Re: Maxwell 31159 cap on eBay

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: Ed Phillips <evp-at-pacbell-dot-net> 

"If a HV pulse cap is not rated for high % voltage reversals, there is
significant corona formation at the edge of the foils.  This corona and
surface tracking begin to heat up the local dielectric area (di/dt) and
will
eventually produce enough heat to begin melting through the dielectric.

Dielectric breakdown at the edge of the foil is the number one killer of
HV
pulse duty caps (according to Maxwell Sr. Engineer Bob Cooper).  Second
on
the list are small voids in the dielectric material, which also produce
corona in the voids due to the different dielectric constant in the void
material, and, again, leading to local intense heating effects that
damage
the dielectric.

Pulse caps not rated for high voltage reversals literally "can't take
the
heat".

Dr. Resonance"

	I understand the breakdown phenomena and also the problem with corona
in voids under AC operation, but don't understand how the construction
of the "high voltage reversal" capacitors differs from the "low voltage
reversal" ones.  Are they built differently or just screened?  When I
was at Hughes Aircraft we had a group building special capacitors for
radar modulator pulse-forming networks (where dV/dT can exceed 10^7
volts/sec and operation is at several thousand pps) and they used to
screen capacitors after construction by applying a high AC voltage and
listening for radio noise as a symptom of corona in unfilled voids.  In
this case they were pitched out because the construction techniques
including vacuum impregnation were the best they knew about.  Does
Maxwell do something like this but salvages the rejects by issuing them
under a different part number?

Ed