Re rolled capacitor plate tabs

>Original Poster: dferguso-at-ebmail.gdeb-dot-com 
>all the rolled capacitor plates i have seen have only one tab or connection
>point on one end of each plate.
>it would seem that having both ends of each plate connected to the
>capacitor terminal would reduce the inductance
>of the capacitor ( i.e. the charge could enter and exit from both ends of
>the plate) , and the resistance of the capacitor
>would be lower since there would be twice the amount of tabs or connection
>leads to each plate. is this idea not used because the flux in the plates
>would be travelling in opposite directions or something?     thanks for
>your opinions- doug

Having connection tabs at either end of each plate would indeed reduce
the inductance of the unit.  This is the idea behind extended-foil rolled
caps.  The only downside is that multiple tabs complicates the
construction geometry of the unit.  Usually both terminals come out of
the same end of the container on simple rolled caps.  If you have four
tabs coming out the same end of the roll, it would be difficult to keep
things from shorting, so you'd probably want to have one terminal at each
end of the roll (two tabs from each plate going to the same end).

However, since the recent thread on this List about the significance of
off-axis primary inductance, I'm not so sure that a cap's ESL matters in
Tesla applications.  It is probably true that ESR is the more significant
loss mechanism (behind dielectric losses of course) and that this is
possibly correlated with ESL, but even this may be minor compared with
gap losses.

Gary Lau
Waltham, MA USA