I have been wrestling for some time with the high divergence of my dye lasers.
The divergence is primarily due to the high Fresnel number of the gain medium. Simply stated, a large diameter gain medium,whether it is a YAG rod or a tube filled with dye allows higher order modes to resonate. Visualizing this geometrically, you can imagine that a large number of random zigzag paths could bounce between the end mirrors before they walked off sufficiently to hit a tube wall. This is one reason ion lasers with their high length to diameter ratios have such low divergence. Their low gain magnifies this advantage.
But, given a practical length limit the high power desired out of pulsed lasers, favors a large volume and that means a large diameter.
What to do?
Telescopic resonators that actually or optically move the end mirrors away from each other, decrease the Fresnel number and improve divergence, but at the square root of the length. They also increase beam intensity on at least one mirror. I tried this...blew up a couple of mirrors
Unstable resonators are counter intuitive, but effectively favor the lowest order mode because of their unstable losses. The light that escapes from this low divergence, sub region does so because of diffraction and acts as a high quality seed oscillation that is amplified in the rest of the full aperture. These are very had to align and the uneven pumping intensity as you move inward from the highly illuminated tube wall towrd the central axis plays havoc with the gain ratios. This never worked for my dyes, but is a very popular technique for CVL's, high energy YAGs and gas dynamic lasers.
Plano-Plano optics are harder to align, but all other things being equal produce lower divergence than more stable resonators, yet this improves the divergence only moderately.
However, I might have nailed it.
I placed a central solid quartz rod in my cylindrical cell which constrained the dye to a thin annular region between the rod and the inner wall of the tube. The divergence has dropped from tens of milliradians to sub millirad the the increased pumping intensity in the remaining dye has caused the efficiency to rise substantially. The trick, is that the cavity optics have to be adjusted to make this work. That is what I succeeded in doing tonight.
And, oh boy.
Nothing for hours, I was about to give up, but decided to change out the lamps and refresh the dye just to be sure and this allowed those telltale speckles to appear from operation just outside alignment. Then the magic began. With each adjustment of the mirrors the beam tightened and the power rose. Each time I thought, well that's great, it worked, I would make a final round robin of the optics and BOOM the power pegged the meter and I had after images. So, again AND AGAIN I lowered the discharge voltage to the lamps. Talk about the agony and the ecstasy.
I stopped without reaching a maximum and am now writing this. It will be difficult to sleep.