Yes. I have a 150 watt CO2 laser for burning and cutting things. My interest in visible handheld lasers is just that, to create the most visible beam at the longest distances physically possible.
In one of our earlier discussions, we hypothesized that instead of starting the optical train with an off-the-shelf 9mm collimator, then using cylindricals to correct the beam shape, then using an expander to lower divergence, one could use an approach that integrates these into fewer steps.
My current plan is to use only two lenses rather than 5-7 as would be the case above. Among other benefits, this greatly reduces reflected power, having a lower number of surfaces. Both lenses will be cylindrical plano convex. One will collimate the fast axis and the other will collimate the slow axis. Since the two axis diverge at different rates, the fast axis collimator will be placed much closer to the diode than the slow axis collimator. This will allow me to collimate each axis at the precise distance such that I have a round/square beam shape.
Further, I will choose a distance for the lenses that also collimates the beam at a diameter that has very low divergence, thus obviating the need for an expander.
I found elsewhere on this forum that 520nm diodes have a 1um x 14um emitter. If we assume this is correct and use the fast and slow axis divergence from the NDG7475 datasheet, then these models below should be quite accurate. Overall, the system consists of: NDG7475 -> PCX -> PCX. It has a final beam shape of 20mm x 20mm and divergence of 0.04 mRad x 0.14 mRad.
Below is an image I just happen to come across while searching for something else. It resembles what I am trying to accomplish, so I guess it is not a completely novel concept. Thoughts?
