable light. That means that with a 250w 15,000 lumen short arc lamp we will deliver 12,000 lumens to the panel I think. I'm still working on this.

I am glad you are! I want to try a few things, if I had a coated glass cylinder and your reflector I could take very lightly sandblasted glass and diffuse the light at that point and then focus the fresnel on the sandblasted glass. (I know you lose light everytime you diffuse, but if we had all the light hitting the diffusing glass would we care?).

I just want to try different ideas, yet at the same time I get sick of trying, so hurry up - lol.

With no facts at all I will take a whack at why the condenser system doesn’t seem to work right.

The most important thing that I think is being done wrong is the spherical reflector does NOT focus on the arc of the lamp but on the condenser lens!

There is an angle of the light reaching the condenser lens from the reflector. You do not set the light to enter straight from the reflector you set it to be as close to the same angle as the lamp. There for the spherical reflector must have a much longer fl than the one we now use. The focal point would have to be 80mm or greater
The condenser lens will "see" two light sources the lamp and the reflector. Maximum light out put will occur at the point that where the difference between the lamp light angle and the reflector light angle is the lowest.


Click to view attachment

The lamp to spherical reflector value should be set to make the angle of light reaching the condenser lens as close as possible to the same as the angle of light from the direct output from the lamp. Then the condenser lens will determine the final angle.
You end up with a light sources that will give a nice even beam spread over the rear Fresnel. I don't know the angle that the light leaving the condenser to cover a 17' LCD maybe around 17deg?

As always the arc image will be enlarged buy the condenser lens so the lamp will have to have a small arc or the rear Fresnel will have be longer.

Please tell me where I'm wrong Flame suit on

The reflector you have drawn there is not a spherical or at least the way you have drawn the rays is more like a parabolic reflector set up. With an arc at the centre of radius, not at its FL, of a sperical reflector an image of the arc will be overlaid on the real arc so ideally all rays appear to be coming from just the arc.
I think I might do a Wiki write up.

DJ

I can’t believe you are referring to real world theory.
I’m only stirring you . I know that perfect theory can give you the ideal and you can work from there.

DJ

I draw like crud If you can do a ray trace then then we will see. The refelctor should be a sphercal one. You offset to one side of the focal on the sphercal.

If you look at this lamp systen I think the refelctor has a longer fl onto the lens not the lamp?

Click to view attachment

I got your back Arizona...

I know what you are saying, you are saying instead of having a short FL that throws the rays back at that lamp arc, lets have a longer FL that collimates the light and shoots it at the condensor lens. I have three 145mm diameter mirrors on the way, one of them might have your name on it (I was also thinking of making three cool looking lamps with them though too).

So many things to try, so little time, so little money. *sigh*

Anyways, not to take nothing away from Brain because I think that is the simplest fool proof design happening. Also, it is probably the cheapest and it is passes IR to boot. I just want to try other means because well... there is nothing like a fool and his money. lol. I guess I can try so I am.

I also bought a 145mm IR filtering optical flat, and I have several more lenses on the way. One shipment came today while I was at work.. don't know which order it was, but I'll pick it up tommorrow.

Click to view attachment
the raytraces is wrong. This is a happy world again;

1)the line coming from the reflector is diverged, this only works with negative focal lenses. But suprisingly the other line coming from the lamp is condensed, this only works with positive focal lenses, no posible to do both things at once; a lens can´t be positive focal and negative focal IN THE SAME AREA at once.

2)negative lens.

3)negative lens.

4)positive lens for the light coming from the ARC and no power lens (glass window) for the light coming from the reflector.

I am not asking for the EXACT ray trace but this is easy if you take care of very easy laws, you can do very accurate raytraces by hand at paintbrush;

this is ideal lens teory;

Click to view attachment
1)light coming from the focal of the lens is going to be paralle outputed. (line number 1)
2)any light going trhow the center of the lens is not going to be refracted (never bended)
3)light coming paralelled, is going to be refracted to the focal lengh on the other side, just as law 1.

You can try new drawing with these laws, good luck

I will never be good at drawing. Can you focus the reflector on the condenser or not? I say yes and this will greatly improve the light output of the system. let's say 30% over the condenser systen done the "old" way

if ou are asking if it is posibleto find a lens so the paraell light (the one coming from the reflector) is going to be condensed (just asking where do you want it to be condensed?) yes, it is posible, but forget about the direct light from the lamp. Both things at once is not posible to handle with the same lens. As you said, there are 2 light sources. No way to control both.

The only way;

a spherical reflector reflecting light to the lamp, (lamp at R) then there is only 1 light source, you can play with lenses to control the light cone, or increase of efficiencies...

Rox you are saying that the light from the lamp or the refelctor won't go through the lens? Yes they will be at different fl but both will go through and both will make it to the fresnel

yes, both will go trhow the lens, but the key is where?

Click to view attachment

consider this light source, so in fact is like there where 2 light sources.

it looks like both are shining the lcd so "ok, it will work" is the first sensation.

now check this arc projection on the triplet system;

Click to view attachment

the virtual arc behinf the lens, is going to be projected at somewhere the triplet (green lines).
the light source 2 is going to be projected at somewhere farther because it is closer (blue lines)
this will result is a lot of lost light (and i guess a huge hotspoting because of the blue lines in the center of the image entering the triplet but not in the edges).

