Counterfeit or aftermarket batteries not held to rigorous standards can cause systems to fail, adversely impacting the original suppliers' reputations, as well as their revenue streams. The highly publicized battery recalls that have cost suppliers hundreds of millions of dollars in recent years help illuminate the significance of the role of battery management in success or failure in the portable systems market today.

The proper care and feeding of your battery subsystem involves being cognizant of a host of battery-related complexities and options for addressing them in the system design process. Not only do system designers need to accommodate varying battery types, address cell imbalance and protect their integrity against counterfeiters, but the realities of the marketplace demand that they do so cost effectively and with the flexibility to respond to changing demands.

So many battery types, so little time...
Battery technology itself poses complications to be managed during the design cycle. The various cell technologies commonly used today, such as rechargeable Nickel Metal Hydride (NiMH) and Lithium-ion (Li-Ion), have different profiles, i.e. different cell voltages and different charging requirements. The multiplicity of profiles makes it difficult to design a single system compatible with different popular battery types.

Another issue related to battery type is the so-called memory effect that has plagued rechargeable batteries, such as Nickel-Cadmium (NiCad) and to a lesser extent NiMH. Memory effect describes a familiar phenomenon to users of portable devices in which a battery that has not been fully discharged prior to recharging fails to subsequently recharge to as high a level and also subsequently discharges more quickly. Memory effect has been a focus for systems designers and battery technologists for years. It is believed that the remedy requires a combination of battery technology as well as intelligence in controlling the charge cycle.

Charge termination also presents several challenges for different battery types. Depending upon the cell technology, termination may need to occur at certain times or under certain conditions. Some cells have timer-related requirements. Yet others can be prone to overheating, so temperature must factor into charge termination. Overcharge can significantly impact battery life and even safety, and for this reason also, termination is a particularly critical system design consideration.

Management of different battery types requires intelligent monitoring and control of the various factors that impact the efficiency and safety of battery operation in a given environment. For example, cell temperature needs to be monitored to avoid overheating. The current profile can be monitored to detect roll-off that occurs when charging is complete. Likewise, voltage can be monitored to detect the so-called "voltage bump" indicating a maximum voltage at which full charge has been attained. In most cases, a combination of these charge termination methods is appropriate.

Because of the many and varied complexities associated with different battery types, a default approach to managing this complexity has been to create a custom or semi-custom battery system for each battery technology of interest. While this approach provides a simple solution for each battery type, there are obvious inefficiencies associated with developing multiple solutions for various technologies, some of which may or may not even see much use in practice.

Instead, a flexible system design that accommodates multiple battery technologies is an important option to consider. A platform-based approach in which a single intelligent battery charging circuit or system, triggered or programmed by a small detection device that identifies battery type, manages charge cycles for multiple types of batteries and affords the most flexibility in system implementation. The platform-based approach gives the system designer the agility to rapidly deploy different battery technologies depending on either end application demands and/or market prices for specific battery types, without the need for redesign. Thus, design effort is leveraged across many potential applications and cell technologies, and return on engineering investment is maximized.

A balancing act
One of the most significant factors that contributes to battery overcharging and hence compromising battery life and risking overheating, is failure to manage cell imbalance. Cell imbalance refers to the different states of charge that exist between different cells in a system. An inevitable attribute of multi-cell systems, cell imbalance occurs as a result of innate differences in charge states, total cell capacities and even cleanliness and uniformity of cell contacts between different cells in a system.

In a system that performs a simple charge of all cells, imbalanced cells will not reach full charge simultaneously, and the unchecked imbalance can have significant negative consequences. Cells at a lower state of charge at the beginning of a charge cycle experience higher voltages during the charge that accelerate cell degradation. Overcharging of cells that achieve a full charge first can induce thermal runaway, introducing the potential for overheating or even ignition. If overcharge mechanisms are in place, imbalanced cells can prematurely terminate a charge. Finally, imbalanced cells can reduce total battery voltage and decrease the efficiency of battery usage.

So, some intelligence must be applied to cell charge management in multi-cell systems so that all batteries attain the same voltage level simultaneously. This involves monitoring, on a cell-by-cell basis, to ensure voltage is equalized across all cells. Once charge is attained on any given cell, that cell must be bypassed so that premature termination and thermal runaway can be avoided.

Several approaches have been applied to the problem of managing cell imbalance. An inductive transformer can be used to transfer charge from the battery to the cell with low charge to balance or even out cell charge. However, inductive transformers are expensive and add size to the overall solution, making them impractical for most consumer applications. Another approach is to deploy a pass transistor in parallel with each cell to enable cell bypass when it reaches full charge. Such current bypass transistors are easy to implement, but lack intelligence, thereby limiting their ability to assure optimal battery management. An external microcontroller is an option for "smart" control of cell balance, but this approach also has its limitations, particularly with respect to current monitoring. By monitoring the current (xV), the total power (charge) going into the battery can be monitored to ensure that each cell is recharged appropriately. A microcontroller, however, measures the difference in pre- and post-charge voltage, a parameter that is only observable at the end of the charge cycle, perhaps after overcharge may have occurred on one or more cells. The voltage of the cell is not a good gauge to monitor battery charging. In monitoring the state of charge, a microcontroller assumes equivalent total capacity for all cells. However, this assumption is often flawed and can result in inaccuracies when gauging cell balance. In order to support current monitoring effectively, microcontrollers require the support of external devices, adding bulk and cost to a system, making them inappropriate and impractical for cost- and space-sensitive applications.

