Capacitive Energy Storage for Magnetic Resonant Imaging Gradient Driver

The magnetic resonance imaging (MRI) gradient driver is a high-performance inverter that provides large currents, >1000A, for the gradient magnetic fields used for imaging. It is challenging to achieve the required performance with the fast transitions on current pulses driven with high voltages,...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE Applied Power Electronics Conference and Exposition (APEC) pp. 695 - 700
Main Authors Singh, Yash Veer, Sabate, Juan, Wang, Ruxi
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.03.2019
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:The magnetic resonance imaging (MRI) gradient driver is a high-performance inverter that provides large currents, >1000A, for the gradient magnetic fields used for imaging. It is challenging to achieve the required performance with the fast transitions on current pulses driven with high voltages, >2000V, for fast imaging. The currents provided by the driver are defined by the imaging procedure and are always changing from zero to maximum current, resulting on pulsating load from zero load to 10 times the rated power. The accuracy for the currents supplied to the gradient coils requires regulated voltages provided to the inverter not to sag below -10% of rated value. To avoid voltage sag and to limit peak power draw from the facility power utilities, the system has capacitors for energy storage at the inverter supply voltages and at the input of the power supplies for the inverter. The paper presents the power architecture for power supplies, inverter (gradient amplifier), and energy storage, and how the energy storage is optimized to avoid excessive demands on facility power feed that have to be designed for peak power. The architecture and energy storage selections reduced the input current and minimized the voltage sagging at the inverter input. The solution was implemented, and the results are presented here when operating standard imaging sequences.
ISSN:2470-6647
DOI:10.1109/APEC.2019.8722223