Two light sources are not posible to control, only 1 is posible to control.

I was going to point out Arizona_Video's incorrect Ray trace earlier but I wasn't sure what happens to light rays when they are further out from the lenses focal point. Rox, would you not agree that it may be detrimental to reflect light back through the lamp? In an ideal world would it not be beneficial to reflect that light on a path where it is not physcially blocked (by the lamp)? I think that is the idea here.

A few points, or ideas you could ponder.

Diffusion. I know it EATS light, but what if we had a faint diffusion right before Arizona_Video's Condensor lens in his drawing? What would happen then?

A Second Lens. What would happen if there was a second smaller lens to collimate only the ARC light so that all light hitting the Condensor Lens in Arizona-Video's drawing was Collimated?

A Negative Lens for a Condensor. I guess that doesn't make sense. Negative Lenses spread the light out so in effect they do not Condense, but from looking at a ray trace perspective it "should" be beneficial but they are not. How come? Why do negative lenses eat so much light when I play with them?

A different use of the Reflector. I have a 50mm diameter concave mirror with a 50mm focal point (I think it was 50mm I forget now). Now if I place a light source at 100mm, the reflection is a point of light at 100mm. If I place a point of light at 50mm the result is collimated light at 100mm. If I place a light at 0mm the result is a spread out light beam at 100mm.

Could we not use a "XX"mm Focal Length mirror (even convex I don't care) and place it right against the lamp causing the reflected light to spread out. In my mind this may do two things. One it will stop light from getting eaten by the lamp. And two, if it spreads out, maybe if everything is the right Focal Length and size it will lighten up the corners of the LCD without brightening the center (something I would like to see).

What are your thoughts Rox?

P.S. Please be kind when you reply, sometimes you are a little bit... not so nice. Not to me, but to someone else here earlier in this thread.

Thanks,
~Luckly

One more thing I meant to go deeper in depth on was the negative lens. I gave my lens calculator information that I had printed out away to another member here so I am working on memory. If a negative lens has a greater focal length than a positive lens (don't worry about the +/- symbol), if a negative lens has a greater focal length it's effect on a postive length lens in an increase in "virtual" focal length.

Bear with me, I am trying to describe this from memory.

The point is, it "should" be beneficial to place a negative lens next to the lamp. The 220mm fresnel would then have to be placed closer to the lamp aswell. But in my theory (I probably need to review it), placing a large negative lens directly in front of the lamp and then moving the 220mm fresnel in closer "should" result in more light getting collimated.

If you want to Ray Trace something, why don't you work out that. Work out the optimal focal length for the negative lens with a 220mm fresnel (and the relative distances between them). I would greatly appreciate it.

~Lucky



EDIT
I think I may be out ot lunch on the negative lens bit. I am too tired to know if I am thinking straight or not, I'm pulling an all nighter because I have to be up in an hour and a half, no sense sleeping now. Anyways, please feel free to correct me if I am wrong on anything.

just a quick question..... why would you go with a 250 watt, not a 400 watt (or both) ? small arcs harder/more expensive the higer the wattage? if you substitute 12000 into Irongeckos lumens calculator as the amount entering the first fresnel you get less than 400 lumens hitting the screen... I hate to be greedy, but why not shoot for 20,000 or 25000 lumens hitting the first fresnel? if its just not cost effective(say , it puts the cost of the bulb too close to commercial pj bulbs) I can understand that, but otherwise its no different than designing a light enigine to gather only 35% of the current ushio...

well, It was not my intention. Maybe my really bad english was again the guilty. Can you tell me what post exactly on this thread was that? just to improve my expresion, I would like to learn what not to say .

about the negative lens, i don´t think it is the way to go, as you say it makes light to spread on a wider angle this results in a closer lamp setup, but you forget a very important thing; the virtual lamp is now closer than real lamp, this will require the whole lamp+lens syatem to place farther. positive lenses is the way we should go, and even in that case, there is doubty effectivity considering LCD size, arc size and rear fresnell size... on our model.

yea, smaller LCD, larger rear focal and shorther arc does notoriously help on a condenser setup efficiency.

What's this. Are you two ganging up on me?

DJ

edit: Lucky diffuse light in our projector is our enemy. Once it is diffuse we have practically no chance of controlling it.

Lucky,
The idea behind a condenser is to use more of the light. It does this by capturing light that would normally miss the fresnel and then converging it to hit the fresnel (as you can see in the first pic). The second pic shows the situation without a condenser. The third pic shows what would happen with a negative lens. A negative lens would diverge the light and result in even more waste. As Rox said, to collimate the light, you'd actually need to move the fresnel further so you'd be wasting a lot of light that could have been used.

Of course all this is based on a point source of light, where a condenser would make a significant difference.

Arizona this is a typical spherical reflector set up. The lamp is at the centre of radius, the FL of this reflector is half way between the lamp and the reflector.

DJ

Ok, now if you move down to your THIRD drawing, as far as the Fresnel is concerned it thinks the lamp is further away than it really is. To get to the point, move the convex fresnel CLOSER to the Negative lens. even an inch apart... Naturally the Negative lens may have to be a fresnel too. Maybe its mute but do you see what I am trying to say?

Thanks for the excellent Ray Traces.

Lucky here's a link to an applet that looks to be the same as what Rym is using.http://webphysics.davidson.edu/Applets/Optics/Intro.html