This article is concerned with a fault detection enhancement method using adaptive amplifier coefficients (a.a.c.) concept. It is developed for linear systems with mixed uncertainties (stochastic and bounded uncertainties framework). The study provides, also, analysis and discussions about the applicability and the efficiency of the enhanced method to several sensors fault error types. Simulations on a vehicle bicycle model (validated by experimental tests on the Renault Megane) are presented to emphasize on the performances of the developed method.