Nowadays, antenna arrays are important tools adopted in a great number of applications including radar, mobile and satellite communication systems, and electromagnetic imaging. Moreover, in these applications, arrays with a high number of elements are ever more requested, which results in a growing possibility of damages in the array. The identification of defective components in array of antennas is really significant due to their applicative use: indeed, faulty detected elements can be fixed, thus avoiding to replace the whole antenna. In this work, a diagnostic technique for planar antenna arrays is presented. This approach enables recovering the eventually defective elements of the antenna under test using far-field data. To this end, an inversion approach established outside the standard context of Hilbertian spaces is used to address an inverse-source problem. A numerical validation concerning simple array antennas has been carried out to study the performances of the approach versus some antenna parameters, e.g., the size of the array.