Component selection and layout for smart-battery packs
Selecting and implementing the key components of today's smart-battery circuits present unique challenges and trade-offs. Analyzing your entire design ensures that it will reliably meet the application's requirements.
By Arun M Sanghani, Motorola Energy Systems Group -- EDN, April 14, 2005
Designing smart-battery packs requires a holistic systems approach. By evaluating the trade-offs between components, designers can develop the best battery pack for a host system. Selecting the proper components requires a thorough review of the specification requirements and component limits of the application because each of the battery components affects the performance of the entire battery. The key benefit of doing the job well is managing the cycle time and cost efficiency for the product-development cycle.
Because of their potential safety concerns, ensuring a reliable and safe design under worst-case conditions is especially critical when designing with lithium-ion batteries. These concerns include overvoltage and undervoltage of the cells and overcurrent of the battery pack. The safety circuit must limit both the charge voltage and the discharge voltage to avoid permanent damage to the cell. The safety circuit also protects against excessive discharge current. This article discusses how each component affects other areas of the battery design, including the pc-board layout. Analyzing a design as a total system helps to ensure that the design reliably meets an application's requirements.
A smart battery comes with specialized hardware that provides current, calculated, and predicted information to its SMBus host under software control (Figure 1). The key components of a smart battery are the connector, the fuse (F1), the charge and discharge FETs (Q1 and Q2), the cell pack, the sense resistor (RSENSE), the primary and secondary protection ICs, the fuel-gauge IC, the thermistor, the pc board, and the EEPROM or firmware for the fuel-gauge IC. The connector, fuse, FETs, cell pack, and sense resistor reside in the high-current path of the battery pack. Two protection ICs and a fuel-gauge IC control the FETs and the fuse; a thermistor provides temperature feedback of the cell pack; and the fuel-gauge IC uses an EEPROM for constants and parameters, such as pack configuration, serial number, and cycle life. All these components need the pc board to complete the circuit.
A typical battery pack may require more than one connector. The main pack connector is the mechanical and electrical part that interfaces the battery pack to the host system. In addition, the battery pack may require a connector between the cell pack and the pc board, or it can mount directly on the pc board using a hot-bar soldering process, although this approach can cause assembly issues in production with failures at the end-of-line tests.
Some critical items for the main pack connector are operating-temperature limits and pin assignment, orientation, and impedance. Connector operating temperature is critical, because many packs now require higher operating currents, which can generate excessive heat within the connector when a user discharges the high-capacity battery pack. Some designers mistakenly establish the pin assignments at the onset of design, but the impedance and the current ratings of the pins on the connector are critical in meeting the capacity performance and short-circuit thresholds. With most designs, designers must determine these parameters based on system requirements, which designers can do only later in a design. A design uses a connector between the cell pack and the pc board, so pay close attention to spacing between the two cell-sense traces for possible short circuits.
Cell-sense lines allow the primary and secondary protection ICs to monitor the health of the cells. The trace impedance of cell-sense lines can become an issue if the cells become unbalanced during charge and discharge cycles with higher currents. The cell pairs are out of balance when one pair charges or discharges more rapidly than another pair, which can happen if the sense lines' impedances cause markedly different current drains from the cells. Thus, the cell-voltage-sense line needs to be as close as possible to the cell to keep cells from becoming out of balance. Any cell tabs and wire interconnects need careful evaluation for the same reasons.
Selecting a fuse
A fuse is typically a three-terminal component that limits the current flow based on the temperature, current, or power across the heating element. Fuse selection is critical because it determines a secondary protection level for the battery pack in most designs. Other factors designers should consider are rated temperature, hold current, trip current, maximum pack voltage, and fuse size. Designers should also know under which conditions the fuse should activate for permanent failure if the specifications omit this information. Some systems must meet the Underwriters Laboratories limited-power-source requirement.
When a fuse trips, it passes the high current across the heating element. The designer can then review the worst-case timing and power to ensure that the fuse will trip only when the designer intends. On the pc board, the fuse should be far enough away from other components, such as hot-bar pads and other high-power parts, that it does not cause mechanical constraints and to prevent any manufacturing-process-related failures or nuisance trips. Working with the manufacturing or process engineer is useful when dealing with pc-board layout and component placement for optimum process and performance.
Selecting FETs
Designers base their selections of charge and discharge MOSFETs on the basis of the on-resistance, gate-to-source voltage, drain-to-source voltage, power dissipation, absolute maximum temperature, current ratings, package size, and cost. All of these factors involve trade-offs. For example, FETs with low on-resistances cost more but offer good thermal performance and battery capacity. Designers need to know the safe operating area of a FET for the worst-case application. They should perform a worst-case analysis of power dissipation and thermal characterization of the FETs under conditions of maximum continuous current and short circuit at the maximum operating temperature. In some cases, a FET's on-resistance may also play a part in total pack impedance. Some protection ICs use the on-resistance of the discharge FET to determine the trip threshold for the overdischarge current. In this case, the layout plays a part in when the FET trips for overcurrent.
To help with proper layout, designers should consider ESD (electrostatic-discharge)-protection capacitors for the FET. A good rule of thumb is to keep ESD capacitors close to the FETs. Designers should rate the body-diode voltage of the FET for worst-case conditions. Connecting both body diodes back to back means that they are reverse-biased. If designers fail to specify this part to handle the worst-case conditions, then the FETs can fail.
