Presented is a simple, low-pressure design for a nitrogen laser using a Blumlein / transmission line arrangement similar to that described by James Small in The Amateur Scientist. The newer design utilizes a laser tube 30cm long, made from milled plexiglas. Capacitors are fabricated from thin (0.020") printed circuit board etched to form two capacitors. The laser operates on either pure nitrogen gas or plain air (both flowing gas) at pressures ranging from 25 to 75 torr. Using plain air as a lasant it was observed that pulse repeatability is poor (about 50% of pulses fail to lase) and asymmetric beam distribution - where the beam was concentrated towards the spark-gap side transverse electrode - was observed, especially at higher than normal operating pressures. With pure nitrogen as a lasant performance was excellent as was shot repeatability and output power was also increased four to five-fold over the use of air. The laser operates at 10kV to 15kV with a maximum voltage of 20kV set by the breakdown limit of the epoxy/glass dielectric employed. The laser, with high peak powers and a 1 cm wide beam, makes an excellent pump source for dye lasers or for exciting fluorescence in materials.Introduction
Likely the easiest of all lasers to construct, this pulsed gas laser produces an ultraviolet laser beam at 337.1 nm. Pulses from this laser last 5-10 ns. The laser itself consists of a simple laser tube with long transverse (parallel to the tube) electrodes and high-voltage capacitors usually made of dual-sided printed circuit board. Expensive optics are not required since the laser operates superradiantly due to its enormous gain! As well, only a basic vacuum pump is required which can pull a vacuum of 25 torr (a commercial single-stage pump or a converted fridge compressor will work here). If built well, this laser will also operate using regular air (which is 78% nitrogen) but at reduced power outputs.
This laser is highly recommended for the amateur looking to construct a laser 'from scratch'. While the average power output will be under 10mW, the peak power of the UV output pulses will be on the order of 100's of kilowatts. On its own it is useful as a source of fast pulses of intense UV light for exciting fluorescence and other uses. It is extremely useful as a pump source for a tunable dye laser.
CAREFUL! Although the output of the laser does not look particularly bright when the diffuse light is viewed as it strikes a piece of paper remember that you are not viewing the laser beam itself (which is in the UV and hence invisible) but rather fluorescence from the material the beam strikes (e.g. all white paper fluoresces). The beam is much more powerful than it appears to be. Safety goggles or glasses with plastic lenses which absorb UV radiation (many inexpensive safety glasses do indeed absorb UV) are highly recommended. Remember, too, that even with goggles one must NEVER attempt to view the actual beam (i.e. by looking at the front of the laser). Goggles or glasses only help protect against accidental exposure, not intentional direct viewing!Two Types of Nitrogen Lasers
There are two versions a nitrogen laser which may be constructed. The first approach, outlined on this page, is a low-pressure design - it is more 'traditional' and requires a vacuum pump. It produces a pulse of 5 to 10nS in duration which has a cross-section of about 10mm by 2mm. The power output and beam shape lend themselves well to pumping a dye laser. The second approach is a TEA version of the nitrogen laser. TEA lasers (for Transverse Electrical-discharge at Atmospheric pressure) do not require a vacuum system at all as they operate at atmospheric (or greater) pressures. Either laser, if built well enough, can operate using regular air as the lasant - a real boon to the experimenter on a budget. In the case of the TEA laser using open air no gas housing is required. The TEA laser is covered in another page on this site. Although the idea of a TEA laser is tempting, since no vacuum pump required, construction this type of laser is by far more 'touchy' than the type employing a vacuum pump. Ultra-fast discharges are required for the TEA laser (10 times faster than a 'normal' low-pressure nitrogen laser outlined on this page) which makes optimization of the transmission-line capacitors far more critical. Further, TEA lasers have a beam profile of only a few mm wide (usually 1.5 or 2mm) making it difficult to use as a dye laser pump source.
