Mutil watt 445nm using beam splitters?

Bluefan said:

The prism trick probably doesn't work, prism have a low dispersion, so the setup would be huge even to combine 2 wavelengths. Efficient gratings would be worth an experiment. It would still make a huge setup, but you might be able to combine more than 2 beams.
In the end the wavelength trick is probably not worth the trouble.

Click to expand...

That's just quitter talk.

JK. I was imagining 10 diodes with about 1-2nm difference between each. Prisms would work but you'd probably need several of them, and it would help if they were made out of diamond. Of course the beams would need to start out close together and the angular change from one beam to the next would have to be small and precise - then the prism(s) wound only need to apply a small amount of shift to bring them together. Using gratings is probably a better idea. As you noted, the setup would probably need to be quite large.

The concept is very similar to chirped pulse amplification except that it's spatial rather than temporal compression.


The more I think about it, the more I'm liking this idea. The technique could be applied to any pre-existing knife-edge setup to improve beam quality. Just add temperature gradient and prisms/gratings, and adjust the convergence-point distance toward the near-field (away from infinity).

It may not be perfect, but any temperature/spectral gradient could be leveraged to compress the size of the beam (along the axis of the gradient), and at least make it smaller. Think of it more as beam-shaping a combined beam rather than beam-combining.

Yeah, let's buy some custom diamond brewster angle prism, I bet those are cheap

I'm very familiar with chirped pulse amplification, I've some some experiments with a Ti:Sa laser. I used both prisms and gratings, and I can tell you gratings work much better.

Let's do the math, we can get 1800lines/mm, giving d(sin(theta_m) + sin(theta_i)) = lambda. Say we use a littrow configuration as exampe. then 2 d sin(theta) = lambda or theta = arcsin(lambda / (2d))
then d theta / d lambda = 1 / (sqrt{1 - (lambda/(2 d))^2} 2 d)
That means the angular spread for 1nm in wavelength is roughly 1mrad. If you could place two beams 1mm from eachother you'd need 1 meter before they hit the grating. 1mm isn't much, and 1 meter is long. A series of dielectric mirrors could fold the path without much loss. A grating in littrow configuration can reach 75% efficiency for p polarisation.
Taking a 2 meter path would give you 2mm separation per diode, having 2nm separation in wavelength between the diodes would also do this. I don't know the temperature dependency of the wavelength, I'll see if I can measure that someday.

It's possible, but I wouldn't call it practical.

andy_con said:

lol does not exist and will not exist.

that makes no sense whatsoever.

to combine 10 diodes you would need to knife edge them together and then use some lens to bring the size of the final beam down

the same way arctos do it

Click to expand...

were is ur safety glass's?

Safety glasses for an RGB laser are pretty much non-existent. The laser on the picture is a bit big for a pointer, so it makes a solid setup that doesn't move. Terminating the beam would be needed, but I'd also recommend shielding the beam path more on the picture. I'd definately not work on the laser (alignment and stuff) without safety glasses, but for a proper stationary setup it's not always needed.

Hi, was following this and wondering where one obtains the cube to combine two lasers into one? Also, once combined do you use another focusing lens to get the beam right? Or are the beams focused first and then into the cube? If afterwards, what type of lens and where would you find it? There is a thread of a handheld 3.5w build on tue forum...beautiful build! Problemos he doesn't explain how it was done. I would like to someday try to build a two diode laser like this but haven't a clue where to start...can anyone here point me in the right direction?

You can get a cube at laser show parts...

First you collimate both lasers, that's using a lens to get a nice straight beam out of the laser diode. Then you combine two beams with the polarising beam splitter.