To be an expander, which takes a collimated input beam at infinity focus and expands it to a larger collimated output beam at infinity, you would obviously need the right concave lens on the input. There is a poor mans work around for this, I found most PCX lenses will collimate to infinity if I mal-adjust the focus of a G2 lens in a pointer to produce a spot the right diameter on the face of the larger lens, and then slide that larger lens back and forth to find infinity focus, or adjust both the G2 in the pointer and the larger lens to find the right positions to collimate to infinity.
If you design the host around the large lens, it's easier. Although you might need to match the laser diode to the lens, if you are given one (like I gave you) and unable to order the perfect diameter and focal length because some diodes have small angles of radiation, others larger, like the NUBM44, this makes some large lenses unsuitable for some diodes, as too much lens aperture is waisted, or the spot size ends up too wide for the lens and you loose some of the light due to being truncated, or cut off. That's when finding the right concave lens comes in handy, you can then make up for that so everything comes out right anyway.
If the lens I gave you isn't the right diameter for a NUBM44 at its focal length, maybe better to use it for a 520 nm green laser diode? Maybe just use them for a collimation lens for a host instead of a real beam expander? It would be less lossy that way anyway. I don't care if only 50% of the aperture is used, but prefer 80%.
trinh hong phuoc is a rich man trinh hong phuoc is a rich man 
