Dopaminergic (DA) neurons display two functionally distinct modes of electrical activity: low- and high-frequency firing. We suggest a new minimal computational model that unites data on these firing modes obtained under different experimental conditions. The model reproduces the separation of maximal frequencies under NMDA synaptic stimulation vs. other treatments. In accord to recent experimental data, NMDA stimulation restricted to the soma effectively evokes high-frequency oscillations in the model. We have also reproduced low- and high-frequency oscillations under blockade of the SK current. Thus, the new model suggests a way that overcomes all major limitations of the switching dominance mechanism for controlling the frequency of the DA neuron. We explain recent experimental facts and make further predictions.