Modeling COTS System TID Response With Monte Carlo Sampling and Transistor Swapping Experiments

Modeling systems containing commercial-off-the-shelf (COTS) parts exposed to total ionizing dose (TID) are difficult due to large parameter variability in the parts as a function of dose. This article provides a rapid approach to predicting the system performance as a function of TID given these parameter variations. A radiation-aware model for an IRF510 power MOSFET is generated from a small population of irradiated parts, and this model is randomly sampled for each of the transistors in an example circuit, utilizing the degraded parameters. The simulations predict the circuit operating in one of three distinct modes, with circuits containing large variability in parts failing in some instances. The simulated data are compared to experimental data using a method that involves interchanging irradiated parts in a socketed printed circuit board (PCB) to produce a large number of different combinations of test circuits. The simulated data correctly predicted the circuit modes of operation, and the part-swapping technique eliminated the cost of building many test boards.