Strategies for conversion from lithium-ion to primary battery power
Rechargeable lithium-ion batteries are widely used in today’s portable electronic devices; particularly those demanding either significant amounts of power or a heavy usage frequency. When designing with lithium-ion, safety should always be on the forefront of the design engineer’s mind to ensure that proper care is taken—especially during charging. Lithium-ion safety considerations frequently result in a battery that is specially chosen for the given application and is tightly integrated with the system. Making future changes to that system can then be challenging and time-consuming.
If this is the case, why would a designer consider changing the power source? More often than not, the reason for converting comes down to a need to reduce cost. This may be to drive increased margins, to be more competitive with substitutes, or to reach a new or expanded customer base at a much lower price point. Converting from rechargeable lithium-ion to a single-use primary battery can be an attractive place to reduce costs. In addition to the price of the battery itself, lithium-ion designs require sophisticated circuitry to closely monitor charging and discharging, which is unnecessary if primary batteries are used.
Costly redundant systems may have even been employed to intentionally over-design for safety. Aside from bill-of-material costs, additional shipping and labour costs may be required to process returns and repairs. In addition to any cost considerations, the power requirements for modern electronic components have continued to decrease, and the usage of an extreme low power microcontroller in applications that were previously well suited for a rechargeable solution may now enable a primary battery solution instead.
In general, primary batteries that are designed for single use offer some inherent advantages over rechargeable lithium-ion batteries in portable devices. For example, if the device is used often, widely available primary batteries can be quickly swapped out for fresh ones with minimal downtime. There is also no need for packing bulky accessories, such as cables and charging adapters.
Challenges with conversion
Any combination of reasons may lead you to consider designing around a primary battery, but you may be hesitant to switch because of the aforementioned highly integrated power supply. It may then be necessary to first prove a primary battery, which fits the device’s form-factor constraints, can even handle its potentially high-power requirements. Although the best prescription for long battery life will almost always be to redesign the power system from the ground up, there may first be a need for an intermediate design that quickly retrofits the existing lithium-ion design, with the various primary-battery solutions under consideration. This design can then be used to quickly and cheaply determine feasibility, while estimating minimum performance.
Designers looking to retrofit an existing lithium-ion-powered device with a primary-battery solution are likely to confront a number of technical issues, and the first one may be voltage. Lithium-ion batteries typically operate between 3.0V and 4.2V; whereas primary alkaline and lithium batteries in the AA form factor typically operate between 0.8V and 1.6V per cell (see Figure 1).
Figure 1: Lithium ion, AA lithium, and AA alkaline voltages during discharge
Lithium coin cells would offer a 3.0V supply; however, they will not be discussed here as they lack the current-sourcing capability that is often needed in applications that have been served by lithium-ion. Therefore, this intermediate design will likely necessitate the use of multiple cells in series, a DC-DC voltage converter, or both, so as to operate within this voltage window. It is also important to understand the range of voltages with which the device will operate, as it is possible that voltages outside of the lithium-ion range could be acceptable.
This leads to the second potential issue that must be addressed—matching the current needs of the device with the sourcing capabilities of the battery and the operating ranges of the DC-DC converter. When it comes to batteries, the amount of current capacity they can provide is closely linked to their internal resistance, which can change throughout discharge, as shown in Figure 2.
Figure 2: AA lithium, AA alkaline and lithium-ion internal resistances
As previously mentioned, coin-cell batteries may be able to meet the requirements for size and voltage, but fall short if more than tens of milliamps of current are needed, due to their high internal resistance that leads to large drops in voltage when higher amounts of current are drawn. Although the data in Figures 1 and 2 depict AA size batteries, the characteristics will be the same for the AAA or AAAA sizes. Choosing the right size for a given application is a balance between available capacity and size.
Once a battery has been chosen that provides adequate voltage, current and capacity, the electronics must also be selected appropriately. Generally speaking, the more voltage boost provided through DC-DC conversion, the less current it will be able to deliver. Thus, it is important to consider the application, the battery and the power-conversion design as one complete system where tradeoffs may need to be made.
next; potential solutions and strategies…
Potential solutions and strategies
The low-power, single-cell design
For low-power applications, a single-cell battery with a simple boost converter can be used to replace the rechargeable lithium-ion battery and associated battery-protection circuitry, leading to lower total system cost. This design can provide long runtimes while keeping size and weight to a minimum. Various 1.5V battery sizes can be used, such as AA, AAA or AAAA. If size is most important, AAAA alkaline batteries provide good runtimes in a very small form factor. If long shelf-life, cold-temperature performance, leakage resistance, or energy maximisation is most important, primary lithium batteries available in the AA and AAA sizes that have a 20-year shelf life, perform exceptionally well even down to -40°C, are leak-proof, and deliver the highest energy at a high and stable voltage, making them well suited for boost-converter applications that require constant power.
A single-cell solution requires a boost converter to step the single-cell voltage range up to a usable, regulated 3.0V or 3.3V system input-voltage range. Readily available boost converters, such as the example from Figure 3, integrate the entire switch-mode power solution, including compensation components, and provide the system with a low-current shutdown (< 1 µA) and low-current standby (<14 µA), along with the ability to deliver up to 100 mA of continuous current during operation. Additional integrated features block the typical boost-converter input to output DC connection and provide recoverable output short-circuit protection, as shown in Figure 4.
Figure 3: Single-cell MCP16251 boost converter
Figure 4: Showing blocking path and current-limit function
The high-power, two-cell design
For higher-power applications, two cells in series can be used, providing longer run time and higher current. As an example, Microchip’s MCP1643 high-current boost converter can regulate up to 500 mA from two cells in series for high-power LED applications, while providing true disconnect shutdown, current limit and short-circuit protection (see Figure 5). A high (1 MHz) switching frequency enables small inductors and capacitors, reducing size and cost. The MCP1643 can also be turned off, minimising battery drain to less than 1 µA. A regulated current source can be developed with a single integrated device, from two primary lithium batteries (see Figure 6). A low, 120 mV Vref is used to maximise system efficiency. Optimum performance is enabled by their high and consistent voltage profile across a wide range of discharge rates.
Figure 5: Two series cells-powered MCP1643 application, shown driving high-power LED
Figure 6: Output current capability of two-cell + MCP1643 solution
The buck-boost, three-cell design
For some applications, the input-voltage range can be greater than or less than the desired output voltage. This can be true for either primary or rechargeable batteries. For example, lithium-ion battery voltage can range from 3.0V to 4.2V throughout the charge and discharge cycles. Three alkaline batteries connected in series will measure about 4.5V when fresh, but as low as 2.4V when depleted. Similarly, three primary lithium batteries connected in series will measure about 5.1V when fresh, but about 3.6V when depleted. The application, however, may require a regulated 3.3V rail, which suggests that a power supply with buck-boost capability is needed for this configuration.
A SEPIC converter is capable of stepping the input voltage up or down, to regulate the 3.3V system voltage. In addition to providing system regulation, the primary-battery SEPIC design doesn’t require protection circuitry or back-to-back P-channel switches to protect the battery. Instead, an isolation capacitor provides input-to-output shorted switch protection and the control IC, such as an MCP1632, can provide short-circuit-output and current-limit protection with minimal external components.
For low-power standby applications, a parallel MCP1700 1 µA IQ LDO can be used in parallel with the SEPIC power train, as shown in Figure 7. With the SEPIC power train disabled, the MCP1700 will provide a “keep alive” voltage rail while consuming 1.6 µA of battery current with no load.
Figure 7: MCP1632 application diagram with 3 series batteries and optional MCP1700 @ 2.0V
next; the 4-cell design…
The Pure Buck, Four-Cell Design
For applications with high peak-current demands or long run-time requirements, additional series cells can be added to increase the input voltage of the system. Increasing input voltage reduces input current, resulting in smaller size and lower cost for the connectors, wiring and distribution switches. A low-power, integrated synchronous buck converter can step down four primary cells, in series, to a nominal 3.xV system voltage, with an input range of 3.6V to 6.4V. The MCP16311 buck converter integrates the entire switch-mode power system, including compensation. This reduces system cost and size while operating over a wide input-voltage range and reducing input current, using an auto-detected low-power mode when the system load is inactive. A parallel approach can be used to further reduce system input current, by using the MCP16311 EN input to shut down switching while powering the system with a low IQ LDO, similar to the MCP1703, during sleep or inactive periods, as shown in Figure 8.
Figure 8: MCP16311 with parallel MCP1703
The intermittent, very high power, battery and capacitor hybrid design
There are many situations in embedded designs that require infrequent, but very high, peak currents to operate. Consider, for example, a portable medical device with a display and an internal motor or pump. Assuming the motor or pump runs intermittently, the current consumed could be several orders of magnitude higher than the typical operating drain. Depending on the load or resistance, the current required from the battery could easily exceed 1A. Consider another example of an embedded device with a high-power radio, such as an 802.11n transceiver or a cellular modem. It may generally draw very low amounts of power, but when the device needs to transmit data over the network, the system needs to allow the radio to power up and transmit, which will briefly consume large amounts of current.
When considering how to power this type of application, it is important to understand the internal-resistance levels of various battery systems. Lithium-ion batteries typically have low internal resistance and are capable of handling very high current spikes. Within primary batteries, the internal resistance is somewhat higher. When alkaline and primary lithium batteries are fresh, they have similar and comparatively low internal resistance. However, as alkaline batteries are discharged, their internal resistance will gradually increase. Thus, when large amounts of current are required, the voltage on a stack of alkaline cells will drop precipitously as the required current rises. Conversely, the internal resistance of primary lithium batteries will remain low throughout discharge, and will provide longer runtimes in applications with high-current demands.
To overcome the voltage drop, especially if alkaline batteries are used, an energy reservoir needs to provide additional current when needed. A capacitor is the natural element to use, and a wide range of sizes and capacitances is available. With values up to and above 1F available in the lithium-ion voltage range, there remain two additional challenges in implementing a solution. First, the capacitor will have very low internal resistance—particularly when empty—so a current limit needs to be placed between the primary cells and the capacitor. Second, introducing a current limit assumes an intentional resistance that needs to be controlled and managed to keep the efficiency high. A microcontroller with integrated analogue circuitry that controls an external MOSFET solves both of these issues. The remaining design challenge centres on sizing the capacitor and MOSFET.
Figure 9 illustrates the blocks needed to control and protect the capacitor, and which items are integrated into a microcontroller. A suitable microcontroller will contain an internal oscillator, A/D, comparators and op amps, so that the external components needed are limited to just the MOSFET.
Figure 9: Block diagram of microcontroller, capacitor and battery
Figure 10: Flow chart of software algorithm
To design the solution, two elements need sizing:
● Capacitor value
● Current limit
For a design example, the assumed current peak will be 1A for a maximum of 500 msec, with a minimum repeat rate of 5 seconds between peaks.
The capacitor needs to keep its voltage value between 3.0V and 4.2V, during the 1A discharge of 500 msec, and recharge in 5 seconds.
Based on the capacitor discharge equation and the voltage assumptions above, the following equation sizes the capacitor:
C = t*Ipeak
A 0.5F capacitor provides the necessary reserve to deliver 1A for 500 msec.
To recharge the capacitor for the next peak current, the circuit must charge the capacitor back up within 5 seconds. The value of R will be determined by the following equation, which is, in turn, based on the charging-capacitor equation and the terminal-charge value:
R = -t/(C*ln(1-V0/Vb))
Solve for t = 5 sec, where C is the capacitor value already determined, V0 is 4.2V, and Vb is the nominal battery voltage. In this example, the R value solves to 3.7Ω. The current range peaks at 406 mA and drops to 81 mA, during the five-second charging window.
The circuit needs to monitor the capacitor voltage. It turns on the switch when the voltage drops below 4.2V, and turns off the switch once the capacitor voltage matches the battery voltage or arrives at 4.2V. Figure 10 illustrates the algorithm for the firmware in the microcontroller. As shown, the firmware complexity is very simple, leaving room for other functions. For example, a smarter version of this firmware could modulate the switch to achieve a constant current over the five-second window, to lessen the impact on the battery stack.
Conclusion
One of the many design challenges facing engineers today is the choice of which battery technology to use. Many designs use lithium-ion cells when the power requirements and the rate of usage are high. Other designs can attain long life with primary cells and take advantage of some of their less technical benefits, including:
● Lower bill-of-material cost
● No charger solution required (external and internal)
● Lower weight
● Instant use without waiting for recharging
Converting a lithium-ion design to primary cells is sometimes an intermediate step in the design process, and can be accomplished with the example solutions described in this article. These solutions are listed in order of increasing current output and complexity. There are, however, many alternative solutions, since the use cases for batteries are as varied as the designs that use them. Designers need to select a solution that is ultimately based on the needs of their end applications. The methods outlined above are meant to provide just a few approaches from which to begin.
About the Authors
Terry Cleveland is Manager, Design Architecture & Applications Engineering, at Microchip Technology.
John Haroian is Senior Field Applications Engineer for the USA Northwest Region, at Microchip Technology.
Adam Jakubiak is Senior Technology Engineer at Energizer, where he assists a wide variety of device manufacturers in selecting the right power source for their application and understanding how to get the most out of the battery chosen.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
