Possible failure mode for rolled caps

Hi ...

A question on Blumlein generators rang a very faint
bell, which just connected.

In the IEEE Pulsed power conference of 1976, in an
abstract on page IIIC-9 is a dissertation on spiral

For those of us using home rolled caps connected
only on one end, we have, in fact, a spiral generator.

The voltage gain when the input is shorted after
charging can be in theory 2 times the number of turns,
and occurs between the inner and outer turn. However
the gap has to go really fast, much less than the
transit time of the spiral.

Normal firing of the gap would be too slow to do much,
since the primary is in series, but when the 'safety' gap
across the cap goes, down in the middle of the roll
appears probably about 10 times the charging voltage,
assuming around 10 turns on the cap plus losses.

Although there is no place to go for this brief pulse, it
nearly has to contribute to ionization, even in oil, and may
potentially help explain why rolled caps have to be so

Maybe another reason to look at MMC's