Michael,

When I first got thinking about EV's I spent a lot of time designing a
charging system and thinking about the best ways to keep AGM batteries
balanced.  The plan was a bunch of small chargers for each battery that
operated from a common supply and a bulk charger on the whole string.  The
small ones would make up for the differences in batteries even to the point
where batteries could be changed mid life of the complete set.
Unfortunately, it got big and complicated.  It however lead me to believe
the balancing circuit only needs to correct for the differences in charging
and discharging the batteries.

I have read about voltage clamps as you describe with your zener diodes.
Basically the zener turns on at a fixed voltage, which would be the float
voltage for the battery.  Once the float voltage is reached, the extra
charge current would flow through the zener and bypass that battery.  

The first problem is the limited zener voltages available and the 5%
tolerance on that voltage.  It would be simple and cheap to make some
adjustable shunt regulators that function like a zener with a chip and NPN
power transistor(s).  With this approach, the voltage could be set exactly
for the batteries requirements and all batteries need not be the same.

The problem with zener shunt regulation is it wastes a lot of power.  If we
are charging a 6 volt battery at 10 amps (to keep the numbers simple), 60
watts is going into the battery while it is bulk charging and about 90% of
that energy is stored.  When the zener shunt regulator has fully bypassed
the battery the 60 watts is wasted in heat in the zener and no energy is
added to the battery.  If there are 24 batteries in the vehicle, 60W X 24 =
1,440 watts of heat as wasted power.  That is a lot of power, heat, even if
it is just for a few minutes at the end of a charging cycle.

What about the mirror image of this approach and charge the string until any
one battery reaches full charge voltage?  This will leave the rest at a
slight under charge.  If we use a low current shunt regulator continually on
the batteries, it will apply some extra load to the fully charged battery(s)
and none to the undercharged ones.  This circuit would be operating 24 hours
so a small current like 50 or 100ma would accumulate to amp-hours in a day.
Next charge cycle this battery need a little longer to charge because of
this extra discharging, that allows others to charge more as well.  As the
days go along more batteries get to full charge and more approach full
charge.  In a week or so, all should be in balance.  As the set becomes in
balance less correction is needed and efficiency goes up.  If we implemented
this at 100ma and 6 volt batteries, the power dissipation would be 0.6 watts
when active.  Over 24 hours that would be 14.4 watt-hours.  If we scale it
up to 24 batteries, it would waste 345 watt-hours maximum in 24 hours if
every battery was overcharged.  I would think it reasonable not every
battery would be over charged and with driving a lot less than 24 hours of
overcharge time would be seen.  Best part is it is simple and cheap.

David Kerzel
Pompano Beach


_______________________________________________
Florida EAA mailing list
listserv@floridaeaa.org
http://www.floridaeaa.org

Reply via email to