In the context of rising energy costs and the need to use new energy sources, works aimed at raising the surface temperature of heat radiators with reduced energy consumption are of particular importance, and it is especially important if these processes are also accompanied by the effects of self-stabilization. Bulk materials do not possess these properties. However, materials whose dielectric matrix is also an active element can provide up to 10 - 30% of the thermal energy that will be released in the material, thereby increasing the surface temperature and without increasing energy consumption. Therefore, the study of composite materials with different matrices is relevant. This article the influence of the matrix material on the electrical properties of composite materials was examined. It was established that the microstructure morphology of resistive materials changes significantly depending on the matrix type. In composites based on matrix AlN, for the entire range of concentrations HfC, conducting cluster is formed with a metallic conductivity. For composite systems Al2O3-HfC and Si3N4-HfC thermoactivated hopping conduction between nearest neighboring states observed. Thus, for materials based on Si3N4 matrix at temperatures up to 300°C observed reduction of charge carriers concentration with increasing temperature. The approximation of the temperature dependence of the electrical conductivity was carried out on the basis of the following possible variants of the nature of the electrical conductivity, namely: jump conductivity (nonlocalized states, localized states in the tails of conduction and valence bands, localized states near the Fermi level), tunneling. It can be assumed that the formation of conductive clusters occurred under the influence of two factors: magnetic field and mechanical loading. When using the AlN matrix, the influence of the magnetic field on the structure formation is smallest. This conclusion can be drawn from the fact that the formed conductive clusters have the appearance of a linear chain structure.