Designers select cells based on their chemistry type, cycle life, storage-capacity loss, shelf life, impedance, capacity at different rates of discharge and temperature, and mechanical and environmental requirements. Lithium-ion cells are available in prismatic and cylindrical form factors, each having different sizes and storage capacities. The mechanical design is usually the driving decision-maker. The cell-pack connection to the pc board requires evaluation for possible thermal issues. The cell pack has tabs that attach to the pc board and make the complete battery pack. The cells within the cell pack should have short interconnections between the cells. A shorter connection reduces heat loss, thus improving storage capacity. Cells usually connect using nickel strips, which manufacturers precision-weld—rather than solder—to the cell terminals or the case. Soldering is apt to apply large, uncontrolled amounts of heat to the battery components, which may damage the separators or the vents, which are normally plastic.
One of the primary functions of a smart-battery pack is to monitor and control cells to protect them from out-of-tolerance ambient or operating conditions. The primary protection IC provides cell-voltage measurements to the fuel-gauge IC and protects against overcurrent and short circuits. The primary IC senses the cell voltages and sends this information to the fuel gauge, which in turn commands the primary protection IC to open the FETs. The secondary protection IC has an independent connection to sense the same cell voltages to provide redundant overvoltage protection in case the primary protection IC fails.
The cell-hookup sequence—the physical connection order of cells to the circuit—is also critical. If the sequences of the primary and secondary ICs do not match, the fuse could activate during the cell hookup because the secondary IC sets off the permanent failure signal. Electronic protection circuits draw current from the battery pack, reducing the effective capacity of the battery pack to supply the desired load. Low quiescent current is therefore an essential requirement for protection circuits.
Determining the state of charge of the battery is the second major function of the smart-battery pack. The state of charge does more than provide the fuel-gauge indication. The battery pack monitors and calculates the state of charge of each cell in the battery to check for uniform charge in all of the cells, verifying that individual cells do not become overstressed. The state-of-charge indication also determines the end of the charging and discharging cycles. Overcharging and overdischarging are two of the primary causes of battery failure, and the battery pack must maintain the cells within the desired depth-of-discharge operating limits. A gas- or fuel-gauge IC determines when to turn off the FETs based on the cell pack's voltage, current, or temperature, and it also counts coulombs to track added or removed capacity. The usual requirement is that the fuel-gauge IC communicates using a standard protocol, such as SMBus 1.1. The firmware of this IC is critical with any EEPROM or data-flash parameters that affect the operation of the pack with the host system. Running separate sense lines to the cells and sense resistor is critical to obtaining accurate voltage and current sensing. Short circuits can produce currents as high as 80 or 90A. The current across the sense resistor can produce a voltage that can damage the fuel gauge.
Designers choose sense resistors based on their rated power and temperature, value, tolerance, and package size. The level of short-circuit current necessary to ensure the fuel gauge's coulomb-counting accuracy determines the resistor's value. Rated power and temperature are critical because, during short-circuit and high-current discharge, this part runs heavy, heat-generating current through the resistor. Designers should pick the lowest value possible without sacrificing the coulomb-counting accuracy of the fuel gauge. Tolerance of the sense resistor is critical because it affects the accuracy at temperatures higher than operating-temperature range, and it should be less than 1%. To minimize the power loss and increase the efficiency of the battery pack, the sense resistor's value must be as low as possible without affecting the resolution of the fuel gauge, which requires a certain level of voltage to the sense resistor. Designers must determine the power dissipation of the sense resistor during a short circuit to ensure that the sense resistor is rated properly for maximum power.
Not just any thermistor
The system's battery charger can use a thermistor to determine starting temperature and prevent charging if the battery temperature is too low or too high. Designers should note the thermistor's resistance versus temperature to determine whether it is acceptable to the host system. The battery charger also uses the thermistor as an external thermal sense that provides input to temperature sense for the fuel gauge. In many cases, the fuel gauge uses its onboard internal temperature sensor. The designer needs to check that the constants in the fuel gauge for temperature correspond to the thermistor specifications. These constants determine the impedance values for the fuel gauge.
The placement of the thermistor on the pc-board layout is critical for ensuring that the thermistor is sensing the proper component for temperature. Check with the pc-board designer to make sure there are no other physical limitations on its placement. Battery packs operate in a limited temperature range. Attempting to use them outside these limits usually results in a permanent degradation in performance or complete failure. The battery specification should therefore stipulate these limits. Note that the actual working temperature of the battery is not the ambient temperature but some higher temperature depending on the heat that the battery application generates and the heat that conduction and radiation remove.
Other guidelines
Critical-part placement and signal routing are important for accurate current and voltage sensing. High-power analog components, such as the fuse, the FETs, and the sense resistor, should be close together to improve thermal performance. Placing copper near the FETs improves the FETs' power dissipation. Voltage- or current-sense lines should be away from any high currents to reduce noise. Routing clock and data lines is also critical for reducing noise. Protection spacing rules also add limitations to how designers place the parts. The heavy charge/discharge-current trace impedance should be minimal for better capacity. Also, negative and positive cell traces need to have short and wide traces from the cells up to the sense input of the protection IC; otherwise, issues could arise concerning cell balancing and reported cell-voltage accuracy.
Flash memory stores the cell-pack parameters, which the fuel-gauge IC can access. The system host can control the charge and discharge operation of the pack by using these parameters. Parameters for charging include the charge-termination voltage, the charge termination, and the current- and charge-inhibiting temperature. Start with default values and then understand the application to optimize the battery pack's function in the system. Create a table and make sure to thoroughly check the firmware, including the default values that the chip suppliers provide. In some cases, the values may not work as you expect, and the battery may have some firmware issues.
Selecting the proper components for smart batteries involves thoroughly reviewing the specification requirements and component limits of an application. Each of the battery components impacts the performance of the entire battery. Taking time to design in the proper component reduces both cycle time and costs and re-spins in the product-development cycle. By holistically approaching battery-pack design and evaluating the trade-offs between components, designers can develop the best battery pack for the host system.
