Charging lithium-ion batteries: not all charging systems are created equal

"Developing Affordable Mixed-Signal Power Systems for
Battery Charger Applications"
By Terry Cleveland, Staff Applications Engineer Scott Dearborn, Principal Applications Engineer
INTRODUCTION
As battery-powered electronic devices continue to become more common and more powerful,
easily adaptable battery charger designs are needed. Using standard components, battery charger
designs can be made simultaneously more flexible and more cost-effective. Mixed-signal designs
simplify the addition of new, unique features to the system, and they allow the addition of
differentiating features.
BATTERY CHEMISTRIES
Many different battery chemistries are used for rechargeable portable applications, including
Lithium-Ion (Li-Ion), Nickel Metal Hydride (NiMH), Nickel Cadmium (NiCd), and Lead Acid
batteries. This article will focus on two of the more popular chemistries, Li-Ion and NiMH,
although the topics discussed apply to the other chemistries, as well.
Li-Ion batteries have the highest energy density of all battery types, making them the most
portable of all rechargeable technologies. NiMH batteries are popular because they are safe and
environmentally friendly. It is possible to design a mixed-signal, universal battery charger to
charge both of these battery chemistries.
BATTERY CHARGING TERMINOLOGY
The rate of charge or discharge is expressed in relation to battery capacity. Known as the “C-
Rate,” this rate of charge equates to a charge or discharge current, and is defined as:
Where: I = charge or discharge current, A M = multiple or fraction of C C = numerical value of rated capacity, Ah n = time in hours at which C is declared. A battery discharging at a C-rate of 1 delivers its nominal rated capacity in one hour. For example, if the rated capacity is 1000mAhr, a discharge rate of 1C corresponds to a discharge current of 1000mA. Similarly, a rate of C/10 corresponds to a discharge current of 100mA. PREFERRED CHARGE PROFILE (Li-Ion and NiMH)
Li-Ion battery chemistries utilize a constant, or controlled, current and constant voltage algorithm
that can be broken-up into four stages: (1) trickle charge, (2) constant current charge, (3)
constant voltage charge and (4) charge termination. Figure 1 illustrates these four stages of Li-
Ion battery charging.
The preferred algorithm for NiMH consists of the following stages: (1) trickle charge, (2)
constant current, (3) top-off charge and (4) charge termination. Figure 2 illustrates these four
stages of NiMH battery charging.
Stage 1: Trickle Charge -- Trickle charge restores charge to deeply depleted cells. For Li-Ion
batteries, when the cell voltage is below approximately 3V, the cell charges with a constant
current of 0.1C maximum. For NiMH batteries, trickle charge conditions weak batteries, when
the cell voltage is greater than 0.9V per cell “fast” charge, or constant current charge can begin.
Stage 2: Constant Current Charge – For Li-Ion and NiMH batteries, after the cell voltage has
risen above the trickle charge threshold, the charge current increases in order to perform constant
current charging. The constant current charge should be in the 0.2C to 1.0C range.
Stage 3: Constant Voltage – For Li-Ion batteries only, constant current charge ends and the
constant voltage stage begins when the cell voltage reaches 4.2V. In order to maximize
performance, the voltage regulation tolerance should be better than ±1%.
Stage 4: Charge Termination – For Li-Ion batteries, the continuation of trickle charging is not
recommended. Instead, charge termination is a good option. For NiMH batteries, a timed trickle
charge ensures 100% of battery capacity use. When the timed trickle charge is complete, charge
termination is then necessary.
For Li-Ion batteries, one of two methods -- minimum charge current, or a timer (or a
combination of the two), typically terminates charging. The minimum charge current approach
monitors the charge current during the constant voltage stage and terminates the charge when the
charge current diminishes in the range of 0.02C to 0.07C. The timer method determines when the
constant voltage stage begins. Charging then continues for two hours, and then the charge
terminates.
Charging in this manner replenishes a deeply depleted battery in roughly 2.5 to 3 hours.
Advanced chargers employ additional safety features. For example, with many advanced
chargers, the charge stops if battery temperature is less than 0°C or greater than 45°C.
CELL VOLTAGE
CAPACITY
CHARGE CURRENT
Time (Hours)
Figure 1: Li-Ion Charge Profile

For NiMH batteries, charge termination is based on a –dV/dt reading of the battery pack, a
+dT/dt (delta temperature versus time), or a combination of both. In this case, temperature
sensing is a possible safety precaution, as well as a termination method.
Figure 2: NiMH Charge Profile

BATTERY CHARGING -- SYSTEM CONSIDERATIONS
To recharge any battery quickly and reliably, a high-performance charging system is required.
The following system parameters ensure a reliable, cost-effective solution:
1. Input Source
Many applications use very inexpensive wall cubes for the input supply. Output voltage is highly
dependent on the wide-ranging ac input voltage, as well as the load current drawn from the wall
cube.
Applications that charge from a car adapter can experience a similar problem. The output voltage
of a car adapter will typically have a range of 9V to 18V.
2. Output Voltage Regulation Accuracy
For Li-Ion batteries, output voltage regulation accuracy is critical to maximizing battery capacity
usage. A SMALL decrease in output voltage accuracy results in a LARGE decrease in capacity.
However, the output voltage cannot be set arbitrarily high because of safety and reliability
concerns. Figure 3 depicts the importance of output voltage regulation accuracy.
Capacity Loss (%)
Percent Undercharge (%)
Figure 3: Capacity Loss vs. Undercharge Voltage

3. Charge Termination Method
Over-charging is the Achilles' heal of Li-Ion and NiMH cells. Accurate charge termination
methods are essential for a safe and reliable charging system.
4. Cell Temperature Monitoring
The temperature range over which a rechargeable battery should be charged is typically 0°C to
45°C. Charging the battery at temperatures outside of this range may cause the battery to
overheat. During a charge cycle, pressure inside the battery increases, causing it to swell. As
temperature and pressure are directly related, the combination of high temperature and high
pressure inside the battery can lead to mechanical breakdown or venting inside the battery.
Charging the battery outside of the 0°C to 45°C temperature range may also harm battery
performance, or reduce its life-expectance.

5. “Battery Discharge Current” or “Reverse Leakage Current”
In many applications, the charging system remains connected to the battery, even in the absence
of input power. The charging system minimizes current drain from the battery when input power
is not present. Maximum current drain should be less than a few microamperes and, ideally,
below one microampere.
DESIGNING BATTERY CHARGERS
Given the aforementioned system considerations, an appropriate charge management system can
be developed.
Linear Solutions
Linear charging solutions are employed when a well-regulated input source is available. An
example of a linear charging solution is Microchip Technology’s MCP738xx Linear Battery
Charger family. Linear solutions, in these applications, offer advantages such as ease-of-use,
size, and cost.
Switch Mode Charging Solutions
For a wide input voltage range, such as the unregulated ac-dc wall cube or the automotive dc
input, switching regulators lower the internal battery charger power dissipation to an acceptable
level.
Selecting Topology
Switching regulator topology defines the organization of the regulator’s switches and passive
filtering components. This difference in organization distinguishes topologies, offering a trade-
off between complexity, efficiency, noise, and output voltage range. Many converter topologies
exist, while only a few are popular to battery chargers in the 5-Watt to 50-Watt range.
Buck Regulator

