A Really Big Laser?

[VDX]

... look at the 4th image with the opened 25W@975nm-diode here: http://www.reprap.org/wiki/Laser_Cutter

They combined three single 9W-chips into a 105µm-fibre by shaping the beams to a flat horizontal bar and 'stacking' them to a square and focussing into the fiber.

The single-fiber-per-diode modules (the opened 5W and 9W diodes) are made by simply fixing the fiber with 5 to 10µ space direct to the emitter without any focussing ...

Viktor

[dsli_jon]

...But the module is probably as expensive as my house.

If you're talking the 'BigBlue' chassis-design, well I hope your house costs more than $10-12K!! If you're talking the '8-up', ya, dunno.. I'm sure 'DTR' can comment on that..

... look at the 4th image...They combined three single 9W-chips into a 105µm-fibre..

Impressive, but.. What is the 'emitter size' / beam-profile of those chips like, vs, say, one of the 3W 445s? I think a lot depends on that 'initial quality (thus, 'achievable spot-size' at the fiber-in..) There are, for example, 'VCSELS' that are *quite* high-power, but the 'native' beam is so good, it doesn't take much more than a ball lens to get it all into a single-mode fiber, I'm sure even a 50µm would be no-prob.. But, these almost 'multi-multimode' Blues? sheesh.. Pretty much like tryin to put oatmeal into a hypodermic needle..

.02
j

[VDX]

... AFAIK the emitter size of the diodes is something like 100x1µm with the max. energy density in the center - so a fiber collecting 90% of the energy should be no problem ...

Viktor

[planters]

Good posts! I'm dancing around, basically three different combining techniques. The Osram or Nichia modules can be used as a convenient mounting block followed by conventional folding/knife edging. A collection of individual diodes could be moved closely together and focused and aimed to converge on a single fiber or each individual diode could be launched into a fiber and then all the fibers are spliced into a single fiber.

I'll leave the first option for now as it is pretty straight forward. I am not sure the multiple diode into a single fiber will work and not because of practical challenges like alignment. A multimode fiber say 100um in diameter will scramble whatever wavefront the diode had to begin with. After hundreds of glancing reflections off of curving walls, the M^2 of the output has to rise. The question I had from above is something like this; if a typical multi-mode fiber has a NA of .2 then its acceptance angle is within a cone with a vertex to base-radius ratio of 5 to 1. That is an input angle of about 12 degrees 1/2 angle. With a FL of 20mm it should be easy to knife edge, say 4 converging diode outputs on a 100um fiber with the knives mounted within that 12 degree cone. But will a 100um, 0.2NA fiber, output be focus-able to the same far field spot as the co-aligned and parallel diodes themselves. I suspect not, because the diodes' emitters at 5um x 60um are much smaller than the area of the fiber exit. Even the combined area of the emitters is smaller than the exit of the fiber. This rise in entropy has to be detrimental.

The third option confuses me. You can't just fiber pigtail a diode whether glued to the face or free space launched and then splice an arbitrary number of these fibers to a single common fiber of the same diameter. This reminds me of perpetual motion. You seem to be getting something for nothing. It's like the problem of using a tapered fiber to concentrate a laser. The back reflections prevent you from achieving an intensity greater than the launching intensity. 10 x 100um fibers, spliced into a single 100 um fiber requires you to taper the area of each of the input fibers to 1/10 the original area...at some point, in some way. If the common fiber were enlarged to 300um to avoid the taper then the beam quality that you could generate from this larger source would deteriorate proportionally.