How to get the most out of your portable product battery

There is a trend towards increasing functionality or performance in the design of many portable products such as cell phones, smartphones, digital media players or digital cameras. This is usually achieved by using more powerful processors and adding more complex analog circuits, but the result is higher power consumption in the application circuit. The increasing power demand can be met by increasing the battery capacity, but this requires larger batteries or improved battery technology. Usually, people do not choose to increase the battery size because the case size is limited. As the current battery technology progresses and the development of new technologies cannot meet the high power requirements at the same size level, more advanced power management circuits are needed. At the same time, the demand for small solutions makes this challenge more difficult.

In the past, in order to obtain the required performance, it was enough to use a few linear regulators. These regulators were directly connected to the battery to generate the required system voltage rails. Many power management units used in portable products only use a few linear regulators to control the power consumption. The typical battery technology used at that time was a 3-cell NiCd or NiMH battery pack. At the same time, these chemistries have been almost completely replaced by single-cell lithium-ion batteries because of their higher performance. As current demands for many applications have increased, some linear regulators have been replaced by more expensive but more efficient buck converters. Some power rails, such as processor cores and I/Os, are often produced this way.

Since linear regulators and buck converters can only regulate the output voltage when their input voltage is higher, if the battery voltage drops below the programmed output voltage, the system needs to shut down. A linear regulator's minimum dropout margin or the dropout margin across the inductor and switch must be added to the output voltage. Therefore, for a typical 3.3V rail from a lithium battery, the typical battery voltage at which the system shuts down is 3.4V. The remaining energy when discharged to 3.0V is not used in this case. Measurements show that the remaining energy in a lithium-ion battery is currently about 10%. This means that any power management solution that can utilize this remaining energy must be able to operate at an efficiency that is higher than the efficiency of the buck converter solution minus 10%. In other words, any alternative solution to using a 97% average efficient buck converter must operate at at least an average efficiency higher than 87% to extend the application’s run time on a single battery charge. This is a significant challenge for many buck-boost converter solutions. The typical efficiency of a SEPIC or flyback solution is in the upper range of 85% of an economically viable solution. To achieve this efficiency, various efficiency enhancement methods such as synchronous rectification have been considered, while the size of such a solution would be larger than a buck converter. In a 4-switch buck-boost conversion, where 2 switches are always switched simultaneously, using this buck conversion would yield the same efficiency (85%) in a very optimized solution. Therefore, using a buck-boost converter does not work from this perspective and has not been considered in the past for this very reason. However,

there are some other challenges. For example, a mobile phone uses high current pulses to drive its RF-PA during data transmission. These pulsed currents can be drawn directly from the battery, which can cause battery impedance and additional voltage drops on the battery connector. This can cause the system voltage monitor to shut down the system when the current pulses occur due to the low supply voltage. LED-based camera flash applications in cell phones, or hard drive activation in media player applications, can have similar effects on the battery. These problems are exacerbated by the increase in battery impedance due to aging or low temperatures. In this case, a buck-boost converter can be used to cope with the voltage drop of the critical system voltage rail. This makes the system more stable and reliable while also allowing a lower battery voltage discharge. In addition to

this, batteries are also improving. Generally, increasing battery capacity is accompanied by the use of a wider output voltage range. For example, with future lithium battery technology, the battery can be charged up to 4.5V and discharged as low as 2.3V. With an intermediate voltage of 3.4V, it can leave a significant portion of the battery capacity unused. There are also battery technologies in the development stage that will work well below 3.4V (such as Li-S).

In this case, buck-boost conversion is definitely needed. A simple way to solve this problem is to generate a higher system voltage rail (such as 5V) that can be used to generate all system voltage rails, which are above the battery cutoff voltage. This can be accomplished by using a large, efficient boost converter and cascaded buck converters. The total power conversion efficiency can easily reach over 90%. Unfortunately, additional boost converters require more space, which is not usually available in portable handheld devices.

Another option is to use a buck-boost converter to generate the system voltage rail directly from the battery. As mentioned above, power conversion efficiency is a key factor in designing a competitive power management solution. Another important factor is the size of the solution. With this in mind, buck-boost conversion solutions based on SEPIC or flyback topologies are not suitable because they require more large passive components and are generally less efficient. A single inductor solution using 4 switches has the greatest potential to meet these requirements. However, in a simple driver solution, where 2 switches are active at any time during operation, using this solution not only sacrifices efficiency, but also increases the size requirements of the inductor and switches due to the higher RMS currents flowing through these components. Actively driving only one side of these switches means that always operating the device as a buck or boost converter achieves the highest efficiency, while lower RSM current also results in the smallest solution size. In this case, buck and boost conversions are accomplished at operating points where both topologies have the highest efficiency. This is shown in the example of efficiency vs. input voltage curves for boost (TPS61020) and buck (TPS62046) converters in Figure 1.

Figure 1. Boost (TPS61025) and buck (TPS62046) converter efficiency vs. input voltage.

Figure 2 shows efficiency versus input voltage for an optimized buck-boost solution such as the TPS63001, demonstrating a perfect implementation of this control method.

Figure 2. TPS63000 buck-boost converter efficiency vs. input voltage

As predicted earlier, when the efficiency curves of the independent boost and buck converters are discussed, the highest efficiency is achieved when the input and output voltages are close. Since this is the most likely buck-boost operating state, the TPS63001 perfectly solves some of the problems that arise in this application. As we can see in Figure 2, the optimized control scheme can achieve efficiencies in the 95% range at critical operating input and output voltage conditions. The TPS63001 can also be used to extend the operating time of applications powered by standard lithium-ion batteries. This goal can be achieved by discharging the battery to 3.0V or even fully charging it at 2.5V, as long as the integrated safety circuit allows. Figure 3 shows the power conversion efficiency of a battery discharged to 2.5V using the TPS63000. Compared to a power solution based on a high-performance buck converter (TPS62046), both converter architectures can extend battery life by 15% when using the same load.

Figure 3 Efficiency of buck-boost and buck converters during battery discharge