A heuristic thermistor model

Negative temperature coefficient thermistors, traditionally characterized by a nonlinear relationship in the resistance-temperature space, are shown to be intrinsically linear when that characteristic is mapped into a space defined by an appropriate transform. Therefore, information formerly conveyed by scores of tabulated ordered sets is contained in a pair of first-order equations which define resistance as a function of temperature and the converse. The necessary transforms and equations are defined; and, procedures for calculating the coefficients from manufacturers data using regression techniques are delineated. The ambiguity inherent in estimating parameters by statistical methods is discussed. Maximum mean ambiguity occurs when the model is applied over a span of ; this figure is reduced to for a span on . The ambiguity continues to decrease as the span decreases.