some projection beam manipulation question

[shoujin]

I have this project where I need to split a video projection beam vertically into red, green and blue beams and superimpose them. My original idea was a color filter in front of the projector, then a mirror and trichroic cube setup such as this to superimpose the split beams onto each other. Any ideas how this can be improved or otherwise be made to work? Thanks.

[NobleGas]

It might help if you explain what end result you're trying to achieve; I'm struggling to understand the purpose.

[buffo]

I have this project where I need to split a video projection beam vertically into red, green and blue beams and superimpose them.

Can you be more specific? Splitting the video into red, green, and blue components is easy using a trichro. That's actually what they were originally designed to to. (They were used in 3 CCD video cameras to split the incoming light into red, green, and blue portions to send to the individual CCD elements.) However, they also work in reverse - that is they can also combine 3 separate Red, Green, and Blue light beams into a single output, as seen here:

My original idea was a color filter in front of the projector, then a mirror and trichroic cube setup such as this to superimpose the split beams onto each other.

This is the part I don't understand. Do you want to separate the beams so when they recombine they aren't perfectly aligned or something? Can you draw a picture? And why does the beam path length matter?

In theory a pair of trichros plus two adjustable bounce mirrors on each beam path between them (so 6 mirrors total) would allow you to change the alignment up or down and left or right for all 3 color separations.

Adam

[shoujin]

This is the part I don't understand. Do you want to separate the beams so when they recombine they aren't perfectly aligned or something? Can you draw a picture? And why does the beam path length matter?

Sorry, please check the added diagram in my first post. As you can see the distance matters as two channels travel longer than the middle one, and so the beams (projections) become bigger and will not be overlayed correctly.

[buffo]

The picture and link helped!

If you add enough bounce mirrors you can make all 3 paths equal length. Reflect the green path down, then back up, and finally straight in to the trichro. For the red and blue paths you should only need 1 bounce, assuming you measure very carefully. Make all 3 paths the same length and it should work as intended. (Assuming your trichro cube faces are large enough to accept the image, that is.)

Adam

[shoujin]

I understand your solution. Thank you. I will see how it works out in practice since I don't have an imaging optics simulator. I also had this idea of using two PBS cubes stacked on top of each other. And if you were to choose which method would you try first, provided this one also should work? PS. In the above diagram the stray red, green and blue beams are ignored but I could filter them out with the same gel filters before they reach the first PBS cube.

[buffo]

Dual Trichros in the top/bottom configuration as shown in your sketch will also work. In fact, it probably would work better, since all 3 colors will experience the same loss (just two bounces off identical mirrors). My idea would have 3 bounces on green and only 1 bounce on the other two colors, which would make the green intensity appear dimmer. Also, your idea is more compact.

You should be able to vary the positioning of the lower mirror for red and blue to ensure that the path length is the same for those colors as it is for the green light.

Adam

[shoujin]

Yeah, that makes sense.

Now to just find some affordable x-cubes. I quoted few chinese companies and thy ask for a 50x50x50mm minimum 100 USD for each cube. Wow, I remember these being cheaper, even sold as unused parts for old projection TVs that never made it in one of those.
Are there maybe different quality grades of trichros?

[buffo]

I quoted few chinese companies and thy ask for a 50x50x50mm minimum 100 USD for each cube. Wow, I remember these being cheaper, even sold as unused parts for old projection TVs that never made it in one of those.

The Fujinon Trichros that I purchased back in 2000 cost me $50 each. Most of the early ones were surplus (new obsolete stock) but some of the later ones I saw for sale were pulls from used equipment. The individual color faces on those units were quite small - less than 1 inch square. So if you're being quoted $100 for new units with 2 inch square faces, that's probably a pretty decent price.

Are there maybe different quality grades of trichros?

I'm sure there are different grades. As a general rule, the small Fujinon units were very high quality when compared to the larger cubes that were more commonly found on E-bay after the supply of the Fujinon units dried up. No idea how that will affect your project though. (In the end, reduced quality = greater signal loss and a reduced intensity on the output.)

Adam

[dkumpula]

I bought a couple to play with about two years ago from SurplusShed somewhere in the US$20-50 range each. They were the smaller ones - about 250mmx250mm. I can't speak to relative quality, but for lasers, the losses on these things were stupidly high. Greater than 50% as I recall, so I quickly abandoned the thought of making a laser projector using them as opposed to traditional dichros.

-David