Self-organization of spontaneous activity of a network of active elements is important to the general theory of reaction-diffusion systems as well as for pacemaking activity to initiate beating of the heart. Monolayer cultures of neonatal rat ventricular myocytes, consisting of resting and pacemaker cells, exhibit spontaneous activation of their electrical activity. Similarly, one proposed approach to the development of biopacemakers as an alternative to electronic pacemakers for cardiac therapy is based on heterogeneous cardiac cells with resting and spontaneously beating phenotypes. However, the combined effect of pacemaker characteristics, density, and spatial distribution of the pacemaker cells on spontaneous activity is unknown. Using a simple stochastic pattern formation algorithm, we previously showed a clear nonlinear dependency of spontaneous activity (occurrence and amplitude of spontaneous period) on the spatial patterns of pacemaker cells. In this study, we show that this behavior is dependent on the pacemaker cell characteristics, with weaker pacemaker cells requiring higher density and larger clusters to sustain multicellular activity. These multicellular structures also demonstrated an increased sensitivity to voltage noise that favored spontaneous activity at lower density while increasing temporal variation in the period of activity. This information will help researchers overcome the current limitations of biopacemakers.