Back to Ultracold Atomic Experiment
Of all the cool projects I’ve been involved in during my school career through my Physics career, I’ve got to say that getting to build a custom laser is the one I’m going to have the most fun talking about. Not that it is the most interesting project ever…but come on, a home-built laser! It should be a requirement for getting a Physics degree.
The laser light itself is supplied by a laser diode, this one is actually a commercial one used in DVD reader/recorder units. The experiment requires the use of light locked to a 671nm wavelength, but this is not in the range of commonly available commercial lasers. Laser diodes actually emit a range of frequencies around a central frequency, which can be shifted by heat and also stimulation of a particular frequency. In fact by shining light at precisely the desired wavelength of light back onto the diode, this increases the diode’s emission at that one frequency, and can narrow the amount of light emitted at other frequencies.
The trick we will use in achieving a 671nm wavelength is to use a Littrow-mount feedback system. Light from the laser is shined onto a diffraction grating, which reflects most of the light. A small amount of light is reflected at precise angles based on its wavelength, (as is expected from the physics of diffraction gratings), and the grating’s first-order reflection is made to reflect back onto the laser diode. By changing the orientation of the grating, a particular wavelength can be made to shine back onto the diode (again, this is explained by the basic science of diffraction gratings – like a shiny CD, wavelengths hitting a grating are reflected at angles proportional to their wavelength), and so select the desired wavelength to stimulate in the diode.
The zeroth-order reflection (essentially the mirror reflection off of the grating) is reflected from a mirror, then out parallel to the original beam path. The grating and mirror are mounted as a unit on a swing-arm in front of the diode laser, so that rotation of the mirror/grating unit can be achieved by a piezo-electric unit. This allows the first-order reflection to be locked onto the laser, while not (largely) affecting the attitude of the outgoing beam.
The parts I have made were all from plans which the professor in charge of the Ultracold Atoms Experiment obtained through a group he had worked with at Berkeley. So producing the whole system was quite straight-forward and took about a week of actual work in total, once all the parts were together.

In the picture above, the diode mount is a modified lens mount with aluminum blocks attached to the front end (the laser emits to the left of the photo). The aluminum block with an angled cut and Allen screws on top will hold the diffraction grating, and the aluminum bar it sits on holds the mirror, as seen below:

The laser diode itself is behind the bluish collimating lens, which can be adjusted by turning the threaded steel ring around it. The black cylinder around this is a commercial diode mount, with a layer of "thermal goop" between it and a modified mounting unit, with more goop between it and the modified lens mount. The thermal paste helps heat get to the laser diode efficiently.


The unit as seen above sandwiches a thermoelectric cooler (TEC, also known as a Peltier device), which acts as a "heat pump" when current is applied to it, by moving heat from one side of its body to the other (the red and black wires attach to it). This is all attached to a massive (about 7-8 lbs.) brass block, and its mass and a rubber sheet below it help to dampen vibrations which might make their way through the optics table. Everything sits inside a housing box, which is lined with Acoustiblock material to keep out sound vibrations traveling over the table.
Back to Ultracold Atomic Experiment