An Electrical Technique for the Measurement of the Peak Junction Temperature of Power Transistors

Abstract

A technique is described which uses straightforward electrical measurement procedures to determine the peak junction temperature of power transistors. To determine the peak temperature, standard electrical measurement techniques are altered to account for the difference between the distributions of the calibration and measurement currents in the active area of the device. For relatively uniform temperature distributions, the electrically determined peak junction temperature is only about 6% or less below the infrared measured peak temperature whereas the standard electrically measured temperature is about 10 to 25% below the infrared measured peak temperature. For severely non-uniform temperature distributions, when only about 20% of the total active area of the device is dissipating power at steady state, the electrically determined peak temperature is within 11% of the infrared measured peak temperature while the standard electrically measured temperature is more than 40% below the infrared measured peak temperature. Device operating conditions for which the junction temperature as determined by standard electrical methods, infrared techniques, and the electrical peak temperature technique equals the manufacturer's specified maximum safe operating temperature are compared with one another and with the manufacturer's specified safe operating limits. It is suggested that the electrical peak temperature technique can be used to generate more realistic safe operating area limits and to determine the validity of specified safe operating limits of power transistors.