Advanced Li-ion battery system charge management and protection

Lithium-ion (Li-Ion) chemistry has become common as the battery technology of choice for many portable devices. This battery chemistry is well known in terms of the required charging algorithms, there are many charge management integrated circuits (ICs) on the market, and its competitive low cost is driving its use in more applications. Now, with a wide variety of power supplies, charging IC products, and system architectures, engineers are busy selecting the appropriate charging and charge protection topology. This article introduces some important system protection methods and advanced charge management to achieve maximum safety, reliability, and system performance.

The many challenges of overvoltage and overcurrent protection

Portable end devices such as mobile phones, media players or GPS systems usually have a dedicated power supply with known supply voltage and current characteristics, and usually have a dedicated plug. The purpose of this is to prevent consumers from using unapproved power supplies or connecting the power supply in reverse. Today, driven by the demand for higher energy efficiency, consumers and standards organizations are demanding a universal power interface, aiming to allow end devices from different manufacturers to use standardized AC/DC adapters or USB connections.

This presents a huge challenge to power supply designers, as consumers now control the power to the devices they plug into. Protection circuitry is now required in front of the charger circuit to protect the system from overvoltage and overcurrent conditions while avoiding serious damage to the end device. More importantly, it eliminates safety risks that could be dangerous to the end user. Because this protection circuitry needs to protect against a “large unknown,” it must include a wide range of input power conditions while ensuring that the device is charged within specified limits. Some examples of real-world applications are USB charging that complies with the China Universal Standard Charger specification and charging from a low-cost, unregulated wall power supply. These can present temporary open circuit voltages above 10V and require the charge management system to still charge to that threshold. This places many specific requirements on the input and output protection circuitry.

Under high board space and cost constraints, the input of the protection solution needs to withstand the highest voltage possible. It needs to pass energy under specified operating conditions and block energy above the expected overvoltage protection threshold (OVP) without damaging the device. Above the solution's absolute maximum rating, it needs to interrupt the "open circuit" to prevent potentially harmful overcurrent from entering the system. Whether powered by a battery or directly from an AC adapter, the output must guarantee voltage levels that do not exceed the subsystem's specifications. Often, subsystems such as voltage protection and handling cannot withstand high voltage input transients using cost-effective, low-voltage process technologies that are sensitive to overvoltage.

Overcurrent protection and current limiting may be worthwhile from both a safety and compliance perspective, with the goal of not exceeding the inrush current limit at power-up or exceeding the maximum USB current specification.

Figure 1 depicts two scenarios for charger subsystem input protection in a single-cell Li-ion battery system. In scenario A, the charging function is primarily performed by a software-controlled charging circuit integrated into a low-voltage power management unit. This is typically part of a highly integrated chipset for wireless handsets, GPS navigation systems, or Bluetooth headsets. In this case, a separate overvoltage and overcurrent protection IC makes sense to add the necessary protection functions.

In case B, the charging function is implemented by a dedicated stand-alone charger IC, which manages the battery charging and dynamic power path control to ensure that the system operates normally - even with a defective, fully discharged battery pack or when the battery pack is removed. System engineers may choose this configuration to minimize the development work of charging control related software. In addition, designers may also want to protect the system from unnecessary and potentially unsafe charging behavior caused by a locked microcontroller. In this application case, it makes sense to integrate overvoltage protection and current limiting functions in the charger IC.

Figure 1 System topology for input protection and charging management

Case A: System protection of analog baseband

Given the required integration and cost level of the analog baseband, IC processes based on CMOS technology are usually used, with a voltage tolerance range of 4.5 to 6 volts (V). Lower semiconductor process voltages generally mean higher digital density and smaller silicon chip size for on-chip power components, such as transistors and diodes. This leads to smaller die size and packaging, resulting in lower total system cost. However, it is usually accompanied by higher overvoltage sensitivity, resulting in electrical overstress.

That is, the output of the protection solution must be guaranteed not to exceed this level under any input environment, including static DC operating levels and transient conditions. Ideally, it regulates the output to a preset limit while accepting a wide input range without damage. In case A, it is assumed that the baseband IC and other subsystem components can withstand an absolute maximum voltage of about 6V, such as linear regulators (LDOs) and DC/DC switching power converters. Therefore, the output of the protection IC is regulated to a nominal 5.5V to allow for the regulation tolerance caused by transient response time.

