The series arc faults have been increasingly recognized as one of the major causes of electrical fires in residential environments. However, detecting series arc faults is becoming more challenging due to the increasing diversity of electrical loads used in daily life. This paper proposes an algorithm for detecting series arc faults in low-voltage AC residential environments. The proposed algorithm utilizes Approximate Entropy as a key feature. However, computing Approximate Entropy within a limited time is constrained by the computational performance of commonly used conventional microcontrollers, which restricts the number of input data points that can be processed. To address this limitation, Maximum Difference and MinMax preprocessing methods are introduced to reduce the number of data points by extracting representative values from segmented intervals. During Approximate Entropy computation, these preprocessed data points are used as inputs. Additionally, key parameters for Approximate Entropy are carefully selected to ensure effective differentiation between pre- and post-series arc conditions across various load types and configurations. The computed Approximate Entropy values for each load and configuration are then used as features in the proposed algorithm. Finally, the algorithm’s performance is evaluated by assessing its series arc fault detection performance and its ability to minimize false detections during normal operation.