DTI considered several different architectures for providing pulsed power at 500 kV and > 250 A to the Next Generation Linear Collider's 3300 klystrons. Under this SBIR, DTI continued development of IGBT-based, modulator designs in response to the requirements of the Next Linear Collider project based at Stanford Linear Accelerator Center.

Efficiency is a key factor in selecting a modulator design for NLC, because at 96 MW, the specified operating power of the hundreds of modulators, an increase of one percentage point in electrical efficiency is estimated to save $5 million over the modulators' ten-year life. Conventional designs use a pulse transformer, which typically reduces overall modulator efficiency by 10%.

In Phase I, DTI investigated two architectures for a "transformerless" modulator. They are (1), a direct hard-switched 500 kV IGBT modulator to support eight klystrons, and (2) a Marx bank modulator that requires a maximum 85 kV DC to pulse to two NLC klystrons.

The hard switch modulator has a stiff 500 kV DC present at all times and utilizes a simple series switch to pulse the load. The switch must be able to open under load (and fault) currents at the end of pulse. The Marx bank uses active switching to erect 500 kV only during the pulse. A bank of capacitors are charged in parallel and pulsed in series. A network of IGBT switches is used to flip the series/parallel connections back and forth. The highest DC voltage existing in the Marx bank system is 85 kV.

As a result of the research in this Phase II effort, DTI selected a Marx bank approach for yet further development of the transformerless NLC modulator architecture.

<<back to Past Research list>>