Methodology for Analyzing Non-Steady-State Thermal Transient Test (T~3) Data

摘要

When the 10 ms constant current pulse duration specified in MIL-STD-1512 Method 206 is applied to a typical 1W-1A EED in T~3, steady-state voltage signal condition usually cannot be established at the end of the pulse for evaluation of the thermal time constant- a key data should be provided by the test. A detailed analysis of the apparent shortcoming using recently formulated analytic model for T~3, indicates some encouraging results: 1) The voltage signal reading at the end of the current pulse may be translated to the quantitative value of thermal time constant by establishing a correlation formulas; 2) The standard deviation of the voltage signal may correlate to that of the thermal time constant. Both methods may be useful to assess historical T~3 data for many heritage EEDs accumulated in past decades using insufficient test pulse durations; and 3) Statistical accuracy of standard deviation of signal voltage as a function of the pulse duration is examined. It indicates that duration on the order of 16 times of the nominal time constant or greater can provide highly accurate results in thermal time constant evaluation; 8 to 16 times of nominal time constant can provide results with small and acceptable errors, 2 to 8 times of nominal time constant can introduce larger errors and needs exercising correction procedures such as illustrated in this paper.