The energy sector is moving into the era of distributed generation (DG) and microgrids (MGs). The stability and operation aspects of converter-dominated DG MGs, however, are faced by many challenges. Important among these, are: 1) the absence of physical inertia; 2) comparable size of power converters; 3) mutual interactions among generators; 4) islanding detection delays; and 5) large sudden disturbances associated with transition to islanded mode, grid restoration, and load power changes. To overcome these difficulties, this paper presents a new large-signal-based control topology for DG power converters that is suitable for both grid-connected and islanding modes of operation without any need to reconfigure the control system and without islanding detection. To improve MG stability, the proposed control structure is realized via two steps. First, an emulated inertia and damping functions are adopted. Second, to guarantee stability and high performance of the MG system during sudden harsh transients such as islanding, grid reconnection, and large load power changes, a nonlinear MG stabilizer is proposed. An augmented converter model is developed and used to design the MG stabilizer via the adaptive backstepping (AB) technique to guarantee large-angle stability and robustness against unmodeled dynamics. Theoretical analysis and evaluation results are presented to show the effectiveness of the proposed control scheme in achieving stable and smooth operation of a MG system in grid-connected, islanding, and transition modes.