Home Return to Main Page

 Notes on Version 3 and Version 4 that lead to Version 5 

RW-MOPA ver 3 was based on two pass experiments done with version 2 with a PORRO HR in the last amp to double back the beam back through both amps and being picked off by a right angle prism. The porro was set lower to reflect the beam back thru at a lower altitude and with a slight tilt to allow pickoff prism to pass first beam and catch return beam. This two pass arrangement generated 3 to 4 times more power. But due to short distances the pickoff couldn't catch all of the beam due to increase divergence in the beam and short angular distances of version two. RW-MOPA ver 3 corrected this by providing a much larger optical bench and therefore required a separate case for the laser head and another for the power supply. At 250 mjoules the two amp two pass was later modified to a two amp 1 two pass and 1 single pass due to damage of right angle prisms as intense hot spots developed and pitted one right angle prism on return from the second amp which produced 190 mj. I found damage had occurred to the laser rod in the second amp (bubble crack appeared). Some of the problem was trying to push 25mj from this oscillator was allowing some transverse modes to return. The motor q switch at the rated speed was just producing single pulse and 25mj was pushing too hard. I now only had effective 3 amps out of the 2 physical ones. The beam made just a single pass through the 2nd amp next to the oscillator instead of a porro to bounce a second pass.

 rwmopa3.gif diagram of version 3. As you can see I had to have patience and a lot of alignment paper. Also a trick by drilling a hole in folded over alignment paper (making it two sided) so I could find and pass the first beam and find the return one.

Overall the double pass had problems and this began the design changes for version 4 as 1/4 inch rods are too narrow for multiple passes and cause possible beam degradation due to non linear optical rod pumping in various radial rod locations or radial gain variation from possible overlap beams


Instead of double pass a new design on four physical stages would be used as a single pass.

rwmopa4.gif diagram of version 4.

So the version 3 laser was gutted and the optical plate reused and reworked and the laser was rebuilt with the new configuration. The motor q switch was replaced with a HR/dye cell and the tank OC element was extracted and placed in a lens tube and gimbal mount. (Details on the HR/dye cell are covered in version 5 application notes). In this new configurations I would give up on motor q switches for the excellent capabilities of saturable absorbers. This would give me better mode control and pulse control. A lot of work done in the sixties and seventies showed the benefits of dye q switches. (they are cheap too!)

I sawed off parts of the tank laser I no longer needed for the oscillator. Also for the amps but kept the fixed right angle prism which was next to the q switch. I rotated the prism upside down on amp 2 to allow it to send the beam to amp 3. Also I cut off the fixed prism mount from some spare tank lasers to use this right angle mount for the oscillator. A bolt threading into the bench and allowed the left/right adjustments and rotating the prism controlled altitude. I used two for the resonator and one more to bend the beam toward amp 1. All left handed beam directions was handled by the prism rotated to it's upside orientation.

This new configuration yielded 8-12mj q switch pulse at 40 nanoseconds from the oscillator. I aligned up the beam pass through the amps using alignment paper and if I had powered up the amps, they would have yielded a total to 190-220 mj output by my calculations. The design had to be folded to fit the case that had been used before. Two cases of the same size held the laser head and the other the power supply. But due to the increase in the number of amps, I needed a larger power supply case to house all the new caps needed. I decided to redesign the power supply to increase the power supply current from 40milliamps to 300 milliamps by purchasing a larger transformer. 3 tank lasers had been taking 15 minutes to charge up and going to 5 lasers would then take 30 minutes and I felt that was too long to be waiting. During this phase, I decided to go ahead and redesign the laser head as well to get rid of extra prisms used in a folded oscillator and amp stages.

 With version 5, I will try to have all rods tilted a little to cause back reflections to land harmlessly on component metal and not go back down through previous stages and hit a front surface of a prisms and come back up again etc. Version 4 could not be adjusted to prevent this since the right angle prisms on the tank laser body would also be shifted when the body was shifted to move the rod surface to a tilt to the beam axis. Reducing the extra prisms by straightening the configuration hopefully would cut down on reflections as I was trying to trace down what the alignment paper showed with version 4. I had expected by not using motor q switches and using a plano-concave resonator and round optics that I would have this perfect circle of TEM00 mode. I found instead a spot with what looked like the mode but it also had these little side lobes: 1 above and 1 below with a faint output. I found this similar to what I experienced with the tank laser from earlier versions (always some little lobe etc) . During my studies, I knew that misalign mirrors can make it hard to eliminate higher order transverse modes and the tank laser used a planar resonator. Plus I had concerns about the changing angles of the rotating prism. Not having a source to compare my alignment paper readings, I continued to try and get that perfect spot. One thing I learned is you can get the occasional tiny dark spot in the lens expanded view of the beam when reading the alignment paper as this is caused by diffraction with a dust particle. The lobes I had no answer for except I always could diminish them by reducing the variable diaphragm aperture and the output power until they are too weak to show up I guessed. Therefore in version 5 my main goal was to eliminate the prisms in the oscillator by not using a folded design and see if that would clear up these lobes. As it turned out it didn't and now it seems to be a by product of extra reflections from some of the optical components in the cavity/resonator arrangement.