Discussion in 'Electronic Basics' started by aman, Apr 23, 2005.

1. ### amanGuest

When I think of capacitor I can see there are two conducting parallel
plates seperated by an insulator dielectric. So as there are opposite
charges on the inside of the plates they attract each other thus
containing the charge.

Here is my question. If there is a perfect insulator used as
dielectric, does it mean that the charge is fully contained and if the
dielectric material gets a little conducting the charge on the plates
gets reduced(cannot be retained fully) and some charge flows in the
dielectric. Am I correct to some extent ?

I am asking this because I am constructing a kind of capacitor detector
which is seperated by dielectric which is water(different kind of
samples with different conductivity and am trying to nuetralise the
harmful free ions in water).

2. ### Larry BrasfieldGuest

Two plates is the minimum, often exceeded.
Yes, although what would be called the charge is fully
contained whether the dielectric conducts or not.
If you know your chemistry, you know that water
becomes something like an insulator when you
get the free ion concentration low enough. Are
you using leakage to guage that concentration?
The capacitance will stay just about the same
unless you have very high impurity levels.

3. ### amanGuest

So doesnt capacitance depend on conductivity of the dielectric ? I mean
wont a more conductive sample of water give a different capacitance
than less conductive water ?

4. ### Larry BrasfieldGuest

No. The effects act in parallel. When you apply an
electric field to a slightly conductive dielectric, some
bound charges move a limited distance then stop,
and some unbound charges drift continously. We
call the first displacement (or capacitive) current.
We call the second leakage. The movements of
those different charges are largely independent.
I don't think you will be able to measure a capacitance
change as you alter the ion concentration until you have
such a large concentration that the dielectric property
of the solute itself is a factor. By that time, measuring
the capacitance would be quite a trick.

5. ### amanGuest

In C= KA/d if K does not depend on conductivity, what makes a
dielectric have different dielectric constant K. What are the
parameters on which K depends on?

6. ### Larry BrasfieldGuest

Usually, 'K' refers to the relative dielectric "constant" and
would be multiplied by the permittivity of free space in
the above formula.
It is affected by how easily bound charges can be
displaced and the density of those charges. For
example, in water, the bound charges are at
sort of opposite ends of the H2O molecule and
displacement occurs as they become polarized
in an E field.

Your question covers a lot of territory.

7. ### amanGuest

So doesnt this imply that if there are free ions in water there will
be a different K from a nuetral sample of water with no free ions ? I
am actually adding chemicals(ionic) to water which nuetralises harmful
ions in water to form a nuetral particle floc. So I need to detect zero
crossing. As i am adding ions into water i need to detect a zero
crossing in going from postive charge to negetive.

8. ### Larry BrasfieldGuest

I don't think so. The free ions will drift under the
influence of an E field, contributing in-phase current,
not lagged current such as displacement produces.
The only way free ions could change the K would be
if there were enough of them to act as a dielectric
and they had a K different than the H2O they displace.
Sorry, but I cannot make sense of that. Maybe it
would help to show your circuit or block diagram.

9. ### Larry BrasfieldGuest

Woops, that should be:
not leading current such as displacement produces.

10. ### amanGuest

I am adding salt to water. I read something in one of the forums below
which confuses me. It also says "adding virtually any contaminant to
water will alter its conductivity. That also ruins the diaelectric
constant . Only pure water will have a measurable diaelectric
constant."

It says that pure water is insulator and adding salt changes
conductivity. I am adding salts to contaminated water to which makes it
an insulator. But if more salts are added water crosses the point at
which it is an insulator(pure) and again starts conducting. So I need
to detect that zero crossing.

But according to the forum in the link above, conductors cant be
dielectric. But what if I use conducting water as dielectric. What
happens. I know dielectric is supposed to support electric field but
oppose current. But if there is leakage current the capacitor can't
hold charge or what ?

Also what is meant by "Only pure water will have a measurable
diaelectric constant."

11. ### Larry BrasfieldGuest

I don't know what a ruined dielectric constant is.
That is gibberish as far as I am concerned.
I don't understand the zero crossing. Pure water
conducts. There are always H+ and OH- ions
present, and they can carry current. What you
will see, if you can reduce free ions by adding
salts to react with already present ions so as to
produce a non-ionic product (a precipitate, I
presume), is a decrease in conductivity which
will approach the conductivity of pure water,
possibly followed by an increase if you add too
much. There will be no zero crossing.

To me, it looks like you should almost forget
If you are using AC for that measurement, it
may be a good idea to measure only in-phase
current so as to ignore the capacitance.
For good conductors, the dielectric properties do
not matter and are difficult or impossible to measure.
A counter-example to that statement can be found
in any electrolytic capacitor. They all leak current,
so, because the material between their plates is a
conductor (albeit a poor one), and taking the above
as true, they cannot also be capacitors. Countless
devices say otherwise.
The question should not be "Can it hold charge?" but
"How long can it hold charge?" Leakage limits the
time that a capacitor can hold charge.
That overstates the case. If we are to believe that,
then adding one ion molecule to a vat of pure water
will turn it from being a dielectric to being a poor
conductor.

12. ### John PopelishGuest

A more accurate statement might have been that adding things to water
that increase its conductivity interferes with many dielectric
measurement methods.

Imagine measuring the value of a capacitor with a capacitance bridge.
The null is nice and sharp and determining the value with high
precision is easy. Now, connect various resistances in parallel with
that capacitor. The bridge null gets shallower and wider as more and
more of the instrument's bridge current detours through the resistance
and less and less of it passes through the capacitance. But
paralleling those resistors did not alter the dielectric constant of
the insulating material in the capacitor.
The key word, here, is "measurable". Using a higher AC excitation
frequency lowers the capacitive impedance while the Resistive
component is essentially unchanged. So measuring conductive samples
is more practical with a higher excitation frequency. The caution is
that the actual value of high dielectric materials, especially liquids
can vary dramatically as the excitation frequency passes through
motional resonances of the polar molecules. So you will need
calibration standards that you can use at the frequency you choose.

13. ### GarethGuest

There will be some ions, but not many. Only about 1 in 500 million
water molecules in pure water break up into H+ and OH- ions.

Pure water is actually a good insulator (conductivity < 0.06 uS/cm)

--

14. ### Rich GriseGuest

What you're actually looking for is a pH meter.

This isn't really related to dielectric constant at all, unless, as
Larry said, the solute has a dramatically different K than water, and
if it were in enough concentration to affect a capacitive readout,
you wouldn't have a capacitor but an electroplating tank.

Good Luck!
Rich