The success of discharge-pumped lasers depends upon achieving some degree of control over the discharge stability. A very large effort has been expended toward this goal by researchers since the early discharge CO2 lasers were developed. With the extensive development of the rare gas halide lasers much more effort has been expended because of the inherent instability in pumping electronegative gases. This paper reviews the major approaches of preionization, the critical issues and the limitations of each.

The physical processes involved in x-ray preionization of high-pressure avalanche-discharge gas lasers are reviewed to determine optimum preionizer parameters. Preionization of lasers such as XeCl, CO2 etc., requires an x-ray dose of ~1mR. More strongly attaching rare-gas-fluoride lasers (e.g. KrF) require stronger preionization (> 10 mR per pulse). X-ray generation by electron-beam sources is discussed and candidate cathode types are reviewed. Typical preionizer pulsed-power2 parameters are 100 kV pulses of 100 - 200 ns duration, electron-beam current density of lA cm-2 and repetition rate of ~100 pps.

Weight and volume scaling of pulsed modualtors requires techniques which involve system design and trades before actual design of the modualtor. This paper covers various aspects of these trades and gives examples of thier implementation.

Classically, pulsed flashlamps have been driven (excited electrically) with a single loop RLC critically damped circuit. The capacitor is chosen for the required energy output, the inductor tunes the peak current for optical output requirements, and the resistor is used to ensure that the circuit is critically damped (that is no current reversal that can extinguish the lamp). A few circuits using several discharge loops, known as a pulse forming network, or PFN, have been explored but again, the current in the lamp is controlled by the passive circuit components. With both circuits, a two state closing switch,such as an SCR, is used to initiate current flow. This approach severally limits the agility in changing the drive characteristics as it is a "point design." To make such a change, the entire discharge circuit must be redesigned. The new excitation circuit utilizes the active region of high power field effect transistors in a Class A or Class AB amplifier circuit to provide a linear control of the flashlamp current. Having linear control will eliminate the need for simmer circuits required when a two state closing switch is used, (the simmer circuit keeps the flashlamp ionized in the interpulse period.) Also, since the optical output of the flashlamp is dependent on the lamp current, the spectral output can be adjusted to enhance wanted frequencies and depress unwanted wavelengths which will lead to greater efficiency which is not the case in a single point design circuit.

This paper describes the design, construction and test results of a thyratron modulator circuit for use with high peak power metal vapour lasers in which the discharge current pulse through the laser is rapidly terminated. The circuit generates current pulses with peak amplitudes up to 300A and fall times of less than 7Ons (90%-10%). The system consists of a thyratron-switched capacitive discharge circuit in which the bypass element is a saturable inductor. Saturation of the magnetic core provides an alternative path for the discharge current. Furthermore, the point in time at which saturation occurs can be controlled by appropriate magnetic core biasing. This paper describes the operation of the circuit.

The use of spiker/sustainer circuitry for exciting excimer lasers is now well established due to its proven high performance. In spiker circuitry reported to date, the rate of rise of spiker voltage incident upon the laser head has been limited by both the turn-on time of the spiker circuit switching element and the spiker circuit inductance. Rise times achieved have generally been limited to between 50ns and 20ns depending upon whether a thyratron or rail-gap is used as the principal switching element. This paper will describe an alternative method of generating high voltage pulses whose rise times are no longer limited by the commutation speed of the principal high voltage switch. These pulses are generated in two stages. An initial high voltage pulse with a rise time determined by the principal switching element and circuit geometry is incident upon a ferrite pulse sharpener. This consists of a coaxial transmission line whose inter-conductor region is filled with saturable ferrite material surrounded by high voltage insulation. The pulse drives the ferrite into saturation as it propagates along the cable resulting in the sharpening of the leading edge of the pulse. The rate of rise of the output pulse from the pulse sharpener is determined by the magnitude of the incident voltage pulse, the rate of rise of the incident voltage pulse, the switching constant of the ferrite material and the geometry of the pulse sharpener. Experimentally it has been found that 60ns rise time, 20kV pulses incident upon the sharpener are reduced to 2ns rise time pulses after propagating along 2m of cable. This corresponds to a dV/dt of 1013 Vs-1. Using drive voltages of the order of 50-100kV with 5ns rise times, generated by a 3 stage Marx bank, voltage rise times in the region of 300ps have been generated using 95cm length of cable. This corresponds to a dV/dt of > 1014 Vs-1. Details of the pulser design are given and its suitability for use in excimer laser spiker circuitry is discussed.