The buck or “step-down” regulator is one popular topology for battery charging applications.
The buck regulator, like other solutions, has the following advantages and disadvantages:
1) Low complexity, single inductor topology. 2) For Synchronous applications, conversion efficiency can reach 90%. 1) The buck regulator MOSFET switch integral body diode creates a path to discharge the battery when input voltage is not present. An additional blocking diode is therefore necessary, adding an additional component and, hence, voltage drop to the system. 2) Buck regulator input current is pulsed or “chopped.” This topology generates high electromagnetic interference (EMI) at the input of the power supply. Most buck regulators require additional input EMI filtering. 3) The buck regulator can only regulate output voltages that are lower than the input voltage. Some applications have a wide input voltage range that spans the necessary output voltage range. This is more common for multiple cell Li-Ion charger applications. 4) A single fault mode (buck switch short) creates a short circuit from input to battery. For NiMH applications where there is no internal battery protection, this poses a safety concern. 5) The buck regulator requires a high side drive (for N-Channel MOSFET switches). This is more complex when compared to low-side topologies. 6) External switch current sensing in pulse width modulation (PWM) controller applications is complex. Limiting switch current is important for fault modes such as shorted batteries or load. Without a high-speed switch current limit, the battery charger can be destroyed during a shorted condition. Figure 4: Buck Regulator Topology
SEPIC (Single-Ended Primary Inductive) Regulator The SEPIC regulator is also a popular topology used for battery charging applications. SEPIC regulators have advantages over buck regulators and other topologies, and a few disadvantages. 1) The blocking diode is built-in to the battery system topology, so no additional 2) The input current pulled from the source is continuous (smooth) when compared to the “choppy” input currents of buck regulators. 3) Input to output is isolated, protecting the load or battery from a switch short. 4) The SEPIC regulator topology has step-down or step-up (buck-boost) capabilities. 5) The SEPIC switch is low side, simplifying the gate drive and current sensing in the 6) The secondary side average inductor current is equal to battery current, enabling the sensing of current not in series with the low side of the battery. 1) Requires two inductors or a “coupled” inductor. 2) Requires a single coupling capacitor, which can be expensive for high power (> 50 Watts, or high voltage (VIN > 100V) applications. Figure 5: SEPIC Regulator Topology

SWITCHING BATTERY CHARGER DESIGN
By partitioning the design into two parts, it is possible to develop affordable, “intelligent” power
systems. Battery chargers are, by nature, mixed-signal systems. For example, the power train (in
this case, the SEPIC regulator) is analog. Turning the power switch on and off at high frequency
requires some type of analog driver circuit. On the other hand, charge termination timers, fault
management, and on/off control are typically digital functions that use timers and programmable
capability.


BATTERY CHARGER SPECIFICATIONS

Input
0V to 4.2V for single cell, 0V to 8.4V for two cell 100 mA (current at which charge cycle is completed)
Features:
Over-Voltage Protection (Battery Removal)
Over-Current Protection (Battery or Load Shorted)
Sense Battery Temperature for Charge Qualification
STRATEGY
Using a two-part approach to the mixed-signal design, first select a microcontroller that is
capable of reading the state of the battery pack (voltage and temperature) and programming the
SEPIC regulator output current. For our example, we will use the PIC® PIC12F683 8-pin Flash
microcontroller.
Next, add a high-speed, analog PWM with a built-in MOSFET driver such as the MCP1630 to
develop the “analog” programmable current source.
DESIGNING A SEPIC-PROGRAMMABLE CURRENT SOURCE
As with all switching regulator designs, the output is controlled by varying either the duty cycle,
or the percentage of switch on-time (Q1, Figure 6). To regulate current going into the battery,
charge current must be sensed. As shown in Figure 6, there is no sense element in series with the
battery. The SEPIC regulator secondary winding, Ls, carries the average output current. The
primary winding, Lp, carries the average input current. Secondary resistor Rs senses battery
charge current. The high-speed, analog PWM reference input programs the desired battery
charge current.
Figure 6: Mixed Signal Battery Charger Block Diagram

