Comparison of statistical distributions for the analysis of GaAs MMIC life test data

Four statistical distributions were fit to life test data of gallium arsenide monolithic microwave integrated circuits devices to predict the median time to failure (t₅₀ lifetime) of each device type at the life test temperatures. These lifetimes were fit to an Arrhenius equation to predict each device´s activation energy and median time to failure at 140 degrees Celsius channel temperature. Results comparing the four distributions are presented and the activation energies are compared to those reported by industry. The median time to failure is useful for verifying device reliability, and the activation energy is useful for predicting the failure mechanism. This information assists chip manufacturers in improving device reliability, and in predicting the life time of the device in various applications. To our knowledge, this is the first report of a comparison of statistical distributions among several types of devices