Re: cap specifications
I'm a bit curious as to these caps's rating
of tesla service ??? --- or any service for that matter.
If the L.C.R. ckt.is in resonance ---
the power factor is 100 % ----or 1 with respect
to the cap.---- power is dissipated by virtue
of dielectric heating ---- in direct proportion
to the dielectric constant --- harmonics notwithstanding.
-at- K = 25 ... 40-- around 25 % + of the energy of the ckt
can be wasted -- at resonance .---
Poly has a K of around 2 --- so around 2% --
Swollen capacitors syndrome ----
Tesla used methods of recirculating the oil
in his high power systems---- one of the purposes being
to cool it,another purpose being to run the rotary spark gap .
In induction heating ---- (high duty cycles - MONSTER POWER !!)
water cooled capacitors are an absolute .
I have a switchable bank of .002 20 kv 22 A.-at-
3 MC & 1MC ----- derated to 15 A. -at- 400 KC that I use
to tune --as well as run my experiments ----- Bullet proof so far ----
they are mica transmitting caps -- no cooling necessary (YET) !
In TC applications the spark gap is a primary voltage limiting
device,(stationary is most easily configured ). the rotary gap,on the
other hand,has a tendency to push the envelope to its max.---
(and beyond) ! -- zappenfrappen !!
I have had good success with High pressure axial air flow.
-- I'm soon to continue my work (time and kids permitting)
on both my stationary and rotary gap with hydrogen as the quenching medium --
I'm wondering if it wouldn't be prudent to drill a hole
in the cap ---- insert a glass tube-seal it in-- top off the oil & use the
"oil manometer" to check the temperature of the system
by the expansion of the oil ??---- Also a safety valve---
Resonant tuning can be facilitated by the implimentation
of such a device designed for that purpose as well .
If there is any gratafication in an explosion of
a capacitor ----- It would be ---- "It started working " !!!!!
(assuming the design frequency was at work --not
some sneaky harmonic ) If one cap was used ----
get 3 more & connect them in series parallel ---
I would appreciate feedback on the above ----
Make it work --- then make it work gooder !