Use of phase-locked loop as a model of neuron-like element is discussed. Parameter space of the model is partitioned into areas of different regimes specific for dynamics of real neurons. Bifurcation mechanisms of transitions between regimes are examined.