From: boris petkovic
Date: Thu, 10 Oct 2002 10:55:01 -0700 (PDT)
Subject: Re: [TSSP] Top Voltage
Hi Paul, > I had in mind that the various factors that have > been mentioned > might combine to drop the breakout field quite a way > below 26kV/cm, > say down to 12 or 13 kV/cm. --- I would hardly imagine something like that in mentioned magnitude.I think (this doesn't necessary mean that I'm 100 % correct) wild variations in temperature,moisture , air pressure,content ( etc,etc) are required for such things.What I agree with is that breakout voltage may vary ,but this is another thing. Notch rise times (k) and difference in polarities between successive rf semicycles might be also important factors (IMO ,more important than decay rf cycles in low Q OLTC secondary).All TBD things. --- > > But with the sphere in place, and Terry reporting > 190kV, that > puts us back to the original observation: the > sphere is not > breaking out until much higher than software > predicts. --- But <50% discrepacy is within reach.That is not that bad.Breakout events are stohastic in nature and something like +- 10% correlation for every shoot is illusion to be ever achieved. --- > > > I remodelled this low sphere to predict a breakout > voltage > of 120kV, based on 26kV/cm. The 190kV breakout > translates > to a breakdown field of 42 kV/cm. --- Has he experimented with "higher positioned" sphere (say 1 feet+ above toroid) and still no breakout with 180-190 kV?Sorry for asking that but I haven't read all tssp post details yet. --- > Maybe the software's wrong. It reproduces tabulated > > sphere-sphere breakdown voltages reasonably well, > and the > surface field strengths from which these are > calculated are > a free by-product of the cap calculations. If the > field was > wrong, so would the caps be too. Maybe I've got the > average > surface charge correct (so the cap is right) but the > minima > and maxima are a way out? The software searches for > the > location of highest surface field and goes off that. --- Profound questions.You would better know to investigate software pecularities since you are the author.How about trying to use E-tesla for gotten structure voltage profile? --- > Is the breakout threshold much higher than we expect > when > we're only dealing with a very few RF half cycles > repeated > every 100mS? -- As some experimental observation suggest,for 300 kHz high k single pulse TC systems the threshold could be more than 2X higher (compared with 28 kV/cm).For 38 kHz OLTC the effect of corona formative time lags shouldn't be that large I guess. --- > > Boris wrote: > > If so,Terry improved OLTC secondary Q (or it was > > improved by itself) and can do somewhat better by > > changing k factor to see how rise time effect > > breakout. > > Yes, or maybe put the small sphere on a CW coil as > the only > top terminal. The breakout voltages could then be > determined > (eg by measuring base current) and we could easily > test if this > changes as a function of burst length and burst > rate. > > Note that with breakout voltages this far out from > the expected > values, there's no immediate pressure to setup a > topvolts probe. > A measure of Ibase is more than adequate to estimate > Vtop. > -- If Terry can run OLTC secondary in CW mode somehow (remember Q=80 now) ,than excellent.Few seconds of CW run wouldn't do any harm to the secondary,but again it is him to decide. regards, Boris __________________________________________________ Do you Yahoo!? Faith Hill - Exclusive Performances, Videos & More http://faith.yahoo.com
Maintainer Paul Nicholson, paul@abelian.demon.co.uk.