There are three basic geometries for multigap thyratrons at the present time, and these are reviewed in terms of the constraints imposed by Paschen's Law. Practical examples of each of the three geometries are given and those thyratron characteristics which are a direct result of the choice of geometry are discussed. Performance data are presented for a two-gap thyratron, the CX 1725, and for two four-gap thyratrons, the CX2024 and the CX2025. The basic characteristics of multigap thyratron commutation are briefly discussed, and evidence in support of a simple model is presented. Finally, the results of a recent comparison between the reverse current conduction charactersitics of a hollow anode multigap thyratron and a double-ended multigap thyratron are presented.

A review of recent progress in the development of the low-pressure (~27 Pa H2) glow discharge pulsed high-power (10~30kV) switch, the Back-Lighted Thyratron (BLT), is to be presented. The BLT operates with a glow discharge and utilizes a simple device geometry. New data on peak current capability, current rise rate, current reversal ability, life, repetition rate operation, and trigger efficiency and comparisons of delay and jitter with different trigger methods (flash-triggered, laser-triggered and electrically-triggered) are reported.

The optically triggered psuedo-spark, also known as the Back-Lit-Thyratron (BLT), is a low pressure plasma switch having an unheated metallic cathode. In this paper, a computer simulation of the BLT is presented consisting of a 2-1/2 dimensional time dependent continuum model for electron and ion transport. The model utilizes both the local field approximation and a beam component for the electron distribution function. We find that switch closure depends critically on the formation of a virtual anode in front of the cathode hole by generation of positive space charge.

There are about twenty common solid dielectrics and about a dozen fluids commonly used in pulsed capacitors. In high repetition rate discharge circuits where very high average and RMS currents can be attained, these materials are reduced to just a few. Dielectrics suitable for this duty must have several specific characteristics. One, small temperature coefficient, that is, the dielectric constant should remain constant over a large range of temperatures. Low dielectric loss is essential for high repetition rate capacitors because, at large RMS currents, substantial heat can be generated in even the smallest of dissipative elements. This is all important in spirally wound capacitors because the only path of heat removal is by conduction out through the foil edges. Pulse capacitor dielectrics should also exhibit constant characteristics with frequency. Since pulse capacitors are exposed to a wide range of frequencies, fluctuations in dielectric constant and loss are not a desired characteristic. Frequency dependence of capacitance and loss further reduce the number of suitable dielectrics for high repetition rate applications. Last, many dielectrics exhibit an electric field dependence, usually given in data sheets as the "voltage coefficient". As with temperature, this parameter is commonly stated as a negative percentage. The magnitude of this effect depends on material properties and mechanical forces. Dielectric constant variations with electric field are determined empirically for the most part. Some spirally wound fluid impregnated capacitors were found to exhibit electrical characteristic changes due to movement of the capacitor plates.

A new concept for fast pulse transformers is described. The transformers have been used in two applications with excimer lasers: a) To produce a short 100 MW/90 kV/100 ns pulse for driving a preionization X-ray source. b) To step up the voltage in a command charging circuit for charging a water pulse forming network within a few microseconds to an energy level of 100 Joule electrical energy per pulse. A report is given on the performance characteristics at repetition rates up to 100 Hz. A coupling factor close to unity and nearly ideal transformer efficiencies have been achieved.

A large variety of laser excitation circuits can be reduced to two simple topologies. This paper describes those two topologies and defines the major components. Alternatives within the major component groups are presented. Two examples are given which illustrate the two basic topologies.

