by dcgillespie » Thu Aug 05, 2010 9:03 am
No doubt it is imperative to identify the exact conditions under which an arc occurs, before any progress can be made as to why. Arcs that occur literally at turn on, after warm up, under light loading, heavy loading, etc. imply different concerns. I went through much the same process years ago when I set out to determine why output tubes arced all too frequently. Identifying the exact conditions present at the time of an event is important to finding the cause.
With new tubes today, you can't assume proper manufacturing conditions, or proper manufacture to a given tube's standards either, so manufacturing source becomes part of the equation.
On the other hand, if a tube is well manufactured, then the issue will either come down to one of excessive current flow or PIV breakdown relative to safe area operation for the tube and choice of filter used.
Regarding the size of the first filter cap, the concern for hot switching is what places the biggest limitations on the size of that cap. When hot switching occurs, that is when the greatest peak currents occur -- hence, the need to limit the cap's size to limit peak current under that condition. It is also why the use of current in-rush limiters are so crucial when the size of these caps are increased.
Dave