Pitfalls when recharging batteries

Gerf Helles, Maxim Integrated Products

Secondary batteries can be simple devices, but improper charging can be devastating to the user experience.

The use of batteries has never been greater, particularly in light of the boom in portable devices such as mobile phones, laptops, camcorders, and MP3 players. The bat­teries are becoming smaller and lighter, even as they pack more and more energy per unit volume.

Nonrechargeable (primary) batteries create electricity from a chemical reaction that permanently transforms the cell. Discharge of the primary cell leads to a perma­nent and irreversible change in the cell chemicals. Rechargeable (secondary) batteries store rather than generate energy.

Charge or discharge current is typically expressed as a multiple of the rated capacity (C-rate). For example, a C/10 discharge current for a battery rated at 1 Ah is 1 Ah/10 =100 mA. The rated capacity of a cell is the amount of electricity it can store (produce) when fully charged under specified conditions. Thus, a battery's to­tal energy is its capacity multiplied by its voltage, result­ing in a watt-hours measurement.

Measuring performance
The cell's chemistry and design together limit the current it can source. Barring the practical factors that limit performance, a battery could produce an infinite current, if only briefly. The main impediments to infinite current are the chemicals' basic reaction rates, the cell design, and the area over which the reaction occurs. Some cells can inherently produce high currents. Short­ing a NiCd cell, for instance, produces currents high enough to melt metals and start fires.

Other batteries
The net effect of all chemical and mechanical factors in a battery can be expressed as a mathematical factor.

No battery stores energy forever. Unavoidably, the cell chemicals react and slowly degrade, causing degra­dation in the stored charge. The ratio of battery capacity to weight (or size) is the battery's storage density (see Table 1).
One question that comes to mind is, why not always choose secondary cells if primary and secondary cells ful­fill the same purpose? Secondary cells have two key drawbacks. First, secondary cells lose their electrical charge relatively quickly through self-discharge. Second,

Every charging operation applies voltage and current in a sequence that depends on the battery's chemistry. Thus, a look at chemistries reveals different requirements to be met by the charger and charging algorithm (see

NiCd cells are charged by applying a constant cur­rent ranging from 0.05C to more than 1C. Some low--cost chargers terminate the charge by means of absolute temperature. Though simple and inexpensive, this method of charge termination is not accurate. A better choice is to terminate charging when the full-charge condition is indicated by a drop in voltage. The -AV phenomenon is most useful for charging NiCd cells of 0.5C or greater. The -AV end-of-charge detection should be combined with battery-temperature measurement as well, because aging and mismatched < voltage reduce cells>

A more precise full-charge detec­tion can be achieved by sensing the rate of temperature increase (dT/dt). This charge-detection method is kinder to the battery than a fixed-temperature cutoff. Charge termination based on a combination of AT/dt and -AV cutoff enables a longer lifecy-cle by avoiding overcharge.

Fast-charging improves charge ef­ficiency. At 1C, the efficiency is close to 1.1 (91%), and the charge time for an empty pack will be slightly more than 1 hr. When apply­ing a 0.1C charge, the efficiency drops to 1.4 (71%), with a charge time of about 14 hrs.

Because a NiCd cell's charge ac­ceptance is close to 100%, almost all energy is absorbed during the initial 70% of charging, and the battery re-mains cool. Ultra-fast chargers use this phenomenon to charge a battery to the 70% level within minutes, ap­plying currents several times the C-rating without heat buildup. Above 70%, the charging continues at a lower rate until the battery is fully charged. Eventually, the battery is topped off by applying a trickle charge in the 0.02C to 0.1C range.

Table 1. Battery storage density

Cell type Nominal voltage Storage density

Lead acid


30 Wh/kg

Nickel cadmium


40 to 60 Wh/kg

Nickel metal hydride


60 to 80 Wh/kg

Circular lithium ion


90 to 100 Wh/kg

Prismatic lithium ion


100 to 110 Wh/kg

Polymer lithium ion


130 to 150 Wh/kg

Charging the chemistries
Though similar to NiCd chargers, an NiMH charger employs the AT/dt method, which is the best method for charging NiMH cells. The end-of-charge voltage depression for NiMH batteries is smaller, and for small charge rates (below 0.5C, depending on temperature), there may be no voltage depression at all.

New NiMH batteries can show false peaks early in the charge cycle, causing the charger to terminate prematurely. Moreover, an end-of-charge termination by -AV detection alone al­most certainly ensures an overcharge, which in turn limits the number of possible charge/discharge cycles.