The EU213 EUREKA Excimer Laser project is aimed at the development of a 1kW average power excimer laser by 1992. In order to realize this goal, part of the UK programme will be to explore the possibility of creating a 1kW laser with a repetition rate approaching 10 kHz. In order to investigate the problems of, for example, gas flow, preionisation and pulse power at such a high repetition rate, an XeCl test bed laser has been assembled at Culham Laboratory: the Compact HIgh Repetition rate (CHIRP) laser. The CHIRP laser is a fast flow discharge pumped,UV corona preionised excimer laser. One of the major considerations in developing a 1 kW system is the electrical input to optical output conversion efficiency and its optimisation through the design of the pulse power system. The pulser circuit presently used uses a low impedance transmission line, and provides a high voltage prepulse for the main discharge. A voltage doubler circuit generates a synchronous pulse for the corona preioniser. Isolation of the prepulse is achieved by use of a ferrite saturable magnetic inductor. The typical stored energy per pulse is ~3.5 J (at 10 kV) for the main pulse,~1 J (at 25 kV) for the prepulse and ~ .2 J for the preioniser. Active switching is by deuterium filled hollow anode thyratrons. This paper discusses the design of the circuit, its operation at pulse repetition rates of ~1 kHz, and prospects of multikilohertz operation.

In continuous wave operation gas lasers operate at mean powers in the multi-kilowatt range and offer high processing rates. But intensities generated at the workpiece are below the damage threshold of many materials employed routinely in the manufacture of engineering components and of metals specifically. It has been found under pulse irradiation at 10μm, that many of the refractory metals exhibit a dramatic increase in the absorption of the incident flux, if the material is raised to a sufficiently high temperature. The absorbed power intensities in the range 1010 - 1011 W/m2 required to induce this high absorption, high reaction pressure heating regime, can be generated by the CO2:N2 system with proper choice of electrical input pulse and gas mixture composition. A high pulse repetition frequency CO2 laser source has been built to investigate the utility of the enhanced absorption technique: the ability to process alloys of copper and aluminium and the refractory metals being the prime objective. This paper describes some of the design and development aspects of a high frequency pulsed power source for this laser system.

We report on an experimental long pulse CO2 laser (25uS at 100) that employed two plasma electrodes and an impedance-matched external sustainer pulser to preionize and electrically pump the CO2 laser mixture. The plasma electrodes, each measuring 7cm wide by 100cm long, were struck on by a 1011 V/S voltage trigger wavefront, and were found to have a uniform arc distribution along the electrodes in argon, but showed severe arc striations in CO2 mixtures. These striations resulted in non-uniform laser preionization, which in turn caused the sustainer discharge to constrict. Consequently the active laser volume was not pumped uniformly and the system did not lase. However, the CO2 discharge did not collapse into a low-impedance arc at energy loading greater than 350 J/1-atm and 25uS pulse duration, indicating that by improving the arc uniformity on the plasma electrodes, this approach can be useful for high energy, long pulse lasers. Finally, important engineering issues like system life, and size and weight scaling for the plasma electrode CO2 laser are discussed.

A two module electron beam source operating over a wide range of output parameters has been designed and fabricated to be used in conjunction with a pair of electron beam sustained CO2 lasers. Each module comprised a grid-controlled thermionic electron beam gun including a compact grid pulser for control of the electron beam, a 250 kV thyratron switched modulator for acceleration of the electron beam, a 1 kHz filament heater and a complex computerized control system. The system was designed to reliably produce 45 μs is wide electron pulses of 150-200 keV energy, operate at repetition rates of 1-10 pps and current densities of 5-20 mA/cm2. Additional parameters are listed in Table 1. The high voltage cathode assembly employs 132 thoriated tungsten filaments distributed over the area of the 250 cm x 10 cm output window. The cathode assembly including the control grids is supported by two high voltage ceramic bushings in a stainless steel vacuum chamber. For acceleration of the electron beam, a pulsed 150-250 kV voltage is applied by a thyratron switched modulator, between the cathode and anode window. The fiber optically controlled grid pulser located in the modulator tank is floated at the applied cathode potential.

This paper describes the development of the Radial Glow Discharge Pulsed CO2 Laser (RCL) Technology. This laser is pumped by a high pressure UV-preionized self-sustained glow discharge. The discharge is between two concenteric cylinders, with radial current flow. Radial UV preionization leaves the ends open for laser power extraction. The 5 kW laser extraction is axial out of the device. The laser uses 2 resonant charge circuits operating in parallel to charge the main discharge PFN and the UV preionization circuit at 50 Hz. The UV preionization circuit drives 8 sets of flashboards in parallel using 1 thyratron.