A fixed point Extended Kalman filter (EKF) model is designed and implemented for modest Battery Management System (BMS) for light EV applications. The standard EKF state equations are re-scaled to a higher resolution and converted to a fixed-point decimal word-length. The coulomb counting (CC) and open circuit voltage (OCV) based methods are integrated via a fixed point extended Kalman filter technique. The updated approach can overcome the inaccuracy of CC and the insufficiency of OCV-based methods. Initial state estimation, process and measurement noise co-variances are estimated based on experimental cell data. The algorithm is able to achieve < 500μs of computation time on the fixed-point micro-controller BMS unit. The validation results based on the continuous and multiple drive cycle tests indicate an error of ≤ 2% in State of Charge (SOC) estimation.