Nitrogen lasers are quintessential devices for anyone wishing to build an actual laser from scratch. Like carbon dioxide lasers, nitrogen lasers can be truly homemade using some of the most basic parts imaginable. Unlike carbon dioxide lasers, nitrogen lasers require little or no feedback via resonator mirrors. In fact, optimal performance can be obtained using only a single mirror.

Homemade lasers offer unique advantages over diode lasers in that experimental possibilities are limited only by the imagination and persistence of the builder. The range of variations in design are seemingly endless. So many variables can be changed or modified, from gas pressure and gas mixture, to voltage and dielectric type. The experimenter can explore ways to improve laser performance as well as attempting to find novel materials or unusual ways to construct the laser. Adventurous attempts to do something different or unusual, be it for the satisfaction of one's curiosity or merely for a challenge, are ideally matched to a laser type that can be build almost entirely from scratch!

A nitrogen laser IS a laser, for at least two reasons:

1. By definition.
• By definition, it qualifies as a laser being "Light Amplification by Stimulated Emission of Radiation".
2. By relationship.
• By relationship, because it is no different from any other laser.

Resonator Mirrors provide at least two functions:

1. Feedback to enable net gain within an amplifying medium.
2. Selective feedback for control of wavelength, or group of wavelengths at which net gain will be obtained.

Any laser (at least that I am aware) should have a net gain if the amplifying medium is long enough. The net gain can also be obtained by increasing the amplifying medium density, as opposed to its length. But this often raises the threshold required for lasing. An example of this is where a nitrogen laser is used to pump a dye laser: the nitrogen laser beam is focused into a line on the dye's surface, using a cylindrical lens. The dye is often so concentrated that it absorbs nearly all of the nitrogen laser beam within less than a millimeter past the surface. This 'dense' concentration of dye molecules results in extraordinary gain (lasing can often occur along a line only 1cm long, with no mirrors required), but a high threshold (a nitrogen laser, by the way, would require mirrors if it were only 1cm long). The threshold is so high that pumping is impractical with a flash lamp (not to mention that only the surface of the dye would be usable), but the peak power of a nitrogen laser overcomes this threshold, and in doing so results in a simple laser requiring minimal or no feedback from mirrors. Here's an example (click on image for larger view):

Solid State pulsed lasers, such as ruby and Nd:Yag, generally require at least two mirrors for optimal performance, but the gain is still exceptionally high compared to that of most CW lasers. My ruby laser requires so very little feedback ( percentage of reflectance) that a piece of ordinary glass can be used as an output coupler. Common helium neon lasers often require output couplers that reflect around 99.9% of the laser wavelength in order to enable the low gain to overcome inherent losses. However, even a HeNe laser would lase without mirrors if the tube were long enough (at least in theory - I'm not sure if there would be a problem with competing laser wavelengths). In part, mirrors therefore serve as a method to artificially increase the length of an active medium. The level of feedback from these mirrors (percentage of output coupler reflectivity at a relevant wavelength, or group of wavelengths) differs from one type of laser to the next.

So the requirement for mirrors with any laser, and the percentage of reflectance (feedback) required for an output coupler, is dependent upon several factors which include:

1. Percentage of gain for a given host (by host, I mean the specific material that generally defines a laser type in which the light amplification takes place).
2. Length of active medium (or volume, if you will: if the length is not significantly greater than the 'width', then there is no way to confine potential amplification to a chosen path (we see an example of this in the atmosphere of Mars, where CO2 in the Martian atmosphere lases at invisible wavelengths when pumped by sunlight).
3. Density of gain medium. I don't think this applies with all lasers, but it does given the examples above (ruby and dye types). Gain can be increased by increasing the density of the active medium as opposed to its length, but does so at the cost of having a higher threshold (at least in my examples).

The difference between a laser with mirrors that reflect most of the light, some of the light, or no mirrors at all, depends in part on the factors mentioned above. From a practical perspective, low gain lasers always require very good dielectric mirrors. Solid state lasers often require good mirrors in order to manage their size and pumping requirements. Nitrogen lasers can be designed without mirrors while still maintaining a reasonable length, due to unusually high gain.

In the absence of feedback, the amplifying length of a transverse electrode channel should be theoretically limited to less than 1 foot (between 9 and 10 inches), in nitrogen at atmospheric pressure. If the electrode channel is slightly wider at one end of the laser, then the laser output from that end appears to be greater. I've never established an angle for this effect, so it's best determined by trial and error. A tapered channel appears to permit an electrode length that is greater than the theoretical limit. I have no explanation for this other than the possibility of an electrical discharge which begins at the narrower end of the channel and proceeds towards the wider. Such a discharge must certainly be equal or close to the speed of light. Study the following photographs which show an electrode channel that is 1 yard in length (.915 meters).