Well I finished my lightbox and I noticed the fan grill I used is not completely symmetric. Therefor my reflector is a about 4-5 millimeters higher placed than my condensorlens! So I should lower it for about 2,5 millimeters to get the best alignment.

However as you may see (see pictures) this is a bit difficult. So I wonder. How 'bad' is it? Does this misalignment really bring me a noticable effect?






Ok just two more pictures, I'm rather proud of the result (first one is taken with a certain angle...). I used some glass wool in the U-profiles to mount the condensor lens. It should give it some room for expansion. The fan-side will be mounted at the side of the enclosure so hot air is sucked out.

My second question is if the distance between the bulb and the condensor can be related to unsharp corners @ the projection, or that the distance between the bulb and the condensor only affects the distance between your light box and fresnel.

The distance between my condensor and bulb now is 8 mm.

My understanding is that reflector/lamp misalignment may result in a fuzzy (more than one) virtual arc and potential fuzzy projection artifacts. There are a variety of methods to ensure proper alignment of lamp and reflector most involving a "visual" approach and not simply a theoretical measured approach. The condenser is more of a device for improving vignetting and possibly gathering a bit more light than a setup without a condenser. Having that distance "wrong" or not optimal would likely just result in a less than optimal image in terms of brightness and/or vignetting.

Edit: I think one must do the best one can given what one is working with (in terms of perfection of alignment). Once you're near the sweet spot for everything I'm sure there are diminishing returns in image quality per millimeter of additional adjustment given the components we typically work with.

If it was me I would straighten out the loops on the ends of the grill. Not straight, but more into a "u" shape. Then use washers along with the nuts. This should give you a mm or 2 in any direction.

I came to my second question because a friend of mine had fuzzy corners and he resolved it by decreasing the distance between his ARC and condensor.



I see, misalignment of the reflector may cause a double or more projections of the arc, also called fuzzyness. Due to this, there could also be a double projection of the image/tft. I get it! Thanks.



To my understanding vignetting is another word for the proces which causes hotspot. So I guess you mean a condensor can decrease the hotspot effect. Thinking about that, I think this improvement is caused by the effect that more light is broken to the edges of the screen while this is not the case for the centre of the image.

So without condensor the centre of the fresnel is closer to the ARC so that part is brighter, which means hotspot. So if I'm correct in this understanding the condensor has nothing to do with fuzzyness, only with (even) brightness. Only the reflector can cause fuzzyness.

It sounds logical. But you never know with optics! There's always more to know



I understand .

Ok I think I understand. I could also turn the fan grill 180 degrees. But than it could be the reflector is 4 mm to low . But I will see and bring along your idea.

Good luck. Tweaking things seems an almost never ending process.

And yes, vignetting is the term we use here for how evenly lit the screen is.

And you're also correct that theory can go out the window in the harsh light of reality.

Durachko and maybe Supraguy (since you educated me) if you read this:

I've been thinking this over. I thought that it is of no importance where and with what angle the light hits the tft, as long as this light gets into the triplet (especially at unsplit, at split the field fresnel must break the light equally). The triplet always creates a focus. So when a misalignment of the reflector creates more than one (fuzzy) projection of the ARC at the triplet, this should have no effect at all to the projection of the image itself...

Right

I cannot answer that from experience but I'm pretty certain ArizonaVideo has an opinion on the "double arc" thing resulting from reflector misalignment. There have been arguments that it actually might help vignetting to smear the virtual arc out by intentional misalignment if memory serves. Optics are a funny animal. Better to have everything "just so" than to take it on faith that your "gut" feeling is right. I know my gut can be very, very stupid.

Well, I think I understand.

The guy I was talking about told me that his 'vignetting' was better due to the fact he placed his lamp 2 mm away from the condensorlens instead of 8 mm.

Therefor there is more light in the corner, and for what he said, also more 'sharpness' because of it.

It's good enough for me .

I have my condenser 2mm from the bulb. I does help with all of the above issues. I know some commercial PJ use a double arc it improve brightness but their reflectors are very different.I know the 3M MP8620 does this. I have not tried this myself but it is something I am considering doing just to see if it helps at all.

I'm currently experimenting with the double arch position of my reflector. I first positioned my reflector very close to my 150w SA, this gave me 135 center white screen lux . Close position made the virtual arch (the arch you see in your reflector) bigger than the real arch and apparently more light went to the triplet. The problem I had with this is uneven focus and more glowies. Next I put the lamp arch positioned at the reflectors radius distance or 2x Focal length, this solved my focus problem completely, cut down glowies but also cut my white screen lux down to 50 lux! At the radius position, the virtual arch should match the real arch in size causing it to hide perfectly behind the real arch casing. That's why the offset position is used to "unhide" the virtual arch so it's rays can travel forward without obstruction. Using the offset position, I get about 60 center white screen lux (small increase), very few glowies and very nice focus. I could use more brightness but I know it will cost me even focus ability. So I'm finding there's always a comprise to get what I value most out of the final projected image. Anyone else have double arch experience?

Thanks

So, clearly it does matter 'how' the light goes through!

Of course, as long as it also gets to the triplet, it will be projected. But it seems, the smaller the ARC, the better the focus. Although indeed, the smaller the arc (or no double ARC) means a decrease of vignetting (virtual ARC is hiding behind real ARC ). So tuning is the key.
> I red this over today, the smaller the ARC, the less the vignetting conclusion might be wrong...

The glowies could be a result of light going through with too much of an angle, because you say in the same PJ you get no glowies with just one ARC image.

The decrease in focus could be the result of the triplet doing a less better job while it has to break more light with bigger angles (or maybe also due to light not going straight through the tft).

At least that's what I think. It's very interesting.

I believe having the virtual and real arch the same size puts there output light starting from the same focal plane so they focus evenly with each other. When the virtual arch gets bigger than the lamp arch, more light gets to the triplet but from a separate focal plane, that leads to focus problems. It may cause wider angles of incidence that muddy the focal plane of the lcd panel.

I'm going to try unsplit fresnels with radius position and see if I get improved glowies and better focus. I would suspect it would improve keeping the lamp arch in the radius position.