Why do I have wings?

[DTR]

I don't know much about beam expanders but you say the beam needs to be slightly defocused when you go into the expander for it to work right? My first thought is are you trying both unfocused(pre focus) and overfocused(past focus). If you are prefocus maybe you are getting some light hitting the rim of the lens barrel but if you are past focus and closer to the diode it may clear that up. Just a thought.

 

[Benm]

This depends. If you have an idea beam expander, the light coming in needs to be focussed to infinity for it to work. All a beam expander does is swap divergence for dot size, basically.

With these single-axis expanders it could be a bit different: ofsetting the focus a tad could lead to a more 'circular' (well, lets just say square?) far pattern. It depends on if the beam expansion matches the difference between slow and fast axes exactly or not.

 

[Alaskan]

The difference with my beam expander setup is it's a special Alaskan beam expander, no one else does it the way I'm doing it that I've seen yet. I'm using a defocused G2 lens in place of an expander lens, that's all, a lazy mans cheat. That way, I can adjust the spot size to match the larger lens on the end of the tube to its aperture and only need to get the focal length right for it to work. I could do the same thing by focusing the beam with a G2 to infinity, then shooting that through a concave or expander lens, then into a PCX lens matched to the concave lens to a specific magnification factor, i.e. a 50mm positive FL lens with a 5mm negative FL lens gives a magnification or expansion of 10X. But thats a lot of getting things just right for it to work, finding just the right lenses etc. By tossing the concave lens and just adjusting a G2 out of focus enough so the resulting spot size matches the diameter of a collimation lens placed a few inches away, this in effect gives me an adjustable expander lens which I don't have to care about its focal length (except that I don't adjust it so far out of focus the sides of the beam from the diode are cut off inside the barrel) and through that produce a far lower divergence output with one less lens compared to a standard beam expander.

I studied beam expanders for a couple of years until this light bulb turned on to just use a de-focused lens instead of a concave lens and it works great for me, except multimode diodes really need corrective optics, otherwise you get this wide stripe or rectangle going through the collimation lens and each plane; the fast and slow axis, each having slightly different focal lengths to adjust to infinity. Without corrective opltics, try as you may, one or the other axis won't perfectly adjust to infinity, although very close.

Placing the PCX cylinder lens on the output of my defocused G2 lens helps expand the slow axis some so it is much fatter, but the output is still a rectangle, as below for my NUBM07E diode. I drew a tighter outline of the size the beam was prior to adding the PCX cylinder lens right on top of the out of focus G2 lens. Being that close, the beam is still extremely small, so will work much like if it were at infinity focus for either the NDG7475T or the NUBM07E, but not the NUBM44, with my particular lens:

Below, might as well put a photo of the cylinder lens I used on the 2nd page next to the other photo. This is the PX-CY cylinder lens I added on top of the output of my NUBM07E 465nm laser pointer to widen the slow axis, it doesn't correct the slow axis all the way, as you can see in the photo above, but helps a lot. The glue I used to fasten the added optic is just rubber cement, so I can easily remove and clean things up later if I want to do something different: