[Prev][Next][Index][Thread]

Re: Large Transformer Question (fwd)





---------- Forwarded message ----------
Date: Sat, 14 Feb 1998 09:55:56 -0500
From: Thomas McGahee <tom_mcgahee-at-sigmais-dot-com>
To: Tesla List <tesla-at-pupman-dot-com>
Cc: seraphim-at-WPI.EDU
Subject: Re: Large Transformer Question (fwd)



----------
> From: Tesla List <tesla-at-pupman-dot-com>
> To: tesla-at-pupman-dot-com
> Subject: Re: Large Transformer Question (fwd)
> Date: Friday, February 13, 1998 11:14 PM
> 
> 
> 
> ---------- Forwarded message ----------
> Date: Fri, 13 Feb 1998 08:09:37 -0800
> From: Joshua Resnick <seraphim-at-WPI.EDU>
> To: Tesla List <tesla-at-pupman-dot-com>
> Subject: Re: Large Transformer Question (fwd)
> 
> >
> 
> Fr. Tom,
> 
> >
> >
> > I am confused. How can 'only one layer of foil' have 'two continuous
> > strips interleaved with polyethelene.
> >
> > I am wondering if maybe your cap isn't really a cap after all.
> > Do an ohmmeter check and see if you have continuity. If you do,
> > then you don't have a cap, but a dead short with some poly for
> > good looks :(
> >
> > What thickness of poly are you using?  It is also possible that you
> > had
> > a poly breakdown and therefore a foil short.
> >
> > NEVER NEVER NEVER fire your poly caps "dry". Surface conduction
> > can cause a 15KV spark to snake along several inches of surface
> > even though it might only be capable of 'jumping' a fraction of
> > an inch. You NEED the oil!!!
> >
> > Hope this helps.
> > Fr. Tom McGahee
> 
> Believe me it is a capacitor, a very big capacitor. I know that I need
> the oil but I figured that I should at least test the system briefly to
> make sure I wasn't wasting my time and money potting the cap.  My
> makeshift gap for the cap test was only 3 mm larger than the dielectric
> thickness so I was sure that no severe breakdown would occur. I also did
> not feel that powering up slowly with a variac would cause any sudden ,
> violent startup conditions. I learned alot from this test including the
> method in which I should bus the cap. in order to distribute current
> more evenly.
> 
> Thank you for clarifying the grounding situation with the xfmrs, that
> cleared alot up.
> Sincerely,
> Joshua Resnick
>              ____
>       /\/\ ((____)) /\/\/\/\                Joshua Resnick
>      /       |  |    \       \       http://www.glubco-dot-com/weaponry
>    /         |  |      \/\/\   \ __  http://ece.wpi.edu/~seraphim
>  /       ____|  |____        \         -  One Room - One Dream -
>         ||||||||||||||         \       -   One Million Volts   -
> 
Joshua,
OK, I have a clearer picture now of what happened. Let me explain:
You built a humongous LOOOONG capacitor with the connections either
at the same or opposite ends. When discharging such a capacitor the
discharge current is greatest at the end(s) [It sums up as it goes]
In a really large value cap this can cause such excessive currents
at the terminal ends that the thin aluminum foil will overheat. 
In the worst case it could actually evaporate! 

This is one of the major drawbacks of long capacitors. This effect
limits the maximum pulse currents that can be delivered. When such
a cap is rolled, there are the additional inductive effects that
come into play, further limiting peak deliverable current.

The flat cap eliminates most of the inductance. Also, the distance
the current has to move before reaching a collection point (buss)
is MUCH smaller. The plates are in parallel, so the current
carrying capacity is GREATLY enhanced. 

Now let's investigate your 'cure'.
You can run the foil out the LONG sides of a long capacitor, and
that helps things a lot, but if you examine the situation closely
you will see that with this arrangement bubble removal is more
difficult. (The bubbles can't go through the foil, now can they?
You could perforate the foil at intervals to help a bit.

One more 'gotcha' that you should be aware of: The amount of HV
along the SURFACE of the poly is greatest at the foil collection
points. A little thought will show you that the foil to poly
tracking distance along the poly surface with a constant 1" foil
to edge of poly spacing is doubled to a 2" tracking length along
the 'free' side, since any arc would have to go out one inch, go
around the poly and then back in another inch.  But at the terminated
ends that same 1" separation gives you ONLY 1" of surface
tracking distance. That is why the termianted ends MUST have at least
double the distance from the foil edge to the poly edge.

That means that if you decide to run the foil out along the LONG
edge of the poly, then you had better reduce the width of the
foil or you will have premature breakdown.

So many things to always keep in mind!

Hope this helps.
Fr. Tom McGahee