Designing a Battery Pack That’s Right For Your Application – Technical Articles

Battery power is a must-have for modern devices—even plugged-in devices frequently have internal batteries to preserve critical systems when the power goes out. In today’s multitude of mobile and portable devices, the AC power cord has been relegated to battery charging.

Many device fires, product recalls, and explosions can be traced to improper battery system design. Also, many end users are frustrated by devices with insufficient battery capacity to make it through a typical day of use. Clearly, there are a lot of marginal battery systems out there in need of improvement.

Part of the problem is that battery capacity values can be misleading. Test methods and testing conditions vary between vendors, so it’s important to pay attention to the details of both. This article will hopefully give you the tools you need to interpret battery specifications and choose the right solution for your system.

Defining Power Requirements

The power consumption of electronic systems can vary greatly depending on application. At one extreme, supercomputers run on megawatts; at the other, modern electronic watches function on 2 to 6 mW. Determining exactly where a design sits in this broad energy spectrum is an important part of defining whether batteries can be used and, if so, what the battery strategy should be.

When defining the system’s power requirements, we need to consider the following:

• Use time.
• Peak current during use.
• Long-term average current (LTAC) over all use scenarios.
• The functional voltage range of the system, including the minimum and maximum voltages needed.

Use Time

Battery capacity is measured in terms of current and time—essentially, the duration over which the battery can produce a fixed current. Capacity values are usually given in ampere-hours (Ah). Ultimately, a battery pack needs to not only provide the correct amount of current—a requirement we’ll discuss shortly—but to continue providing it for the duration of the system’s use time.

Use time is straightforward—it’s simply how long you should be able to use the device before its battery requires recharging or replacing. Here are some examples:

• Use time of a cell phone: >20 hours.
• Use time of a smoke detector: >12 months.
• Use time of any ZigBee device: >2 years (mandated for certification).

For things like wireless mice or handheld remote controls, specifying the number of clicks between battery refresh events is also a possible approach. To determine the necessary battery capacity, we multiply the Ah used per click by a number of clicks we’ve defined as the minimum before battery replacement. The devices need to consume negligible current between click events for this to be valid, however.

Peak and Long-Term Average Current

Determining peak current and LTAC requires examining the transient current of the device (Figure 1). This can be done in the lab with a test fixture. Alternatively, an accounting of all active circuits can be manually estimated.

Figure 1. Four different power budget current profiles.

Depending on the system design, many different profiles of current consumption are possible. Case A, which shows a steady-state current, is the simplest scenario. Because the level of current is constant, the LTAC and peak current are equal.

Case B is common for many single-purpose systems. In these systems, brief bursts of high-current activity are separated by long, low-current sleep periods. The LTAC will be highly dependent on the time periods of active versus sleep and their respective currents. The peak current occurs during the active period.

Case C and Case D are typical of many multifunctional systems with mixed current use profiles. In both situations, various things turn on and off intermittently.

Prior to building hardware, we can base a first-pass estimate of the LTAC on the system’s high-current devices and their expected amount of active time. Electromechanical devices—motors, servos, and solenoids, for example—tend to be the biggest power consumers, but display lights, high-performance field-programmable gate arrays (FPGAs), and multi-core microprocessors can also dominate a power budget.

Voltage Range

The minimum and maximum operating voltages of the system electronics depend on the power system design. If battery power feeds directly to active circuits, the operating voltage is limited by those circuits. Many power systems use some form of voltage regulator up front, allowing a wider functional voltage range.

With power requirements defined, a battery can be fitted to the electronics. The voltage range of the battery also needs to be considered, as we’ll soon see.

Battery Discharge Versus Functional Voltage Range

Figure 2 shows a typical battery discharge profile with a fixed-current load. The voltage output of the battery slowly descends in a nonlinear manner as the battery discharges.

Figure 2. Change in battery output voltage over time.

Let’s break this diagram down.

• The discharge profile starts from A, which illustrates either a fresh, single-use battery or a fully charged device where the battery is topped off by the charger.
• Discharge quickly reduces the voltage to a region (B, C, and D) where a more linear discharge profile occurs.
• Discharge is considered complete at the knee of the curve (E).

Depending on its chemistry, discharge beyond E can sometimes damage the battery. Many rechargeable battery management systems protectively cut off at E to avoid this, though single-use batteries are often used down to F.

As the battery discharges, a progressively lower voltage is presented to the system. If the system can function down to E or F, it’s making full use of the battery’s energy supply. If the system stops functioning in the C to D range, the full capacity of the battery isn’t being used.

Improving Performance

To make full use of the battery capacity, the system electronics need to have a voltage range that includes the battery pack’s minimum and maximum voltages. If the system electronics shut down before hitting the battery’s V_MIN, you can:

1. Bring V_MIN up by using batteries in series. The number of batteries will depend both on the battery voltage and the minimum voltage needs of the system.
2. Enable the electronics to function at lower battery voltages by using a low-dropout (LDO) regulator or a boost converter in the power system.

Discharge Behavior of Batteries

Simple Ah ratings and nominal voltages can be very misleading. To make sure you select the right battery setup, it’s useful to look a bit deeper at how a battery discharges during use.

Battery discharge profiles can provide an expedient way to design a suitable battery pack. The curves in Figure 3 show the discharge profile of a typical AA battery for five different currents.

Figure 3. AA alkaline battery discharge current vs. use time.

These curves display Ah ratings between 0.9 Ah and 1.9 Ah. The vendor-stated capacity for this particular battery was 2.7 Ah, which should give you an idea of how misleading these values can be.

