@tomleslie Well, first, the problem what i try to solve is a reduction of the Hogdkin-Huxley Model of the neuroonal dynamics to a Cortical Neuron Model, in order to check the reducton of the computaqtional cost.
In the Cortical Neuron Model i need to introduce a spike from a threshold voltage to a peak voltage, and from this peak voltage to a rest voltage, i achieved this spike with events, but a instantaneus spike, really what i need is that the time from the threshold voltage to the rest voltage be a determinated refractory time in order to improve the accuracy of the model (the same number of spikes and at same time), that is what i don't achieve with events. Really the important is the time of integration betwen the rest voltage and the threhold voltage, off of the spike, being the spike a stereotiped form, barely a triangle.
I think that with parameters and control fow statements is easy, but reading the help on events, i though if is posible to specify a ODE system with three equations: the Cortical Neuronal Model, and the spike divided into two straigh lines expresed as a differential equation, the upward line [diff(u(t),t)=260/(t_refractory/2)] and the downward side line of the triangle, with idem expression with a negative pendient, (now i am not sure if this is a system of ODE's), and suitch into diferents equations dependenig the voltage u(t). That is what sggested me the expresion "...separated into disjoint cases dependent on the values of the values of particular triggers...).
Really, if it is posible, what i don't know is why express, write this, gramatically, in the command dsolve.