Re: additional transformers [rolled caps]

From: 	Robert W. Stephens[SMTP:rwstephens-at-headwaters-dot-com]
Reply To: 	rwstephens-at-headwaters-dot-com
Sent: 	Thursday, July 31, 1997 4:05 PM
To: 	Tesla List
Subject: 	Re: additional transformers [rolled caps]

> From: 	Peter Electric[SMTP:elekessy-at-macquarie.matra-dot-com.au]
> Reply To: 	elekessy-at-macquarie.matra-dot-com.au
> Sent: 	Thursday, July 31, 1997 6:15 AM
> To: 	Tesla List
> Subject: 	Re: additional transformers [rolled caps]
> Tesla List wrote:
> > 
> > From:   richard hull[SMTP:rhull-at-richmond.infi-dot-net]
> > Sent:   Wednesday, July 30, 1997 1:27 PM
> > To:     Tesla List
> > Subject:        Re: additional transformers
> > 
> > >Just though I'd throw in my two bobs worth after having blown up a few
> > >rolled poly caps!
> > >
> > >All of mine failed at the edges (even being under oil) so I reckon the
> > >secret is to put big overlaps between the plastic and the Al plates i.e.
> > >at least an inch or more. Also here the thick poly sheet is bloody
> > >expensive! so use multiple sheets of 8 mil concreters plastic,
> > >preferably the clear stuff but if all else fails use black plastic with
> > >engine oil! It ran ok on my coil for more than 3 months.
> > >
> > >Cheers,
> > >
> > >Peter E.
> > >
> > >Peter,
> > 
> > I recommended leaving a full 2" edge separation metal to edge of dielectric
> > in all rolled capacitors!
> > 
> > Richard Hull, TCBOR
> Thanks for the info on that, I will go 2" on my future overlaps. It
> seems strange to me that sparks seem to travel much further along the
> surface of an insulator, no matter how clean it seems to be. Maybe this
> is due to microscopic impurities on the surface of the plastic?
> Another little trick that you probably have already figured out is to
> insulate the input connections to the plates by rolling a few layers of
> poly around them.
> Cheers, Peter E.
Peter, Richard, All,

I had the same problem with a rolled poly cap I made once, and I 
*was* using a 2 inch surround from the foil to the edge of the poly.  
Dismantling the cap after the failure showed these burns distributed 
all along the edge of the foil outwards into the poly.

I had a chat with a capacitor engineer at Condesnser Products about 
this phenomenon and was told that this is why HV capacitors are 
generally not successful when one tries to apply more than 8-10 kV 
across a single stage.  He explained that the energy exchange between 
the plate and the local surface of the dielectric becomes too violent 
and just completely overstresses the dielectric causing it to 
break down.  Placing additional thickness of poly does not 
significantly reduce this violent surface area exchange so you will 
still get failures no matter what you do if you try to run a single 
stage cap at 15 or 20 kV.