Power-management issues are fundamental concerns in the design of portable electronic equipment. They affect all elements that make a design portable, influencing device performance, run time, size, and weight. While power management encompasses many aspects of circuit design relating to power conversion and consumption, the topic always leads back to the power source—typically a battery. The portable device's performance depends heavily on the capabilities of available battery types, which very often are secondary (or rechargeable) cells.
For applications that are portable in the sense that they can be carried around, designers generally select one of three basic battery chemistries—NiCd, NiMH, and Li-ion. These batteries have found their way into cell phones, pagers, camcorders, PDAs, notebook computers, power tools, and a variety of other applications. Choosing from among the three chemistries is very often a tradeoff between size, performance, and cost (Table 1). When minimizing the size and weight of the battery is paramount, though, high energy density makes Li-ion the chemistry of choice.
Li-ion doesn't just offer the highest energy density. It also provides moderate to high power density, the highest charge retention, no memory effect, and a wide operating-temperature range. And, it isn't considered hazardous for use on airplanes or in disposal. Some additional safety concerns are associated with Li-ion cells, and charge termination requires a high degree of accuracy. But overall, charging Li-ion cells can be simpler than charging nickel-based cells.
Li-ion cells were originally developed by Sony for use in cellular phones. These cells were cylindrically shaped. The 14500 model, an AA-size cell, had dimensions of 14 (diameter) by 50 (length) mm. The 20500 model was 20 by 50 mm. Sony then developed the higher-capacity 18650 cell—now a popular size—as well as other cylindrical sizes tailored to various applications.
Prismatic Li-ion cells, which have a flat, rectangular shape, were developed in response to a need for thinner batteries that better fit cell phones, notebook computers, and other consumer portables. Cells that are 8 to 10 mm thick have become widely available, and it has been possible to obtain over 1000 mAh of capacity in a prismatic measuring 34 by 48 by 8 mm. More recently, 6-mm cells have arrived—albeit with battery capacity closer to 700 mAh—to satisfy the requirements of thinner products.
Industry efforts now focus on shaving cell thickness to 4 mm and below, making Li-ion cells as low-profile as many chip-sized components. To achieve this goal, battery developers are optimizing existing Li-ion cell designs, which rely on a liquid electrolyte. They're also developing new Li-ion cells, commonly known as Li-polymer, which exploit a polymer-style electrolyte.
In the established Li-ion technology, cells consist of negative (anode) and positive (cathode) electrodes based on intercalation compounds—materials with lithium ions mixed in. The anode starts as pure carbon but then has lithium intercalated into it on the first charge. The cathode is a lithium metal oxide where the metal is either manganese, cobalt, or nickel-cobalt. Energy storage and release depends on the interelectrode flow of lithium ions through a liquid electrolyte . The electrodes, which are wound in a "jelly-roll" fashion, are packed in a steel or aluminum can that provides internal stack pressure.
Li-polymer batteries replace the liquid electrolyte with a polymer in gel or solid form. The polymer electrolyte provides the required electrode stack pressure, so the metal can is no longer required and it becomes possible to enclose the cell in a foil pouch. A laminate of aluminum foil and plastic, the pouch occupies less space and weighs less than the metal can. The foil pouch permits greater room for active material, and it can be readily formed in a shape closely tailored to the space available in the application. Moreover, the foil can reduce the cell weight by 50% or more.
As prismatic Li-ion cells get thinner, the liquid electrolyte cell runs into difficulties. Somewhere in the neighborhood of 4-mm thickness, the metal can begins to take up too much space. Above this value, cell makers can play tricks like squeezing the jelly roll into a smaller can. Examples include the 6-mm liquid Li-ion cells PolyStor plans to introduce this year (Table 2). In contrast, the same company went with a polymer electrolyte design in developing its 4-mm-thick Li-ion cells, because at this size the can walls (0.4 to 0.5 mm thick) would occupy 20% to 25% of the cell's volume (Table 3). On the other hand, the foil pouch has a thickness of less than 0.15 mm.
The benefits of Li-polymer technology show up in production where assembly is simplified. Those steps associated with construction of a drawn metal can and its associated header are eliminated. Header assembly, laser welding the header to the can, sealing the can (with a ball welder) after the electrolyte fill, and shrink wrapping the can are no longer required.
