Topband: Homebrew capacitors

[Tom W8JI]

This post is worth reading and saving for homebrewers. Not nearly enough of 
us understand capacitors this well.


> KK1L noted that there must be either an overlap or a gap in the sleeve
> that wraps the inner tube. A gap would create a "hot spot" where the
> breakdown voltage would be reduced, while an overlap could cause the
> inner tube to bind. This could be avoided if one could obtain
> dielectric tubing of the required dimensions or if one could "weld"
> the dielectric to seal the gap. I don't anticipate having a problem
> with the possibility of a small gap in my application, but two
> possible solutions to this condition come to mind.
>
> First, one could wrap the dielectric material twice around the inner
> conducting tube. While there would still be a gap, it would not be
> entirely an air gap, which would increase the breakdown voltage. Of
> course, one would also get 1/2 the capacitance per unit length, since
> the dielectric would be twice as thick - unless material half the
> original thickness could be obtained.
>
> Second, one could cut a slot down the length of the outer tube and
> center the gap in the dielectric material in the slot. In this way,
> the potential at the gap would be reduced according to the width of
> the slot. Here, again, the capacitance per unit length would be
> reduced, but less so than in the first solution.

Other issues are any sharp points or edges in conductors will greatly reduce 
breakdown voltage. This is often an operating time issue, because years of 
cornoa will slowly eat away the dielectric. This is why some "capacitor" 
stubs, even though apparently very conservative, will fail over time. I've 
seen 50kV insulation fail at 5 kV just from having cornoa from a sharp edge.