To investigate charged defects in gapped materials by first-principles calculations, they must be charged by either adding/removing electrons or compensating donors/acceptors. The former approach is more common, but it is not without drawbacks. We tested the latter method for a collection of model systems consisting of charge-compensated point defects in diamond (NV/SiV-centers, substitutional nitrogen/phosphorous/oxygen/sulfur donors, and substitutional boron/beryllium acceptors), comparing the geometrical and electronic properties of the compensated defect pairs with those of the individual defects in charged supercells. We find that the charging by explicit donors/acceptors works well and can be advantageous if properly designed, although interpretation of the results can be challenging. In this archive, we share the final optimized geometries of all studied structural models (in CIF format).