Re: Czpacitor Question- Repost

["Jim Lux" <jimlux-at-xxxxxxxxxxxxx> (by way of Terry Fritz <twftesla-at-xxxxxxxxxx>)]

Typically, you start with what is called "partial discharge" testing.
Search the web and you'll find some links to the folks who make the
equipment that does this, with explanations of how it works.

The idea is that you drive the device under test with a high impedance
source (like a NST), and look for the corona or partial discharges that
occur before actual puncture.  Typically, an oscilloscope, coupled by a
suitable HV capacitor (which can be pretty small.. a few tens of pF,
probably) to look for the high frequency "hash" characteristic of the
discharges.


Once you've characterized the breakdown voltage of your device, the next
thing would be to measure the loss.  (actually, you might want to do this
one first)  Probably the easiest way to do this is to put it in a tuned
circuit and measure the Q. Another way is to run a fast edge signal through
a resistor into it (say, from a fast CMOS driver.. like a 74ACxxx or
something) and look at the waveshape with a scope.

You'll also need to characterize the self inductance.

Once you've got the loss measured, you can make a good estimate of the
power (RMS current) handling capability, using some ball park estimates for
the thermal mass and conductivity.

These measurements are all straightforward, in principal, but might be
tough to do with a high degree of precision.

Tesla list wrote:
> 
> Original poster: "Duke, Ronn (CCI-San Diego CCC)" <Ron.Duke-at-cox-dot-com>
> 
> Hi Terry and fellow coilers,
> 
> This question is for all you 'seasoned coilers' out there. I have a new idea
> for a rolled poly cap, and I wanted to test it to destruction to see if my
> calculations for rating are correct and to see how much abuse it would take.
> The question is: What would be the best way to set up a circuit to test this
> cap using a 15/60 transformer?  I plan on using a variac to control power
> and a HV probe to monitor input to the cap. How many spark gaps should I
> use? I assume some sort of protection filter circuit would be in order to
> save the transformer.
> Thank you for your help.
> 
> Ronn
> (Teslamanian Devil)