The development of a digital prototype using a digital signal processor (DSP), which can perform two algorithms to diagnose faults in the stator of induction motors is shown in this work. The first algorithm uses artificial neural networks to estimate the negative sequence current, which is used as a fault indicator. The second algorithm uses the negative sequence impedance as a fault indicator. The general structure of the prototype and its implementation are presented in detail. Experimental results by online measurements using a 3 HP induction motor are also included.