Liquid-Dielectric Capacitor

Hi all,

After posting about using a liquid-dielectric yesterday, I decided to actually
try the setup I have...but before I list the results I will describe the
actual capacitor.

The vessel/tank that holds the conducting plates and liquid-dielectric is 
actually an old 29gal L/H fish tank in which I used to raise/breed octopus 
(blue-ring and California brown, if anyone cares).  The material the tank is
constructed of is clear cast acrylic.  It has 2 sets of inlet/outlet ports, 
1 set of these ports takes water to the cyro/thermal mechanism, which me and
my girlfriend created, it controls the temperature of the water or liquid-
dielectric.  The other set of ports used to go into a identical tank where
UV sterilization, massive filters and water purification took place, but know
they are just connected to a small adjustable-flow water pump (about 750gal/hr 
max) which is used to circulate the liquid-dielectric.  I have two types
of different conducting plates, aluminum flashing and some copper plates (they
came from old printing plates that a friend of mine cut into 10" x 10" squares
at a thickness of about 0.0625 inches).  I made a frame assemble out of some
spare acrylic pieces and 1/8" rigid plastic tubing.  The tubing was cut into
11" long sections (about 100 of them), them 50 were placed one each side of
the tank by using the acrylic strips (spare ones) and clamping the tubes in
between the strips using nylon screws (or PVC screws not exactly sure what they
are).  This frame is kinda hard to explain, but in essence it sits flush on the
bottom of the tank and has recesses that keep the conducting plates about 0.5"
off the floor of the tank.  The plates are then slid down in between two pieces
of the rigid tubing (for a plate spacing of 0.125" seperate the plates by 1
tube, for spacing of 0.250" seperate the plates by two tubes, and so on), this
allows for fairly easy configuring and adjustment of the capactor, thus it is
capable of quite a range of capacitances.

The two end plates were then connected to the HV posts which I had made in the 
tank, by drilling a hole in the wall on opposite ends of the tank.  Messing 
with some spare PVC, large ceramic inslators, and good sized brass bolts, I made
some pretty good HV output/input posts (although one leaks alittle have to 
re-seal it).  I then installed a small PCV valve on the bottom front of the tank
so I could drain the liquid dielectric with ease.  I installed the aluminum
plates, and added the NH4 (Ammonia) liquid-dielectric (approx. 98.8% purity).
After I double checked all of the plates and connections.  I then used a digital
thermometer to check the temperature of the liquid-dielectric (was about 24C)
and then started the cryo/thermal mechanism and circulation pump all going well
so I decided get the temperature of the liquid-dielectric down to 20 C, that to
worked perfectly.

Time for a test run, I started slow on the power and in increments
(9kV-at-30mA, 12kV-at-30mA, 12kV-at-60mA, 12kV-at-120mA, 15kV-at-60mA, 15kV-at-120mA,
15kV-at-240mA, max.
15kV-at-480mA), the capacitor appeared to handle fine (was getting ripping sparks
from the gap, sounded and looked pretty normal).  The LD NH4 seemed to hold
up pretty well although some severe heating did occur at 15kV-at-60mA and above,
which wasn't really a problem just compensated with the cryo/thermal mechainism.
The only noticable corona losses were at the posts, which was I fixed by
spraying on some telfon coating.  The test setup seemed to check out ok, for a 
prototype, the rating was for 15kV-at-240mA at a capacitance of roughly
0.04uF (just rough calculations):

Dielectric Constant:  22 (at 20C) NH4
Plate Dimensions   :  10" x 10" (Al)
Plate Spacing      :  0.250" (2 tubes)
# of Plates        :  20

Didn't measure the capacitance yet, I wanted to try using the copper plates.
At higher power inputs (1800W and above) some bubble formed around the plates
especially the end plates (most likely nitrogen or hydrogen), have to test the
gas to find out exactly what it is.  I have decided to make a hood/cap to put 
over the tank to collect the gas for testing (see if it in toxic or dangerous
levels) and also to determine how much and pressure it is produced at.  After
I ran the initial tests I stopped using NH4 for the LD and switched to 
distilled H2O (K=79 approx. at 20C), but after I drained the tank I inspected
the aluminum plates and there was a little corrosion (discoloration) on them.
I them removed all the plates and inserted 12 new plates at a spacing of 0.5":

Dielectric Constant:  79 (at 20C) H2O (distilled)
Plate Dimensions   :  10" x 10" (Al)
Plate Spacing      :  0.50" (4 tubes)
# of Plates        :  12

I then fired this setup the same as the NH4 setup and there were some noticable
differences: (1) During all tests gas production at the plates occured, 
                 increasing in effect as power increased (as expected)
                 (large amounts of hydrogen and oxygen were being produced,
                  quantities you really would not want near HV arcs (SG) )
             (2) Coronal effects were alittle more pronounced in this setup,
                 although at tolerable levels (teflon spray helped out)
             (3) H2O LD heated up much faster than the NH4, but also was easily
             (4) Seemed (probably just me) to make louder arcing at the spark

I concluded my experiments here for the night.  After I do some calculations
and further research on this (relearn my forgotten chemistry), I am going to
run some more tests, this time taking actual measurements on the capacitor.
Then if I do not find any faults or problems I will try it out on the 6" coil.

I typed this in a hurry so I probably left some stuff out, any comments or 
questions anyone...

While this is not exactly as easy to move around as other capacitors it is not
impossible (if you drain it first :).  Configurability of the capacitor is 
rather nice for testing different setups.

If I am overlooking something with this capacitor please let me know...


| Timothy A. Chandler                ||   M.S.Physics/B.S.Chemistry     |
| NASA-Langley Research Center       ||   George Mason University       |
| Department of Energy               ||   Department of Physics         |
| FRT/Alpha - NASALaRC/DOE JRD/OPM   ||   Department of Chemistry       |
| CHOCT FR Designation #82749156/MG09||   OPC-EFC                       |
|                 Private Email Address:  tchand-at-slip-dot-net               |