|
Tom: I don't believe that arcing across terminals is what really determines
voltage ratings of at least carbon composition and possibly metal film resistors.
The reasson is that for a given voltage rating, terminal strips of the same outer material of the same spacing have higher voltage ratings. For a given size (spacing) epoxy transistors and diodes have enormously higher ratings.
I suspect that for carbon composition reststors the voltage rating has to do with possible internal damage due perhaps to tracking. In any case, experience says that the actual usable voltage for one and two watt carbon composition resistors of high resistance starting about 1947 or 48 is enormously higher than the ratings. The two watt ones easily take 2 kV.
I've used 2 watt 1 meg carbon resistors pulsed to 2.5 kV run continuously for months and on and off for 20 years with zero failures. Ditto for 1/2 watt resistors at 1 kV. One circuit generates 20 kV pulses with 800 pS rise times using 22 series connected ordinary transistors (yes, all one type and all only two near-sequential batches from one manufacturer) in avalanche mode with the bases driven by pulse transformers with resistor bias networks. The pulse rate was up to 10 kHz. I was stupefied when the idea worked at all, but after 20 years with zero failures in numerous systems .....
They ran continuously to keep the temperature constant as they were driving crystal polarization rotators that needed to keep the resonant frequency constant to better than 0.5 ppm (and rotators, while being quartz just like radio crystals, are an abysmally bad temperature quality cut).
|