This is only a preview of the January 1996 issue of Silicon Chip. You can view 22 of the 96 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Articles in this series:
Articles in this series:
|
New from Smart Fastchargers, this nicad and NiMH charger
caters for a wide range of battery voltages and capacities and
uses the patented Reflex charging method. It has eight buttons
to set the rate of charge, a rotary switch to select the battery
voltage and a LED bargraph to indicate the cell voltage. An
audible beep, at one second intervals, gives an indication that
the main charge is still in progress.
Recharging nicad
batteries for long life
Nickel cadmium and nickel metal hydride
batteries are widely used in all sorts of portable
equipment but they often don’t last long before
they must be replaced. One solution is to use
“burp charging” which is claimed to provide
many thousands of charge/discharge cycles.
By HORST REUTER*
Battery powered equipment is undeniably practical – lightweight, portable and small, with no cables to drag
around. But there is a price to pay for
that convenience. Rechargeable bat
teries are costly to buy and often don’t
last long. The problem can usually be
traced back to the type of charger used.
Unfortunately, most of the nicad
chargers supplied with appliances at
10 Silicon Chip
present require manual termination;
ie, the user has to switch off the charger
and disconnect the battery. This makes
it practically impossible to avoid
overcharging the batteries, thereby
reducing their life expectancy. The key
to long life lies in the charging method.
In this article, a battery is defined as
consisting of one cell or several cells
connected in series. Internal cell im
pedance is defined as the sum of the
resistance of the internal connections
and plates (both constant) and the
degree of difficulty the ions encounter
passing through the separators and
electrolyte (variable).
Nicad or NiMH?
From an environmental point of
view it would be an advan
tage to
change to NiMH batteries. Nickel Metal Hydride (NiMH) batteries are made
without cadmium and are therefore
less damaging to biological systems. At
present, they have about 20% higher
energy density than nicad cells (AA)
and produce no memory effect (more
about memory effect later).
However, typical NiMH cells have
a higher internal im
pedance than
nicad cells; 50mΩ instead of 10mΩ
for 1200mAh cells. As a consequence,
NiMH batteries have lower maximum
discharge currents. This means they
are only suitable for low current ap
pliances like handheld radios. The
maximum discharge current is 3C for
NiMH AA cells and 2C for NiMH button cells, whereupon the cell voltage
drops to approximately 1.1V. “C” is
defined as the current that equals the
rated battery capacity. For example,
charging a 1.2Ah battery with a 4.8A
current is a 4C charge. The same 4.8A
current applied to a 4.8Ah battery is
a 1C charge.
In practice, the useful discharge currents for NiMH batteries are limited to
less than 1C (the cell voltage remains
above 1.2V). For currents above 1C,
nicad batteries are superior. Fig.1 is a
comparison of the load characteristics
of one 1200mAh AA size NiMH cell,
one 600mAh AA size nicad and one
1200mAh sub-C size nicad cell. The
load was only applied for 5 milliseconds.
The tests showed that a fully charg
ed 12V 1200mAh nicad battery as used
in power drills delivers a maximum
of 11.9V with a 10A load. Even a
12V 600mAh nicad battery delivers a
maximum of 10.9V with a 10A load.
However, a 1200mAh NiMH battery
with the same load delivers only 7.65V.
NiMH batteries also differ from
nicad cells in that the chemical reaction during charge is exothermic; ie,
the charging process produces heat.
The chemical reaction in nicad cells
is endothermic; the reaction absorbs
heat. However both battery types
produce some heat during the main
charge cycle because of internal impedance and both produce heat when
overcharged. Over
charging creates
heat and gas but does not produce
any further energy storage in the cells.
The heat produced during the
main charge in nicad cells due to cell
impedance is absorbed in the endo
thermic reaction. In NiMH cells, the
cell heating due to internal impedance
is added to the heat of the exothermic
reaction. When NiMH batteries reach
the overcharge region, they are therefore hotter than nicads.
All available NiMH cells I have
tested vented at around 43-45°C case
temperature – much lower than for
nicads. This means that charge termination at high charge rates is critical
and cannot safely be achieved with
delta V termination chargers. The case
temperature should not exceed 40°C
Fig.1: a comparison of the load characteristics of a 1200mAh AA
size NiMH cell, a 600mAh AA size nicad and a 1200mAh sub-C size
nicad cell. The load was only applied for 5 milliseconds. This clearly
demonstrates that the higher internal impedance of NiMH batteries
limits their usefulness in delivering high currents.
for NiMH cells and 45°C for nicads.
Delta V termination utilises the
voltage drop at the beginning of the
overcharge region of the cell voltage
curve (see Fig.2). The magnitude of
this voltage drop is generally not as
well defined in NiMH cells as it is in
nicad cells. It depends on factors like
charge current, ambient temperature,
cell impedance, cell capacity, etc.
The situation can be worse in battery packs. Several unmatched cells
may cause the battery voltage to reach
only a very shallow peak if some cells
reach their individual peaks while
others are still charging. Even if the
Fig.2: delta V termination utilises the voltage drop at the beginning of the
overcharge region of the cell voltage curve. The magnitude of this voltage drop
is generally not as well defined in NiMH cells as it is in nicad cells. It depends
on factors like charge current, ambient temperature, cell impedance, cell
capacity, and so on.
January 1996 11
the nickel hydroxide to
nickel oxyhydroxide.
This process is reversed
during discharge.
If each cell in the battery pack is discharged
completely and then
charged, the individual
crystal sizes on the cell
plates remain unaltered.
However, if the cadmium is not completely
converted back into cadmium hydroxide during
partial discharge, on
the following charge
the cadmium hydroxide crystals will clump
together, forming larger
crystal structures.
Although it is not yet
fully understood how
Fig.3: the patented Reflex or “burp” charge
this happens, scanning
method consists of a positive charge pulse
followed by a high current, short duration
electron micrographs
discharge pulse. This is quite different from
of batteries with and
other chargers which have an essentially
without memory effect
pulsed output but no discharge pulses.
clearly show the difference in crystal sizes.
The net result is that we
charger circuit is able to detect a very
are left with a smaller, less reactive
small voltage drop (<10mV) at the very surface area and therefore reduced
start of the overcharge region, the fact capacity.
remains that we are already operating
The clumping of crystals is mostly
in the overcharge region.
a slow process but is cumulative.
Overcharging is not acceptable if we However, as we will see later, it is
want to achieve maximum battery life reversible.
for nicad cells. For NiMH cells it can
End point voltage
be dangerous if used in combination
with high charge currents. It can lead
The usual strategy to prevent memto venting and consequent loss of ory effect is to discharge each cell to
capacity and in extreme cases to cell
1.1V or 1.0V, a level where very little
explosion, due to a build up of gas useful energy is left. This is only partly
pressure.
effective with single cells and with
If NiMH cells are charged with new and well matched cells in battery
delta V termination char
gers, then
this has to be done at the rate the
manufacturer recommends, typically
C/10 (120mA) for 1200mAh cells. At
this rate, any heat produced during
charging and overcharging will be
safely dissipated.
packs. Not all cells in a battery pack
will age equally or charge and discharge equally at different operating
temperatures. In the end, some cells
will only be partly discharged when
others are deep discharged.
At the final stage of the battery discharge, a sudden substantial voltage
drop occurs. This can lead to reverse
charging of the weakest cell in a battery pack of more than 12 cells and
will still cause a clumping of crystals
in all cells (except the weakest cell)
during the next charge. The magnitude
of memory effect in each cell depends
on the depth of discharge.
Unlike some other types of cells,
nicads can be totally discharged and
then even shorted to avoid the memory
effect but not without reducing life
expectancy. The life expectancy of all
types of batteries, including nicads,
is partly dependent on the depth of
discharge.
Hence, a total discharge will reduce
life expectancy (up to a factor of 10
in cases of frequent total discharge).
Totally discharging a battery to 0V –
unlike discharging a single cell – is a
sure recipe for extremely short battery
life due to cell voltage reversal.
Shallow discharge, less than 25% of
total capacity, makes for long battery
life but creates the conditions for memory effect. A 1.1V or 1.0V discharge
voltage is only a compromise, not a
magic value.
Freezing cells
Another strategy to combat memory
effect, the practice of freezing batteries to break up the clumping of the
crystals, creates mechanical stresses
in the cells. This can also lead to
Memory effect
Let’s look at the major problem
of nicad cells: memory effect. This
is caused by charging a partially
discharged battery and enhanced by
slow charging and high operating
temperatures. During charging, the
negative plate loses oxygen and converts cadmium hydroxide to metallic
cadmium, while the positive plate goes
to a higher state of oxidation, changing
12 Silicon Chip
Fig.4: the essential characteristic of the Reflex charging method is a
high current charge pulse, followed by a short rest period and then
an even higher discharge pulse for 5ms. The battery voltage is then
measured before the next charge pulse.
This is the view inside
the prototype from
Smart Fastchargers. It
uses a total of three PC
boards and can charge
batteries at a rate of up
to 9A.
reduced life expectancy since a high
degree of mechanical precision goes
into the production of today’s high
capacity cells. It is also a time consuming method, since all cells in the
battery have to be slowly warmed to
above 10°C after freezing for efficient
fast charging.
All these are makeshift solutions.
The problem should be tackled at the
roots, by using a charge method that
will reduce crystal size in batteries
where crystal clumping has occurred
and avoids crystal clumping during
the charging of partially discharged
batteries.
Another area that needs improvement is the small number of recharge
cycles suggested for most nicad batteries. In the case of some hand-held
radios, the batteries are supposed to
have only 300 recharge cycles. Batteries for other appliances are rated for
500 and 1000 cycles.
Theoretically, 5000 charge/discharge cycles are possible over a
minimum life span of 10 years. One
power hand tool manu
facturer advertises 3000 cycles and 10 minutes
charging time. This is achieved by
using advanced charger technology
and fast charge batteries. 3000 cycles
represent approximately 6.5 cents per
cycle as compared to 55 cents per cycle
for the hand-held radio batteries (at
presently quoted prices).
Another problem is the excessive
time required to charge nicad and
NiMH batteries with delta V termination chargers: generally between one
hour for fast charge nicad batteries and
15 hours for standard nicad batteries
and NiMH batteries. Only in exceptional cases, as with some chargers for
battery powered tools, is it possible to
achieve charge rates of less than one
hour for nicad batteries.
Burp charging
One overseas company has designed
a fast charger that achieves an amount
of recharge cycles close to the theoretical limit. This patented charger,
well proven in industrial and military
applications, is used to charge aircraft
batteries, emergency standby batteries
for hospitals, etc and operates fully
automatically. It automatically detects
the type of battery (nicad, NiMH,
lead-acid, etc), battery capacity and
voltage and adjusts itself accordingly.
These complex chargers use the
patented Reflex or BURP charge method. This consists of a positive charge
pulse followed by a high current,
short duration discharge pulse. This
should not be confused with pulse or
switchmode chargers which switch
the charge current on and off but do
not apply a discharge current – see
Fig.3.
By using a charger circuit with the
patented Reflex method incorporated
in a licensed integrated circuit, it is
possible to obtain a dramatic increase
in the charge/discharge cycles of nicad
batteries, to at least 3000 cycles if
reasonable care is exercised. There is
no need to run appliances until the
batteries are flat to avoid the memory
effect. It is now possible to recharge
the batteries after each use. Partial dis
charge, as opposed to full discharge,
will significantly increase the life of
the batteries.
A microprocessor calculates and accurately terminates the applied charge
by evaluating the inflection points on
the charge voltage curve. The termination point varies according to the
charging characteristic of the battery;
it occurs just prior to the transition
into overcharge (see Fig.2).
The circuit provides a fast charge,
preceded by a series of soft start charge
pulses. Then, if the battery is left in
January 1996 13
Fig.5: the timing for soft start, fast, topping and maintenance charges. The
charge/discharge pulse combination for the topping and maintenance modes
remain the same as for the fast charge cycle; only the rest time is changed.
the charger, the fast charge will be
followed by a topping charge and a
non-destructive indefinite maintenance charge.
All of the above can be done by
one charger with an adjustable output
current sufficient for batteries of
7000mAh capacity at the 1C (1 hour)
charge rate or for 1900mAh capacity
batteries at the 4C (15 minute) charge
rate, taking the charge efficiency into
account. To fully charge a battery,
approximately 20% more charge than
has been withdrawn has to be put back
into the battery if charged at or above
C/10 at 20°C.
The charge efficiency of batteries
depends on charge current and ambient temperature. High or very low
ambient temper
atures and/or low
charge currents decrease the charge
efficiency; in extreme cases to a point
where the battery cannot be fully
charged.
Soft start
Batteries can exhibit a high impedance during the initial stages of
charging. The resulting voltage peak
can be interpreted by the processor as
a fully charged battery.
However, with the soft start cycle, at
first only short duration current pulses
are applied to the battery. Starting at
200ms, the pulse width is gradually increased to approximately one second
in duration. This gradual increase in
pulse width takes place over a period
of two minutes to avoid voltage peaks.
Fast charge
During the main charge cycle, each
positive current pulse is followed by
a discharge pulse, as shown in Fig.4.
The discharge pulse is 2.5 times the
amplitude of the charge pulse. After
the main charge, if the battery is left
on the charger, it will be fed a topping
14 Silicon Chip
charge. This charge is at a current low
enough to prevent cell heating but high
enough to convert all active material
in the cells to the charged state.
Due to higher temperatures and
gas bubbles (see explanation further
on), 100% charge cannot be achieved
with fast chargers. Standard constant
current chargers create heat and gas
bubbles on the cell plates during
charging. This results in less than 90%
efficiency.
This version of the Reflex charger
is approximately 95% efficient, since
the termination method largely avoids
cell heating and the charge/discharge
pulse sequence removes most of the
gas bubbles from the cell plates. The
2-hour C/10 charge tops up the battery
if the time is available or 100% capacity is required.
Maintenance charge
After the full charge and topping
charge, the C/40 charge compensates
for the internal self-discharge of the
battery, at the same time preventing
dendrite formation and maintaining
the crystal structure. The battery can
remain on the charger until used –
there is no time limit. This charge
cycle can be useful in standby applications, as in security installations.
Fig.5 shows the timing for soft start,
fast, topping and maintenance charges.
The charge/discharge pulse combination for the topping and maintenance
modes remain the same as for the
fast charge cycle; only the rest time
is changed.
The removal of gas bubbles from
the cell plates during charge keeps the
cell impedance low, reduces operating
temperature and allows higher charge
currents for nicad and NiMH batteries.
The following charge times can be
achieved: fast-charge nicad batteries
in less than 15 minutes at the 4C rate,
standard nicad and NiMH batteries in
less than one hour.
As well, memory effect in batteries
can be eliminated. This works even
when the battery no longer holds any
charge. It requires a minimum of three
complete charge/discharge cycles. A
typical case in practice involved a
4.8V 600mAh cellular phone battery
pack. This had only 20% of its stated
capacity, after it had been used over a
period of six months with the supplied
charger. After five charge/discharge cycles, it had recovered to approximately
95% of capacity.
The possibility to rejuvenate shorted nicad batteries is also a feature.
Whenever a nicad battery has been
stored charged and has then slowly
self-discharged over a very long period
of time at an elevated temperature, or
has been charged at a low current over
a long period, as in constant current
trickle charging in standby applications, crystals on the cell plates can
form crystalline fingers, or dendrites,
which can propagate through the plate
separators and across the cell plates.
In extreme cases, these crystalline
dendrites can partially or completely
short-circuit a cell. Such cells can be
rejuvenated by this charger.
Charger circuit
Fig.6 shows the block diagram of
a charger using the patented Reflex
charging method. The charger covers
a battery voltage range from 1.2V to
13.2V at charge currents from 0.1A
to 9.0A. The central part of the battery charger is basically a reduced
instruction set microprocessor (RISC)
to handle the complex calculations for
the charge termination point.
The microprocessor uses an
analog-to-digital converter (ADC)
with 300µV resolution to convert the
battery voltage, normalised to one
cell by the input attenuator VR1. The
ADC is followed by a filter to limit
the effects caused by battery voltage
jumps and ADC noise and to eliminate
Fig.6: the block diagram of a charger using a RISC microprocessor programmed
with the patented Reflex charging method. The charger covers a battery voltage
range from 1.2V to 13.2V at charge currents from 0.1A to 9.0A.
any large aberrations in the battery
voltage curve.
The microprocessor controls the
charge, topping and main
tenance
modes. One input of the microprocessor controls the charge rates (1C
or 4C) and is linked to the bank of
push- buttons for selection of charge
current.
One input resets the microprocessor
to repeat a charge cycle or to charge
shorted cells. In this case, the reset
button has to be activated until the
LED “cell voltage” display indicates
acceptance of the charge current.
A battery voltage guard circuit
avoids automatic charging of shorted
batteries. This is necessary since the
current required to kick start a shorted
battery varies from case to case and
should be controlled manually.
Another detect circuit avoids the
automatic charging of batteries with
a voltage or more than 2V per cell.
This condition is due to high internal
impedance, as found in new batteries
that have not been cycled and in some
batteries which have been stored for
several months. Charging these batteries would cause excessive heating.
The DC input to the charger can
range from 11.5V to 28V, depending
on the number of cells in the battery
to be charged. Essentially, this is a
minimum of 2V per cell plus an additional 2V. Hence a 6V battery (5 cells)
requires a minimum of 12VDC to the
charger while a 12V battery (10 cells)
requires a minimum of 22VDC.
Safety cut-off
In case the voltage sensing for end of
charge does not work there is a timeout circuit which is set for 72 minutes
at the 1C rate and 18 minutes for the
4C rate. In addition, there is a heatsink
temperature sensor to interrupt the
charge as a safety measure in extreme
hot weather conditions.
The microprocessor controls three
output circuits and two LED indicators. The charge circuit is a switch
mode current source, adjustable
from 0.1A to 9A with VR2 (a bank of
pushbutton switches). The discharge
circuit is a pulsed constant current
sink adjusted to between 0.25A and
22.5A (2.5 times the charge current).
During the main charge cycle, a
small piezo speaker emits a brief tone
once a second, synchronised to the
discharge pulses. This is a convenient
audio cue to tell the user the battery is
still in the main charge sequence. The
tone control on the front panel actually
adjusts the volume, so that the tone is
not obtrusive.
Other details of the operation can
be gleaned from the block diagram.
By this time this issue goes on sale,
the charger will have been released
for sale. For information concerning
availability and price, contact Smart
Fastchargers, R.S.D. 540, Devonport,
Tas 7310. Phone/fax (004) 921 368.
*Horst Reuter is Technical Manager
of Smart Fastchargers.
January 1996 15
|