Dichros PBS cubes etc?

[rfourt]

oki i get it now-

i thought it would split it either way you light it .

thnx.

[James Lehman]

RGB to RGBV... that's tricky as you're expanding the gamut and adding completely new colors. I guess the sensible thing would be to let the V-channel help with the otherwise problematic purple line, as well as add a little punch to blue and white.

As for the function to do this, V=f(R,G,B)... that's maths beyond what I'm comfortable with. The reason I hate triangles (and circles) is trigonometry.

I was thinking of making a large palette of colors, that would "look" more like the real rainbow, using RGB, so it could be displayed on a computer screen. Purple would be at the high end, which in RGB is really half lit magenta. If we look at that palette as being a new continuum of hues, we can color match to it with RGB values and then turn that match into RGBV values. So a nearest match of purple in RGB would be fake violet, which would traslate to real violet laser. In RGBV, magenta isn't really a hue. It would most likely translate to violet with some RGB mixed in to bring it closer to white.

The coolest thing to do would be to set up a 6 volt battery and (4) 100K pots so you could manually adjust all four colors and measure the voltages. I think you could find a nice spectrum in there somewhere. Turn that into a linear continuum and make a fake version of it that looks OK in RGB. Translation is done by lookup table.

Of course, I'd have to have a four color solid state laser projector to do this! (hint hint)

James.

[buffo]

For what it's worth, Pangolin has support for a violet line in the color balance wizard. One could assign the 405 nm diode(s) to that channel. It wouldn't be perfect, but it would probably work better than just using the regular blue modulation output.

One of these days I'll probably try that just to see how it looks. But that's low on the priority list right now; too many other projects competing for my time!

Adam

[rfourt]

oki.. after seeing some companys adding 405nm to there list of dichros im getting more and more closer to building this.

Although the schematics have changed abit.
Although i managed to drive one of the diodes up to 180mW yestarday its not something i will go for. planing on running them between 90-110mW

So here is te new schematics. The old part is still there but this is for the combingin of 405´s.

My question is now, what i should use where the blue- "??" marks are`?


Thank you.

/Rickard

[buffo]

You can't mix more than two beams of the same frequency into a single beam of the same size. Even then you need the beams to be polarized. You mount one so the beam is horizontally polarized, and the other so the beam is vertically polarized. Then you use a PBS cube to mix them, and you get a beam of the same diameter and roughly double the power.

If you want to combine the beams from more than 2 diodes of the same wavelength, you need to "stack" the beams very close together using mirrors. You use the "knife-edge" mirror trick - the first beam (top, in your picture) gets reflected down at a 90 degree angle. You position another mirror at the next diode such that the first beam *just* misses the edge of the mirror but the next beam catches the very edge of the mirror and is reflected down so that it's as close to the first beam as possible. You continue this for all the diodes in each column, resulting in a beam profile at the bottom of each column that looks like 4 dots stacked up. (Obviously this requires either adjustable mounts on both the mirrors and the diodes, or *very* precise machining of the fixed mounts beforehand.)

Then you combine those beams together on the horizontal run, except that this time you stack them horizontally. So you'll end up with a beam profile that looks like a 2-D grid of dots. And yeah, it will be significantly wider than the beam from a single diode.

Note: A dichro is used to mix beams of *different* wavelengths into a single beam. You'll need one when you want to mix the combined 405 nm beams with the 473 nm beam of your other blue laser. But you don't need one in the upper section where you're mixing all the 405 nm diodes. It won't do you any good there. Dichros are useless when the incident and transmitted beam are the same wavelength.

Here's a thread that has a good picture showing what this knife-edge mirror trick looks like when you use a whole *bunch* of diodes. He's got 12 red beams stacked up on each side of the projector. Then he uses a large PBS cube to mix the horizontally polarized ones with the vertically polarized ones, for a total of 24 beams. (Well, actually I think it's only 23 beams, since one diode is missing on the left.) A final collimator helps tame the beam diameter somewhat, at the expense of higher divergence.

Adam

[rfourt]

ah oki... i get it now. wow i guess i have some alignement to look forward to then. hehe.,

Thnx alot for the info.

I got alitle confused as i have seen another picture like that from "Arctos" when they use PBS cubes. ???

here the link.
http://www.photonlexicon.com/forums/...1&d=1218526482

Why are they using PBS cubes only for mirroring.. isnt it better with an propper mirror instead?

[decix]

A problem I am experiencing with 405nm is that reflection on scanner mirrors varies with deflection angle.
Example: I draw a full circle at max. scansize... on my canvas the upper area appears very dark while the lower part is brighter.
Obviously those scanner-mirrors have a problem with this wavelength.
Anyone seen this yet? I am using Raytrack 35+ from jmlaser.com

[buffo]

I got alitle confused as i have seen another picture like that from "Arctos" when they use PBS cubes. ???
here the link.
http://www.photonlexicon.com/forums/...1&d=1218526482

Hehe... That picture causes a *lot* of confusion, which is why I didn't link to it in my post. People look at that beautiful Arctos projector, and they see all those square objects and immediately assume that they are PBS cubes. But they're not. They're actually mirror mounts! If you look very closely, you'll see that there is a first-surface mirror attached to each one.

A problem I am experiencing with 405nm is that reflection on scanner mirrors varies with deflection angle.

That's interesting... And I have to admit that I'm not entirely surprised to hear of it. 405 nm is very close to the limit for visual light, so you're probably getting close to the cutoff for the mirror coating as well. So "angle-tuning" effects become more pronounced.

Some results of 405nm in a RGV...

Very nice pictures!

An issue with the 405nm is that the spot size is very small so it tends to sit inside of the 650nm. If one could combine 2 of these and then defocus it slightly, perhaps it would have enough energy to fill up the space of the fatter beam an provide a more continuous color effect along the path.

This would be a good arguement in favor of using the knife-edge trick discussed above to get more power. If you had the space (and the money) to install 4, or even 8 high power 405 nm diodes, you'd end up with a larger beam that would more closely match your wide red beam.

Adam

[stanwax]

Brilliant
well done for trying and great for getting results. I know what you mean about the shades it makes with green and red from my experimenting - weird colours but really nice.

Rob

[p1t8ull]

The colours are great, but the beam profiles and alignment in those pics look superb... top job!