Higher-discharge currents generally produce a lower Ah rating, so vendors often state the Ah rating at a very low discharge current to get a better capacity number. As you can see, it’s important to seek further details regarding capacity testing and, if possible, examine a discharge curve plotted for a current value similar to the LTAC of your system.

It’s worth emphasizing that battery capacity isn’t a single, rigid number. To avoid battery damage, we need to include some margin and consider peak current along with LTAC. Table 1 shows how capacity varies with load current.

Table 1. Battery capacity vs. discharge current for single-use alkaline batteries (AAA through D).

The Importance of Temperature

Current isn’t our only consideration. As an electrochemical reaction, battery performance is very temperature-dependent. The thermal performance of the battery and the temperature of the application environment are therefore important design considerations. Consider Figure 4, which shows a drastic reduction in performance—from more than seven hours to less than one hour—as temperature drops.

Figure 4. Battery discharge curves showing use time over a range of temperatures.

Many vendors test with warm batteries (25 °C to 30 °C) to produce better capacity numbers. When designing your battery pack, you’ll need to take into account not just the battery itself, but also the temperature of your operating environment and how it differs from the testing environment.

Case Study: Wireless Thermostat

To cement what we’ve learned so far, let’s examine a case study. Figure 5 shows an example of the energy and lifetime calculations for a ZigBee wireless thermostat. This low-power device is a good application for single-use batteries.

Figure 5. Example requirements/specifications for a ZigBee wireless thermostat powered by single-use batteries.

In this example, a thermostat wirelessly reports over the network every 60 seconds. The device goes live for 250 ms, using 9.3 mA. It then sleeps, using 4.2 μA, for 60 seconds.

The LTAC is 43 μA, providing a use time of over seven years for alkaline AA batteries and over three years for AAA batteries. The ZigBee standard requires that devices have a battery life longer than two years, so both types of battery are providing acceptable use time.

The thermostat’s operational V_MIN is 1.9 V. However, the minimum voltage of an alkaline battery at full discharge is only 0.8 V. Using two batteries in series is a suitable solution.

Two batteries in series provide a combined minimum voltage of 2 x 0.8 V = 1.6 V, meaning that the batteries aren’t fully discharged when the system stops functioning at 1.9 V (0.95 V per cell). However, there’s ample margin on the two-year requirement if using AA batteries. If desired, a more careful analysis can be done using 1.6 V as the bottom of the discharge cycle.

Since this ZigBee thermostat spends most of the time in a micropower (4.2 μA) sleep mode, using a buck-boost converter to access the full energy of the battery isn’t recommended. The buck-boost converter would need to run all the time, greatly increasing the current in sleep mode.

High-Current Parallel Batteries

To achieve an appropriate use time, it may be necessary to combine batteries in parallel. When doing so, the current-resistance drops in the wiring need to be kept balanced so that the batteries provide equal currents to the load. The interconnect resistance and where it’s placed can affect battery load in high-surge-current scenarios.

Keeping the battery currents balanced for both high-current discharge and charging requires some attention to detail, as Figure 6 demonstrates.

Figure 6. Five different configurations for high-current battery connections.

Case 1 illustrates a common, if less than ideal, way of placing two high-current batteries in parallel to increase capacity and surge current capability. When a high-current surge is demanded by the load, the battery currents I_a and I_b won’t be equal because extra resistance is present from R_wire on both sides of the left-hand battery. In high-current situations, the resistance can be significant. Case 2 (good) fixes the surge current imbalance by making sure each battery has a single R_wire in series.

Case 3 is an extension of the problem in Case 1 to three batteries that will experience three different surge currents. Case 4 solves the problem by using tie-together bus bars to rebalance the resistive paths to all three batteries. Case 5 balances the resistive paths to all batteries without needing a bus bar tie point.

Finally, when combining batteries, keep them matched. For best performance, use the same age, type, model, and vendor.

Configuring Your Battery Pack Safely

Lead-acid and alkaline batteries must be individually purchased and manually configured into a connected array. However, battery packs that use lithium-ion (Li-ion) or nickel-metal hydride (NiMH) cells are generally assembled by a specialized manufacturer.

Due to safety issues with Li-ion cells, many manufacturers are reluctant to offer unprotected batteries on the open market. Instead, a common approach is to sell unprotected batteries only to recognized manufacturers of battery packs so that the devices are properly outfitted with safety circuits, charge balancing, and support electronics.

Many battery pack vendors offer a multitude of preassembled battery array modules with battery management and safety electronics built in. Some of these come with safety regulatory certifications, which can be a cost and time saver.

Safety testing is important. A large number of cell phone battery fires have been traced to the mechanical enclosure of the battery not allowing enough room for battery swelling. Careful electromechanical design review and safety compliance testing should keep your product off the evening news.

Key Takeaways

We’ve covered a lot of information in this article. Here are the most important things to take away from our discussion:

• Battery specifications aren’t standardized and are inconsistent between vendors. Consequently, designers need to carefully read the specification details and test conditions to determine meaningful battery capacity numbers.
• System power requirements can be estimated or measured by determining the long-term average current (LTAC).
• By fitting the battery discharge voltage profile to the power requirements of the system, full use of the battery capacity can be achieved.
• The chemistry of batteries can differ widely, causing corresponding differences in performance.
• Battery capacity changes with load and is not a fixed value. Higher current use results in lower capacity performance.
• Battery capacity also changes with temperature. Typically, cold batteries have lower capacity.
• Safety testing can prevent major problems down the road.

Further information on battery chemistry selection, charging methods, and supporting electronics can be found in the book “Applied Embedded Electronics” by Jerry Twomey.

All images used courtesy of Jerry Twomey