Quote:
Generaing high-power green light
The key to the Danish diode is a novel tapered design. A tapered laser-diode is used as a pump source for second-harmonic generation – a process that results in over 9 W of nearly diffraction-limited power in a narrow spectral region. The light is then frequency-doubled in a single-pass configuration using a periodically poled MgLN crystal, which leads to 1.58 W of output power at a wavelength of 531 nm.
End quote.
Without reading the paper, this is my best guess. This is a MOPA with a new twist.
MOPA is a master oscillator, power amplifier. The group starts with a normal laser diode at ~1040 nanometers. Its mounted on a long heat sink, and butted against that diode is a second diode. The second diode does not have mirrors cleaved into its surface, and is tapered in one axis like a funnel. The second diode is a long chip amplifier, and driven by the beam from the master laser, boosts the beam power.
The second chip has a very high index of refraction compared to air, and thus acts as a waveguide for the amplified light. The end of the chip is very tiny, and there is a high electric field where the light exits the taper.
This field is high enough to drive the doubling crystal without the normal resonant cavity required to achive non-linear conversion.
To fit a normal doubling crystal at the tip of the taper would require a very tightly controlled dicing and orienting procedure for a material like kTP or BBO, and all of us know that in low cost lasers, KTP varies all over the place in performance. So when you cannot find a doubling crystal in nature, you look to making one. You need a crystal that you can orient in place, on the mounting slab, and consistantly. This is where periodic poling comes in.
Periodic poling is a simple sounding procedure, but getting it right in practice can be tricky. The blank crystal is polished and the optical faces and coatings are polished and applied, respectively. A custom chip is made
with metal stripes set at a spacing of a few hundred nanometers, in relation to the wavelength you are working with. The crystal, with the chip placed against it, is raised in a oven to a point above its Curie temperature, where the domains in the crystal lattice can be pushed around. A high voltage is applied to the metal strips, with a polarity change between the electrodes, ie +-+-+-+- and this voltage is kept on while the crystal is cooled. This sets up the right atomic structure for doubling to occur, and the manufacturer has precise control over the alignment, but the process is tricky and the setup is only woth doing if your making thousands of crystals, or need one really special crystal for a megadollar process. As
The MOPA-PPO scheme is very useful if your making tens of thousands of lasers, but there is a caveat. Paraphrasing the character Goldilocks, in Robert Southey's 1837 classic, '
The Three Bears", "The financial porrige must be just right", Normal DPSS yags are made by hand for the most part by gluing parts in place with a micromanipulator and low cost labor. Ole Jensen's technique is going to require specialized robotics to be done on a large scale,as the parts are measured in microns, not millimeters.
Thus this method is going to required a "Killer" application, needing 10s of thousands of units per year, to justify the cost of setting up the plant to make these. That application would be medical, custom lighting, or laser video. The small device size allows the little known, but difficult, third beam combining technique,"Coherent Combining" and thus power would be scalable in arrays while maintaining good beam quality.
For our purposes, this device has two advantages, Within reason, you can make a wide variety of wavelengths. Blue would be the next logical step, and red would be easy. The second advanatage is there is no upper storage limit to interfere with rapid modulation, and within reason, you could modulate the amplifier very rapidly.
Periodically poled materials can do more then double, They can also do parametric techniques, where you can make the sum or difference of a pair of lasers, and you can end up with a tunable laser , or wavelengths that you cannot do with a classical diode, such as yellow. Or even mutiple wavelengths off the same laser.
The classical disavantages of diodes are well known to this forum, and perhaps there will be issues with divergence from the tapered field.
We all know how fast diodes can be destroyed, and on a side note, Pangolin should perhaps send them some Lasorbs to get in on the "ground floor" so to speak.
Steve
Originally Posted by Laserman532
and it is a VERY rare event when one of our lasers roll off the table
Reply With Quote
:
