Stupid idea? Focusing several 445s

Don't why I'm even suggesting this. Just something crazy I've thought of while sitting @ work.

I've got this 10-15W fiber laser system complete with cooling etc. The laser exists the fiber in the most perfect circle....although not collimated.

Now, I know it's a bar type laser that is somehow focused into some optics then into the fiber.

Would it be possible to remove the laser bar and focus several 445s in this optic system which would then exit the fiber? Even with all the loses one could get some nice watts.

I don't really have any use for the high power 808nm laser. I'd happily rip the bar out to make a high power 445.

Good idea but im pretty sure all of the optics would be AR coated for 808nm which would result in some serious losses at 445nm. Im not too sure if fiber optic cables are AR coated/built for certain wavelengths but i've got a feeling they are, that would be another large amount of power lost. Aligning all of those diodes would take some time, but I bet it could be done.
Even with the looses I would love to see a 5w+ fiber coupled 445nm laser

If the losses ive seen with 808nm coated prisms are any indication,
you may see approx 30% or more in loss.

If anything you could play with the idea, but may be a good idea to
replace the fiber's with ones meant for this wavelength.

Im no expert, but I think its not only the AR coating that would
be an issue here... IIRC the size and dimension of the fiber itself
also plays a role. Id try it with one LD, then see what kind of losses
you end up with first.

I have a few pieces of fiber from 836nm communication modules, but
have never tried it. Knife edging to me is the best route in combining
445. The flat beam profile just screams COMBINE ME!!! :san:

You'd have to know the fiber NA and core diameter to make a good gues of the losses. That's probably difficult to find out, so see if you can find info on the diode bar. The emitter size on those may be standard and if the emitter is larger than a 445nm diode's emitter you have more chance of succes.
The fiber input end probably has 19 fibers in it at 500um spacing, I don't think you can separate those. Focussing 19 laser straight in such a small space would be difficult.

What kind of system is it now in? An intact fiber coupled 808nm diode bar is very usefull for end pumping lasers, that's the whole reason they are fiber coupled. It would be a shame to rip such a unit apart for a project with a lot of risks.
I'd love to have such a 808nm fiber system, I'd make a kickass yag laser, maybe even double it to 532nm. Probably a bit more expensive and involved project, but far more epic and realistic then a 445nm fiber coupled laser with a terrible beam quality. Try collimating the 808nm to see, even in a matched system the beam qaulity is very low, but good enough for a pump beam.

@Bluefan

You do have a good point. Perhaps I shouldn't kill the SDL 808 system. It's still fully operational with red aiming laser etc.
Maybe what I should be doing is stripping my redundant 800mw 532 and pumping 10W 808 into it

The 19 fiber thing is interesting. You saying that these coupled 808's are routed through a number of fibers then into one fiber?
I still have a number of 445s to play with. Just looking for another option than the knife edge way.

Yeah, you need to have a more specific fiber size to allow passage of the 445 beam. You can't take just any fiber, (particularly fiber made for 808nm) and jam 445nm light down its guzzle. It would be an interesting project to try, but I warn you: Getting a beam focused down into the tip of a fiber is harder to do than it sounds. You may simply end up destroying the ends of the fiber by burning them with 445nm, making them unusable.

These are multimode fibers, so the focus isn't that difficult with a proper xyz stage. The reason these things use 19 fibers is because it's easy to stack 19 fiber in a sort of round pattern:
At least that what I read somewhere, I can't find where it was.

I plan to pump a crystal set of a 300mW laser with it. That laser probably had something like a 2W pump diode, so I won't shoot in the whole 20W.