It seems there's no available < for well works that algorithm dt -dV>charging NiMH batteries under all conditions: new or old, hot or cold, and fully or partly discharged. For that reason, don't charge an NiMH battery with a NiCd charger unless it utilizes the dT/dt method for end-of-charge termination. And because NiMH cells don't absorb overcharge well, the trickle charge must be lower (about 0.05C) than that recom­mended for NiCd cells.

Slow-charging a NiMH battery is difficult, if not impossible, because the voltage and temperature profiles associated with a C-rate of 0.1C to 0.3C don't provide a sufficiently ac­curate and unambiguous indication of the full-charge state. The slow charger must therefore rely on a timer to indicate when the charge cycle should be terminated. Thus, to fully charge a NiMH battery, apply a

1. Constant-current, rapid charge of constant-voltage about 1C (or a (CCCV) charging is rate specified popular in high-end by the battery portable equipment, maker), while monitoring both voltage (AV = 0) and temperature < dt) (dT>termine when the charge should be terminated.

Whereas chargers for nickel-based batteries are current-limiting devices, chargers for Lilon batteries limit both voltage and current. The first Lilon cells called for a charge-voltage limit of 4.10 V/cell. Higher voltage means greater capacity, and cell volt­ages as high as 4-2 V have been achieved by adding chemical addi­tives. Modern Lilon cells are typi­cally charged to 4.20 V with a toler­ance of ±0.05 V/cell.

Full charge is attained after the terminal voltage reaches the voltage threshold and the charging current drops below 0.03C, which is about 3% of Icharge (see Figure 1). The time for most chargers to achieve a full charge is about 3 hrs, though some linear chargers claim to charge a Lilon cell in about an hour. Such chargers usually terminate the charge when the battery's terminal voltage reaches 4-2 V. That kind of charge determination, however, charges the battery only to 70% of its capacity.

A higher-charging current doesn't shorten the charge time by much. Higher current allows reaching the voltage peak earlier, but then the top­ping charge takes longer. As a rule of thumb, the topping charge will take twice as long as the initial charge.

Because overcharging (or overdis-charging) a Lilon cell can cause it to explode, safety is a major concern in proper charging for Lilon batteries.

The battery industry has seen significant growth due to the latest stream of portable devices. With so many hand­held products using batteries as a power source, battery manufacturers have devoted extensive R&D to make their batteries safer and lighter and with higher energy densities. That has been the predominate driver for the increase of lithium-cell chemistries at the expense of either NiCd or NiMH. A key advantage Lilon has over NiMH relates to "memory loss." Unlike the NiMH, a Lilon battery can be recharged at any point during its discharge cycle and can efficiently hold its charge, which is more than twice as long as a NiMH battery.

Another advantage of rechargeable lithium technology is its weight—it offers three times the energy density of a NiCd battery but with half the weight. Furthermore, new chip technologies combined with the latest software allow for the maximum extension of power use in the latest notebook computers, portable instruments, and cel­lular communicators.

However, care must be taken when charging a Li-lon battery in the end product, because the battery can be eas­ily damaged if the wrong voltage or current is applied. To deal with these concerns, 1C makers have designed spe­cialized charger ICs that ensure the battery is charged fully and safely under varying environmental conditions.

Squeezing a fully featured high-power charger into one 1C has always come with tradeoffs. To fit everything in a small package, most chargers included a few specialized features but at the expense of other generally desirable features.  < voltage reduce cells < for well works that algorithm dt -dV>

Nevertheless, the latest generation of chargers shares new control architectures to combine as many desirable features as possible. They support batteries up to 28 V, at charge rates up to 4 A, with efficiencies higher than 95%. Charging voltage and current accuracy are typically 0.7% and 4%, respectively. Furthermore, these pulse-width modulation (PWM) controllers, with 300- to 500-kHz synchronous architectures, can achieve 98% duty cycles, providing excellent low-dropout performance as well as continuous switching down to zero charge cur­rent. Lastly, these PWM controllers are designed to suppress the audible noise from ceramic capacitors while also allowing the use of popular inductor values for small size at high currents handling this type of storage cell. As a result, commercial Lilon packs con­tain a protection circuit that provides all electronic safety functions re­quired for applications involving rechargeable Lilon batteries: protect­ing the battery during charge, pro­tecting the circuit against excess cur­rent flow, and maximizing battery life by limiting the level of cell depletion (see Figure 2).

Protection modes

If the cell voltage sensed at vdd ex­ceeds the overvoltage threshold vqv for a period longer than the overvolt­age delay tOVD, the protection 1C should shut off the external charge FET and set an OV flag. The dis­charge path remains open during overvoltage. The charge FET is re-en­abled (unless blocked by another pro­tection condition) when the cell volt­age falls below the charge-enable threshold VCE, or discharging causes vdd"~vpls to be 8reater than vqq.