All of the can assembly equipment is replaced with simple vacuum-bag sealing equipment. According to Marc Juzkow, director of marketing at PolyStor, this great simplification of the packaging process will make it easier and less expensive to reconfigure the cell production line for cells of various sizes and shapes. Initially, the new process will cost more than the standard Li-ion production. Eventually, however, Li-polymer will be much cheaper to produce than liquid Li-ion.
Several different systems are being used to construct Li-polymer cells. In some cell types, materials are wound into the jelly-roll shape, flattened, and then encapsulated. Meanwhile in others, the electrodes are stacked elements. The stacked approach allows construction of Li-polymer cells as thin as 1 mm or less, while permitting a simple scaling up of cell thickness to achieve higher battery capacities. One Li-polymer design developed at Matsushita Battery Industrial Co. (Panasonic Batteries) is based on a 0.5- to 0.7-mm stack consisting of two positive electrodes (a lithium-cobalt-dioxide cathode), one negative electrode (a graphite anode), and a gel polymer electrolyte. Battery capacity varies linearly with each stack contributing 100 mAh. This approach yields energy density comparable to liquid Li-ion cells (Table 3, again).
Another vendor, Thomas & Betts Corp., Memphis, Tenn., claims energy densities higher than Li-ion using a stacked Li-polymer construction. This technology, which also allows cells as thin as 1 mm, offers energy densities ranging from 275 to 340 Wh/l and 155 to 170 Wh/kg.
Using a solid polymer electrolyte design, Battery Engineering has developed a range of cells as thin as 0.7 mm with 100-mAh capacity. The company's bipolar stacked construction not only permits very thin dimensions, but also high-voltage output. In these designs, a proprietary separator material inserted between stacks lets the cathode of one stack and the anode of another be placed back to back (one above the other) in the stack. Along with the 0.7-mm cell, the company offers a range of sizes and thicknesses (up to 6.3 mm) with capacities up to 4000 mAh.
Some cells use a polymer separator between electrodes. Others, such as those designed with the Bellcore method, rely instead on an electrolyte that doubles as a separator. The separator provides electrical isolation between the anode and cathode, but it conducts the flow of ions between them.
Lithium polymer—whether in gel or solid form—is generally considered safer than liquid Li-ion. Li-polymer cells don't leak when they're punctured. As a result, simplifications (or elimination) can be made to the in-pack battery protection circuitry.
Li-polymer cells have already been designed into cell-phone applications like Ericsson's P28, a device that weighs just 80 g, counting the battery. Several other vendors are in the process of introducing their versions of the Li-polymer technology (Table 3, again). While it seems that Li-polymer cells will eventually replace liquid Li-ion cells, there should be some healthy competition between the two cell types in the short run. First, the polymer batteries usually do not yet match the liquid Li-ion in terms of energy density. Second, the metal can affords the Li-ion cell greater durability, which some designs may require. Third, the initially higher cost of Li-polymer may be a deterrent.
These considerations will be measured against the desire for thin size and light weight. Such factors will induce some designers to cut back on their demands for Li-ion capacity. Juzkow explains that for 6-mm Li-ion cells to be accepted in the cell-phone application, they must provide more than 700 mAh of capacity. Yet 4-mm Li-polymer cells that fall short of this goal will be acceptable because they're so much lighter than liquid Li-ion models.
The transition to Li-polymer technologies is an evolutionary process. While some companies are trying to bring Li-polymer cells to production right away and improve cell performance over time, others are taking a more cautious approach by introducing interim Li-ion technologies. NEC Electronics' combination of a traditional liquid Li-ion chemistry with laminate foil packaging is being used to develop a cell that is only 3 mm thick, yet maintains the volumetric energy density of existing Li-ion technology. The cell contains a lithium-manganese cathode, rather than the more common lithium-cobalt type. Other cells will be produced using this design, including a 6.5-mm cell with 264-Wh/l energy density (Table 2, again).
According to Michael Hasegawa, senior marketing manager of the Energy Devices Group at NEC Electronics, lithium manganese is more robust on overcharging and does not combust if the cell is punctured. Consequently, the company feels confident in putting its liquid Li-ion cell in a laminate package. He says that by using this hybrid Li-ion technology, it's possible to build cells as thin as 1.5 mm in various footprints with some decrease in volumetric energy density in the very thin form factors.