The protection circuit input takes into account the "normal use" environment, the tolerable transient conditions that occur during normal operation when charging, and the abnormal transients that the system needs to be thoroughly protected.

Charging via USB power is very popular today, with the USB 2.0 specification specifying a nominal VBUS operating voltage of 5V, a minimum voltage of 4.75V, and a maximum voltage of 5.25V. These power sources may be a computer USB port or USB hub that complies with the USB specification. However, they can also be powered from a regulated AC/DC wall adapter that "mimics" the power behavior of a USB port. The normal use environment here refers to the 5V nominal voltage. However, the emerging Chinese Universal Standard Charging Adapter Specification, which intends to simplify the use of power for USB port and adapter-powered end devices, requires uninterrupted charging during VBUS power transients of up to 6V, which exceeds the rated voltage of the analog baseband and can cause damage. For such applications, an input OVP level of 5.85V is recommended, beyond which the protection circuit will cut power to the system. If below this input OVP threshold, the output will be regulated to a safe 5.5V. That is, operation to the OVP level is ideal and is considered an acceptable transient environment. Transient conditions above the OVP level are considered abnormal and require system isolation. Assume that the system is operating in protection mode. State-of-the-art integrated protection circuits can withstand overvoltages up to 30V and recover from this state. When this level is exceeded, additional circuits such as Zener diodes can be added to protect the OVP IC above the 30V level.

In the case of unregulated, low-cost AC/DC wall adapters, the nominal adapter voltage may be specified as 5V under load conditions. However, due to the inherent characteristics of these low-cost adapters, open-circuit voltages as high as 10V may appear during initial connection under no-load conditions, which can immediately destroy the low-voltage chipset. In this case, a protection circuit with a 10.5V OVP threshold is used to protect the circuit output from regulation to a safe 5.5V. Depending on the input power supply, the correct solution with the appropriate OVP level needs to be selected.

Figure 2 shows the voltage protection of this type of linear regulation mode overvoltage protection solution.

Figure 2 shows the voltage protection of this type of linear regulation mode overvoltage protection solution.

The output of the protection circuit is held at 0V for input voltages from 0V to the undervoltage lockout (UVLO) level (IC operating level). Above VUVLO, the output voltage tracks the voltage drop caused by the input voltage being below the RDS,on of the protection FET of the protection circuit, even though the input voltage is below the regulation voltage VO(REG). The output is regulated to 5.5V when the input voltage is between VO(REG) and VOVP, which is the threshold for permissible transient conditions. If the input voltage rises above VOVP, the protection FET Q1 is turned off, removing the output power. The response must be fast, and the FET must turn off in less than one microsecond. This state is signaled to the host system via a FAULT signal. When the input voltage returns to below VOVP minus the hysteresis voltage Vhys(OVP) but still above VUVLO, the protection FET is turned on again after a tON(OVP) deglitch time to ensure that the input power has stabilized.

The solution’s over-current protection (OCP) threshold is programmable with a resistor for ease of use. If the load current is about to exceed the IOCP threshold, the device limits the current during the tBLANK(OCP) blanking period. If the load current returns below IOCP before the tBLANK(OCP) time has expired, the solution continues to operate. However, if the over-current condition persists for tBLANK(OCP), Q1 turns off for a recovery period of tREC(OCP) and a fault signal is sent. The FET turns on again after the recovery time and the current is again fully monitored. Whenever an OCP fault occurs, an internal counter increments. If several OCP faults occur during a charge cycle, the FET is permanently turned off. The counter can be cleared by removing and reapplying the input supply or by restarting the device. To prevent the input voltage from experiencing a spike caused by the input cable inductance, Q1 can be slowly turned off to achieve a “soft shutdown.”

A more stringent battery overvoltage protection can also be implemented that sends a fault signal for each battery overvoltage event.