If the voltage sensed at Vj-,D drops below the undervoltage threshold Vuv for a period longer than the un­dervoltage delay tuyD, the 1C shuts off the charge and discharge FETs and sets the UV flag. When the voltage rises above Vuv and a charger is pres­ent, the 1C turns on the charge and discharge FETs.

If the cell voltage sensed at VDD drops below the depletion threshold Vsc for a period of tSCD, the 1C shuts off the charge and discharge FETs and sets the SC flag. The current path through the charge and discharge FETs isn't reestablished until the volt­age on PLS rises above VDD-VOC.
If voltage across the protection FETs (VDD-VpLS) is greater than VQC for a period longer than tQ^p, the 1C shuts off the external charge and dis­charge FETs and sets the DQC flag. The current path isn't reestablished until the voltage on PLS rises above vdd-voc-

If the 1C exceeds a predetermined temperature TMAX, the device imme­diately shuts off the external charge and discharge FETs. The FETs aren't

How much time before recharging your battery pack?

A common question when people think about battery-pack performance is: "How long until I need to recharge?" It generally doesn't matter whether it's a notebook, a data-logger, an MP3 player, a PDA, a cell phone, or a digital camcorder.

Consumers tend to compare a battery's life to a car's gas tank; how­ever, it's not that simple. The behavior of a car's driver significantly in­fluences fuel consumption. But a battery is affected by how you treat it, including the discharge rate level and frequency, how many times it has been cycled, and its environmental situation.

Today's leading semiconductor vendors offer sophisticated gas-gauge 1C algorithms that require precise analog measurement and computing capabilities as well as detailed battery characterization. Most current al­gorithms are based on measuring the cell voltages and monitoring the charge and discharge current, a technique known as coulomb counting. Looking at the cell voltage, especially with Lilon batteries, the discharge curves are relatively flat, making it difficult to determine remaining capacity.

Computing the remaining capacity based on a constant load directly correlates to the age and temperature of the battery pack throughout its life. The temperature is easily measured, but age is a different story. In simple terms, age means that the impedance and capacity have changed due to chemical reactions.

The typical capacity change after 100 cycles is about 3% to 4%, but the impedance level is nearly double. In addition, today's gas-gauge al­gorithms compensate for aging effects using certain amounts of charge and discharge to determine cycle counts. The algorithm then provides a rough estimate based on these assumptions, but the results can be un­reliable.

Under varying loads, the battery's impedance plays a bigger role. If the load increases, the voltage drop at the battery's internal resistance in­creases and switches off mechanisms to protect the battery and triggers deep discharges. Protection ICs used in Lilon packs disable further dis­charge, and the unit will stop operating, even if usable capacity remains.

Variable loads also make it hard to predict the remaining runtime. Us­ing an update for the capacity, often referred to as a "learning cycle," can help better understand usable capacity.

Current-based coulomb counting solutions enable such updates but also require certain conditions to function properly. If such a learning cy­cle doesn't occur for a period of time, or never appears, the failure rate increases up to 50% over time. With an updated system, the failure rate can be as low as 2% to 3%.

New developments are on the horizon to identify the battery's imped­ance as the main contributor to the capacity uncertainty in accounting for aging and fading. Innovative algorithms will be integrated into future gas-gauge ICs to improve how designers account for these aging effects. But the main benefactors of such battery management improvements will be the consumers turned back on until two conditions are met: Cell temper­ature drops below tmax and the host resets the OT bit.

Charging at extreme temperatures

Efforts should be made to charge at room temperature. Nickel-based batteries should only be fast-charged between 10° and 30°C. Below 5°C and above 45°C, the charge ac­ceptance of nickel-based batteries is drastically reduced.

2. Typical application Lilon cells offer reasonably good diagram for a lithium- charge performance throughout the cell protection 1C is temperature range, but below 5°C the illustrated charge rate should be less than 1C.

NiMH chargers can accommodate NiCd batteries, but not vice versa. Chargers dedicated to NiCd batteries will overcharge a NiMH battery. The cycle life and performance of nickel-based batteries are enhanced by fast-charging, because it reduces the mem­ory effect due to the formation of internal crystals. Nickel- and lithium-based batteries call for different charge algorithms. Lilon batteries need pro­tection circuitry to monitor and pro­tect against overcurrent, short circuits, overvoltage and un-dervoltage, and excessive temperature.

Gert Helles is a senior field applications engineer at Maxim's battery and thermal management group, based in Hadsten, Denmark. He holds a BSEE from the University of Southern Denmark. The company, based in Sunnyvale, CA, can be reached at (408) 737-7600 or WWW.maxim-iC.com.