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

Re: Res cap size charge



Original poster: Bart Anderson <classi6-at-classictesla-dot-com> 

Hi Luke,

I guess I have to disagree. The time constant doesn't fall apart. The time 
constant is very relevant to cap and transformer selection for breaking at 
a particular rate at a particular voltage. The voltage at any point in time 
can be derived with the time constant. Yes, of course the voltage is 
changing (it's AC). The time constant isn't based on the voltage, it is 
based on the capacitance and impedance. For accuracy, it helps to compute 
the complex impedance, but even a bulk impedance is very helpful. 
Regardless, it's "when" to look at the time constant that is relevant and 
helpful. The breakrate is an obvious time to look. Suppose I had a cap of 
5uF and used a little 9/30 NST to fire across a 9kV gap at 120 bps? The 
answer via the rc time constant (and in real life) is "you can't" because 
the voltage is only 71 volts at the cap. It's the time constant that makes 
this information available before you go to all trouble only to find out 
had you equated the cap voltage in the first place, you would have seen 
that 71 volts isn't going to do much for you.

This is how the time constant is "useful". Yes, there may be issues with 
resonant rise, inductive kick, but when that is the case, the complex 
impedance in those situations must be used for precision numbers, but even 
a close approximation is still very useful in the design stage.

Take care,
Bart

Tesla list wrote:

>Original poster: "Luke" <Bluu-at-cox-dot-net>
>
>I was looking around the list archives and making up a spread sheet for
>joules of the cap after charge time of 1/4 cycle and voltage on the cap
>and number of rc time constants in the 14 cycle etc............
>
>When I ran across a post on the archives pointing out the obvious. :(
>The voltage is always changing so the rc time constant thing kind of
>falls apart.
>
>Thanx for the info.
>
>
>Looking at the infor on the first link you sent I ran across something I
>would like some opinions on.
>
>Seems a good value for c would be 1.357 times c res.
>That is where you hit the 120bps mark.  Going below this obviously gets
>too close to c res and going above this slowes down the bps while also
>giving less real power on the cap.   Why is the rule of thumb them 1.6
>times c res for a static gap?
>Luke Galyan
>Bluu-at-cox-dot-net
>
>-----Original Message-----
>From: Tesla list [mailto:tesla-at-pupman-dot-com]
>Sent: Wednesday, January 21, 2004 7:29 PM
>To: tesla-at-pupman-dot-com
>Subject: Re: Res cap size charge
>
>Original poster: Terry Fritz <teslalist-at-twfpowerelectronics-dot-com>
>
>Hi Luke,
>
>At 06:41 PM 1/21/2004, you wrote:
> >I know it is not recommended to use a resonant size cap and that is not
>my
> >intent.
> >
> >Using ohms law for the impedance of an NST of 15KV 60mA gives 250k ohms
> >Using the capacitive reactance formula there the capacitance at 60Hz is
>
> >0.0106mfd
>
>Yes.
>
> >
> >Using 250k ohms and 0.0106mfd the RC time constant would be 2.65ms
> >Since the time to charge a cap is ¼ of the wave the time for cap
>charging
> >at 60Hz is 4.17ms.
>
>Nope.  It get's real messy now...  Although the time constant thing
>gives
>sort of comfortable numbers at first.  It is all screwed up in this
>case.
>
> >This works out to the cap going through 1.57 RC time constants for each
>
> >charge cycle.
> >This is about 75% charged and 75% of 15KV is 11.2KV.
> >
> >So using a resonant sized cap will give max voltage of 75% (11.2KV in
>this
> >case).
> >
> >Does this mean setting the spark gap (static) to fire at 11.2KV would
>give
> >the same voltage to the system as if the gap were set to fire at 15kv?
> >Since the 15kv from the NST would fire the gap but only 11.2KV would
> >discharge from the cap into the primary.
> >
> >I know that the amount of energy is not the same as the amount of
> >voltage.  Energy is will be related to joules which is dependent on
>both
> >voltage and cap size.  I am only talking about the amount of voltage
> >present not the amount of energy.
>
>When F = 1/ (2 x pi x QSRT (L x C)) , or, 250k = 1 / (2 x pi x 60 x
>C),  all kinds of factors start dividing out and things "theoretically"
>get
>explosive as infinite amounts of energy are transferred.  But in
>reality,
>resistances and losses keep the mess under theoretical control.  At that
>
>point, simple equations we learned at school start to fall down and we
>need
>to crank the mess through the computer to get results that reflect the
>"real world".  Here is a nice thing that deals with all this here, that
>a
>surprising amount of theory is based on, but nobody realizes it :o)))
>
>http://hot-streamer-dot-com/TeslaCoils/Misc/NSTStudy/NSTStudy.htm
>
>You "might" also want to see:
>
>http://hot-streamer-dot-com/TeslaCoils/OtherPapers/TeroRanta/CurrentLimitedT
>ransformers/NSTModel.htm
>
>http://hot-streamer-dot-com/TeslaCoils/OtherPapers/TeroRanta/NSTCapMatching/
>ResonantCapacitorMatch.htm
>
>but this is getting into areas that makes my brain hurt too -at-:o)))
>
>Cheers,
>
>          Terry
>
>
>
> >
> >Luke Galyan
> >Bluu-at-cox-dot-net
> >
>
>
>
>