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

Re: joules confusion sort of



Original poster: "Gerry Reynolds" <gerryreynolds-at-earthlink-dot-net> 



 > Original poster: Harvey Norris <harvich-at-yahoo-dot-com>
 >
 > Isnt Cp irrevelant to the BPS rate, and any small
 > capacity will naturally exibit a 120 BPS rate by the
 > proper gap spacing?

For static gaps, once you set the gap properly, Cp will determine the BPS.
Of course, if you leave Cp fixed and reduce the gap setting the BPS will
change as well.

 > I am still at a loss as to how the capacity correlates
 > to the BPS rate?

If you go to a smaller cap, it will take less time for it to charge and the
spark gap to fire.  Once the energy is gone and the SG quenches, it will
start charging again.  If small enough, it will get sufficiently charged by
the next peak.  If the charging cycle does not result in SG firing, energy
will slosh back and forth between the cap and the transformer inductance
while more energy is put into the system.  Eventually, the AC voltage on the
cap will reach its steady state AC level and cease rising (if SG doesn't
fire before hand and rekindle another transient response).  The voltage on
the cap will be the sum of the forced (60 Hz) and transient responses.  Any
one transient response will die out in time and there may be many transient
responses going on at any one time.  A well firing system will probably
always have some transients going on.  The BPS can be higher or lower than
the 120 AC peak rate (can have multiple firings in a half cycle and/or you
have half cycles that don't fire).

 > With a SRSG at 120 BPS (called pps as in
 >  > presentations per secend)
 >  > and timed right, a Cp of 3.2 * Cres can be almost
 >  > fully charged (hence
 >  > approximately 900 watts are can be processed.

 > Isnt this a more complicated issue where we need to
 > find the RC time constant of the tank capacity value
 > to know if the cap can be fully charged in one half of
 > the AC input voltage signal?

The term RC time constant keeps coming up as if the cap was being charged
thru a resister.  Im sure there is a history to this that I'm not privy to.
I'm in search of a better way that does take resonant charging into account.
It is this effect, I believe, that seems to cause divergence between the "RC
time constant" prediction and what I measure for cases of Cp >= Cres.  Power
processing apears to be much higher than what the predicted BPS  would
suggest.

The actual charging circuit is with an LRC network where the R (effective
winding resistance) is probably not significant compared to the inductive
reactance of the transformer.  For a synchronous rotatary spark gap (SRSG)
where the firing is well controlled (frequency and phase wise) the energy
that is stored in the xformer's inductance can be utilized (inductive kick)
to get a larger value cap charged "fully" before the next presentation.
Please see richieburnett's web site for theory of operation.  He does a
great job explaining this.

IMO, the static gap system is harder to analyse because of chaotic like
nature of its behavior.  From what I've read, SRSG's are generally better
behaved and easier to analyze.  I will build one before long and then we
will see :-)))

Gerry R