In testing Discrete Event System, an important topic is determining the final state of the machine after the application of a test. Synchronizing and homing sequences have been proposed in the sixties to solve the problem using Mealy machines. A synchronizing sequence drives an implementation of a given model, seen as a black box, to a known state regardless of its initial state and the outputs. In this paper, we investigate how to determine synchronizing sequences using systems represented by a class of synchronized Petri nets. We propose an approach that can be applied to (not necessarily strongly) connected nets. Regardless of the number of tokens that the net contains, a synchronizing sequence may be computed in terms of the net structure, thus avoiding the state explosion problem.