From: Bert Hickman
Date: Tue, 19 Jun 2001 07:33:16 -0500
Subject: Re: [TSSP] Some Considerations
Marco, Malcolm and all, Was thinking about this last night - a suspended capacitive divider as Marco describes OR one going through the center of the coil should work as long as the voltage stress is kept sufficiently low. The presence of the divider will unavoidably disrupt the local E-field around the top terminal as well as capacitively loading it. If we use low k materials around the divider this should be controllable, and if the voltage is kept below the point of corona or breakout around the divider (due to increased voltage stress at the triple point of metal contact/air/housing) we should be OK. Since we really don't need to capture any DC component, a simple capacitive divider should work as long as corona can be prevented and as long as its LF response is sufficient. Perhaps a chain of vacuum capacitors immersed in oil and housed in a low dielectric constant (polypropylene or polyethylene?) pipe? By immersing it in the more uniform E-field seen at the top or bottom of the toroid, the E-stresses can be made more uniform along the length of the divider. The problem with a pure capacitive divider versus a compensated divider is the changing response characteristic versus frequency, but allowances can be made for this. I have also seen professional dividers that use ceramic capacitors in the chain - have often wondered how they got around the effects of the capacitance changing as a function of applied voltage (a common problem for high-k ceramic dielectrics). -- Bert -- -- Bert Hickman Stoneridge Engineering Email: bert.hickman@aquila.net Web Site: http://www.teslamania.com Marco Denicolai wrote: > > Hello all, > > Malcolm Watts wrote: > > Where were you going to position the divider? Off-axis to the coil? > > We are talking about a 2-3 m high "tower": it is usually connected to > the target with an aluminium rod, supported by a rope hanging from the > roof (roof is maybe 10 m high). > > > Also, I wanted to eliminate as far as possible any loading effects. I > > was thinking of resistance of several GOhm. > > I believe capacitive dividers are really purely capacitive (not a > compensated resistive divider). > > >If we could get some > > accurate measurement under disruptive conditions but not necessarily > > at an arbitrarily high power level I think it would be reasonable to > > extrapolate the result under non-breakout conditions at least as it > > is possible to accurately quantify the secondary energy under those > > msmt conditions. > > If we recall that "disruptive" doesn't necessarily imply streamer > formation (with streamer loading, uncertainty, and all what follows), we > can run the measurement at a medium power level, thus avoiding extra > streamers leaving the aluminium rod. > > > > > > It would be also a matter of choice to retune the coil to compensate for > > > the tuning ratio shift or just to ignore it, as that is modeled too. > > > > It seems to me that this is a chicken and egg situation. We want to > > verify models so incorporating a model as part of the verification > > process seems self-defeating. Yes? > > The capacitive divider is a fairly simple device. It should be possible > to make a trustable model of it in five minutes. At least to a degree of > precision good enough for our purposes. It's a device used for precision > measurements, not a home-made low-budget toy like I have got at home. > We are basically looking for its capacitance (I guess), not its > resistive part. > > Regards > > -- > _____________________________________________________________ > > Marco Denicolai Senior Design Engineer > Tellabs Oy tel: +358 9 4131 2769 > DSL Products mobile: +358 50 353 9468 > Sinikalliontie 7 fax: +358 9 4131 2410 > 02630 Espoo FINLAND email: marco.denicolai@tellabs.com > _____________________________________________________________
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.