This degree of thinness will extend traditional Li-ion technology into the Li-polymer realm, further blurring the boundaries that distinguish the two approaches. Such distinctions about what's inside the battery, though, are secondary to designers. As Hasegawa explains, "It's not about polymer, it's about weight reduction and thinness." Nevertheless, hybrid approaches like NEC's are still considered stepping stones to true Li-polymer designs. And, the company intends to introduce Li-polymer cells that can compete favorably in performance with Li-ion. These cells will likely be in production in 2001.
With a thickness under 4 mm, the LT5A from GS-Melcotec is another cell that crosses over into Li-polymer territory by packaging standard Li-ion materials in a laminated film case. In safety tests, this cell has survived a number of destructive conditions, including nail penetration, crushing, oven heating, short circuit, and overcharge. Even so, this cell series is considered a step toward Li-polymer. The knowledge gained in the LT series will be put to use in developing the anticipated LY Li-polymer series.
Toshiba America Electronic Components has introduced a cell that falls on the other side of the hybrid technology fence. The ALB363562 is a 3.6-mm-thick Li-polymer cell that incorporates some Li-ion techniques, but details of its construction are sketchy. It incorporates a lithium-cobalt-oxide cathode and a mesophase-pitch-based carbon fiber anode, which are said to be similar to the electrodes used in the company's standard Li-ion cells. As for the other aspects of its design, Ritch Russ, sales and marketing manager at Toshiba, indicates that it's "not completely gel and not completely liquid." Yet it still resembles the gel types, since it won't leak if it's cut.
The cell distinguishes itself by its high drain rate—battery capacity drops off just a few percent at discharge rates up to 2 C—and a resistance to swelling. As a result of its safety performance, Toshiba expects to eliminate the battery's protection circuit, which adds about $1 to the cost of each cell. This is a step that other vendors of Li-polymer cells also are taking.
Battery Charging The task of charging the latest generation of Li-ion and Li-polymer cells remains essentially the same, requiring a constant current/constant voltage (CC/CV) scheme. However, new circuit architectures and trends are emerging to satisfy the Li-ion cell's demand for high-voltage accuracy (at least ±1% in the CV or "float" mode), while addressing the key considerations in charger design—heat management, size, and cost.Chargers are available as complete modular solutions from battery and power-supply vendors. They're also available as chip-based designs using Li-ion specific charging ICs from the likes of Maxim Integrated Products, Linear Technology, and Analog Devices. Li-ion battery chargers can be divided along the lines of circuit architecture (linear versus switcher versus pulse) and physical implementation (embedded versus external).
Typically, the charger circuit receives regulated (or loosely regulated) dc power from an external ac-dc power adapter—or perhaps a cigarette lighter adapter in automotive products. In some cases, the external supply may simply be a transformer and rectifier. In a linear charger, output from the adapter goes to a simple linear regulator circuit that often provides the smallest and least expensive solution. Unfortunately, the linear charger's low efficiency can generate significant amounts of heat. This can pose a problem in a compact embedded design and force designers to adopt a low charging rate. A switch-based charger costs more and requires more components, including a sometimes bulky inductor, but its high efficiency limits heat generation.
Pulse charging is an alternative method that achieves high efficiency without an inductor. It sends a stream of high-current pulses to the battery until the end-of-charge voltage is reached. This method provides faster charging rates (greater than the usual 1-C rate) without generating much heat within the charger. An IC implementation offered by Maxim Integrated Products relies on the use of a current-limited wall-plug adapter to set the cell-charging current.
Pulse charging is a new technique, so it is still gaining acceptance. Maxim's Karl Volk, senior corporate applications engineer for cell-phone power supplies, notes that there is some concern that pulse charging temporarily raises the cell voltage above the end-of-charge voltage. Some worry that this may affect battery life expectancy. Volk, however, observes that most battery manufacturers are comfortable with pulse charging of Li-ion cells.
Battery-charging implementations vary greatly from application to application, but certain approaches are gaining ground. Ed Myszka, director of strategy and business development at Motorola Energy Systems Group, Lawrenceville, Ga., notes a trend away from external charging systems toward integration of the charger within the portable device. In some applications, the external circuitry can be reduced to just a transformer. Myszka also points to an increasing use of switcher-based chargers as the costs of such solutions come down with advances in circuit integration.
