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

Re: bottle capacitors




From: 	ghunter-at-mail.enterprise-dot-net[SMTP:ghunter-at-mail.enterprise-dot-net]
Sent: 	Monday, November 10, 1997 3:36 PM
To: 	Tesla List
Subject: 	bottle capacitors

I've been experimenting with capacitors made from 750ml wine
bottles.  I found a case of empty bottles at a "boot" sale in East
Anglia for the princely sum of two pounds (that's roughly $3.20). 
They are clear, colorless glass, with straight, parallel sides ending
in abrupt, round shoulders.  I covered them with foil up to the
junction of the neck & shoulder.  I filled them with very strong
salt water topped off with USP mineral oil.  I used 12" x 1/4"
threaded brass rods through plastic champagne stoppers for top
terminals.

I measured the output of my Transco neon transformer at 15,200v and
32ma with a Fluke multimeter.  This is at 50Hz, as I'm using UK
mains power and a stepdown autotransformer.

I found that individual  bottles could not withstand 15kv.  Two of four 
bottle caps I tested failed instantly with a loud "snap" as the 
voltage punched right through the glass.  I have plenty of bottles, 
so I continued testing.

I found that two bottles in parallel drew 19ma of current from the
neon.  Three caps pulled 25ma.  Four caps pulled the max rating of
32ma.  Apparently, two or more parallel bottles pull enough current
to drop the voltage to a level that can't puncture the glass.  I put 
two bottles in series (for about 950pf) and ran them for several 
minutes with no failures.  I switched the wall current on and off 
several times to create voltage spikes, but they didn't fail.

The bottle caps have much higher capacitance than I expected.  All
of the www Tesla design software I have tried predicted much
lower capacitance for my bottle caps than I actually measured.  The
real-world value seems to be around 1900pf per bottle, with a spread
of 1770pf to 2015pf.  The puncture voltage is also much lower than I
expected.  I wouldn't have believed that 15kv could break them down
if I hadn't seen it myself.  On a broken bottle I measured the
sidewall glass thickness to be from 1/16th to 3/32nd of an inch (the
glass is a bit wavy).

I'm going to employ six or eight of these on my 4 1/2" dia. Tesla 
coil (3 in series with 3, or 4 in series with 4).  It will be 
interesting to see if they can hold up to the harsh, flywheeling 
voltage swings of a primary tank circuit.

Greg