[TCML] capacitor charge time?
jhowson4 at comcast.net
jhowson4 at comcast.net
Sat Oct 24 11:52:02 MDT 2009
Thanks for the explanation Steve
But I am still a little foggy
for the capacitor charging the input current decides on the time. which is what that equation gets me.
suppose i have a 12kv 60ma supply and a cap of 9nf
if i just let that charge all by it self the charge time is 2ms.
but wait i am using an ac supply so 1/f=T and the period is .017ms and a quarter of that is .0042ms meaning that my capacitor does charge in one quarter of a wave length.
but lets look at the other side now my coil has a resonant F of 328khz so its period is 3us and a quarter wave is .76us.
so really my cap only have 76us to become charged before it is forced to discharge; completely changing the first paragraph.
the charge after charging by the power supply could be 46nC thus the voltage across the cap at this time would be umm this cant be right 5v. Q/C=V 46nC/9nF =~5v
Steve this makes no sense to me
does this mean that the discharging of the capacitor is chaotic and does not follow the resonant frequency
or am i missing another critical idea that makes is all work out.
even in your explanation how can something charge on a ms scale at ma and discharge on a ns scale at ka.
John "Jay" Howson IV
----- Original Message -----
From: "Steve Ward" <steve.ward at gmail.com>
To: "Tesla Coil Mailing List" <tesla at pupman.com>
Sent: Monday, October 19, 2009 12:37:49 PM GMT -05:00 US/Canada Eastern
Subject: Re: [TCML] capacitor charge time?
In the case of a constant current source (or even otherwise as long as you
can do some calculus)
Q = C*V = I*t
Assuming the charging is limited purely by R is not valid in most tesla coil
charging systems. Generally its limited by some inductive impedance at
60hz. The impact of this impedance on charging the capacitor is
non-obvious, particularly as the capacitor is discharged at various phases
along the 60hz mains cycle. Using spice would be a good option to gain more
insight. Most tesla coilers dont bother to dig this deep and just pick a
ballast (inductor) that gets their average charging current within an
To get at your original idea... the resonant behavior of the tesla coil is
damped and long gone before you really even start to charge up the capacitor
again. You are dealing with time scales and power levels that are on very
different magnitudes. The capacitor charging is in the 10's of mA and mS
range, while the resonant discharge is in the 100s of Amps and 10's of uS
range. Interestingly enough, the DRSSTC is a solid-state approach to
driving a TC where the tank capacitor is always in a state of being "charged
up" by a special power source that can keep up with the 100's of Amps and
can switch polarities along with the resonant frequency of the system,
somewhat along the lines of your idea in question.
On Thu, Oct 15, 2009 at 6:26 PM, <jhowson4 at comcast.net> wrote:
> Hey guys.
> I was explaining the tesla coil to one of my physics friends (we are both
> sophmores) and he asked a question that i could not answer.
> I was asked if it would be a good idea to match the time required to charge
> a capacitor to (1/4) of the resonant frequencies period. with the general
> idea that the capacitor would be completely charged when the resonant wave
> form was at a max or a min, thus maximizing efficiency.
> sounds like a good idea to me.
> But when we went to try and do an example calculation we realized that our
> standard RC capacitor charging equation would not work because our current
> output from the transformer was fixed and would only decrease after the cap
> charging current became less than or equal to our source current.
> and this spawns my question. how does one calculate the charge time of a
> capacitor with a fixed source voltage and current.
> does the capacitor ignore the fact that the transformer is current limited
> and draw that initial huge current anyway. or is there some other special
> equation that accounts for a limited source current and fixed source
> or are we both just missing something.
> t=-CR ln(IR/V) derived from I=V/R e^(-t/RC)
> thanks for the help
> Jay Howson
> Tesla mailing list
> Tesla at pupman.com
Tesla mailing list
Tesla at pupman.com
More information about the Tesla