Thomas Szepesi, product line director for the Power Management Group at Analog Devices, Santa Clara, Calif., notes that in the past there was great interest in building universal battery chargers that could handle multiple battery chemistries. But that interest has diminished for a number of reasons. For one, many types of end-user equipment are being designed exclusively for use with Li-ion batteries, and the battery packs used are commonly application specific with shape, thickness, and connector styles tailored to a specific design. For products such as notebook computers, this means the battery design may even change from one generation of product to the next. Another factor is that fewer battery packagers now offer compatible packaging across different chemistries.
Equipment vendors may be changing batteries from one design to the next. But in many cases, they may stick with the same power adapter, which not only supplies power for battery charging but also may power the device for extended periods of time when an external source is available. Available in wall-plug or desktop enclosures, these supplies generate single-voltage outputs ranging from a couple of watts up to about 70 W. These units offer a number of advantages over embedded power supplies.
Taking the power adapter circuit out of the equipment being designed allows for a smaller equipment design. It also transfers some of the burden in obtaining safety and EMI approvals from the equipment vendor to the power-supply vendor, and it simplifies requalification of a product when changes are made outside the supply.
As was the case with battery-charger circuits, linear versus switcher is the main distinction among power adapters. Linear-style adapters have a reputation for being low-cost but bulky ("bricks"). These supplies are still popular in many applications that require just a few watts. But switch-mode power adapters, with their smaller size, lighter weight, and greater efficiency, are gaining ground overall despite the fact they may cost more at the lower power levels. Additionally, emerging standards for power-supply efficiency provide further motivation for adopting switch-mode adapters.
In many consumer and business applications, customers will accept the additional cost of a switcher (perhaps a 20% premium over a linear) in exchange for its small size and light weight. It also offers advantages such as a universal voltage input and more flexible packaging options. Switchers can be made sleeker and less obtrusive, whereas linears are normally limited to a boxy shape because of their transformers.
This last point is important as more portable applications demand an adapter that blends into the design. Responding to this need, power-supply manufacturers such as EOS Corp., Camarillo, Calif., are getting more creative with their packaging (see the figure). Still, the main goal in adapter design is to make it smaller.
With this goal in mind, power-adapter vendors are developing some very-high-density switch-mode adapters with ratings of 10 W/in.3 and higher. According to Jim Schultz, executive vice-president of sales and marketing at EOS, such gains make it possible to replace a brick-sized linear supply that normally weighs 5 to 8 lb with a switch-mode version that weighs just 5 oz. Despite its small size, such a supply could produce 45 W of output.
These high power densities are made possible by the use of highly efficient circuit architectures that reduce requirements for heatsinking, through use of ASICs that reduce component counts, and by the use of high-quality capacitors and other passive components. For example, EOS reports that it achieves greater than 92% efficiency using a resonant-conversion switching technique.
Steven Willing, marketing director for imaging and mobile electronics at power-supply manufacturer Astec, indicates that design-automation tools are another element making higher- density adapter designs a reality. These tools let power-supply designers simulate electrical, thermal, and EMI performance so they can fine-tune their designs.
Nevertheless, there's a price premium when it comes to purchasing power adapters with the most power per unit volume. Willing cites as an example two 70-W switch-mode desktop adapters that he recently quoted. A version of this supply with just under 10 W/in.3 was priced at 25% to 30% above that of a low-density brick-shaped version with 2 W/in.3 Besides the tradeoff between size and cost, size can be traded for other parameters. A larger adapter design allows for better management of heat, which results in greater reliability.
New standards for green electronics also will impact adapter selection. To reduce the energy wasted by external power supplies that remain plugged in, the European Commission is drafting specifications for allowable standby power. In creating products for the European market, Willing advises designers to consider the current draft of specifications as guidelines (Table 4). Although these guidelines are still at the committee level, a formal agreement is expected shortly. Barring improvements in the linears' efficiencies, these specifications should lead to a phasing out of linear-style power adapters in Europe.
Such standards can be expected to provide added incentive for development of smaller, more efficient adapters. These advances, when combined with improvements in Li-ion batteries and associated charging circuitry, will help make portable equipment more portable while expanding the performance capabilities of these products.