Why PWM?

Page 3
8 Pheasant Run | Newtown, PA 18940 USA | www.morningstarcorp.com | info@morningstarcorp.com
Ability to Recover Lost Battery Capacity:
According to the Battery Council International, 84% of all lead acid-battery failures are due to sulfation.
Sulfation is even more of a problem in solar systems, since “opportunity charging” differs significantly
from traditional battery charging. The extended periods of undercharging, common to solar systems,
cause grid corrosion, and the batterys positive plates become coated with sulfate crystals.
Morningstar’s PWM pulse charging can deter the formation of sulfate deposits, help overcome the
resistive barrier on the surface of the grids, and punch through the corrosion at the interface. In
addition to improving charge acceptance and efficiency, there is strong evidence that this particular
charging can recover capacity that has been “lost” in a solar battery over time. Some research results
are summarized here.
A 1994 paper by CSIRO, a leading battery research group in Australia (reference 1), notes that pulsed-
current charging (similar to Morningstar controllers) “has the ability to recover the capacity of cycled
cells.” The sulfate crystallization process is slowed, and the inner corrosion layer becomes thinner and
is divided into islands. The electrical resistance is reduced and capacity is improved. The paper’s
conclusion is that pulse charging a cycled battery “can evoke a recovery in battery capacity.”
Another paper, a Sandia National Labs study in 1996 (reference 2, attached), summarizes testing of a
VRLA battery that had “permanently” lost over 20% of its capacity. Conventional constant voltage
charging could not recover the lost capacity. Then the battery was charged with a Morningstar
SunSavercontroller, and “much of the battery capacity has been recovered.”
Finally, Morningstar has been testing for capacity recovery. An attached graph (reference 3, attached)
shows how a battery that was “dead” recovered much of its lost capacity after extended charging with
a SunLightcontroller.
After the test was set-up, for 30 days the solar lighting system produced virtually no lighting since the
system went directly into LVD each night. The battery was very old and about to be recycled. Then, the
load began to turn on longer each night as shown on the graph. For the next 3 months the battery
capacity steadily increased.
This test and other capacity recovery tests are ongoing at Morningstar.