Can someone please explain to me

[Stiffler]

Why the fast axis is so hard to tame??? If beam divergence = beam angle (this might be an incorrect assumption) then it should be as simple as a single curved lens or mirror; What am I missing?? If Clay Paky can get a nearly parallel beam from a non coherent omni directional light source, why is it so hard to do with the 445s?

-Eric

[-bart-]

Yet it may look like rocketsience as some people want it to look like. Collimating laserdiodes is not. Having it available to the public cheaply may be another thing.

What clay paky did in their sharpies is really a lot more complicated and requires more custom optics than collimating a simple laser diode.

No offence, but most of us here are a bunch of amateurs that only have trial and error to go by.
Optical design engineer is a profession, and there countless softwaretools available to aid him.
Working out a collimating solution for a broad stripe diode is a piece of cake for someone who has some little experience in Zemax.

[mixedgas]

The problem is laser shows are not profitable enough to enable one to buy Zemax or other optical software. The software has a limited market and is very expensive. The diode laser's mode shape is just complicated enough to make doing a mathematical ray trace by hand, nearly impossible.
Not too many individuals are going to shell out 5 grand for a one year license for a piece of software.

So everything gets done by trial and error, instead of proper custom optics getting made.

Optical modeling is very expensive. For a past employer, I wanted raytrace software for a new led project. Our product sold for 5 thousand dollars a unit. We would have had to sell about eight units just to pay for the software and the led-laser add on module.

It comes down to lack of demand, plus the fact that the 445s are not a very good diode. They are not intended to be. If they were good, like a single mode diode, the interference fringes would show up in the video image, which is un-desirable.

Steve

[Stiffler]

Bart-

I downloaded the trial version Zemax, just to check it out.

Steve-

That is completely understandable, but with all the hobbyists here dropping thousands of dollars on lasers & modules just for the love of lasers, you would think one of them would have bought the software. Now I know the 445s are crappy, but they are CHEAP; Do you think they could be tamed using computer modelling? Surely if this is the case it would make sense for someone here to get into the custom DIY optics business.

-Eric

[badger1666]

Why the fast axis is so hard to tame??? If beam divergence = beam angle (this might be an incorrect assumption) then it should be as simple as a single curved lens or mirror; What am I missing?? If Clay Paky can get a nearly parallel beam from a non coherent omni directional light source, why is it so hard to do with the 445s?

-Eric

what 1mrad or slightly less not good enough ??? still better than any green any of us could afford so why bother getting it any smaller.
the only one that could do with tightening up and cleaning up is the mitsu p73
that said i think either could be improved, its not easy to set up and there are no lens the right fl on the market that i can find to do it.
they would have to be custom made, i would love to find two plano convex cylinder lens one @ 2mmfl and one 6mm fl setup as +
and see what results the mount would be a nightmare to make and setup

[kiyoukan]

well there is a much better way to correct the p73 but doing a diy grin lens is a bit of a pain.
i have seen before someone had access to optical software before with the 445nm diode specs in and a best case lens setup but the lenses cost around 90$ uncoated and coated around 135$.
as mentioned the greens are now the factor in our beam specs.
as the decently priced ones are around 3mm 1.5mrad or 4mm 1mrad.
there is a solution that i have had in my head for about 1yr now but i dont yet have the means to test it.
3d printers have come down in cost to even make it possible to use these for production.
so if someone can get access to a good dlp resin 3d printer let me know.
and no not printing the lens just a very special holder for some lenses.
that can only be made with a dlp 3d resin printer.

but i tired to do the math and found myself so confused it was not even funny. so i stuck to trial and error.
even sent test lenses out to a few people to test but the current setup was never beat to make the prices worth it.
even still even our reds are running at 4mm by about 1.1mrad

[planters]

This subject of diode beam correction comes up every so often and the solutions are always a trade off. If the final result is to have a decent amount of power then the trade off here is many higher quality single mode diodes that need lower order correction, so a lot of simpler and cheaper lenses or a few higher power multimode diodes that need fewer, but more complex lenses to correct their higher order aberrations. I don't think you need to go to Z max @ this point because characterizing the diode output is the first step. Beyond focus and astigmatism which are currently pretty well neutralized with the spherical and cylinder optics used here you get to spherical aberration and off axis coma. These two aberrations can be minimized by using longer FL optics and dedicating single lens trains to each diode to avoid the inevitable off axis position of compound beams ( @ the cost of a larger beam line) . Going even further really needs an interferometer to analyze the higher order aberrations which can then be input to produce a custom wavefront corrector. This CAN be done. I believe Zygo can do this.

Interestingly enough, I believe the quality of the diode beam is already near the limit of the non laboratory laser projector environment. Now, please believe me, I am not bragging here and I only say this to make a point. The P73 diodes in my projector have about 1/2 the divergence commonly reported on this forum because each diode's rotational position is adjusted independent of it's X/Y position and the individual cylinder lenses in each pair are in turn free to rotate independently (badger might remember this issue). The point is execution. The Flex mounts don't drift much, but they do drift at the scale of 100's of urad and unless your base plate is several inches thick the alignment of the lasers and maybe even the diodes in a larger module degrades depending on how the projector is supported. Also, I suspect I am not the only one who looks at a projector's individual lasers and tweaks the focus of a previously "perfect" collimnating lens and notices it gets just a little sharper.

I think the scanner is the weak link here. If larger mirrors can be driven then the divergence of all the lasers can be reduced (or of coarse the power increased with more beams). I also noticed the large mirrors on my EM 4,000 series are not flat, even @ rest and so final focus of each laser has to be done thru the scanner. This might apply to the smaller mirror models as well. Finally, when you measure the spot size of one of your lasers in both the X and Y dimensions and using the scanner, measure the thickness of a horizontal (for the Y) and vertical (for the X) line in the ILDA test pattern, the lines are always wider. The scanner tracking/jitter/mirror flex has contributed to the "divergence" and superior optics will not reduce that added degradation.