A system for automated noise parameter measurements on MR preamplifiers and application to high B(0) fields

NMR Biomed. 2014 Aug;27(8):926-38. doi: 10.1002/nbm.3138. Epub 2014 May 29.

Abstract

A noise figure and noise parameter measurement system was developed that consists of a combination spectrum and network analyzer, preamplifier, programmable power supply, noise source, tuning board, and desktop computer. The system uses the Y-factor method for noise figure calculation and allows calibrations to correct for a decrease in excess noise ratio between the noise source and device under test, second stage (system) noise, ambient temperature variations, and available gain of the device under test. Noise parameters are extracted by performing noise figure measurements at several source impedance values obtained by adjusting an electronically controlled tuner. Results for several amplifiers at 128 MHz and 200 MHz agree with independent measurements and with the corresponding datasheets. With some modifications, the system was also used to characterize the noise figure of MRI preamplifiers in strong static magnetic fields up to 9.4 T. In most amplifiers tested the gain was found to be reduced by the magnetic field, while the noise figure increased. These changes are detrimental to signal quality (SNR) and are dependent on the electron mobility and design of the amplifier's semiconductor devices. Consequently, gallium arsenide (GaAs) field-effect transistors are most sensitive to magnetic fields due to their high electron mobility and long, narrow channel, while silicon-germanium (SiGe) bipolar transistor amplifiers are largely immune due to their very thin base.

Keywords: MR preamplifiers; electron mobility; gallium arsenide; indium gallium phosphide; magnetic field; noise figure; noise parameters; silicon-germanium.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Amplifiers, Electronic*
  • Artifacts*
  • Automation*
  • Magnetic Fields*
  • Magnetic Resonance Imaging / instrumentation*
  • Magnetic Resonance Imaging / methods*
  • User-Computer Interface