Page 144 - Lindens Handbook of Batteries

BATTERY DESIGN 5.21

that reduces the charge current is enabled until the temperature or cell voltage increases above the
preset thresholds. Similar level-based conditions are often used for charging.
More sophisticated calculation can be required for accurate gas-gauging to represent battery
state-of-charge, state-of-power (the battery’s ability to provide high discharge currents), or state-
of-health (battery cycle life information). Although some chemistries can provide relatively use-
ful gas-gauging information by monitoring the open-circuit voltage of the battery, most lithium
rechargeable chemistries require more sophisticated approaches.
Calculations transform any measured data through the use of simple or complex algorithms,
depending on the host device application’s requirements and the chemistry used. Prior knowledge
of battery characteristics, such as capacity at various discharge loads and temperatures, charge
acceptance, self-discharge, etc., are required to determine future battery performance. Early bat-
tery electronics used simple linear models for these parameters, which severely limited accuracy
in predicting the battery’s performance. As noted in the descriptions in various chapters in this
Handbook, battery performance is often very nonlinear. Self-discharge, for example, is a complex
relationship influenced at least by temperature, time, state-of-charge, and other factors. Further,
the performance of even those batteries using the same chemistry varies with design, size, manu-
facturer, age, etc. A good calculation engine and algorithm will account for these relationships
and help assure safe, reliable operation.
Calculations can also be used to maximize the performance from the battery pack during
actual use by considering calculated values, such as cell impedance along with voltage, current,
and temperature measurements. Techniques that operate the cells to the edge of their performance
envelope require precise measurements but also well-known models of the cells’ characteristics
and performance under various usage conditions. Processors to perform these measurements and
calculations in real time, while under heavy loads, can maximize the performance obtained from
the cells in the battery pack. High-end power tool products and hybrid and electric vehicles often
use such sophisticated calculations.
As with monitoring and measuring, properly matching the calculation requirements with the
battery chemistry’s needs and the end-application requirements is critical to a high performance,
low cost, reliable design.
3. Communication. Just as measurements may vary between exact values and threshold monitor-
ing, communications can range from detailed measurement data over a communications bus to
a single line “go/no-go” signal that indicates that the battery pack is operating outside of preset
limits.
Battery packs have for years used a single interface line to represent the temperature of the
battery via the voltage on the line. The voltage is a representation of the pack temperature with
a negative-temperature-coefficient (NTC) thermistor. The resistance of an NTC temperature
sensing device located in the battery pack is monitored externally, often by the charger. Low
resistances represent high temperatures and vice versa. Nickel-based chemistries often use this
signaling approach to detect end-of-charge via a change in the rate of rise of the temperature.
This same approach can still be utilized by chemistries that do not exhibit any temperature
changes with full charge. For example, a common technique with lithium rechargeable chemis-
tries is to simply mimic the temperature of a “hot” pack, which signals the charger to terminate
charging.
When more information is to be conveyed between the battery, charger, and host device, a
digital interface, such as the Inter-Integrated Circuit (I2C), Serial Peripheral Interface (SPI), DQ/
HDQ, 1-Wire, or Systems Management Bus (SMBus) protocol, is often used. These are stan-
dardized data communications interfaces with low power characteristics well suited for battery
applications. The electrical and data protocols are defined and available in many prepackaged
parts for use in battery packs, chargers, and end-equipment devices. Automotive battery systems
may utilize Local Interconnect Network (LIN) or Controller Area Network (CAN) bus interfaces
for additional robustness.
Information that is often communicated between the battery and charge includes the required
charge conditions, such as maximum charge current, maximum charge voltage, and perhaps maxi-
mum temperature to initiate charge separately from a maximum charge continuation temperature.