Collamination

[Fenechlaser]

A simple stupid question, but it struck me why this idea was never mentioned.

Can wide emitter cavity diodes like the mitsubishi diodes at 638nm be tamed using cylindrical lenses? I'm not referring to the ones used for the 445's but other ones in order to replace the spherical collaminator lens normally used in lasers.

A short FL cylindrical lens say 4mm would be used to focus the fast axis while a longer FL cylindrical lens say 8mm would be used to focus the slow axis i.e. the fast and slow axis are focused SEPARATELY with the cylindrical lenses being oriented perpendicular to each other... Lenses FL and lens thickness can be changed so as to be best suited for the diodes.. Obviously there would be a difference in divergence and the beam would be square and there is a great chance of a strange beam shape, but would such a setup work?

[planters]

If the emitter was infinitely small and the only difference between the fast and slow axis was the divergence, then even a radially symmetric lens would work well. The faster diverging axis fortuitously encounters the lens at a more powerful radius than the slower axis allowing both fast and slow to be corrected by a single lens. Incorporating an aspheric radially symmetric lens could be in theory perfect by eliminating spherical aberration. One way to think of a real world emitter with finite dimensions is as an array of infinitesimal points with only the single center point on axis. The rest of the emitter suffers from off axis aberrations such as astigmatism, comma and spherical aberration (focus). The larger the emitter, the greater the aberrations. If space is not a constraint then independently adjustable cylindrical collimators incorporating aspherical elements would allow the most compensation, But this is probably too bulky, unstable and expensive to be practical. Oh well...

[Solarfire]

Something like this here? On a scale of 1 – 10, with prisms = 1 then drlavas cylinders = 4-5 and this is around 8-10. Not to mention the costs for lens and custom hardware. Very hard to setup!

[Fenechlaser]

Something like this here? On a scale of 1 – 10, with prisms = 1 then drlavas cylinders = 4-5 and this is around 8-10. Not to mention the costs for lens and custom hardware. Very hard to setup!

Yes something like that. I know that setup would be very difficult, but for machining it is no problem for me since I do have a milling machine available. I just wanted to check out whether this idea has ever been tackled, and it appears it has actually been

About the costs, If there is sufficient interest and experiments show that such a setup would be practical with certain diode types, why not make a custom order and a group buy then? I know there is more to that, but with a sufficient quantity price would come down. Like with the 445, I would be willing to pay $60 for such a custom lens pair. And then why not develop a mounting kit for these lenses for those who don't have a milling machine or the 'willingness to invent'... Just saying that although not on such a difficult level, it has been done before with the 445...

@ solarfire: Details please I really like the way the whole system has been set up Is this yours?

[Solarfire]

I developed this mount back when everyone was testing and searching for the ultimate method for collimating and correcting the 445nm diode. This works pretty well but the lens selection depends on knowing the exact initial divergence of the raw diodes slow- and fast-axis. Due to the fact that these parameters can greatly vary from diode to diode means that you would have to determine the divergence for the diode at hand in order to be able to pick the proper focal lengths to collimate the diode (Cylinders have fixed focal lengths). So that pretty much makes it impossible to do any GB or even find a “one works with all setup” unless all diodes met up to tight tolerances what divergence is concerned.

2θ1/2θ2 = 10°/40° = f1/f2.

[Fenechlaser]

Once again, I'm impressed by your builds solarfire

I'm no diode expert, but what would be the difference between diode divergences for mass production diodes? The only thing that would change would be the beam profile, from square to rectangle. And from what I've seen around, square beams are not such a big problem provided they have a small divergence.

Optics that were suitable for the diode in question (the mitsubishi one; slow axis div. typ. @ 7deg, fast axis div. typ. @ 35, ratio being 1:5) would cost around $200. Lets say that total cost of diode and optics is $300 without driver cost and machining/materials expenses, cheaper optics would be better for your wallet.

Did you use off the shelf optics or did you have them custom made for your diode?

[catalanjo]

Funny how semantic mistakes produce amazing results sometimes.

Collimation is what is being discussed here ... BUT "collamination" is really a much better and more elegant description than "knife edging" when it comes to "laminating" long narrow diode outputs.

It's a tad Ikea, but by far the best description of a quad blue beam I have ever heard ..................................... Collaminated blue.......

Lamination + Collimation = Collamination...... Brilliant!

Cheers