Measurement method
A clear explanation of how I am trying to measure breakdown voltage
can be found on this page,
Measuring the dielectric strength of dielectrics by
Zoltan Losonc. You basically place a 3mm thick piece of dielectric
material between a brass cylinder and a brass disk both of specific
dimensions. Note that the dimensions of my brass disk differ from
that on the above webpage in that its diameter is 77mm instead of
75mm.
The following are pictures of my setup.
The cylinder is on the left, the disk is in
the middle and the dielectric material is on the right. The wire
for the disk is not shown as it is taped to the bottom of the
disk and is going through one of the holes in the green plastic box.
 |
|
The cylinder on the dielectric which in turn
is on the disk (hidden by the dielectric).
 |
|
|
|
View of the entire test setup. At the bottom
of the picture from left to right is my DMM, the control box for the
power supply and an ammeter.
 |
|
The dielectric is the same varnished cotton talked about
here.
Results
The following are the results of a sequence of tests. Note that in all
tests 8 sheets of varnished cotton were stacked for a thickness of 3mm.
| Test No. | Voltage at breakdown | Breakdown voltage |
|---|
| 1 | 28,300V | 9.43MV/m or 240V/mil* |
| 2 | 27,500V | 9.17MV/m or 233V/mil |
* A mil is 1000th of an inch.
I slowly turned up the voltage until it reached just below the breakdown
voltage. I couldn't turn it up any further because the varnished cotton
has a low resistance and is actually conducting a little. In the videos
of the DMM readout as the tests were being done the voltage fluctuates
by about 200 volts just below the breakdown voltage for a while as I turn
the voltage up and then reachs the values recorded in the table above and
drops from there. It droped to around 26,000V and then the DMM started
acting up. The dielectric made a cracking sound and a purple
corona appeared between the bottom of the cylinder and the top of the
dielectric (i.e. it had punctured). There is now a small brown hole in
a part of the dielectric which was under the edge of the cylinder.
A lot of hissing was heard during the tests.
The ammeter connected in series with ground was maxed out throughout.
Note that in correspondance with Zoltan, he suggested I test to see
if the voltage was being drained by the dielectric or by the air.
He suggested using another material for the dielectric and seeing
if something similar happens. I tried a 5mm thick piece of ceramic
tile with about the same face area as the varnished cotton (see picture
below). I was
able to bring the voltage all the way to about 50kV before the ceramic
dielectric broke down. While increasing voltage there was the occasional
clicking sound (not sure from where) but the ammeter showed no
significant current flow. That is, until the breakdown occured.
This means that with the varnished cotton, the voltage was being drained
by the dielectric and not by the air.
Testing breakdown of ceramic tile.
 |
|
