Do the Dr. Lava optics correction lenses provide any beam expansion/mrad improvement?

[Cyparagon]

With just a spherical lens, one axis is collimated (generally the slow axis - the long dimension at aperture and the short dimension at a distance). The fast axis (the short dimension at aperture and the long dimension at a distance) remains a problem because it diverges faster and you get a line. We minimize this by expanding the fast axis so its divergence decreases. divergence is inversely proportional to beam size.

It seems to me that you should be able to get one axis perfectly parallel and thin with the Aixiz lens, and then use one of corrective optics to converge the other axis down to a small diameter, and then the other corrective optic to stop the convergence of that axis and make it too parallel.

You're going the wrong direction. They could be used like that, but see bold above. You would end up with a tiny beam, but terrible divergence on both axes. The large emitter size and the multimode nature won't let you get away with having good divergence with a beam that small.

 

[rhd]

Thank you!

This is making sense now. The piece I was missing was that divergence is inversely proportional to beam size. I didn't realize that you couldn't have a very thin beam, with good (minimal) divergence. I knew that in practice this was more seldom seen as a beam characteristic, but I always just presumed this to be a consequence of the quality of the laser source and the quality of optics, as opposed to an actual relationship between the two measures (and an inverse one at that).

So to bring this inquiry back to the practical application that inspired it -

I'm building an RGB with a tight 532 beam, and a very nice 640nm HL63133DG diode. I was initially going to mix it with a 405 so that all three colours would have a similar beam shape.

I decided that I wanted 445 for it's rich blue, but of course this brought along a mis-matched beam shape relative to the R and G. I was thinking that the corrective optics were the answer, but I'm now thinking this to be an incorrect approach. If these optics will just end up expanding the fast axis to create a more square beam, but with a larger general width throughout, then I'm not really getting the 445's characteristics closer to the R and G. In fact near the source of emission (or at least the beam's exit from this RGB labby), I'll have a less pleasing mix, because both axes of the 445 will be wider than the RG, as opposed to just one.

Does that reasoning track?

 

[benmwv]

I would try using the cylindrical lens on the 445 and then regular beam expanders on red and green. Then you would have a square and two circles all about the same size instead of a square and two tiny circles.

 

[Cyparagon]

Don't waste your time with getting all the lasers the same size/divergence. You won't notice any anomalies unless you're looking for them. It will help to correct the blue's axis, however.

 

[rhd]

Don't waste your time with getting all the lasers the same size/divergence. You won't notice any anomalies unless you're looking for them. It will help to correct the blue's axis, however.

Fair enough. This isn't a projector either, though who knows what I may use this for in the future. This is purely an academic endeavor. Having failed at creating two and three colour builds in the past, I've finally splurged for proper adjustable mounts, and I'm using decent diodes. So while I'd love this to look as good as possible, there's no reason that the beams have to be perfectly identical.

Though a nice single-mode 445 would be wonderful right now! But at $200 for 50mW of single-mode blue, I don't think it's a viable option for this particular foray into RGBs

 

[Cyparagon]

How about $59?

 

[HaloBlu]

You have considered clipping the 445nm beam to cut the squareness down to round & perhaps reduce the width as well?
Sorry if its been mentioned & I just missed it.

 

[rhd]

How about $59?

Wow! Great Find! Right under my nose this whole time? I've been tracking official Opnext (And Osram) suppliers for pricing on these, and it has always been $200 +. I may give this a shot!

EDIT: Finding a 3.8mm module is still pretty tough

You have considered clipping the 445nm beam to cut the squareness down to round & perhaps reduce the width as well?
Sorry if its been mentioned & I just missed it.

Clipping? I am completely unfamiliar with this process. Is it what I'm envisioning? (literally placing something on either side to block part of the edges?)

 

[HaloBlu]

Vision confirmed. Tho for better roundness you would want a metal plate with a hole a bit smaller then the width of the 445 beam. It would get clipped all around the edges.

I believe I first saw clipping mentioned as an unintentional but useful effect of 3-element aixiz on 405 & 445. Clips any 'wings' & may be clipping the line on 445s. If your using a 405-G-1/G-2 try the aixiz first.
3-element aixiz is better then a clipped G-1/G-2.

 

[rhd]

Vision confirmed. Tho for better roundness you would want a metal plate with a hole a bit smaller then the width of the 445 beam. It would get clipped all around the edges.

I believe I first saw clipping mentioned as an unintentional but useful effect of 3-element aixiz on 405 & 445. If your using a 405-G-1/G-2 try the aixiz first.

I'm going to test this out! Presumably I can use a small aluminum disk, and drill a hole through it?

 

[HaloBlu]

Yo here we go... http://laserpointerforums.com/f65/445nm-lens-comparisons-51456.html#post710729 I think its mentioned elsewhere too but the way Jayrob has it covered they probably don't add much if anything.
Check out Jayrob's 'small aperture sleeve/retainer' @ Update2 of post 1.