Another limitation of using microcontrollers becomes evident as the number of cells to be balanced increases. If the number of cells exceeds the I/O and analog channel capacity of the microcontroller, it becomes necessary to group cells or, alternatively, add or upgrade controllers. Managing cells as a group results is coarser granularity of control, and ultimately leads to battery degradation and other cell imbalance problems. Adding or upgrading controllers restores granularity of control, but can increase system cost and/or size significantly.

Cell balancing in today's systems requires flexible, intelligent control and monitoring of both voltage and current. An architecture that charges batteries on a cell-by-cell basis is critical. This requires intelligence of individual cell properties, close monitoring of cell charge via current and voltage levels, and flexibility to control a varying number and type of batteries. Programmable devices with on-board voltage and current monitoring capabilities can afford significant flexibility and extensibility as well as integration economies.

Protection from imposters
Counterfeit batteries have cost systems companies dearly, both from the standpoint of loss of reputation as well as from financial losses. Counterfeit batteries are not held to the same standards as those from reputable manufacturers. When deployed in systems, these batteries can fail and can even cause system problems or safety hazards that may not be traced back to the battery that was originally at fault. This results in an image problem for the original manufacturer, and even financial liability. In addition, counterfeit batteries are often sold at lower cost, undercutting the battery revenue stream for the original supplier whose superior batteries are more expensive.

In order to prevent the erosion of their reputation and revenue streams by counterfeit batteries, systems providers have employed a number of battery authentication strategies with varying degrees of success. Unique physical form factor was a common technique, but has fallen from favor because it is easily compromised. Electrical battery IDs offered an improvement, but still did not offer comprehensive security protection. An encrypted handshake is much more difficult to crack, and seems to be the most effective authentication technique today.

There are two common approaches to implementing encrypted handshakes for battery authentication. Some systems engineers implement the secure code in multiple discrete components, offering off-the-shelf convenience, but at the cost of increased system size. Others create specialty security chips that offer integration advantages but are inflexible if/when a code or system change is needed. Yet another approach is to secure the battery code in a mixed-signal programmable device. The programmable option offers the integration and footprint advantages of a custom implementation, but the flexibility to realize both standard and proprietary coding schemes and make changes without expensive and time-consuming redesign.

A flexible battery management solution
Achieving a single, cost-effective battery management solution that addresses the multiplicity of available battery technologies, complexities of cell balancing and increasingly sophisticated authentication techniques seems a daunting task. With programmable system chip (PSC) capabilities (Fig 1) such as those embodied in the Actel Fusion PSC solution, though, such a solution is feasible.

1. Features of a mixed-signal PSC.
(Click this image to view a larger, more detailed version)

PSCs integrate mixed-signal analog capabilities, flash memory, and an FPGA fabric needed for all the various elements of an overall battery management solution. The flexibility and configurability of a programmable array make it possible to support multiple battery technologies and up to 10 cells with a single PSC. A current-sensing resistor provides current monitoring capability, a diode/transistor junction keeps temperature in check, and digital elements support the control functions for cell balancing and termination. Differentiating system features such as proprietary authentication, communications and data logging are also managed flexibly and easily with a programmable architecture.

A fully-integrated 4-cell battery management implementation is shown in Fig 2. In this particular implementation, two chips and several external components have been realized in a single device that greatly simplifies the system board and the development process.

1. Basic 4-Cell smart battery implementation shows how a mixed-signal PSC integrates analog monitor and digital control capabilities, reducing component count, board space and cost.
(Click this image to view a larger, more detailed version)

Additionally, a PSC implementation offers cost reduction and system optimization beyond battery management. By integrating more functionality onto a PSC, overall system cost and performance can be improved further. For example, battery management can be implemented alongside power sequencing, power management, voltage detection and monitoring, and even thermal management functions in a fully integrated device. The capabilities embodied in a mixed-signal, programmable architecture are fully extensible within the entire system design space.

Summary
The proper care and feeding of your battery subsystem is no easy job. There are many complexities to manage. The most effective battery systems have the flexibility to support multiple battery technologies, the intelligence and capability to keep cells balanced and operating safely, the sophistication and features to authenticate batteries from credible sources and the agility to adapt to a changing environment.

Mixed-signal PSCs offer a low-cost, full-featured option for the implementation of these systems, as well as integration benefits that extend beyond the battery management realm. Organizations that take advantage of this technology are well positioned to maximize return on their development investment and respond more rapidly to the needs of their marketplace.

Ravi Pragasam is Senior Manager, Product Marketing, at Actel Corporation. Ravi has nearly 20 years experience in the semiconductor industry. Prior to joining Actel in 2004, he held various marketing positions at AT&T, Lucent, Atmel, Philips, and Xilinx. He holds a master's of engineering degree in electrical and electronics engineering from Madras Institute of Technology, India, and a master's of science in electrical engineering from Kansas State University.