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

Re: Tank Capacitance: what is the limit

Subject:      Re: Tank Capacitance: what is the limit
       Date:  Tue, 6 May 1997 08:10:01 +1200
       From:  "Malcolm Watts" <MALCOLM-at-directorate.wnp.ac.nz>
Organization: Wellington Polytechnic, NZ
         To:  tesla-at-pupman-dot-com


Hi Barry,

<snip>
> spark gap.  The question I am purplexed by is
> whether it is more efficient to transfer energy
> with many beats from a high Q primary system,
> incurring spark gap losses which are highest
> at the transitions, or with one or less beats
> from a low Q, big capacity, low turn, primary
> system which incurs less spark gap transition
> losse but bigger spark gap power loss from
> the higher current?

Difficult to quantify without knowing the amounts. But why should a 
lower current primary suffer more beats than a high current one? The 
gap is actually harder to forcibly extinguish in a high current 
primary.
    Incidentally, low frequency running for a given k should make 
forcible quenching easier because the gap is quieter for a longer 
period between trades. I'll stick my neck out and suggest that if you 
notice no difference in quench between setting the system at the 
loss-corrected value of a magic k and any other k, your forcible 
quenching is not working.

> The secondary is a classic resonator with a variable
> resistance (streamer), skin effect resistance,
> other interturn capacitive esr losses, and the
> variable capacitance of the streamer.
> With many beats the energy lost to heating the
> wire and ionizing the air will be greater.  With a
> single pulse would the energy loss be less with
> the result that more energy goes to the streamer?
> Or could it be that the streamer requires multiple
> pulses to reach maximum length?

Multiple in terms of primary cap charges from the transformer. It 
also takes several cycles of oscillation per cap charge to establish 
strong ionization. You can see this on the scope - the secondary 
never loses its power in just one cycle.

> > Have you ever tried feeding a single pulse at
> full power into the primary with a trigatron
> or field distortion swinging cascade switch?

No. Just single-shot using an ordinary gap.
 
> Reference equation from the Radiotron Designers
> Handbook, fourth edition, page 410:
> 
> I = E * r / ( r**2 + (w**2)(L**2))  at resonance
> 
> Q = w * L / r = 1 / (w * C)
> 
> I = current in parallel resonant circuit (primary)
> E = voltage applied to capacitor
> r = resistances in primary (nonreactive)
> w = 2 * Pi * frequency
> L = inductance
> Q = energy input per unit time / energy lost per unit time
> C = capacitance

Malcolm
 
 
> Barry
> 
>  ----------
> From: "tesla"-at-pupman-dot-com-at-PMDF-at-PAXMB1
> To: Benson Barry; "tesla"-at-poodle.pupman-dot-com-at-PMDF-at-PAXMB1
> Subject: Re: Tank Capacitance: what is the limit
> Date: Monday, April 28, 1997 6:39PM
> 
> <<File Attachment: 00000000.TXT>>
> Subject:       Re: Tank Capacitance: what is the limit
>        Date:   Mon, 28 Apr 1997 16:29:46 +1200
>        From:   "Malcolm Watts" <MALCOLM-at-directorate.wnp.ac.nz>
> Organization:  Wellington Polytechnic, NZ
>          To:   tesla-at-pupman-dot-com
> 
> 
> Hi all,
>         A comment on Fr McGahee's post:
> 
> > While it is generally true that the primary is more than one turn of
> > wire,
> > as the secondary becomes LARGER in DIAMETER, it is often advantageous to
> > use larger capacitors and fewer turns in the primary.
> 
> I usually run at lower frequencies and keep primary Q up by keeping
> Lp large. I hear things like "lower frequency implies lower Q" a lot.
> That no doubt comes from observing that Q = X/R. Let's have a closer
> look at that:
>                   Q = (2 x PI x f x L)/R
> 
> and we know that f = 1/(2 x PI x SQRT(L x C))
> 
>      Let's halve f in three ways, first by making C four times larger.
> That leaves L unaffected which indeed causes Q to dive in half since
> f has halved.
>      Now lets achieve the same thing by doubling L and doubling C. Q
> is now the same (as it should be because the L/C ratio is preserved.
>      Finally, let's quadruple L. Despite f having halved, Q is
> actually doubled.
> 
>      With reference to the primary, we know that "R" is a non-linear
> resistance. In fact I maintain it goes down as gap current goes up.
> If the voltage across the gap when it is conducting is reasonably
> constant, then gap losses double as the current is doubled. Despite
> gap resistance doubling as current is halved, it loses less power
> because it roughly follows a VxI dissipation law.
> 
> Malcolm
> 
> > That is just what
> > Tesla did on many occasions. As you change one parameter, such as the
> > coil
> > diameter or the size of the tank capacitor, it is sometimes necessary to
> > break some of the rules that NORMALLY work quite well. Never be afraid
> > to
> > experiment to find out what ACTUALLY works best in THIS particular
> > setup.
> > The common wisdom is often all too common and not much wisdom.
> >
> > In reading Tesla's Colorado Springs Notes I am kept contantly chuckling
> > at
> > the way he did things that EVEN went against his OWN recommendations (of
> > just a page or two before!) You HAVE to experiment and improvise in this
> > business. What at first appears to be an absolutely minor change can
> > have
> > an absolutely major effect on overall circuit performance. I have
> > learned
> > not to just blindly accept what others say is the BEST way of doing
> > anything, but to experiment on my own and try and understand WHY some
> > things work and some things don't *in this particular case*.
> >
> > Fr. Tom McGahee
> 
> >
> >
> 
>