Okay guys ... AZ, Dazzla, SJ et al

Right now the best setup for a light engine supplying a rear fresnel is the brightest smallest point of light you can get. But, could it be that some of these bulbs are getting a small enough arc size and a thin enough housing that we can either A) put two or more next to each other or somehow maximize more of the light coming out of a single bulb?

I have read a lot of patents posted and whacky ideas about how to get more lumens, so I almost feel bad throwing another one on the pile, but I don't remember this being discussed before.

What if you have small arc lamp like the 150 W powerball or even up to the EYE 400W Clean Arc in a horizontal position (sounds dirty) with the standard placement of a reflector behind and a condenser in front. But then, take it another step further. Put ANOTHER reflector at a 90 degree angle to the first reflector (maybe below the bulb) but still facing and focusing the light back into the arc, but this time from a perpindicular axis to the standard setup.

On the opposing side of the lamp enclosure, snug a 45 degree slanted fs mirror (small), such that it reflects the arc of the lamp without occluding the path of the standard light. See attached pic. From the perspective of the rear fresnel, that arc image would appear about 1.5 x lamp radius more distant than the standard arc. If that lamp radius is very small, though, it could very much merge together with the real arc and a soft focus on the fresnel.

Question is, is this worth testing/pursuing? If I were to put the focal point of fresnel midway between the real and virtual arcs, would I expect to see a light improvement? Could you even make the small mirror concave and magnify the reflection, reducing the perceived size/distance of the reflected arc?

What the lens see is 2 light source. The closer you get, the better the image is. From your description, I don't think you can get it close enough to have a good image. You can calculate them actually. From there, you can compare to a biggest bulb used around here, and I am sure you get an idea.

True ... I currently use an MH400 in my pj, which is about the biggest damn thing I can think of... yet still, I can get pretty precise images. Now that said, maybe if I switched to a 150W powerball, I might think about how I could have stood the mushy image of before (heh).

Anyone know what bulb has the smallest arc in a 400W high CRI 4600-6500k range? With a small envelope?

Wow, when I pulled this thread up, that drawing had me wondering what a couple of eggs cracked onto a womans personal toy had to do with projectors!

LOL T15, now ribbed for her pleasure.

Hmm, okay been noodling this whole thing some more, and playing with the focus on my current projector. One thing I notice about my bulb placement is that I have it WAY too focused. I can see every little blemish on my current lamp. So, I started defocusing a little bit, and it looks a lot better.

Given that experimentation, I can know that if I keep my arcs within a 40mm square package or so, I can defocus them for a good even image on the wall. Here is my latest design, this time possibly incorporating one of the newer super bright xenon headlights from http://www.xenonlink.com/product_info.php?products_id=98. Motorcycle kit comes with everything you need except 12V power for $76 US.

The sticky wicket here is going to be coming up with small enough spherical reflectors in small arc segments.

Of course, I guess if that is possible, then two bulbs in a cross might be as well, with the shells right up against each other at the center of a the spherical radius of the reflector. Twice as expensive, probably close to twice as bright, and possible with the cool running xenon's (maybe).

Edit: Time to put my money where my mouth is. I just ordered a motorcycle kit (single lamp) to begin testing with. We shall see!

It’s a nice idea but unfortunately we have to consider the rest of the optics involved. By using the mirrors on the sides you are effectively creating two virtual arc images and to make matters worse they aren’t in the same plane as the real arc. What you’ll get at the triplet is what you start with only magnified. I’ve added two virtual arcs in your diagram to show you what I mean.

Click to view attachment

The same problem will happen if we use a parabolic reflector, it will create a large virtual arc. Commercial projectors can use parabolics because they are using collimated light (although they do suffer a bit from this problem, just not to the degree we do). If we were to use a large parabolic, as big or bigger than the LCD, and remove the rear fresnel then we could start to use the off axis light. We’d still need to solve the void problem though.

DJ

Right, I understand about the virtual arcs, but it is very similar to a misaligned reflector. If I had the choice between no reflector and a slightly misaligned one, I might choose the latter, because the brightness is such an improvement in viewing over the exactitude of the singular arc.

I don't know for certain, that is for sure. But I think if I put the focal point of the fresnel right in the middle of the triangle formed by the virtual/real arcs, I should get a smooth smear of usable light going through the panel and hitting the triplet. The way my current setup is behaving leads me to think this. With the signal off to my LCD, and before tweaking, I get a dead on clear hotspot image of my arc on the screen. A little wiggling of the light placement and the rear fresnel, and it smears it out nicely.

I would TOTALLY agree this is unworkable for a large envelope lamp, and probably for anything even as big as a T15, but I am curious about the really tight arcs like the 150W ceramic and the xenon HIDs. I have one of the xenon's on order. I will post pics, measurements, etc., when that comes in.

Dazz, about those ceramics ... do you know if anyone has tested the permissiveness of the arc core on those? It seems like the light would radiate from the surface, but not travel THRU the core, which would make spherical reflection a lot less efficient. Any thoughts?

Edit: Oh, and what would you think about overcoming the different planes issue by using side mirrors with a slight spherical shape to magnify ... bring the virtual arc closer?

I have been looking at this from a different angle, so to speak. I have been researching retrofitting the current bulbs in commercial projectors to take advantage of bulbs like this.

In an uncanny bit of serendipity, I recognized the site you posted immediately, as that is where my hunting led to as well.
I think that these bulbs could be used inside of a reflector designed for them which would be lower cost to produce that than a model designed for a t15...Im thinking that the parabolic reflector needed to focus this would be about the size of a Legg's Egg container, possibly a very similar shape as well.

SJ

Perhaps a burned out UHP enclosure like this?

Ooooooo! Look what YOU found!

In theory, we should be able to replace the bulbs in some of the ohp units with a retrofit unit like you ordered. BTW, your shopping skills are extraordinary...

I think the tricky part might be getting the proper voltage to them, but I can wait to see it. There was a member on one of the Chinese forums who built one using a similar method, but I cant remeber seeing how he addressed the voltage.

hmm well given the wattage, I plan to just use a 12V rail on a computer power supply. Seems like that would work?

Speaking of interesting internet shopping, this place looks cool

I have been examining that UHP a bit more. I looks like they use a parabola, but they get around the dark spot problem by only running an extremely thin wire back through the rear of the lamp, with another running perpindicular to the side. I have never seen either of these two "lines" on my Samsung, so it seems like that issue can be dealt with, depending on the lamp filament.

With the Xenon HID's ... I wonder how challenging it would be to remove the lamp portion from the plastic base/housing? Then perhaps wire it similarly to a UHP within a 10.6" diameter mirror polish parabola?

oooo The Xenon HID came in. Arc length = 7mm Wowzers, tiny. Now going to have to figure out all that wiring.

"Oh bother!" - Pooh

The actual void is caused by the lack of reflector where the arc vessel penetrated the rear of the reflector. Adding to the problem is the physical size of the vessel obstructing the rays. The commercial light engines alleviate the problem by using a small vessel size to reflector diameter ratio, creating a smaller void to start with. Then they use other magic to blend/smudge the void out. One approach is to use a thing called an integrator. It uses a couple of lens arrays to create multiple copies of the reflectors image then it jumbles them up and re-combines them back into one reflector image. The jumbling or shuffling is what effectively smears the void away. Another approach is a light pipe, usually used with elliptic reflectors. A similar thing happens, the multiple reflector images are created by the walls of the light pipe.

DJ

Thanks, Dazz. The light pipe work that you did earlier was outstanding. The UHP lamp I have been dissecting, though, appears to have a reflective surface even underneath the lamp, except for a tiny wire piercing the bottom.

The Xenon is a winner btw! Excellent light output rivalling the 400W football. It also runs VERY cool ... more like a bulb from around the house. The color of the light appears very acceptable (at least from the 6000k). More in my plog.

Just got an email back from the xenon folks. Bulb life between 2200-2600 hours, 3250+ lumens. They were a little unclear about the lumens, since they have different temps on their lamps. These are automotive folks, so they don't answer the regular lamp questions too often.

So, $76 for ballast+lamp, ~2400 hours, 7mm arc ... and at least for me, with a reflector and condensor gives a very viewable experience with a 15" panel, no lexan and a pro lens. More to come when I test with arc mirrors and multiple lamps. Wish I had a luxmeter.

What I have learned so far using small arcs:

1) Multiple arcs in close proximity, either real or virtual, have little discernable negative effects regarding blurriness or focus, when the arcs are small.
2) Plane reflections suffer in visual quality as light passes back through the non-arc portions of a bulb.
3) The use of small reflectors to the sides of a cylindrical small arc envelope are a useful technique for gaining brightness while not obscuring the primary path from arc to reflector and back to panel.
4) Slight distances in depth or width of the arc "formation" do not affect the image greatly, except where light is deformed by varying angles of incidence through multiple lamp envelopes.

My plog contains some of the testing results.

One new technique I have been thinking through is "chaining" together arcs with a double-convex lens. Light from the rear arc would reflect back from a spherical reflector, add to the forward flowing light, then be refocused through another arc and added to the light proceeding from that arc on its way to the LCD panel (utilizing a condensing lens). See pic attached.

Dazz, have you ever done a ray diagram on something like this?

No I haven't, I don't know why

Let me think about it for a bit.

Do you have a reason to think number 3 adds anything? Maybe a photo of the arc/s through the triplet.

