The limited fault current tolerance of inverters in AC Microgrids demands the necessity of faster and accurate fault detections. At the point of common coupling, the occurrence of various symmetrical and unsymmetrical faults degrades the performance and robustness of the inverter-interfaced local controllers. To achieve faster fault detection in inverter-based AC microgrid, this paper proposes a combined technique that includes two well-known signal processing techniques such as teager energy operator and Hilbert-Huang transform. The combined principle is called a Teager-Huang technique, which can detect different line faults using the teager energy of the Hilbert-Huang based empirical mode decomposed signals. The fault detection technique is verified by creating faults at the inverter connection point to the grid, using MATLAB/SIMULINK and PLECS.