Joss Institute Projects:

The nitrogen laser was discovered in 1963. As far as I can recall, it was the first ultraviolet gas laser, the first pulsed ultraviolet laser, and possibly the first-ever ultraviolet laser. It puts out short pulses of light at a wavelength of about 337.1 nm, a little shorter than the wavelength of an ordinary “blacklight” but not quite short enough to be described as “midwave UV”. This light is not visible, and it is even more dangerous than the light from small visible lasers. In addition, the laser operates on high voltage, so you should ONLY attempt to built it if you are prepared to exercise appropriate safety precautions. (I will list a number of these as we proceed.)

The nitrogen laser has very high gain. Excited nitrogen gas amplifies so well, in fact, that nitrogen lasers can usually operate without any mirrors at all, though especially with low-pressure ones it is common to use only a single mirror. (Using a mirror at one end of the laser lets you get all of the output at the other end, instead of having it divided between the two ends. With TEA nitrogen lasers there is another way to accomplish this, which I discuss in one of the videos.)

[Note: There are a few people who claim that the lack of mirrors means that a nitrogen isn’t a real laser, because it doesn’t involve a Fabry-Perot resonant cavity. That view is defective in two ways. First, the word LASER is an acronym that stands for “Light Amplification by Stimulated Emission of Radiation”, in this case, optical radiation: photons; there are also masers, which amplify microwaves rather than light. Neither version of the acronym includes any mention of resonant cavities, and any claim that a resonant cavity is necessary (for whatever mysterious reason) is ridiculous on the face of it. It is generally acknowledged, in fact, that there are quite a few naturally-occurring masers in outer space, which obviously do not involve resonant cavities. In addition, it is entirely possible to build a nitrogen laser that operates inside a resonant cavity; it just doesn’t happen to be necessary.]

Every laser must be “pumped”. That is, energy must be provided to it in an appropriate form. Some lasers are optically pumped. This can be accomplished with flashlamps, arc lamps, other lasers, other kinds of lamps including incandescent lamps, or even the sun. Others are driven directly by electricity. There are even a few chemical lasers. It all depends on the active material and its requirements, and on various other criteria that we aren’t going to get into here because there is neither space nor time. (If you are interested, there are various resources on the Web. I will link to some of these at the bottom of the page.)

All or nearly all nitrogen lasers are pumped by electrical discharge, usually in fairly pure nitrogen gas; but a nitrogen laser can even operate with air as its active medium, and that is what we will start with. I will note, however, that air is a lousy laser, and that if you can get or produce nitrogen that contains 0.5% oxygen or less, you will get much better performance. (I demonstrate the difference, using the laser design on this page, in one or more of the videos.)

If you charge up a capacitor to some nominal voltage and then connect it to an inductor (a configuration that is often described as a “tank circuit” or, more formally, an L-C tank circuit), the voltage will change in a characteristic way. People who deal with electronics often say that such a circuit is “ringing”. Because all real circuit elements have some resistance, and because the circuit is radiating some of its energy into space, the oscillation decreases in amplitude with time.

Here is a simulation of what you might see on the screen of an oscilloscope if you had a modest amount of resistance in your circuit:

[If there is less resistance the ringing decays more slowly, and it goes on longer. There is a formal way to describe this characteristic, a so-called “Factor of Merit” of resonances, which is called Q in electronics. Very loosely speaking, the Q of a circuit is the number of times the voltage crosses the middle line of the diagram before the height (or depth) of each peak is less than about 1/3 of the height (or depth) of the first one. The Q of the resonance shown in the diagram above is somewhere around 3, which is low.]

A capacitor in this type of circuit can “ring down” to a negative voltage that is a large fraction of its initial charge voltage. This is termed “voltage reversal”; in the diagram, the reversal is 50%; a highly resonant circuit can achieve a reversal of well over 90%. [Just for reference, and for the sake of comparison: if you are building an organic dye laser that is pumped by light from a flashlamp, you want the Q of the circuit that drives the flashlamp to be a little less than 1. That is, you do not want any voltage reversal at all. There are several reasons for this, including the fact that many capacitors are damaged by reversal, especially at high voltages.]

If we position two capacitors so that there is only a narrow gap or channel between them, we can do something interesting with voltage reversal. First, we charge the two caps simultaneously, so that there is essentially no voltage across the channel. Then we short-circuit one of them. As it rings down, the voltage across the channel increases quite rapidly. In a design like the one I am using for this laser, it probably takes only a few dozen billionths of a second for the channel voltage to become larger than the initial charge voltage. (That is, at the time when the shorted capacitor rings down into negative voltage.)

In principle, the voltage difference between the two capacitors can easily get as high as about 1.9 times the initial charge voltage; but if we have designed and built the laser correctly, at some point a discharge will form in the channel. (If things are not quite right we will probably see white sparks instead of a nice even glow, and if the channel is too wide we may not see much of anything.)

Here is a schematic diagram of such a circuit, with a free-running spark gap as the switch: