Power Management: Active Cell Balancing Improves Power Delivery in Large Li-ion Battery Packs

Some consumer applications require a single lithium-ion battery (such as a cell phone) or three cells in series and two cells in parallel (such as a laptop). This has led to the need for higher power, higher capacity, and more robust battery packs. Installing cells in series increases the voltage , while installing cells in parallel increases the capacity. These battery packs range from six cells in a laptop to hundreds of cells in an electric car, presenting battery designers with many new design challenges.

These large capacity batteries require advanced management to ensure a high quality design. Appropriate temperature, voltage, and current measurements must be considered. As Li-ion battery packs grow larger, more attention is required to thermal management, battery pack reliability, battery life, and cell balancing. In fact, as the number of cells required in a battery pack increases, the temperature, capacity, and series impedance differences between battery cells become a significant issue. This article will focus on the impact of these differences and how to control them in battery design.

Problem: Battery status does not match

The job of a battery is to store and deliver energy to its host. We want to store and get as much energy into and out of the battery pack as possible. The main thing that prevents a multi-cell battery pack from doing this is the cell impedance. Let's look at how that affects the delivery of power to the battery host.

In a Li-Ion battery pack, there are some predefined minimum and maximum voltages that each series cell is allowed to reach. This is a safety feature controlled by the IC in the battery pack, see Figure 1A. As long as each cell remains within the overvoltage and undervoltage disconnect ranges, the battery pack can be discharged and charged. If one cell reaches any of these thresholds, the entire battery pack shuts down (undervoltage), leaving the battery pack uncharged when it should be available to the host (see Figure 1B). In addition, it does not allow the charger to charge the battery pack as much as it should (see Figure 1C) (overvoltage).

Figure 1: The impact of cell imbalance on battery capacity usage.

There are many reasons why batteries are unbalanced:

* Non-uniform thermal stress

* Impedance variable

* Low battery capacity matching

* Chemical differences

Some of these causes can be minimized through cell selection and better battery pack design. Even so, the main cause of cell imbalance in all the upfront design work is non-uniform thermal stress. Temperature differences from cell to cell cause changes in impedance variables and chemical reactions. This creates temperature differences, and the cells are exposed to these differences for a long time (see Figure 2*). This is a notebook FL IR graph that shows the extent of temperature differences, even in consumer electronics applications. The self-discharge rate of a lithium-ion battery doubles for every 10°C increase in temperature. A characteristic of lithium-ion batteries is that the internal impedance is a function of temperature. Cooler batteries exhibit high impedance and therefore a larger IR drop during charge or discharge. This resistance also increases with exposure to high charge states and high temperatures for longer periods of time and longer charge cycles.

Solution: Cell Balancing Technology

Due to the impact on energy supply and the danger of overcharging lithium-ion cells in series battery applications, cell balancing techniques must be used to correct imbalances. There are two types of cell balancing techniques: passive cell balancing and active cell balancing.

Passive Cell Balancing Technology

A passive cell balancing method known as “resistor leakage” balancing uses a simple battery discharge path to discharge the high-voltage battery until all cell voltages are equal. Many devices offer cell balancing capabilities in addition to other battery management functions.

Li-ion battery pack protectors such as the bq77PL900 are used in many cordless battery-powered devices, power-assisted bicycles and mopeds, uninterruptible power supplies , and medical equipment. The circuit functions primarily as a standalone battery protection system using 5 to 10 series-connected cells. In addition to many battery management functions controlled through the I2C port, the cell voltage is compared to a programmable threshold to determine if cell balancing is required. If any particular cell reaches the threshold, charging stops and an internal bypass is activated. When the high-voltage cell drops to the recovery limit, cell balancing stops and charging continues.

Figure 3

Figure 4

Cell balancing algorithms that use only voltage divergence as a balancing criterion have the disadvantage of overbalancing (or underbalancing) due to the effects of impedance imbalance (see Figures 3 and 4). The problem is that the cell impedance also causes voltage differences during charging (VDiff_Start and VDiff_End). Simple voltage cell balancing does not distinguish between capacity imbalance and impedance imbalance. Therefore, this type of balancing cannot guarantee that all cells receive 100% of their capacity after a full charge.

One solution is to use a battery fuel gauge, such as the bq2084. They all have improved voltage balancing techniques. Because impedance differences between cells can mislead the algorithm, it only balances near the end of the charge cycle. This method minimizes the impact of impedance differences because the IRBAT voltage drop becomes smaller as the charge current tapers to the termination threshold. In addition, this IC also makes the balancing decision based on all cell voltages, so it is a more efficient implementation method. Despite many improvements, the need to rely solely on voltage levels limits the balancing operation to high state of charge (SOC) areas and only works when charging.

Another example is the bq20zxx family of battery fuel gauges, which use an impedance tracking balancing approach. Instead of trying to minimize the effects of voltage difference errors, this fuel gauge calculates the charge required for each cell to reach a fully charged state (QNEED), see Figure 5. This balancing algorithm turns on the cell balancing FETs during charging to provide the required QNEED. This type of battery fuel gauge can easily implement a QNEED-based cell balancing scheme because both the total charge and SOC are relatively stable and available during the monitoring function. Because cell balancing does not distort the cell impedance differences, it can work independently of the battery charging, discharging, and even idle state. More importantly, it achieves the best balancing accuracy.

Figure 5: QNEED-based cell balancing.

Since passive cell balancing techniques using integrated FET solutions have limited balancing capabilities, cell differences or imbalance rates can exceed cell balancing. Also, due to low bypass currents , it can take several cycles to correct for typical imbalances. Designing some external bypass circuits using existing components can enhance cell balancing (see Figures 6 and 7). In Figure 6, the internal balancing MOSFET is turned on first when a cell is decided to be balanced. This forms a low current path through the external filter resistors connecting the cell terminals (cell 1 and cell 2) and the IC pins . When the internal FET gate-source voltage develops across the resistor, the external MOSFET is turned on. The disadvantage is that neighboring cells cannot be balanced quickly and simultaneously. For example, if the neighboring internal FET is turned on, Q2 cannot be turned on because there is no current through R2.

Figure 6

Figure 7

Figure 7 shows a recent example of passive cell balancing. It is a low-cost, single- chip battery fuel gauge solution. Unlike the previously described battery fuel gauge solution, this IC does not have internal cell balancing, but requires a similar external bypass circuit to complete the balancing. However, since the balancing implementation circuit is an open drain internal to the IC, it can balance several cells simultaneously, including adjacent cells. This balancing circuit uses a modified voltage algorithm, just like the circuit shown in Figure 6. However, the external FET driver in Figure 7 describes a more effective cell balancing method.

Active Cell Balancing

Since 100% of the excess energy in high-energy cells is dissipated as heat, passive balancing is not the preferred method during discharge. Active cell balancing uses capacitive or inductive charge shuttling to transfer charge between cells, which is a very efficient method. This is because the energy is moved to where it is needed rather than being discharged. The trade-off is more parts and cost.

The patented bq78PL114 PowerPump cell balancing technology is the latest example of active cell balancing using inductive charge transfer. It uses a pair of MOSFETs (N-channel and P-channel) and a power inductor to create a charge transfer circuit between two adjacent cells.

The battery pack designer sets the imbalance threshold between the series cells. If the IC measures an imbalance that exceeds this threshold, it enables the PowerPump. Figure 8 shows a simplified buck- boost circuit using two MOSFETs (Q1 and Q2) and a power inductor. The top cell (V3) needs to transfer energy to the lower cell (V2), and the P3S signal (operating at about 200kHz and 30% duty cycle) triggers this energy transfer, which then flows through Q1 to the inductor. When the P3S signal is reset, Q1 turns off and the inductor energy level is at its highest level. Because the inductor current must continue to flow, the body diode of Q2 is forward biased, completing the charge transfer to the cell at position V2. Note that the energy stored in the inductor is only slightly lost due to its low series resistance.

Figure 8: Cell balancing using PowerPump technology.

Given the varying length and capacity of the series connected cells, there are some limitations on transferring charge. One consideration is how far can we move the energy before we no longer get energy delivery optimization? In other words, how far can we move the charge before the inefficiency of the converter outweighs the benefits of balancing the cells? Using an estimated efficiency of 85% in our testing, the PowerPump only transferred energy to less than 6 cells away. But the important point is that, regardless of efficiency, "area balancing" must be achieved before the entire battery pack can be completely balanced.

In addition to these obvious advantages, the benefit of PowerPump cell balancing technology is that balancing can ignore individual cell voltages. This means that if you decide to transfer charge between two cells, it can be done during any sequence of battery operating modes (charge, discharge, and reset). The transfer can be done even if the cell providing the charge has a lower voltage than the cell receiving the charge (for example, a lower voltage caused by lower cell resistance during charging or discharging). Compared to "resistance leakage" balancing, less energy is lost as heat.

There are three optional balancing algorithms:

* Terminal voltage (TV) extraction

* Open circuit voltage (OCV) extraction

* State of Charge (SOC) extraction (pre-balancing)

TV extraction is like the previously described voltage passive cell balancing. As shown in Figure 4, TV balancing during charge does not always produce a balanced charge toward the end of discharge. This is due to the cell impedance mismatch we mentioned earlier. OCV extraction compensates for the impedance difference by estimating the OCV based on the battery pack current and cell impedance measurements.

Fig. 9

Fig.10

SOC extraction works in a similar way to an impedance tracking device, it calculates the exact charge level of each cell and transfers energy between the cells so that the cells are balanced at the end of charge (EOC) (see Figure 9*). Looking at the discharge OCV graph (see Figure 10*), we pre-balance each cell to an offset voltage that reflects its capacity. A few percentage points of difference in capacity can make a huge difference in the lower middle of this discharge curve. If we know 1% to 2% of the capacity, we can have a very close match at the end of discharge. This is the area at the end of charge and end of discharge where you want to effectively get the cells optimally balanced using active balancing techniques.

Compared to traditional passive balancing techniques, PowerPump technology can better correct cell imbalances because higher balancing currents can be controlled by changing component values.

In laptop computers, the effective balancing current is typically 25 to 50 mA, which is 12 to 20 times that of internal bypass balancing. Taking advantage of this, active cell balancing can correct the cell imbalance within one cycle (95% of the time).

In larger capacitive cells, the PowerPump technology results vary even more. Consider the length of time a pack can be balanced when using voltage passive balancing. The only cell energy level is the positive balance that occurs during the discharge portion of a charge cycle. Therefore, only a few percent of the time in the life of a large pack allows for balancing. As a result, many pack designers choose to balance at 1 amp currents, or even 10 amps or more. This creates many thermal issues, as well as the cost of large FETs. These design barriers can be minimized if true uninterrupted balancing is possible with the PowerPump.

The choice of external components determines how much you balance the current. The peak inductor current is determined by the battery voltage, inductance, and on-time. The average current from the power battery over the entire cycle is equal to 0.5x(peak current)×duty cycle. In normal extraction mode, the duty cycle is 33%. For example: using a 15uH recommended inductor, and assuming a peak current of about 460mA, we get an average current of 75mA from the power battery. This 75mA current can be present for a long time. This keeps the entire system in a balanced state, so we exchange the most energy at the end of charging and discharging.

The question keeps coming up, “So how much balancing current do I need?” No one likes to hear the answer to this question, “It depends on several things!” First, know how much imbalance you expect to leak over time. If your system has a 5% imbalance after a 1-hour discharge of a 20Ahr pack, you have a lot of energy to move. The PowerPump FETs and inductors need to be sized accordingly. Alternatively, use the SuperPump option of the latest firmware. It allows you to have a larger duty cycle to move energy during normal mode when certain measurements are paused. As mentioned earlier, cell quality and thermal control are important prerequisites in determining how much balance you can achieve.

One safety benefit of active cell balancing is that we can track how long a cell has been in use. We can track the net draw of each cell, defined as the positive value drawn into the cell and the negative value drawn from the cell. If the net value of a cell is too high, it will result in too much energy being drawn from other cells, indicating that it is a bad cell. This is a component of the SOH calculation, similar to other parameters such as cell impedance and fully charged capacity.

Conclusion

Emerging battery technologies that focus on safety and longevity often feature advanced cell balancing and effective thermal management. Because new cell balancing technology tracks the balancing required for individual cells, the life and overall safety of the battery pack have been improved. Balancing the cells at each cycle avoids improper use of the cells, which is often the cause of more imbalance and early cell aging. The increasing diversity of battery chemistry, structure, and applications requires battery pack designers to upgrade their technology as well.