Diode differaction vs astigmatism

The far field pattern of a laser diode shows an ellipse/line and it seems the accepted explanation for this is that the source is asymmetrical. The thin aspect of the junction causes that axis to have a higher rate of diffraction. However, while that is surly true, nevertheless the expanding beam which is collimnated in a typical pointer/lab laser is focused by a radially symmetrical lens. The optimal focus/collimation must be at best a compromise between the two expansion rates. Is this reasonable? If so maybe some improvement in spot size and shape do this astigmatic component could be achieved if the output of the initial collimnating lens were defocused slightly to allow a secondary lens in the beam path to be inserted to achieve best focus and then slightly tipped to counter the astigmatic component for the diode. Any thoughts?

Sounds like you're looking for this:
Beam Shaping with Cylindrical Lenses

Thanks for the reference. It's a good summery, however the concept of lens tilt is appealing in that it avoids the use and cost of two additional lens elements that would probably each require 5-axis mounts as well as rotation adjustment. They would of course give a tremendous amount of beam control, but do we need that much in that standard colimnation lenses do a pretty good job already? I'm also interested in that a single lens might be used to compress one axis in an array of multiple diodes while simultaneously improving each beam in that same axis.

planters said:

Thanks for the reference. It's a good summery, however the concept of lens tilt is appealing in that it avoids the use and cost of two additional lens elements that would probably each require 5-axis mounts as well as rotation adjustment. They would of course give a tremendous amount of beam control, but do we need that much in that standard colimnation lenses do a pretty good job already? I'm also interested in that a single lens might be used to compress one axis in an array of multiple diodes while simultaneously improving each beam in that same axis.

Click to expand...

I think a cylindrical lens would be better suited to correcting one axis of an array.
You can expand the slow axis with a concave cylindrical lens without disturbing the fast axis. Or compress the fast axis to match the slow axis with a convex cylindrical lens. I don't think any expensive mounts are required for that.

If you are correcting one beam, the tilted lens could work well enough.

If the "standard colimnation lenses do a pretty good job already" then why go through the trouble of doing this to begin with?

By the way, what kind of laser are you trying to correct? Diode I assume?
What wavelength?

They are twelve A140 445 nm diodes. The trouble may be minimal in that I am already installing beam compressing telescope to reduce the size of the array prior to passing it through a 10mm aperture DT scanner. The native divergence after leaving the Axaiz lenses will get worse by a factor of 4. Any improvement in beam quality is desirable and due to the small amount of correction I'm looking to accomplish I had considered a pair of matched convex/concave cylinders with adjustable spacing to add/subtract a small amount of astigmatism and rotation to adjust the angle. But, am I just getting too complicated? The spacing on the table needs to be set and if additional elements are going to be added it would be best to decide now. Thanks