Lithium Iron Phosphate (LiFePO4) batteries were introduced a decade ago and, in recent years, have begun to be adopted for mass production. LiFePO4, also known as LFP battery systems, are especially suited for applications such as electric vehicles (EV), hybrid-electric vehicles (HEV), electric bicycles, and power tools due to their naturally low impedance, resistance to thermal runaway, and ability to withstand a high number — about 2,000 — charge/discharge cycles.
One trend in applications that use LFP batteries is relatively high discharge-continuous and pulse currents compared with applications using conventional Li-Ion and Li-Polymer batteries. Additionally, applications that use LFPs typically involve multiple cells in series and parallel.
Some LFP charge-management systems are developed with microcontrollers and an external power train to charge the battery at a higher constant current or voltage. Other Li-Ion charge-management systems are either modified with a voltage divider or designed to charge at a higher reference voltage than the manufacturer's specifications recommend. Whatever the case, simple charge-management systems for low-power LFP batteries have not yet been widely developed. We will show designers how to affordably develop these types of systems, using LFP battery-powered radio-controlled (R-C) toys and UPSs as examples. The LFP Battery Chemistry Depending on the manufacturer, nominal operating voltages for LFP batteries range from 3.2 to 3.3 V. Most manufacturers require 3.6 V as a charge-reference voltage. LFP batteries have a similar charge algorithm to Li-Ion batteries in that they are charged with constant-current and constant-voltage (CC-CV) methods.
These methods begin with a set constant current. Most LFP batteries allow a charge rate of more than 1 C for a short amount of time. Once a reference voltage is detected, the state machine moves to the constant-voltage method. While a charge-management controller maintains the battery voltage, the charge current decreases until the minimum set-current is reached. Some battery manufacturers prefer a fixed minimum current, such as 50 mA, while others use a percentage ratio of battery capacity, such as 0.1 or 0.05 C. For proper operation, refer to the product datasheet or consult with the battery manufacturer for the maximum charge-current rating.
Fig. 1 illustrates a CC-CV algorithm that was implemented by a state machine in the charge-management controller. When the battery's presence is detected, the controller verifies the battery condition by voltage. When this voltage is above the preconditioning threshold, the battery moves to fast charge-current mode. Depending upon the design, the fast charge current can be either set by the resistor or a digital signal. When the reference voltage is met, the charge current decreases with the battery-voltage regulation until the charge is terminated. When errors such as overtemperature or overvoltage conditions occur, batteries can short, prompting the system to output a signal using a simple LED or LCD display to warn end users. If a recharge feature is required, the charge-management system monitors the battery voltage and only applies a charge when necessary to prolong a battery's life cycle.
Fig. 2 shows a typical charge profile of an 1,100-mAh-rated, single-cell LFP battery from a highly integrated, stand-alone linear charge-management controller. Typically, the maximum continuous discharge current for LFP batteries ranges from 5 to 30 C, and some batteries can produce more than 100 A of pulse current. Therefore, besides automotive and power-tool applications, R-C toys and UPSs can benefit from LFP batteries.
C-rate is the theoretical capacity of a battery as determined by the amount of active materials in the battery. It is expressed as the total quantity of electricity involved in the electrochemical reaction in coulombs or ampere-hours. The ampere-hour capacity of a battery is directly associated with the quantity of electricity obtained from active materials.
R-C CAR DESIGN EXAMPLE
R-C cars have traditionally been powered by gas engines or batteries. Battery-powered R-C cars are easier to control, and are therefore more popular. The required voltage range for battery-powered R-C cars is typically 6 to 9.6 V, which requires 5 to 8 NiMH battery cells with a nominal voltage of 1.2 V each. A fully charged NiMH battery may have a voltage of 1.5 V or higher. In this design example, the goal is to replace a 5S1P, 1,100-mAh-rated NiMH battery pack with a 2S1P, 1,100-mAh-rated LFP battery pack, where S is the number of batteries in series for the battery voltage and P is the number of batteries in parallel for available capacity. In the Table, we see that using an LFP battery enables the R-C car to be powered by a lower-power solution. This data, however, is only for reference; actual values may vary depending up the manufacturer, operating environment, and battery model selected.
In addition to their small size and light weight, longer cycle life is an attractive characteristic of LFP batteries. Fig. 3 shows a simple design solution for charging a two-cell LFP battery. A common 9-V dc power source is selected for its low cost and wide availability. For linear solutions, fewer voltage differences between source and load translates into less dissipated power.
