slit spatial filter

[kecked]

arrangement1.jpg

Interesting. Skip the relay lens and just do a single axis. Wonder if the noise is mostly in the slow axis we are expanding.
Planters method is for sure the best way but this would reduce the optics train length and might reduce most of the really bad noise.

Play time!

[logsquared]

I think the "noise" is in the fast axis and gets condensed down to the "wings" when the slow axis is corrected. Interested in the results you get when using 2 cylinders instead of sphericals. I would imagine they alignment would be harder due to having to rotate the cylinders. The focused spot would be way more of a line. So, maybe it could be "cleaned-up" better than with sphericals.

[catalanjo]

I agree with Woody on this, but I think going to a Cyl lens first in front of diode will only be worthwhile if we can get closer than 1mm from the LD output and I still haven't figured out how to place and fix the "lens" in question. .

At the moment we are just correcting fast axis not over correcting to get a slit which might be filterable in this manner which means we are talking very very very short focal lengths or alternatively using three cyl lens just for fast axis (big losses).

Mr P has listed a whole bunch of things that could go wrong with this type of setup on thread " http://www.photonlexicon.com/forums/showthread.php/25512-fiber-optic-beam-combining", ........I suppose it boils down to whether all these different aberrations, are sufficiently recognizable in terms of beam shape so that each one he has listed, can be corrected individually just by looking at the results in the far field.

Cheers

[kecked]

Tried this out in one axis. Not worth it. Tooooo hard to align. Tried it in two axis with double slits. Result is same as easy method. Move on.....

[planters]

I agree with Woody on this, but I think going to a Cyl lens first in front of diode will only be worthwhile if we can get closer than 1mm from the LD output

Distance has no benefit, optically. You can make the set up more compact, but the precision will have to increase as well and a lot, as it is the FL to the second power.

The focused spot would be way more of a line. So, maybe it could be "cleaned-up" better than with sphericals.

The shape of a line better matches the slit, although the focus from a typical diode is not round. It is just a miniature version of the far field spot and that is pretty close to the square hole from four blade edges. However, the slit will have to be adjusted more carefully in rotation to align with the stripe. You don't have to worry about that too much with two axis filtering.

If the major concern here is quality and space then I would recommend filtering all the colors at once as kecked proposes (a little compromise here).

But, use high quality achromatic lenses of the longest FL that you can accommodate. Increase the strength of the magnets to hold the blades more firmly. Replace the hole in the mount through which the beam travels with a long slot that can be screw tightened at one end and install high ratio levers just beyond the far edges of the other blades to allow more precise movement of these blades into the beam.

Alternatively, use some folding mirrors to tightly zigzag a longer path between the lasers on the deck. This can even involve retro-reflections and more reflections on a single mirror. The effect of working with longer FL lenses is huge. Alignment and beam centration on the lenses becomes less critical, chromatic aberration shrinks, off axis aberrations reduce and the focal position of the slit and the relative precision of the blade placement and its selectivity improves...to the square of the FL. I would go this route.

[kecked]

There is a method used in IR spectrophotometers where you reflect the beam back and forth in a crystal so as to increase path length and sensitivity. You could do the same with two mirrors and bounce the beam a dozen times to increase the path length. This could let the beam shrink to fit scanners better or allow the use of much longer fl lens as planter suggested. I did try 50mm lens and it was real easy to align. I would not try shorter. PS: the achromats worked well in combination with a 1mm pinhole. Power dropped from 1W to 800mw and good luck finding a color fringe. Now to assemble this kludge into something working.

What do you call accurate chromium an achromat of course. And how do you clean them? at a laundromat! I guess that's not a very precise process definition.