The lasers described on this page are of a 'normal' (i.e., low-pressure) design and are a better bet for construction and utility than TEA lasers. This design is quite 'forgiving' in construction (except for the choice of PCB for the capacitors - more on this later). From the constructor's point of view, the only additional equipment required is a vacuum pump and even that need not be expensive as only a moderate vacuum of 25 torr is required.Basic Nitrogen Laser Structure
The laser tube itself  consists of a housing containing two transverse, flat electrodes parallel to the output beam. A suitable tube may be fashioned from two strips of thin copper or brass shim stock 30cm long cemented with epoxy into a split black ABS sewer pipe or two plexiglass slabs acting as a vacuum housing. Capacitors may be made of either dual-sided PC board etched to form two capacitors or fabricated from aluminum foil and plastic sheets. The critical factor in the construction of this laser is minimizing inductance in the discharge path to generate the incredibly fast pulse needed for laser action to occur. The self-inductance of a capacitor of this type is linearly proportional to the thickness of the dielectric  and so keeping the dielectric thin keeps the discharge fast. In my original laser I used very thin PC board (0.015") for the capacitors as it has a higher capacitance (and hence energy storage) as well as lower inductance (and hence faster discharge) than normal PC board (which is 0.045" thick). I have tried, and failed, to use thicker board myself and have heard from a number af amateurs who have also tried the same but no one seems to be able to make thicker board work. Be sure, then, to obtain the proper (0.015" or 0.020" thick) board or make your own capacitors from flat plates and sheets of dielectric (there are numerous other amateur designs on the web demonstrating the construction of capacitors from aluminum foil and plastic film - see the LINKS page on this site).
The laser gain is extraordinarily high, so much so that a single pass of light down the laser tube amplifies radiation enough to produce a powerful output beam. No mirrors are required for this laser - this is what is called a superradiant laser. The output beam simply passes through a thin glass microscope slide to exit the laser tube. This also eliminates mirror alignment problems which plague homebrew laser builders (and professionals alike) with most other types of lasers. The addition of a rear mirror will, however, boost power over 250%. In my case I've used simple, inexpensive, first-surface mirrors inside the vacuum housing of my first laser itself as a rear reflector with good success. If you put the mirror inside the vacuum housing like I did be sure to have adjustment screws on the mirror which are accessible outside the vacuum housing so that it may be aligned properly.
A high vacuum is not required for this type of laser: a vacuum of 25 to 75 torr is sufficient. This is easily produced from an old sealed refrigerator compressor used in reverse. In the past I have acquired compressors from scrap yards which work well for this purpose. Choose a compressor from a small fridge which has only two metal lines (intake and output) and be sure to scavenge the start relay as well (usually mounted on the side of the compressor itself). Carefully cut the lines to release pressure (these days the refrigerant gas is usually already removed for environmental reasons) before cutting the lines. Leave 10cm of line protruding from the compressor to allow attachment of rubber tubing. The only concern with these compressors is the probability of oil contamination as refrigerant oil is backstreamed into the laser - if possible, obtain a single-stage vacuum pump as these do not have the same issues, otherwise consider a trap to minimize backstreamed oil vapour from entering the laser channel which will necessitate periodic cleaning of windows.
These lasers are usually supplied with flowing nitrogen gas from a compressed gas tank. Air, which is 78% nitrogen anyway, may be used if the laser has been built well and has enough optical gain but this will reduce the output power about five-fold (as well as reduce beam quality - see notes below). Standard-grade nitrogen of the type used by welders is quite suitable for this laser so more expensive high-purity gas is of little benefit. An oxygen regulator of the type used by welders can be used if an adapter from a nitrogen-bottle fitting to an oxygen-regulator fitting is used (I did this years ago. It worked well and was quite inexpensive). The output from the regulator is set to less than 5 psi and a needle valve is used to regulate gas flow through the laser tube. Pressure is regulated in this manner: the vacuum pump is connected directly to one port of the laser tube, and the inflow of gas regulated by the needle valve to control tube pressure. Another inexpensive option is the use of liquid nitrogen as a gas source. This is fine for short term operation but tends to boil off quickly whether being used or not.
To begin, see the  Scientific American Article in the June 1974 edition in the Amateur Scientist column which describes an easily constructed laser of this type. This article is also contained on the CD compilation of the Amateur Scientist available on the web. This is required reading before attempting to construct such a laser.Important Principles Of Operation
The design of this laser is obviously not like other gas lasers such as the ubiquitous HeNe. The sole purpose of the two-capacitor Blumlein switch arrangement and transverse electrodes is to dump energy into the lasing volume in an incredibly short time frame.
By now you should understand population inversion (If not, start with a good basic reference ). To maintain lasing action there must be a higher population of the active medium (nitrogen gas molecules in this case) at the upper lasing level than at the lower level. The problem here is the lifetime of these levels: the upper level (at about 9eV) has a very short lifetime of about 20 nSec  while the lower lasing level (at about 6eV) has a much longer lifetime of 10mSec. The only way to get lasing action to occur is to quickly populate the upper level ensuring population inversion occurs . This is done by a massive electrical pulse where electron collisions cause the preferential population of the upper energy band first (Were it not for this effect of being able to pump the upper energy level first, this laser would not work at all). After about 20nSec though this population of molecules at the ULL will decay to the lower level (a metastable state) where it will stay and so lasing action will quickly cease after the electrical pulse. Nitrogen lasers are hence self terminating in the same manner as the pure neon laser. To make matters worse Nitrogen molecules at the lower state absorb UV light strongly. The pulse length of the low-pressure Nitrogen laser, then, is limited to about 20nS (the point where population inversion is no longer possible since half of the molecules in the upper energy state have decayed to the lower state) and likely somewhat shorter than that since the ULL lifetime is, by definition, the point where 50% of the excited molecules will decay to a lower energy state - at some point before that time has expired lasing action will terminate - the ULL will decay to the point where the inversion DN is insufficient to generate enough gain to overcome losses in the laser. Incidentally, the lifetime of the upper-lasing level is dependent on the pressure in the laser tube. As pressure rises, the lifetime shortens  according to:
This behaviour is an important factor in the design of a TEA nitrogen laser where the pressure is 760 torr (one atmosphere) and so the lifetime is about 2ns. In a low-pressure laser like this one, the lifetime of about 20ns relaxes the need for extreme discharge speed (meaning a Blumlein transmission line which is not completely optimal will still work).
Being a molecular laser, in which the active species is the N2 molecule, energy levels (ULL and LLL) are vibronic levels and are defined by both electron energy state as well as vibrational energies. Because multiple closely-spaced energy levels define each band (the ULL and the LLL), this leads to a relatively large spectral bandwidth for this laser's output of about 0.1nm. Since the laser is superradiant, wavelength selection is ineffectual.
The time scales involved should bring to light the importance of a very low inductance discharge path for currents in the laser. Any stray inductance (including the self-inductance of the capacitors and inductance of the spark gap itself ) will slow the discharge time and hence stop lasing action. The discharge must proceed in considerably less time than the ULL lifetime to ensure efficient pumping. Capacitors must be very low inductance indeed. In epoxy-glass PCB material, inductance increases as thickness of the dielectric  so the thinnest possible material must be used. If 0.015" PCB is used the inductance will be low enough to allow the laser to work but the board will only take 1kV per 0.001" of dielectric thickness so the capacitors can be run at a maximum of about 15kV. If 0.015" (or 0.020" at thickest) PCB is unavailable, the next best option is likely to fabricate capacitors using flat metal plates or foil and thin dielectric. This arrangement also allows for changing the dielectric should it fail due to the high voltages across it (which _does_ happen with dull regularity). A number of amateur laser enthusiasts have successfully done this (see the Links section on this site).
Of course the Blumlein/Transverse electrode arrangement is not the only one which will work  but it is the most efficient at pumping the entire lasing volume in the shortest possible time. Among other things, a fast risetime discharge inhibits arc formation in the discharge and instead encourages a voluminous "spread-out" discharge between the electrodes - required for laser gain. Nitrogen lasers belong to a class of lasers called self-terminating lasers which include several other gases and many metal vapours. Other self-terminating lasers such as the copper-vapour laser have much longer upper-energy level lifetimes so the time for discharge in those lasers can be much longer. Copper-vapour lasers usually do not use the transverse electrode arrangement. Still other self-terminaing lasers such as Neon with a green output at 540.1nm have shorter upper-level lifetimes so their discharges must be even faster than that used for Nitrogen (1.5nS as opposed to 10nS).
Finally, the E/p ratio (the ratio of the electric field to pressure) of the gas must be considered. Nitrogen lasers have been reported to operate over a wide range of E/p values however it is generally acknowledged that the optimal E/p is around 120 to 150 V/cm-torr. For a low-pressure design with a 1cm gap and an operating voltage of 15kV, the optimal gas pressure should be around 100 torr. Observations made using these two lasers show a higher operating E/P of 160 to 360 V/cm-torr.Details Of My Original Laser (Circa 1983) ...
My original Nitrogen laser, built in 1983 and used exclusively as a 'pump' for a dye laser (see the dye laser page on this site for details), was fashioned after the design by James G. Small which appears in Scientific American  (Indeed, it was designed, essentially, by Small during a private talk in Albuquerque NM in 1983). In my laser I used two thin (0.015") dual-sided printed circuit boards as capacitors. Ordinary PC board used for circuits is much thicker and will not work for a laser of this type since it's self-inductance is too high which leads to a slow discharge (Discharge times must be under 10nS for this laser to work). It would have been possible to fabricate a capacitor from of plates of copper or alimunim with a mylar (drafting film) dielectric but the PC Board certainly simplifies construction. PCB this thin is difficult to obtain and I was lucky to find it at Active Surplus on Queen St. in Toronto. Each board was about 15cm by 30cm with the 30cm side being parallel to the tube. The lower plate of each capacitor was connected together via a piece of copper. The upper plates, on the top and bottom of the entire assembly, were soldered to the tube electrodes and a spark gap was soldered to the top capacitor.
The insides of the entire laser are pictured here. The electrodes of this laser were simply made of single-sided PC board which were soldered to the capacitors along their entire length for good connection. The electrodes (single sided) protruded into the vacuum enclosure which was made of a 1.5 inch piece of black ABS drain pipe (DVW) split down the centre. The entire tube was held together with epoxy cement. The rear of the tube contained a first-surface mirror right inside the tube itself with adjustment screws protruding from the rear. The front of the tube was cut at the Brewster angle for 337.1 nm and sealed with a piece of plexiglas. The laser beam exited the tube through a hole in the plexi covered with a quartz microscope slide acing as a window (UV does not pass through plexi so a window is required). Additional sealing of the tube was accomplished using melted vacuum wax.
This laser used flowing nitrogen gas both as coolant and to purge the tube of leaked and evolved gases. As I recall there were a number of leaks in the tube but this hardly affected power output as the purity of the lasing gas was not all that critical. Nitrogen gas was supplied from a small welding-tank (Q size) and was of welding grade. An oxygen regulator from an Oxy-Acetylene torch was used (with an adapter allowing it to connect to the oxygen regulator's fitting) to decrease the tank pressure to about 5 psi. This low pressure gas was then leaked through the laser via a needle valve fashioned from a propane torch head. The other end of the laser tube was connected via a hose directly to a single-stage vacuum pump. A converted refrigerator compressor would easily supply the required vacuum for this laser however I had a commercial pump available and used that instead. Although at the time it had been reported that regular air may be used instead of nitrogen (with decreased power output) this was never investigated with this particular laser.
Power was obtained from a small neon-sign transformer rated at 9000V at 10mA.
The laser used an open-air spark gap for a trigger which was later enclosed in styrofoam
to keep the noise down! A variable autotransformer was used to control the laser.
When running, the output of the nitrogen laser is estimated at about 200kW with a pulse width of about 5nS. This pulse of UV light is well suited to excite most organic dyes for laser use. The nitrogen laser is a superradiant laser which can operate without cavity mirrors, although the inclusion of the rear mirror does boost output power by a factor of about three-times. Mirror alignment was easily done by watching the laser spot on a white card placed a few metres away. The alignment screws were adjusted until two spots on the card became one brighter spot.My New N2 Laser (Circa 2001) ...
I had constructed this laser using machined plexiglas for a laser tube as shown below. The electrodes are made from 0.008" brass shim-stock (a strip of brass 12" in length by about 3" wide available from machine-shop suppliers - in Canada the brand name was 'Papco'). The tube is sealed with silicone. The seals have proven sufficient to hold operating pressure for about 24 hours. This arrangement is a huge improvement over my previous efforts as the wide clamping area afforded by the plexiglas tube keeps the electrodes quite straight. In the old design the long electrodes were held in place by being epoxied to the relatively thin (8mm) cross-section of the halved ASB pipe. Here there are several centimeters of plexi clamping the electrodes which are unsupported for only a few millimeters before the discharge gap. This improved alignment leads to a well-formed output beam and very high optical gain. High gain is an asset when attempting to use regular air as the lasant.
The capacitor is fabricated from a single piece of 0.020" PCB obtained for about
$5 from Active Surplus in Toronto (on Queen St. W.). Although 0.015" PCB is generally
used (as it has 33% lower inductance) it has a nasty tendency to breakdown. The
slightly thicker PCB holds-up much better and has proven to tolerate over-voltage
well so far. Remember though that the increased inductance will slow the discharge
and may affect performance, especially if a gas with an extremely short upper lasing
level lifetime such as neon is used. The total piece measures 30cm by 61cm (12" by 24").
The capacitors are arranged as follows: First, etch a strip around the perimeter
of the PCB 2 cm on the top side and 1.5cm on the bottom side. The reason for the
difference is to avoid having both edges align. This helps reduce E fields at the
edges which cause failures (It was found with the first laser that the PCB usually
breaks down first at the edges where top and bottom copper areas stop and overlap - it is
hoped that by not overlapping the edges the electric filed will be lower and hence the
tendency to fail at that point).
Next the cap is laid out as per the photo. The spark-gap side of the cap is etched as a right-
angled triangle with one side (parallel to the tube) being 26cm (the length of the electrodes)
and the other side 20cm. This shape is calculated to encourage a travelling-wave
down the tube at the speed of light . The distance from the gap to the tube
is calculated according to (1-1/e)*L where e is the dielectric constant of the PCB
used (roughly 4 for epoxy-glass board) and L is the 26 cm length of the tube electrode.
A 2 cm gap is etched parallel to the tube axis to separate the two capacitor segments
and the larger capacitor on the side opposite the spark gap consists of whatever is
left - 26cm by 30 cm in length. The tube is mounted by soldering the thin brass
electrodes (which protrude from the tube) to the copper-clasdding of the PCB. The
tube is soldered down the entire distance of the electrode being careful not to
overheat the board which weakens it and causes the cladding to peel! Soldering time can be
minimized by applying flux and pre-tinning the surfaces to be soldered with a thin
layer of solder. The surfaces, both tinned, are then brought into contace and reheated to
cause solder to flow and the junction to be made. After soldering the
tube to the board the top of the PCB should be coated with urethane varnish to keep corona near the edges
of the capacitor down (this occasionally leads to flashovers around the side of the
A photo of the completed tube showing the channel which was machined
into the plexiglas using a 3/4" wide carbide router bit. Cap-head machine
screws (6-32) are used to hold the halves together. The ends are capped by two pieces
of plexiglas, also 3/8" thick, sealed to the tube with silicone. Thin windows
made from microscope slides cover the tube ends and a rear mirror will be added
externally later (but are not required for operation).
Thin microscope cover slides are used as windows at each end of the tube. Look at the edge of the slide to determine the suitability for use with UV radiation: if the glass appears green it will not work - look for glass which appears clear white. Vacuum connections are made to two holes drilled into the end-caps. 1/8" copper tubing is inserted into the holes and sealed with clear silicone.
The spark gap is housed in a box made from three pieces of 3/8" plexiglas with a 1/2" hole drilled down the centre. Electrodes are two 10-32 brass bolts each of which has a rounded end (round surfaces allow the spark gap to have a lower self-inductance and a more predictable breakdown voltage). Each brass electrode is attached to the laser capacitors using strips of brass shim stock 25mm wide and soldered to the capacitors directly. The gap is open to ambient air via a small hole drilled into the assembly. This hole was made to allow the provision to purge the gap with pure nitrogen later if desired. A nitrogen purge of this spark gap is a good idea since air contains oxygen which quickly produces ozone - the sickly-sweet smell of ozone is quite prominent after the laser has been operated for a few minutes.