The input voltage will be high enough for dropout and, when designing with a linear charger, the worst-case condition is minimum voltage in constant-current mode. With the MCP73223 charge-management controller example shown in Fig. 3, the charge current (I CHARGE) is 1,100 mA and typical supply current (IQ) is 0.7 mA. IQ is much smaller than ICHARGE, so it was omitted from Eq. 1.
PDISSIPATION =(9V-4V)-×1.1A = 5.5W (1)
Temperature = P ×θ JA = 5.5W × 43°C /W = 236°C (2)
The MCP73223 comes in a 3- × 3-mm DFN package and has thermal resistance of 43°C/W. The worst power dissipation that may occur in a linear-charger system is at the lowest battery voltage during fast charge. The 5.5-W power dissipation output in Eq. 1 will result in an increase of 236.5°C over ambient temperature in Eq. 2. This is not a temperature at which most systems can operate.
Therefore, the thermal-regulation function will kick in to ensure that the charger IC does not generate excessive heat. The thermal-foldback set point is 85°C for the MCP73223 charger, and it will only resume full charge speed when the temperature is reduced to below this point.
For additional safety, there is a second thermal shutdown at 150°C. With a charge rate of 1 C, the typical charge time for an LFP battery is about 65 min. (Fig. 2). The environment and power-train design topology may result in a different charge time.
When larger battery capacity is required, larger cells are available, such as those rated for 2,300 mAh or multiple cells in a parallel configuration. With this same charger topology, it will take longer to complete each charge cycle.
Fig. 4 shows a block diagram of a dual-cell LFP battery-charging cradle that can charge a single or multiple cells. While the MCP73223 can be used in standalone applications, the cradle's LCD display allows users to read the battery's charge status and other information.
DC UPS DESIGN EXAMPLE
A UPS contains an energy-storage unit, which is essentially a battery that powers the UPS in the event of a power outage. This energy-storage unit can also protect devices from power surges, spikes, etc. UPSs typically use a sealed lead-acid (SLA) battery chemistry; however the SLA battery's chemistry and power density make UPSs heavy and bulky.
Commercially available UPSs provide an ac output voltage produced by a dc-dc converter powering a dc-ac inverter. However, it is possible to design a dc-output UPS that is embedded with an electronic system. Dc-output UPSs are more efficient because they do not require a dc-to-ac inverter; instead, they only need an ac-dc converter.
Fig. 5 shows the block diagram of an internal dc UPS — whose output is 12 Vdc — that can be used for a small server or computing system. A power-management circuit will be available to convert 12 V to additional power rails, such as 5 or 3.3 V.
The example load requirement in Fig. 5 is 360 W, assuming a 90% efficiency from the boost-conversion circuit. A dc-output UPS has a single ac-dc converter, enabling an efficiency of 84% to 92% or higher. In contrast, an ac-output UPS has a dc-dc converter plus a dc-ac inverter so its efficiency is lower.
The higher the efficiency, the longer the UPS can operate from its battery power source. Therefore, efficiency and battery run-time are critical for UPS designs.
In addition, the time for the UPS to react to a power failure is important, and a few milliseconds can make a difference. The dc-output UPS can respond faster because it has less circuitry.
The reason for adding an internal UPS is to store necessary data and system settings before shutting down after a power failure. When power resumes, the data and systems that were present before the shutdown will remain in the same state. In Fig. 5, a dual-cell, 2,300-mAh LFP battery is selected as the energy-storage device.
You can calculate the required input power for a 90%-efficient ac-dc power supply driving a 360-W dc load:
360 W/90% = 400 W (3)
Although the LFP battery may be capable of 3.6 V per cell, a 3.2-V-per-cell nominal voltage is applied in this example. A dual-cell LFP configuration outputs a nominal 6.4 V, which yields an output current of:
400 W/6.4 V = 62.5 A (4)
Ideally, a 2.3-Ah battery charges at:
2.3 Ah × 60 mins = 138 A per min (5)
Under ideal conditions, this design will charge the battery at a rate of 138 A per min. And, it will take 2 min., 12 s to achieve the total of 62.5 A. That should be sufficient for 1 min. of dc UPS operation.
However, that battery run-time is not realistic because the battery's impedance may increase due to aging, environmental temperature, lifecycle, and how the battery is abused or treated. For a UPS with a 12-V output, tolerance is ±3%, which reduces the 12 V by:
12 V × 0.03 = 0.36 V (6)
Therefore, the dc UPS output should have a minimum voltage of:
12 V - 0.36 V = 11.64 V (7)
Additional LFP batteries in a parallel configuration can help maintain proper output-voltage levels. Although charge time typically increases proportionately with battery capacity, this is not a problem for UPSs. With the UPS design approaches presented in this example, designers can ensure proper system shutdown without losing important data.