Brett D. Smith

Performance characterization for an excimer laser solid-state pulsed power module (SSPPM) after 20B shots

An experiment has been designed to characterize a solid-state pulsed power module (SSPPM) during the initial manufacturing cycle and then repeat the same characterization measurements after the module has gone through several sequences of 10B shots of normal operation in an excimer laser. The goal of such an experiment is to determine what, if any, degradation occurs during these extended periods and to assist in the development of expected module lifetimes that can then be used to estimate the cost of operation of the overall excimer laser. Initial component and subassembly measurements include the capacitance and Q of energy storage capacitors; the inductance and Q of bias, charging, and energy recovery inductors; the B-H characteristics of magnetic cores; insulation breakdown strength; connection resistance; and the general physical appearance of the unit. Operational measurements also compare the efficiency of each pulse compression stage, the repeatability and accuracy of diagnostics, thermal management parameters, and the recovery and on-state characteristics of the silicon-controlled rectifiers (SCRs) and diodes. Each of these items is monitored before testing and after each sequence of 10B shots has been completed. Results of the experiment are described.

Timing compensation for an excimer laser solid-state pulsed power module (SSPPM)

For certain applications, it is critical to minimize variations in the (throughput) timing between trigger and the output pulse of a magnetic modulator. A circuit is described that maintains a relatively constant delay over a large operating voltage range (600-1150 V) and temperature range (25 /spl deg/C-65 /spl deg/C) range, The circuit operates by sampling the charging voltage and magnetic switch temperature just prior to the start switch trigger. Those parameters are then used to calculate the appropriate amount of delay to add into the low-level trigger chain to ensure that the delay stays constant over the voltage and temperature operating range. Although other approaches can be conceived and implemented, this particular design is relatively simple and inexpensive and meets the desired performance goals. Data presented show that the ideal correction function is nonlinear in nature and, as a result, simple linear approximations are limited in their ability to minimize the timing variations. Improvements to the original circuit use a multiple, piece-wise, linear approach in order to obtain better performance. The results are that an initial timing variation of almost 3 /spl mu/s has been reduced to a total variation of less than 100 ns.