MIXED-SIGNAL DESIGN
By using the MCP1630 as an analog PWM and driver, a “programmable” SEPIC current source
is achieved. The PWM and driver provide the analog current regulation, MOSFET gate drive,
and high speed over current protection. The PIC12F683 microcontroller sets the SEPIC power
train switching frequency (500 kHz) and programs the SEPIC constant current.
The PWM and driver utilize the microcontroller hardware PWM to set the SEPIC switching
frequency and maximum duty-cycle. The hardware PWM frequency is equal to the SEPIC power
train switching frequency, while the hardware PWM duty cycle sets the maximum SEPIC power
train duty cycle. A 500 kHz pulse with a 25% duty cycle out of the microcontroller hardware
PWM sets the SEPIC switching frequency to 500 kHz with a maximum duty cycle of 75%. A
standard microcontroller I/O pin generates a software-programmable reference voltage using a
simple R, C filter. This programmable reference programs the constant current SEPIC converter
to a precise charge current.
At the non-inverting input (Vref), the programmable reference voltage sets the amount of battery
charge current. The MCP1630 PWM output duty cycle (Vext) adjusts until the voltage at the
Vref input is equal to the voltage at the FB input of the error amplifier. By adjusting the voltage
at the Vref input, the battery current adjusts accordingly.
The PWM and driver are capable of driving the MOSFET at frequencies greater than 500 kHz,
while monitoring the SEPIC switch current using an internal high-speed (12 ns typical)
comparator. If the switch current is too high, the PWM duty-cycle will terminate, limiting the
battery current.
Finally, the charge current is adjusted based on information such as battery voltage and
temperature, received from an analog-to-digital converter (ADC).
To develop a constant voltage charge phase, the microcontroller A/D converter reads the battery
voltage and updates the programmable current source (SEPIC) to maintain the battery voltage at
4.2V. This occurs at a rate much faster than the rate at which the battery voltage changes when
subject to a constant current.
For Li-Ion applications, the charge cycle terminates when the current necessary to maintain the
battery voltage at a fixed 4.2V reduces to some percentage of the battery C-rate (100 mA). This
is set using firmware and is easily changed for different battery manufacturers’
recommendations. In a typical analog charger, this termination charge current is a percentage of
the charge cycle current, so that it cannot easily be changed.
For NiMH applications, the fast charge cycle terminates when one or both of two conditions
occur -- either the battery voltage remains constant or drops with time, or the battery pack
temperature rise is higher than a predetermined value. When fast charge terminates, a slow,
timed trickle charge can begin.
The development of cost-effective, “intelligent” power systems is possible by combining an
analog PWM and driver with a standard microcontroller.
An ADC input and battery pack thermistor together sense battery temperature. By reading the
voltage at the “TEMP_SENSE” input, battery temperature can be determined.
The interruption of the PIC12F683 code when the sensed battery voltage is too high achieves
over voltage protection (OV). The SEPIC converter shuts down in less than 1 us, with minimal
voltage overshoot occurring at the battery terminals.
The SEPIC converter diode prevents any path for battery discharge back to the system charger.
The only quiescent current draw on the battery is from a battery voltage-sensing path, typically
less than 5 uA.
OPTIONAL FEATURES
In addition, the use of a single microcontroller and multiple high-speed analog PWM modules
enables the addition of charger bays for multi-bay applications, as well as out-of-phase switching
techniques and input power budgeting features.
Firmware such as the items listed here increase system precision because they enable the
calibration of the Li-Ion termination voltages and charge currents.
CONCLUSION
By using a mixed-signal approach to developing battery chargers, battery charger designs can
take advantage of the best of both the analog and digital worlds. Employing a mixed-signal
approach enables high frequency operation (500 kHz), high-speed protection (12ns current sense
to output), and minimizes the size of filtering components. In addition, the programmable digital
features of the system enable appropriate determination of stage of charge and set charge current.
Because it facilitates the programming of settings and currents, firmware enhances new battery
charging methods. This approach differentiates one mixed-signal design from another. This type
of design is not limited to Li-Ion and NiMH batteries, and it leaves the door open for the
programming of future rechargeable technologies into the system.

Note: The Microchip name and logo and PIC are registered trademarks of Microchip Technology Inc. in the U.S.A.
and other countries. I2C is a trademark of Royal Phillips Electronics of the Netherlands. All other trademarks
mentioned herein are property of their respective companies.

References
1. David Linden, Thomas B. Reddy, Handbook of Batteries, Third Edition (New York: McGraw-
Hill, Inc., 2002).
2. http://sanyo.com/batteries/lithium_ion.cfm
3. Low Cost Li-Ion Battery Charger User’s Guide, DS51555
4. PIC12F683 Data Sheet, DS41211.
5. MCP1630 Data Sheet, DS21896A.
6. MCP1630 Li-Ion Multi-Bay Battery Charger User’s Guide, DS51515.
7. Microchip Application Note: AN960, DS00960

Source: http://www.microchip.com.edgekey.net/stellent/groups/designcenter_sg/documents/market_communication/en027883.pdf