[TCML] cap rms current

[Carl Noggle]

These Maxwell caps are bulletproof.  I bought a 0.05uF, 50kV cap about 
20 years ago knowing that it had a leak around the top terminals, and it 
finally gave up the ghost about two weeks ago, after much service in my 
TC, 1800 watts from three 9/60s in parallel.  I got a nonleaking one at 
the same time and it is still going strong.





> trying to start/move the thread
>
> Jay,
>
> I have a cat # 31885, .1 ufd, 75 kV rated Maxwell
> that I believe is also rated at a "mere" 25 amps RMS,
> but I run my Green Monster coil (10 kVA pig fed
> and overdriven up to 20 kVA) with it as the sole
> primary cap and I have never been able to notice
> any external warmth on it above ambient temp,
> even after several minutes of non-stop 15 kVA +
> running.
>
> David Rieben
>
>
> ---
>
> Assuming your pig is 14.4kV, at 15kVA it's roughly 1 Amp RMS "into the 
> coil"
>
> Now, the question comes up of "what's the effective RMS current in the 
> cap".  I like to back of the envelope these kinds of things by doing 
> an energy balance kind of computation.
>
> Let's, say, for argument that your coil runs at 100kHz, and takes 
> about 10 cycles to go from "break" to "quench", or, call it 100 
> microseconds.
>
> So, during a 8.3 millisecond half cycle, the cap is "charging" at 1 
> amp (from the pig) for most of it, and then, for 0.1 millisecond it is 
> discharging ALL that stored energy.
>
> That is, in round numbers, it spends 99% of the time charging and 1% 
> of the time discharging.
>
> We know that energy goes as the square of either voltage or current, 
> so the RMS current during discharge is going to be 1/10th of the RMS 
> current during charging (since we already know that the energy flow 
> ratio is 100:1)
>
> therefore, I would assume that you're running an effective RMS current 
> of somewhere between 10-15 Amps, well within the rating of your 
> Maxwell cap.
>
>
> Somewhere back in the archives, there's a better calculation of this.  
> I integrated a variety of damped sinusoids (linear and exponential 
> decrement).  It would be pretty easy for someone to set up an excel 
> spreadsheet and do it by numerical integration, as well.
> _______________________________________________
> Tesla mailing list
> Tesla at pupman.com
> http://www.pupman.com/mailman/listinfo/tesla