Novel approach to laser waveguide design

A project log for Homebrew space exploration program

New laser, DIY chip fab, hollowed out mountain & space ship, terraform Venus with orbital atmospheric heater, backyard nucleosynthesis

ganzuulganzuul 06/19/2014 at 22:065 Comments

Since I could be on to something here the following maybe-formal text describes what I believe really is a new invention:


Inside a laser waveguide where laser light is 'folded' in a zig-zag pattern using mirrors, the intensity of the beam is increased by effectively lengthening the optical resonator with the mirrors. This is commonly done with multiple mirrors where the beam bounces once per mirror, which results in a requirement to align a multitude of mirrors to a high degree of accuracy. Even in large industrial lasers this alignment is done manually, often by a man with a wrench, which is costly and time-consuming.

Invention and some theory:

None of the patents for waveguide lasers I have reviewed mention that; putting one pair of parallel mirrors at nλ ± ½λ and at a small angle to them another pair of parallel mirrors at nλ means that only the nλ pair causes light amplification, as the other pair interferes destructively. This lets the beam that is stabilized by the nλ pair of mirrors be folded a large number of times between the pair of mirrors at nλ ± ½λ.

Prior art, and such:

I assume that the reason the industry has not adopted this approach lies in the perceived difficulty of maintaining a pair of mirrors at a set half-wavelength from each other while the entire device flexes, expands and contracts from thermal expansion. It is however possible to have this degree of accuracy and control for laser light; which is especially true at the longer, mid-infrared wavelength of the CO2 laser where it can be done cheaply.
Piezoelectric devices respond with small movement to high voltage, which translates to a ultra-fine precision when the voltage is well-behaved. Usually piezoelectric actuators built for the purpose cost hundreds of dollars, but according to my interpretation of the degree of control needed for laser mirror alignment of a CO2 laser can be had for a mere pittance.

Conclusion / claims:

My plan, (should some soulless greyface attempt to claim this invention as their own...) is to actuate the mirrors of a CO2 laser with piezo disks and to drive the actuators with e.g. a 8051 microcontroller. To alleviate friction the mirrors can be mounted on an air cushion. The distance between the mirrors is constantly monitored,e.g. by a laser beam power meter, and the output of the sensor is fed back through the automation system to adjust the voltage of the actuators and keep the laser's power at the desired power level.


PointyOintment wrote 07/07/2014 at 04:09 point
My reaction when I read the invention paragraph: "Wait, what? Is that what I think it is? Wow!"

If the mirrors aren't extremely heavy, I imagine you could mount them directly on the piezo actuators, without an air cushion/maglev/whatever else. See the low-cost AFM (though it's not fully open-source):

  Are you sure? yes | no

ganzuul wrote 07/08/2014 at 18:06 point
Thank you so much! That will probably save me weeks of research! :D

  Are you sure? yes | no

PointyOintment wrote 07/09/2014 at 23:36 point
Another idea: Maybe you could monitor the distance between the mirrors by doing interferometry using another laser beam with a different wavelength. I have no idea if it would be feasible (with the primary laser's lasing medium in the way), but I thought I'd give you the idea just in case.

  Are you sure? yes | no

ganzuul wrote 07/12/2014 at 13:42 point
I like it. I have been thinking about an exhaustive search pattern and some simple machine learning to get the calibration done, but perhap more sensors is the right solution for an experimental setup.
Should interferometry turn out not to be the thing to do, then certainly laser diodes can be used in different ways too.

A CO2 gas laser actually has a bell-curve of evenly distributed frequencies, peaking at 10600nm, all corresponding to minute, quantum variations in the electron shells. When you get to a gas pressure of 25 atmospheres this Bell curve is smooth. The variation between the upper and lower end of frequencies is much less than half a wavelength, so destructive interference centered on 10600nm should still attenuate the 'side-bands' enough for lasing to be prevented. At higher powers some form of mode-locking might be needed here...
Still, we're dealing with more than a single frequency natively. Perhaps that can be turned into an advantage, like you say, instead of a disadvantage? =)

  Are you sure? yes | no

ganzuul wrote 07/12/2014 at 13:53 point
ED: Got a hold of conflicting information:

Still, they show the band distribution as far less than half a wavelength so the original principle remains unchallenged. Presumably the gas mixture influences this too.

  Are you sure? yes | no