From: "Terrell W. Fritz"
Date: Sun, 24 Jun 2001 14:23:43 -0600
Subject: Re: [TSSP] Direct Voltage Measurement of Topload Voltage for
HI Paul, Excellent point! The dimensions would indeed have to be increased to control the internal field strengths to what the oil could tolerate. I have also been looking into the standard capacitor stacks. Reading between the lines of the Ross Engineering notes suggests that these stacks may not be all that sensitive to external fields anyway. Any idea what a maximum capacitance load and the minimum resistance of a probe could be without affecting the Tesla coil too badly? Maybe 10pF at 500M ohm? Ross has a 240kV 6.2pF 480M ohm DC-1MHz 10,000:1 (VD240-6.2-C). It sounds like they can make up anything you want. I wonder if it is cheaper to make your own or just buy them from Ross? I should ask the price of this one to get an idea... I would have to find a retail source for doorknob caps and high voltage resistors... We may have SF6 laying around work somewhere. Cheers, Terry At 12:23 PM 6/24/2001 +0100, you wrote: >Hi Terry, > >There's a problem with it - the E field gradient close to the >0.125" inner conductor considerably exceeds the oil breakdown >threshold: > >Within the coax, E(r) = Q/(2 * pi * epsilon * r * L) > = C * V/(2 * pi * epsilon * r * L) volts/metre > >where a <= r <= b, and L is the length 0.122m > >so setting r to the inner conductor radius a, > > r = a = 0.125 * 0.0254 = 3.175e-3 > >E(a) = 5e-12 * 500e3/(2 * pi * 2.2 * 8.854e-12 * 3.175e-3 * 0.122) > = 52.7 MV/m = 1339 kV/inch > >Working the other way, the minimum radius for the inner of a 5pF coax >cap at 500kV terminal voltage would be > >200e3 * 39.37 = E(a_min) > = 5e-12 * 500e3/(2 * pi * 2.2 * 8.854e-12 * a_min * len) > >so a_min * len = > > 5e-12 * 500e3/(2 * pi * 2.2 * 8.854e-12 * 200e3 * 39.37) = 2.6e-3 > >For a length 0.122m that gives a min radius of 2.1 cm, giving a 20cm >radius to the outer. There might be a similar problem with the top rim >of the outer electrode, which might need to be rolled over to form a >nicely radiused lip. > >The coaxial capacitor is appealing because of its inherent insensitivity >to anything but the voltage on its input terminal. Perhaps a co-linear >cascade of coaxial segments forming a stack of series connected caps? >The dimensions would be a lot more favourable, but it would be quite >hard to make I think. > >Cheers, >-- >Paul Nicholson, >Manchester, UK. >--
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.