The vacuum system for prototype testing was an old Cenco single-stage pump, a Bourdon-type
mechanical vacuum gauge, and a needle valve to regulate the tube pressure. This pump
can achieve an ultimate vacuum of only 1 Torr however in the initial test setup little attention
was paid to leaks so the ultimate vacuum is only about 10 Torr (still more than sufficient
for this laser). In preliminary tests air was used as the lasant as pure nitrogen was not available at the time of the test.
The needle valve was at one end of the tube (visible in photos of the laser
as a silver valve at the front of the laser) and the pump/gauge at the other to
allow air to flow through the laser. There were many vacuum leaks in the system which
was hastily connected. This is unimportant since air is the lasant anyway however these
leaks must be sealed properly if pure nitrogen or neon is used with this laser
(especially if neon is used - it is not a cheap gas!).
At first the laser was powered with a simple power supply consisting of a 7300V, 5 mA transformer whose input was regulated with a variac and whose output was connected to the laser via a high-voltage diode (long-barrel TV type) and 50K Ohm resistor in series (Details here). This allows a maximum voltage of about 12kV across the laser which was enough to allow lasing however the laser was seen to laser asymettrically: the beam appeared much brighter on the (parallel) side of the tube opposite the spark gap. Increasing the voltage by widening the spark gap and using a tripler between the transformer and the laser improved performance drastically! The original spark gap was set to about 5-6 mm. When the tripler was added that gap was widened to 8-10 mm. One must be careful not to exceed about 20kV as the capacitor PCB is 0.020" thick (the dielectric itself as measured with digital calipers) and so can withstand about 20kV.
The laser has been completed with preliminary tests run using ordinary air as the lasant. The laser ran successfully using air alone! Using pure nitrogen it was found that the unit produces about twice the output (estimated visually) as well as improves greatly in shot-to-shot stability. No rear mirror was installed for either of these tests although including one will certainly increase output power.
The laser tube is seen here firing. The coil of wire in the center of the photo is the charging inductor placed between the two capacitors.
In this close-up view 'hot spots' can be seen in the discharge. It was found that these were reduced with the use of pure nitrogen gas but not entirely eliminated, esp. when the tube pressure was higher than optimal.
Identifying lasing was simple as seen in the photographs below. When the laser fires, a white business card places 30cm away from the front of the laser glows diffuse violet but no formed beam is seen. When the laser is operating, a very bright and well-formed beam is seen on the card (although the beam is actually UV, the paper used to manufacture business cards has fluorescent dyes in it to make it 'whiter than white'. It is this fluorescence that is observed here). It was found that when very low voltages are used (about 10kV) the beam is asymmetric and much brighter on the side of the tube opposite the spark gap. As the operating voltage is increased the beam becomes more uniform across it's width. Tube pressure was found to have a similar effect. When the tube pressure was too low (below about 15 torr) the output beam was also asymmetric. Optimal tube pressure (with air) was found to be about 60 torr. Increasing the pressure above about 100 torr caused the laser to stop lasing. It is interesting to note that each photo was a capture from a single frame of a video camera. The beam intensity is indeed quite high and a single laser pulse can be easily seen as a bright spot on the card.
Although it works, using plain air as a lasant was found to result in several
- Shot-to-shot operation was found to be erratic. When the charge-rate was set so that the spark gap fired about 2 Hz it was seen that only 50% of the time the gap fires was a laser beam present.
