Thermal stir:
The warmer base provides a hot spot to induce a thermal updraft
in the center bottom of the container.
Warm water rises, reached the top, spreads out and contacts the
cooler sides of the container and sinks back to the bottom.
Ideally, the amount of heat the water absorbs will equal the amount
of heat shed as the water contacts the container sides and it
won't get warmer overall.
The result is a doughnut shaped stir pattern that works quite
well and isn't so vigorous as to make silver particles impact
the electrodes.
Too fast a stir velocity makes silver particles impact hydrogen
bubbles on an electrode and they get trapped in the surface tension
of the bubble making a semi solid / semi conductive coating [
A "beard"] that grows and grows into the direction of
the water current.
The downside to that system is that the water can cool, spread
and sink before it gets to the top if the container is tall.
Placing a chimney over the warm spot increases the velocity of
the updraft and keeps it centered and channeled so taller containers
can be used and still achieve the desired stir pattern.
A possible downside to that is thermal buildup in the chimney
base.
Temperatures over around 110 deg F makes particles collide with
enough energy to merge.
Easy Solution: Use a cooler bulb, 4 watt rather than 7 watt.
Upside..no moving parts to ever break or wear out. [And it
looks pretty glowing in your room]
Accidental upside: The CS almost completely stabilizes as it's
being made. [Conductivity drop the day after is reduced by about
80%]
More recently discovered nicety: Using that thermal chimney setup
with a slow cycle low voltage AC generator [S.W.A.P.] reduces
sparklies and suspended electrode crud chunks better than any
other system tried. [Not quite eliminated, but dramatically reduced]
Magnetic:
The more recent stir system is like a lab stirrer where a magnet
spins under the container and couples with a magnet inside the
container.
In order to keep the stir velocity down and prevent the bearding
effect, a low RPM motor is used. No commercially available lab
stirrer will spin that slow.
The stir pattern is a cyclonic one...a slow vortex that stirs
from the bottom [where the electrode aren't] and up.
After years of experimentation on LVDC systems, I've found
stir speed the
be very important.
About 20 to 30 RPM seems ideal and is enough even for gallon
sized batches.
If it's not a gear motor, going that slow can be tough. Motors
don't like
to run slow.
Using a pulse width motor controller helps.
Stirring from the top down, you need a lot more RPM to reach
the bottom,
but that's too fast for the top.
Stirring from the bottom up is much better.
Magnetically coupled lab stirrers are great, but none are made
that go
slow enough. [Except the ones I make and I make them because I
can't buy
one anywhere]
So, why isn't fast good?
What happens is this: [Grey Fuzzies]
Hydrogen bubbles compress on the side of the electrodes that face
the
current and get really sticky.
Silver particles in that water current impact the hydrogen bubbles
and
get caught in the surface tension of those bubbles and form a
semi solid
skin that is semi conductive.
The bubble stabilizes on the electrode and won't bubble off,
the skin
[being a semiconductor] grows another bubble and the process continues
till
you have a big grey beard of crud growing into the direction of
the water
on round electrodes or just behind the edge in an eddie current
on flat
electrodes that drops off into the water when trying to remove
the electrode.
The bubble structures that don't fall off, pop when meeting
the surface
tension of the water and all that silver transfers to that spot
making a
big silver slick.
Slowing the water down prevents all that.
Both systems work well and one hasn't superceded the other,
but the magnetically coupled system will stir more water at the
proper velocity and that velocity can be easily adjusted with
spinner size.
Downside: It does have moving parts that could eventually wear
out. [and it's more expensive]
Stirring from the top down requires a high water velocity to
make the vortex reach the bottom in tall containers.
Stir too fast where the electrodes are and you get a lot of bearding.[The
dreaded grey fuzzies]
It also exposes the stir motor to water vapor that tends to corrode
the motor bearings after a year or so. [As I found out the hard
way]
With the magnetic system, that's not a problem at all.
Some people use bubblers and they work OK. Bubbles go from
bottom to top quite nicely.
Downside: Any air borne contaminants dissolve into the water quite
well.
High levels of sulphur dioxide air pollution can be a big problem
with silver. [tarnish]
A small problem is carbon dioxide acidifying the water a little
and possible light sensitive silver carbonate formation.