SELF

32

and S.B. Karavashkin

In parallel with the study of pattern of voltage change in the monotonous sections, we undertook a thorough study of the sections of voltage surges, to corroborate the premise of oscillation pattern of voltage change in these sections. As we saw above, we already have registered the fact that voltage surges in these sections are bipolar. Now using the higher resolution of horizontal sweep and possibility to vary smoothly the delay time between the oscillograph start and beginning of the process of contacts partition, we yielded the developed oscillograms (see Fig. 14) showing that these sections can include several over-oscillations unseen in general oscillogram of process because of grounded slow sweep.

a)

b)

Fig. 14. Sweeped oscillograms of the voltage surge sections showing the oscillation process of voltage change during these periods of developing showering arcs: a) L = 1,5 H, C_p = 0,001 F ; b) L = 1,5 H, C_p = 0 F

Taking into account Fig. 14, return again to Mills model shown in Fig. 4 . Draw our attention again to the arc breaks after each monotonous section that he introduced in order to substantiate contact erosion that occurs because of short arcs, of which we mentioned when considered oscillograms in Fig. 7  . We see that at the moments when Mills premised the arc process in the gap, there factually occurs the high-voltage oscillation process. But separate arc breaks between the intervals of showering arcs are, none the less, present in the pattern of contacts opening, though they appear not so often as Mills expected. In particular, in our oscillogram, Fig. 7 and in Germer's oscillogram, Fig. 4, they clearly are present. As we revealed during long-lasting study with different values of parasitic capacitance and load, the arc breaks of relaxation appear with growing switched power and parameters of load. And, just as we showed in Fig. 7, with these parameters growth, at the initial stage of unstable discharges there can be several breaks. The voltage in these sections is near-zero (we did not register the exact value because is was too little in comparison with the scale of vertical deflection of oscillograph beam). In the end of unstable discharges there can appear a break, but in it the voltage is (300 - 400) V, as Mills [2] also has registered (see Fig. 4).