white beam with NO dichros?

[CountFunkula]

...you could always just cut out the middleman and go with an Ar-Kr beam

[dsli_jon]

...you could always just cut out the middleman and go with an Ar-Kr beam

Lol, Yes! Yes, this will work!

- RGB, all into one 'spot', right into the fiber!

funny...
j

[planters]

I have some large diameter (1mm ) power delivery fibers used for pulsed dye lasers ie. 590nm. I'll try launching the co-aligned beams of an RGB projector into this and see if the scrambling produces a more homogeneous output. Also by readjusting the dicros it should be possible to cause slightly misaligned beams to converge onto the entrance aperture of the fiber to see if the acceptance angle would allow a practical, non-dichro, purely knife edge approach. However, the quality of multi diode beams and what I suspect the fiber will demand in terms of acceptance angle ( in other words relatively long focal length focusing lenses ) might cause a lot of loss.

[Phredy1]

Assuming that somehow fiber can be used for the visible light spectrum, and we can use it for show lasers - why couldn't we use fiber for beam shaping? If you were trying to get a nice round beam from something like a 650nm diode, why couldn't you shape the input fibers to match the beam shape from the diode ( === ) and then when it comes out of the other end of the fibers it would be round?

Also for Marc's idea, if you could separate 1/3 - 1/3 -1/3 of the bazillion fibers and input R, G & B each into each 1/3 wouldn't you have an output of a RGB stripes? Then if you could randomize all the fibers somehow, you would basically have something similar to pixels? I.E. white light output with all 3 lasers on?

Ya Marc, in my mind it seems like it should work too....

[dsli_jon]

why couldn't we use fiber for beam shaping?

Yes, you can somewhat 'clean-up' a diode-beam, but you still will-have to collimate a 'flashlight' at the OC... (output coupler...) you'll also have a challenge with far-field convergence, if all your source-beams are not dead-perfect on center, going in... which they *can't-be* in this proposed scenario, since they're coming in from multi-angles.... and w/o the 'fine-tuning' positionability of summing-dichros for each-color on an MM1, you'd likely drive yourself mad... it's even a challenge to find the ultimate all-capturing 'sweet-spot' with an ion, where all lines are inherently-aligned flawlessly - as you tweek the 'tuning', on your IC (input coupler...) you can watch reds or far-blue fall right-off in power, yet stlll have a fairly decent 'beam'... and if you simply increase fiber-diameter, to give yourself more 'slop', you just have that-much unwieldier of a beam on the OC-end... a 1mm fiber would give you a rather *monster* beam (certainly n/g for graphics-werk...)

..so, with all this effort / loss, etc - you're still better off simply summing with dichros and shooting direct into yer galvos...

imho...

Now, if we *could* have an RGB 'all-in-one' all on the same die, well, then we'd really have something that might 'stand a chance' of 1/2-way decent output... albeit, lossier than a simple lens-array at-aperture, and you'd still have the OC collimation to deal with... *sigh* one can dream...

j

[Stoney3K]

I guess it would be possible to launch a few R/G/B beams into a fiber with the right numerical aperture, there's one simple problem:

The physical laser modules will probably be in the way, causing the beams to be spaced too far apart.

Isn't that easily solved by using a threesome of mirrors on kinematic mounts and folding each beam path in half? As Mixedgas said, the problem can be overcome, but you'd need an optical table of a meter long. It's easy enough to fold that optical table in half or in three bits, isn't it?

Laser modules are difficult to make any smaller, but mirrors and mounts can be tiny.

[Things]

'scuse the crude MS paint drawing, but I was under the impression that if you did this, the beams would still come out going completely different directions.



Also, you'd have the problem of re-collimating all the beams, since you'd have multiple focal lengths to deal with?

[Xytrell]

They don't exit as collimated beams, they exit as a blob - similar to the output of a raw diode. The different focal lengths could be fixed with a special lens.

[laser_freak]

I thought achromatic doublets are designed to fix the issue caused by the difference in the refractive indices of different wavelengths. Wouldn't the different focal lengths not be collimated the same even with an achromatic doublet?

[mixedgas]

Mixedgas can overcome the problem for 4-6 sources and handle the uneven color as well. He even has the documentation on how to do it. Let he or she with a few thousand dollars knock on Mixedgas door, and we'll talk.

Steve