[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] JavaTC Calc. Programm



Hello Fritz, All,

I hoped Bart would chime in, and explain how the static-gap is handled in the Java TC model (what he did :-) ). If I'm not wrong, it seems to be a reasonable linear model, to give a raw estimate of plausible gap performance (a way I'd do it myself too). However I'm not shure, if it was possible, taking in consideration (and to what degree) some of the really hard to model effects, which nevertheless play an important role in the functioning of a static gap in a TC. Three of those effects, interacting with each other, are:

1.) Chaotic (within limits!) firing of a static gap in TC use: the experimental BPS is almost never the one estimated with a design program, 'cause of the items following. Most of the time it will be higher in a well performing TC. Simulation offers a better guess. The measurement of BPS was an interesting topic on this list: search the archives for "BPS measurement". Gary Lau has pioneered the view of the static gap for it's chaotic behavior, by doing important Microsim/Pspice simulations.

2.) Partial (slight or strong!) (50/60Hz) resonant rise of charging the primary cap, from the NST, even if the cap is LTR. This depends of cap uF value, and leak-inductance behavior of the NST (see 3.)). The cap charging peak voltage can easily be qite a bit higher than the NST nameplate HV voltage, which represents 100% in Java TC. This is dangerous for the NST life, but obviously gives better TC performance. I'm quite confident in assuming most of the well performing NST/static gap TC's are charging the cap higher than the 100% of the NST nameplate.
(see and digest Richie Burnett's site).

3.) Current-drawn-depending effects of the leak-inductance of the NST, as a cause of saturation in the magnetic shunts --> which leads to so called "ferroresonant" effects. If i.e. you have an LTR cap, the planned resonance of the primary cap loading system will be _below_ 50Hz (you designed the cap for 1.57 resonant). If now you begin to feed the primary system, and reach the limit, where the NST shunts start saturating, which lowers the leak-inductance, and approaches the primary cap-load tuning to the 50Hz resonant value, you will experience a higher resonant voltage rise (2.)) and power draw. More current is drawn... and the cat begins to bite it's tail, till some balanced point is reached.

These effects are very hard to model in a design program, 'cause no analytical formulae are available for it (with the exception of pt.2.) ). Ferroresonace effects were disputed on the TCML, but no estimation of the effects seem to be available. (Maybe David Mekers magnetics-simulator might be a tool for estimation, in specific cases). My little UBTT-Twin 15/60 NST draws about 1.6kVA from the line, instead of the 900VA calculated by nameplate figures, which has to do with the above effects.

I recommend to read Gary Lau's investigation about chaotic static gap firing:
http://www.laushaus.com/tesla/
http://www.laushaus.com/tesla/gapsim.htm
...please also read the last "late and important footnote", regarding NST saturation effects, very carefully.

The wonderful site of Ritchie Burnett also is a "must" for the topic:
http://www.richieburnett.co.uk/tesla.shtml
...and specifically for the static-gap
http://www.richieburnett.co.uk/static.html
pay special attention to the graph:
http://www.richieburnett.co.uk/static8.gif
and it's description.

Hope, I was able to support you a little

Cheers,

       Kurt


Fritz W. Egli wrote:
Dear Gentlemen,

I woud like to gain some more understanding conc. JavaTC
calculation-design programm.
Following 3 points in the file  STATIC SPARC GAP OUTPUTS  are of my
interest for a clear view:

XXXX (peak volts) = Charging Voltage

XX,X (%) = Percent Cp Charged When Gap Fires

XX (BPS) = Breaks Per Second

Can anyone please explain me the exact meanings of above terms
and the best target-figures (range) to be achieved in a good, sane TC
design -?

Please in short, simple terms, as my language is not English.

Thank you, kind regards,
Fritz W. Egli


_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla