Thermal resistance measurements on monolithic and hybrid Darlington power transistors

A method for measuring the thermal resistance, Rθ, of integrated power Darlington transistors is described that is based upon the emitter-only switching technique. It is shown that for specified measurement conditions this method can be used to measure the average thermal resistance, Rθ(l+2), of the input and output transistors of the Darlington pair and is therefore applicable to production-line monitoring of the thermal characteristics of the Darlington. It is shown that although a direct measurement of the thermal resistance, Rθ (2), of the output transistor cannot be made for most Darlingtons, an indirect determination of Rθ(2) can be made using the emitter-only switching technique and a simple equation. Comparisons of the differences between the infrared-determined thermal resistance and the electrically-measured thermal resistance of Darlingtons and of discrete power transistors illustrate that the accuracy of the Darlington measurements is comparable with that achievable for discrete devices. It is also shown that measurements which attempt to use the collector-base voltage of the output transistor of the Darlington as a temperature-sensitive-parameter, either alone or in conjunction with an output commutating diode (when present), are too sensitive to the magnitude of the measuring current to be reliable. Also, the use of the base-current screen for hot spots, as used for discrete transistors, is marginally applicable to integrated Darlingtons, but it is shown that the emitter-base voltage sensed during emitter-only switching is quite effective for detecting hot-spot formation.