- At lower voltages and non-optimal pressures the beam profile was non linear and lasing was confined to the area around the electrode opposite the spark-gap side of the tube (see photo above)
- Power output was low and difficult to optimize - unusually high voltages were required (20kV plus) to make it lase reliably
When later tested using pure nitrogen - derived from liquid nitrogen by dipping a 2m long polyflow tube into a dewar of LN2 - these effects disappeared. BTW: liquid does not flow into the laser ... during it's 2m path through the tube the LN2 changes back into a gas long before entering the vacuum system. The gas source could just as easily have been from a tank with a regulator. Using pure nitrogen the laser was found to operate optimally at a pressure of 75 torr. Pulse repeatability was drastically improved when using pure N2 - every pulse resulted in lasing action! Power output was increased, and the beam profile was observed visibly to be consistent across - required when used as a pump for a dye laser. As well, parameters such as firing voltage and operating pressure were much more forgiving than when air was used. Any voltage between 10kV and 20kV was found to produce laser output and any pressure between 25 and 100 torr would work. Currently, the laser is housed in a laboratory and is supplied with nitrogen from a large tank. A regulator on the tank reduces the pressure to 20 psi. It is then sent through a needle valve and flows into one end of the laser housing. The other end of the laser houseing is connected directly to a single-stage vacuum pump with a Bourdon (dial) vacuum gauge attached. The power source is the same as the prototype (a small sign transformer powered through a variac) with the secondary (7.3kV) boosted in voltage by an old colour TV tripler to a maximum of about 20kV. The output from the tripler passes through a 50K 10W wirewound resistor to regulate charge current before connecting to the top plates of the Blumlein laser capacitors. With pure nitrogen gas (flowing)the pulse repeatability and power output are excellent and rivals many commercial lasers. It is an excellent pump source for dye lasers.
It may be noteworthy that during prototypical tests neon gas was tried in haste (540.1 nm green line). Neon did not lase, at the time thought to be likely due to the fact that there were many leaks in the crude vacuum system. The nitrogen laser line was seen the entire time (Indicating a major leak leaving enough nitrogen in the tube to lase). This not surprising given that many TEA nitrogen lasers operate with a mixture of only 2% nitrogen and 98% helium! The presence of neon did, however, improve the stability of the air laser by acting as a buffer gas in the same role as that of helium in many commercial TEA lasers (Lasers such as the Lumonics excimer lasers which can operate as nitrogen lasers with a suitable gas mix - see the TEA lasers page for details). During later tests the vacuum system was improved and a major leak around one of the end windows sealed. The laser is sealed well enough that it will still lase 24 hours after a pumpdown once sealed (i.e. the needle valves closed). Pure neon was tried again but failed to lase. This is likely because the upper lifetime of neon is only about 1.5 nS: an order of maginitude shorted than nitrogen. Combine the thicker 0.020" PCB used (higher inductance, slower discharge) with this short lifetime and it can be understood why neon will not work in this laser. A laser designed for neon might be similar to TEA lasers which have sub-nanosecond discharge times (or even closer to a hydrogen laser which has similar lifetime dynamics to that of neon).
For a view of TEA lasers including a commercial unit, see the TEA lasers page on this siteImprovements and Updates
It has been brought-up by Milan Karakas (you can find his page on my LINKS page) that this laser is far from optimal for a number of reasons. The basic design principles are sound enough but the geometry of the laser is in question. One major question is the assymetric design of the laser where the capacitor on the 'spark gap' side is much smaller than the other. Assuming that the impedance of the spark gap is much higher than the transmission line (which it most likely is), each side of the laser should be equal in terms of size! Indeed, I can't say that the geometry of this particular design has yielded any significant gains over the previous design which was symmetrical. Owing to the fact that transmission line design is the most critical for a TEA laser, I shall put comments to this effect there. Suffice it to say that for a low-pressure design such as this the simple approach is likely the best: make the unit symmetrical, use a single gap, use thin PCB, and it will likely perform. To truly optimize the laser one must model it as a true transmission line.
Aside from a Blumlein, alternative designs  are possible in which energy from a primary, low-inductance capacitor is dumped across a discharge tube with a smaller capacitance across it. This design provides a lower driving impedance for the laser channel.
Ideas on the use of a Marx-Bank Generator
Here is an idea which was suggested by a reader from a University in Germany ... Some types of lasers, such as a nitrogen or neon laser, require an extremely fast electrical discharge to function. In the case of the nitrogen laser operating at low pressures a discharge time of about 10 nanoseconds or less is required. An excellent method of developing such a pulse is a Marx-bank generator. Such devices are capable of producing pulses of 100's of thousands of volts with discharge times of one nanosecond or less!
The basic Marx-bank consists of multiple high-voltage capacitors which are charged in parallel then discharged in series. I had constructed one years ago using ABS plastic plumbing for capacitors and plumbing TEE's and carriage bolts for spark gaps. The entire apparatus could produce ultra-fast pulses of energy at about 150kV.
