This research connects two progressive branches of modern measurement. Digital signal processing and self-calibration are frequently used methods in the field of high-accuracy measurement. However, there are only few applications up to now dealing with both of them. Signal processing and self-calibrating methods are applied together in two fields: in the development of multiplying digital-analog converters and high-accuracy power sources. Based on previous experiences it is promising to utilize the signal model (resonator) based approach. Multiplying digital-analog converters are frequently used in electrical instruments, but their linearity features are limited due to the technology. One possibility is to improve the linearity is the application of self-calibration. The determination of the correction terms necessary for calibration can be associated with a certain testing problem, which can be solved by a signal processing method (Walsh-Hadamard transform). If the correction terms are already known, the calibration procedure can be designed, resulting in a high-linearity digital-analog converter. The Virtual Power Source (VPS) is a new concept in the field of calibration of electrical power meters. Recently this indirect method is used instead of the direct one, since it shows better performance. Since the VPSs have high-power output, it is indispensable to use electrical transformers (voltage and current transformers). The aim of the research is to develop such a high-accuracy power source. Furthermore, it is also important to acquire experience with self-calibration of such non-linear instruments as transformers.