Case B: Comprehensive protection and charging functions

In the case of considering the charger function in a separate IC, similar overvoltage protection and current limiting functions should be implemented to protect the system from overvoltage peaks on the DC power line while allowing the use of low-cost, unregulated wall power supplies. The charging IC must also perform correct output voltage regulation without exceeding the many limitations of the subsystem. In addition, the charger solution now needs to manage many functions, such as USB current limiting and power path management, to ensure compliance with the standard and system startup without violating all operating conditions. Figure 3 shows a USB standard charging implementation with integrated OVP and input current limiting.

Some of today’s advanced solutions are able to power the system while charging the battery simultaneously and independently. This reduces the number of battery charge and discharge cycles, enables proper charge termination, and enables the system to operate with a defective or no battery pack, such as in production test environments. It even allows instant system power-up with a fully discharged battery. Input current monitoring and limiting are key to meeting USB standards. In many applications, the input source for charging the battery and running the system can be either an AC/DC adapter or a USB port. Dynamic power path management (DPPM) shares the supply current between the system and battery charging and automatically reduces the charge current as the system load increases. When charging from a USB port, input voltage-based dynamic power management (IDPM) reduces the input current if the input voltage drops below a threshold that prevents the USB port from crashing. The power path architecture also allows the battery to compensate for system current requirements when the adapter cannot provide the peak system current. Our standard is to use a constant current, constant voltage (CCCV) charging scheme and battery preconditioning using precharge and temperature qualification.

Input voltage protection

Similar to standalone protection solutions, the charger should be protected from high input voltage damage. When the input voltage is above VOVP for longer than the deglitch time, OVP turns off output regulation and interrupts charging. At OVP, the system output (OUT) is connected to the battery and the charger sends a differential input supply signal. Once the OVP condition disappears, a new power-up sequence begins and the charger is reset.

Dynamic power path management with input current limiting and output voltage regulation

In some systems with power path management, the charger has an output to power an external load. This is important in systems that are not directly connected to the battery. This output is active as long as power is connected to the charger input (IN) or the battery input (BAT).

Figure 3 Linear, USB standard charger with dynamic power path management

By connecting a power source such as an AC/DC adapter or USB port, the DPPM circuit continuously monitors the input current. The charger output can be regulated to 200mV above the BAT voltage. When the BAT voltage drops below 3.2V, OUT is clamped to 3.4V. This allows proper system load startup even with a discharged battery. The current flowing into IN charges the battery and powers the system at OUT. To include a variety of applications, the charging circuit needs to have selectable current limits of 100mA (USB100) and 500mA (USB500) for charging through a USB port, as well as a resistor-programmable input current limit that adjusts to different AC/DC adapters. The input current limit selection is controlled by the state of the EN1 and EN2 pins. When using the resistor-programmable current limit, the input current limit is set by the resistor value connected from the ILIM pin to VSS. When the IN power source is connected, the system load is given priority. Both DPPM and battery compensation modes are used to maintain the system load. Figure 4 depicts an example of DPPM and compensation mode.

Figure 4 DPPM and battery compensation mode (VOREG = VBAT + 225mV, VBAT = 3.6V)

Input voltage based DPM (IDPM) can be used for current limited USB port operation. When EN1 and EN2 are configured for USB 100 or USB 500 mode, the input voltage is monitored. If VIN drops to a certain input voltage threshold, the input current limit is reduced to prevent the input voltage from dropping further. This prevents the charger circuit from destroying an improperly designed or misconfigured USB power supply. When the sum of the charge current and the system load current exceeds the maximum input current set by EN1, EN2, and ILIM, the OUT voltage drops. Once the voltage at the OUT pin drops below the DPPM threshold, the charge current decreases as the OUT current increases to maintain the system output.

When no power is connected to the IN input, OUT is driven entirely by the battery. In this mode, the current into OUT is not regulated. However, a short circuit is required to prevent the system from overloading the charger.

in conclusion

Lithium-ion battery chemistry is well known for use in consumer applications. Many charging solutions go beyond just managing constant voltage and constant current charging. Today’s charging management circuits are faced with a multitude of protection functions, from input overvoltage and overcurrent to battery overvoltage. Dynamic input management and power path management are required to ensure compliance with various power interface and charging standards and enable the use of electronic devices in a variety of battery application environments.