Condition monitoring (CM) involves the collection of high frequency, instantaneous data for feature extraction. At present, many systems use separate modules to monitor one specific component or subsystem, with one particular type of detection technique. Clearly, it would be beneficial to monitor parameters associated with a variety of components to better identify potential failures of both individual components and of the system as a whole. This paper presents a novel architecture for condition monitoring of wind turbines. The system is developed around a reconfigurable control and data acquisition system, for which hardware modules can be configured for a particular set of signals, thus tailoring the system to specific monitoring tasks. Wavelet analysis is used for signal recovery by means of feature extraction in order to reduce the amount of data transmitted and stored by the CM system. An automated calculation of the Lipschitz exponent, a measure of local signal regularity, is proposed to infer the location, duration and severity of the faults. A FPGA (field-programmable gate array), embedded in the system has been utilised, allowing critical signal processing tasks to be undertaken for real-time monitoring purposes. The proposed algorithms are tested and validated using simulation data from a Simulink/SimPowerSystems model of a DFIG-based wind turbine. In turn, the system can also be configured as a real-time simulation of a grid-connected wind turbine on the same hardware platform in order to test control and protection systems onsite for wind turbines.