Voltage Multiplier(4x) problems

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

Original poster: "Hydrogen18" <hydrogen18-at-hydrogen18-dot-com> 

I just completed a 4 stage voltage multiplier made with 12 kilavolt 350 ma 
diodes and 10 kilavolt 500 pf capacitors(dielectric). It works fine with a 
2500 VAC input from a mot other than a few problems. I get crackling corona 
everywhere and ionic force between clipleads but I was aiming for that.

I will occasionally get arcs from the core to the primary. I'm guessing the 
10 kilavolts somehow makes it way backwards and ends up somewhere on the 
core then arcs to the primary because 120 volts looks alot like 0 when you 
have a 10 kilavolt supply. It doesnt seem to effect anything negatively, 
but I am concerned that the insulation may finally fail and the 120volts 
will connect to the core. Am I right or wrong about why the core arcs to 
the primary?

I connected one of the 10 kilavolt capacitors(bought 50, used 8 for the 
multiplier section) across the output of the multiplier to give me more 
burst amperage(doubled the spark length). For some reason the leads will 
arc over on it if I do not discharge it often enough. I'm guessing the 
voltage builds up enough then finally arcs. What I dont understand is why 
the heck the ones in the multiplier circuit dont also arc, it would seem 
they should since they are exposed to 10 kilavolt stresses as well. Any 
ideas? I am now using 20 kilavolt 500pf doorknobs without any problem.

Another problem is trying to figure out whether the capacitors are failing 
or not. I thought I heard one in the multiplier hissing, but it was near 
where I had a clip lead clipped on, so I'm guessing it was corona "hiss". 
What does a failing capacitor sound like and what are the visual indictations?

Thanks for any help you can provide.