Re: capacitor failing

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Steven Ward by way of Terry Fritz <twftesla-at-qwest-dot-net>" <srward16-at-hotmail-dot-com>

NO NO NO!!

The voltage stays the same when caps are in parallel.  Think about it, the 
poly sheet is still 60 mill thick, but now you just have more surface area, 
so greater capacitance.  When caps are in series, they gain a higher voltage 
rating, but loose the capacitance, because each added cap is like a thicker 
dialectric.  The thicker the dialectric, the lower capacitance you have 
given the same surface area for both situations.  So your thinking was 
incorrect, you still had a 24kv cap, but with 4x capacitance of the single 
unit.

Steve Ward.


>From: "Tesla list" <tesla-at-pupman-dot-com>
>To: tesla-at-pupman-dot-com
>Subject: capacitor failing
>Date: Tue, 13 Aug 2002 21:52:36 -0600
>
>Original poster: "Laurence Davis by way of Terry Fritz 
><twftesla-at-qwest-dot-net>" <meknar-at-hotmail-dot-com>
>
>(apologies for the simple question)
>my homemade cap failed. pe/aluminum in mineral spirits.
>60mil poly in 10 6 mil sheets with roofing aluminum for plates.
>
>there are four caps in the tank. the caps are connected in parallel.
>assuming 500v/mil dielectric strength for PE, that gives 30kv rating.
>some have said 400v/mil for PE, which gives 24kv rating. whatever.
>each cap at 7.2nf.
>
>I had thought the cap was rated at a minimum of 24kv.
>
>(since i've already proved i'm weak in the ohms law area :)
>
>the cap (since it was wired in parallel) was THEN actually rated
>at 6kv.  24kv / 4 = 6kv.
>
>am i correct that it was actually 6kv instead of 24kv?
>
>following that the voltage rating is an inverse relationship
>to capacitance...  well sort of.
>
>an example to be clear:
>3 capacitors.  100v -at- 100uf.
>
>in series...   300v -at- 33uf
>in parallel...  33v -at- 300uf
>
>in series...   ctot = ((c1 + c2 + c3)/3)/3
>                vrat = c1v + c2v + c3v
>
>in parallel...  ctot = c1 + c2 + c3
>                 vrat = ((c1v + c2v + c3v)/3) / 3
>where c1v = c1 voltage rating.
>
>larry.
>
>
>