From: Paul
Date: Sun, 13 Oct 2002 23:33:52 +0100
Subject: Re: [TSSP] Top Voltage
Boris wrote: > 1" sphere will have much higher surface field > (multifold) for given voltage than 1" thick wire > (which is long ,"perpendicular" to ground plane (at > average height say 10 m). In support of Boris's answer, the familiar bathtub shape applies to the long horizonal wire over ground plane, (both distributed C and charge distribution) so the surface field over most of the conductor may be quite a bit less than you'd expect just by its radius. Malcolm wrote: > The absolute losses in the secondary ground path are increasing as > base current goes up. Too early in the morning for me to look from > the point of view of percentage losses. Does this make sense? Not entirely, although the absolute losses are going up, the Q should remain constant, or increase slightly. The slight increase comes from the assumption of a constant saturation voltage of the IGBTs, so that the VI loss in the devices becomes a smaller fraction of total loss as the current goes up. As Vfire is raised from 25V to 300V we would expect the Q to first rise as primary device loss becomes less significant, then would start to fall as streamer loading takes over. Seems that we may be seeing that fall start as early as 100V? Wonder if if rises before that. Terry, have you looked in the dark for signs of early breakout? Oh well, too late at night for me at this end. I'll have to recompute things using a revised secondary Q, but it's not going to make more than a few % difference, I think. -- Paul Nicholson, --
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.