OK first problem. To collect enough rays from the arc the DCX lens will need to be close to the arc so its FL will need to be short. A short FL means that the curvature of the surfaces will be large so the ray at the periphery will tend to reflect back rather than refract through the lens. A possible solution would be to use two PCX lenses with the plano surfaces facing the arcs. I don’t think PCX lenses can have a FL shorter than their diameter so they may need to be DCX lenses with one surface having only a slight curve. Or use Aspherical lenses.

DJ

For #3, I did some testing a few nights ago and again last night. I used little square craft mirrors that you can get from Michaels. Because of the side wire on the xenon, I was only able to put in one side, but it did add brightness and it did not obscure the image. My plog has a screen shot from last night when I used a single side mirror with a measuring spoon reflector and one bulb. The colors and brightness aren't on par with my DVD shots, though. The PC as a DVD player gives a ton more options than the cable box through the Ebox.

Given a small arc and an already good setup of reflector and condenser, I see no reason folks should not add an angled small mirror to either side for slightly distant/angled increased brightness. My luxmeter should be here today, so this weekend I might get some % benefits going.

You might try this from surplusshed: http://www.surplusshed.com/pages/item/pm1101.html $38 + $5 flat shipping

That condenser lens you got from me came from a set very similar to this one (but this one has a smaller diameter and a shorter focal length).

Here’s the problem I see with mirrors to the sides of the arcs:
These are from simul8r’s “Simflector” tests
Click to view attachment
Notice the reflections don’t fit through the triplet. Add a pre-con to this and things get even worse.

Here’s another photo of the arc and reflected arc images at the back of the triplet.
Click to view attachment

I find that looking back through the triplet is by far the best way to understand what is actually happening.

Hers another photo that Fulcrum posted:

Click to view attachment
In my opinion this set up won’t work at all. A standard arc only just fits though the triplet aperture. Line A shows the size of the arc. Now look at the distance between the reflected images, line B. There’s no way that these reflections are going to pass through the triplet unless the triplet is huge or the rear fresnel's FL is very long compared to the front Fresnel FL (negative magnification).


Sorry to be so negative, it's just that I can't see side mirrors being useful unless they are extremely close to the arc.

DJ

Not a problem, Dazz! I am a bit negative on these things too, but rather than theorize, I want to put some legs on em.

I have already tried a single xenon with one side reflector, and that fits through the triplet no problem. I will try to get a rig set up today and take a picture both of the arc image passing through the triplet and from looking down the barrel.

I’m all for practical tests. Although I may come across as all theory I do test things, some times
I use theory to look at a problem first then test it. On the odd occasion I test things and then work the theory out.

The size of the Xenons put them in a slightly different category. Being so small allows you to do things with them that larger lamps can’t. So you may get an increase in light with the side mirrors but you may also get just as much of an increase, maybe more, if you were to use a well matched pre-con and reflector.
I’m really going to have to get a Xenon to try it out

Keep on experimenting though you may find something. Then I’ll have to figure out the theory for it

That's the reason Tesla hated Edison. He said a theory should be thoroughly examined and perfected and then when you build a prototype it will work right the first time.

When Edison said his famous line about "I didn't fail with the lightbulb, I found 1000 ways how to not make a lightbulb" Tesla said that he should of only had to make one

Just some more theory to go along with your observations, maybe it’ll eliminate a few 100 prototypes

Here’s a diagram of just a reflector and how much of it is collecting useful rays.
Click to view attachment

Now here’s one with a pre-con added:
Click to view attachment
The pre-con collects more of the direct rays from the arc. And the reflector now collects more rays from the rear of the arc. So it’s a two fold gain.

There’s only a small area at the fresnel’s focal plane that can pass rays, I think you mentioned it as 30mm. Any rays that pass though this area will pass through the triplet. Using a mirror on the sides you can get an increase in rays passing through the triplet. As long as these rays appear to be coming through the 30mm aperture.
Here’s diagram showing the mirror and the rays ithat are reflecting forward. The green dotted lines show where the virtual arc image will be and that they can pass through the aperture.
Click to view attachment
The rays from the rear of the arc are still not being collected like the pre-con set up.

DJ

Stop being a trouble maker, I don’t hate weldonjb

What would happen if you put 2 xenons 50% inside of the FL of the reflector, then used the focal piont of that reflector as the set piont of the precon?

so you bought the first xenon bulb with a ballast, but you are saying you should be able to just buy the bulb itself and plug it into a standard pc power supply?

If that's the case once I find a bright enough replacement (w/ reflector/precon) for my 400w football my existing lighting setup is going on the auction block .

No, sorry if there was confusion. You need a ballast for each bulb. $76 for ballast plus bulb, but only $30 for the bulb replacements. That is unless the electronics gurus know how to make a single ballast that can supply all three?

ahhh man....oh well..It looks like I'll go the single 150w MH route .

so after I make the announcement that I don't have the first idea what I'm talking about...would something like this work with 2 bulbs (if so could be incorporated to use 4 bulbs)