It as suggested that a simple laser could be built by putting a glass plasma tube  in the centre of such a Mark-bank making a simple coaxial arrangement with high-voltage capacitors surrounding the laser tube. This would be a compact arrangement and one which is almost optimal in terms of reduced inductance and hence fast discharge. Such an arrangement could easily pump nitrogen as well as other gases to lase.
A few key references:
 A simple pulsed nitrogen 3371A laser with a modified Blumlein excitation method
J.G. Small and R. Ashari
Review Of Scientific Instruments, Vol 43, Number 8, August, 1972, pp.1205
Describes a simple laser using acetate sheets as capacitors
 An Unusual kind of gas laser that puts out pulses in the ultraviolet
Scientific American, Amateur Scientist Column, June 1974
A simple nitrogen laser easily constructed by an amateur. Based on Small's design in the above reference.
 Compact high-power TEA N2 laser
Review Of Scientific Instruments, 49(9), Sept 1978
A compact TEA laser which required operates at atmospheric pressures and hence does not require a vacuum pump. Uses multi-segmented electrodes.
 Spark gap power switching circuit for ... plasma gun
Glenn Kuswa & Charles Stallings
Review Of Scientific Instruments, Vol 41, No 10, 1970, pp. 1429
Details construction of low inductance (7nH) spark gap switch which may be used for N2 lasers
The poor man's nitrogen laser
David Phillips and John West
American Journal Of Physics, Vol 38, No. 5, May 1970
A small, simple N2 laser which does not use a Blumlein design but rather a capillary tube. Produces low powers which may not be suitable to pump a dye laser however the laser is easily constructed from parts available in the plumbing section of a hardware store.
 A simple, high power nitrogen laser
D. Basting, et al.
Opto-electronics, 4 (1972) 43-44
An optimized and compact high powered nitrogen laser which uses two flat-plate capacitors (top and bottom). Reported powers are 1.2MW for a 30cm tube. Also reported is the use of air instead of nitrogen gas.
 Travelling wave excitation of high power gas lasers
John Shipman, jr.
Applied Physics Letters, Vol. 10, No. 1, 1 Jan 1967
A long laser of Blumlein design. Includes data on using neon gas as a lasant.
Fundamentals of Light Sources and Lasers
ISBN 0-471-47660-9, John Wiley & Sons, 2004
A general reference covering UV lasers in chapter 10
A Short High-Power TE Nitrogen Laser
Ernest E. Bergmann & N. Eberhardt
IEEE Journal of Quantum Electronics, August, 1973, pp.853
An alternative design to a Blumlein employing a discharge capacitance and a dumping capacitance.
Pulsed molecular nitrogen laser theory
Edward T. Gerry
Applied Physics Letters, Vol. 7, No. 1, 1965, pp. 6
Examines excitation mechanisms & saturation power output of N2 lasers
Along the lines of the Nitrogen laser is the Neon laser. When pure neon gas is used in a nitrogen laser a green line at 540.1 nm may be produced. The output power of this line is about 1/5 that of a nitrogen laser. There is one trick though: the discharge time must be ten times faster than the already fast nitrogen laser! I have attempted to lase this line in the above laser and it did not work: likely because my discharge was indeed too slow. Likely a laser built for neon would resemble a TEA laser (which does have a discharge time of about 1nS). Neon will lase on several other transitions as well - see the Neon Lasers Page on this site for details.
Details of this laser may be found in the following references:
The 5401A pulsed neon laser
IEEE Journal of Quantum Electronics, March 1967, p.134
This is the best reference I've seen. It describes a neon laser which otherwise resembles a nitrogen laser
Observation of a super-radiant self-terminating green laser transition in neon
Leonard, et al
Applied Physics Letters, 15 Sept 1965, Vol 7, No. 6, p.175
Asymmetric visible super-radiant emission from a pulsed neon discharge
Clunie, et al
Physics Letters, 1 Jan 1965, Vol 14, No. 1, p.28
See the LINKS page on this site for homepages specifically on the nitrogen laser