I know this subject has been brought up many times but after nosing around commercial projector sites to see what the technology is doing, i noticed that the reflectors in many of them use a reflector that is more rectangluar than circular. The corners are still rounded off but the shape is more akin to a tin of mackerel..
I realise that the technology pushing these PJ's is different and they are not looking to cover a 15" LCD panel with light, but the shape of this reflector (see pic ) made me wonder about it's effectiveness at spreading light out over our dimly lit corners..

Anyone else thought about this or am i treading the wrong path?
Click to view attachment

I've always figured this is exactly what's needed. For any given lamp of a specific arc geometry there will be a pretty much optimum reflector design. In the world of CNC machines and high-powered computing I'm certain a near perfect reflector design could be produced. Only makes sense, eh? Perhaps it'd still take a few iterations to reach the endpoint in the design but yeah - a "rectangular-ish" reflector makes great sense to me.

good to know i'm not alone in my thinking!

when you consider that the light thrown to the screen usually ends up as a bright cirle with slightly dimmer edges, it makes sense to me that we need to throw that light from a different shape....

These reflectors are almost all basically parabolic reflectors which are the most common used in commercial projectors. A parabolic IS a circular reflector (basically a sattilite dish) that reflects the light strait out, meaning the shape of the reflector will basically denote the shape of the light if you were to shine it on a wall (like a spotlight). To use a parabolic, the reflector has to be the same size or larger than the LCD, so space is a concern. All these manufacturers are doing is taking the basic circular shape, and cutting the sides off to make it fit as a retangle in the box. When you cut the sides off a cone, you leave a flat plane one the sides like you see in your picture. Also by forcing it to fit in a rectangle with the bulb recessed, they reduce light spillage, although it reduces the efficiency of the reflector. You'll never get 100% out of a parabolic with the sides cut off. It's a trade off.

If you want to see a parabolic in action on a LL projector, check out Luckyme's WUXGA Plog.


~heli0s

#1292

You mean like this one.

Yup.

Agreed. However, given the constraints of a small arc lamp of fixed geometry, a desire/need for a non-parabolic reflector, and no other lenses between the lamp and rear fresnel, would you not agree that there must exist a reflector (or reflectors) of complex geometry which would make more efficient use of lamp output than either a spherical or parabolic reflector? That's what I was trying (not in a very good way ) to say in my previous post here. Oh, I'm also suggesting that such a reflector would not be placed at the focal point of the rear fresnel but instead closer to the rear fresnel such that the four-sided, optimized pyramid of light emanating from such a light engine would cross the plane of the rear fresnel in the best possible way. Just seems a rather interesting problem to think about. Certainly such an idea has arisen in these forums previously.

What you're talking about would have to be overly complex, including probably a combination of different reflectors in one to catch the light in different spots and reflect differently. I could probably be done, but I think unessessary. And I'm not sure of the efficiency something like that would be. It might produce hot spots where you have more or less light concentrated in certain areas. Even the parabolic has this problem with shadows from the bulb.

If looking for a very efficiently, small reflecting light engine, you'd almost have to use lenses combined with something like an elliptical reflector. For example, concentrating all the light to a point, and using the appropriate lense to expand and distribute it evenly accross a freshel. You could even do the same thing with a small parabolic (which is what most commercial projectors do only they go from bigger to smaller, where we COULD do the reverse), but you loose some light and efficiency. Using the lense to control the light should give you a much more even light distribution than relying on a complex reflector. I think some others are already experimenting with these types of lenses, but I don't know if they are concentrating the light first which I suspect would help.


Here's an example of two types used in commerical projs:

Agreed . . . again. I still think it's a conceptually "clean" idea that's fun to think about and "complexity" - if defined strictly in terms of number of components - would be at a minimum with such a system. Plus, efficiency - if defined strictly in terms of total amount of light hitting the rear fresnel divided by total lamp output - would be very, very high, indeed. Perhaps I'm full of beans though. Alas, I have neither the equipment, time, nor energy to pursue such a bit of wizardry. After all, can't argue with the results of many projectors built "here" thus far! (And if were worth doing from a manufacturing and/or cost-effective standpoint it'd probably already have been done.)

maybe some kind of hybird shape?

like this

Click to view attachment

arizona - if anyone is going to come up with some freaky winner of a hybrid reflector my money is on you!