Brains and Brawn

The Power of Smart Batteries

Legions of wireless products and industrial appliances have built-in rechargeable batteries. Embedded intelligence can minimize user intervention, reduce charge time, and dramatically increase battery life by optimizing power management. Are you designing a battery-powered product? It’s easier than ever to make it battery friendly and hassle free.

Perry S. Marshall, Perry S. Marshall & Associates

photo

Have you ever been typing away on your notebook computer when it suddenly shuts off without warning? Not only is it annoying, it’s extremely counterproductive. Nobody likes an unpredictable power source.

Making power sources predictable is what smart battery technology is all about. How many minutes of power remain before your PC must be plugged in again? If the hard disk starts up, will enough current be available to finish the task? Fortunately, portable computers can answer these questions on a continuous basis. But what you might not know is that your PC is probably not determining the state of the battery by itself. Instead, sophisticated electronics and sensors inside the battery provide that information.

 
Terminology

Cell: This is the smallest chemical unit in a battery. Batteries are generally constructed with several cells connected in series, parallel, or a combination of the two.

I 2C Bus: This two-wire network was developed by Phillips to transmit data between low-speed devices and ICs.

OEM Battery: This is sometimes called an embedded battery; it is not removed during the life of the product.

Smart Battery: This hardware-enabled (and possibly software-enabled) battery provides detailed information to its host about its present state and history so that optimum charging and discharging can be achieved.

Smart Battery Charger: This type of charger optimizes its charging characteristics based on communications from a smart battery.

SMBus: The System Management Bus is an implementation of the I2 bus with different voltage levels. It describes communication protocols, device addresses, and electrical requirements so that information can be shared among power components and the host.

The ideal smart battery monitors its own states; manages charge and discharge rates; and in the extreme case, requires only an unregulated power supply. It has a communication mechanism for passing status information between itself and the host. Different battery chemistries have widely divergent charging requirements, but a device that powers a smart battery neither knows nor cares because the battery takes care of itself.

You might also hope that the battery could be charged instantly, hold enormous reserves of power, sustain an infinite number of charge cycles, and be environmentally friendly all at the same time. While no battery is that good, embedded intelligence moves battery performance much closer to these ideals.

Nonlinear and Finicky Batteries

Most semiconductors, transducers, sensors, and passive components have well understood thermal characteristics and fairly uniform behavior over a wide temperature range. But a battery is very nonlinear. It’s the most temperature-sensitive component you can buy, with significant changes taking place at temperature extremes, especially high temperatures. Nickel-metal-hydride (NiMH) batteries canbe permanently degraded, or even destroyed, just above 65 C; lithium ion charging should be cut off below 50 C.

Not only are batteries sensitive to heat, but they also generate heat when charging and discharging. The percentage of heat loss increases with charge or discharge current. Also, the relationships between voltage, current, and temperature can change drastically over the charging cycle (see Figures 1-5).

figure

Figure 1. The voltage and slope curves show inflection and zero slope points for a constant current charge (fast-charging NiMH/NiCd batteries). Just beyond the inflection point but before the zero slope point, the battery is fully charged, and further charging is detrimental to the battery. (Courtesy of Galaxy Power.)

figure

Figure 2. The sharp inflection point at 55 min. indicates a point of rapidly diminishing return (fast-charging NiMH). Note that the fluctuations in the slope prior to the inflection are due to ±0.5°C variations in ambient temperature. (Courtesy of Galaxy Power.)

figure

Figure 3. These curves show battery temperature as a function of charging time. Inflection points once again signal the rapid approach of the optimum time to stop charging the battery (fast-charging NiMH). (Courtesy of Galaxy Power.)

figure

Figure 4. This shows the calorimetric heat generated for three charge methods: galvanostatic (constant current), pause (brief rests), and burp (brief discharge pulses and rests). (Courtesy of Galaxy Power.)

figure

Figure 5. This shows the end of charge gas generated for three charge methods: galvanostatic (constant current), pause (brief rests), and burp (brief discharge pulses and rests) using a flooded (unsealed) NiMH battery for ease of measurement. The charge current amplitude is the same for the three charge methods.
(Courtesy of Galaxy Power.

Things You Have to Worry About

Charging the Battery. If you study Figure 1, you’ll see that the voltage slope levels off as the battery approaches a peak voltage. Fast charge must end at or before this point. Beyond that, the battery voltage and voltage slope start falling. This is the point at which the battery is in overcharge, and it acts more and more like a resistor.

Overcharge heat generation causes a drop in battery voltage at a rate of -2 mV per cell per °C for both nickel-cadmium (NiCd) and NiMH batteries. If you have to charge the battery beyond the zero-slope point to determine that it’s full, you damage the battery and reduce its life span (see Photo 1).

photo

Photo 1. The primary inhibitor of battery life is surface corrosion. Every charge cycle adds some corrosion to the battery. The photos show the difference in corrosion levels resulting from four different charging methods. These scanning electron micrographs show a metal hydride electrode surface at 100X magnification after one charge (A), 400 galvanostatic cycles (B), 400 pause cycles (C), and 400 burp cycles (D).

According to applications specialist Galaxy Power in Valley Forge, PA, its patented method determines the precise time at which a fast charge should end, without pushing the battery beyond that critical limit. This is based on a dynamic model that takes into account the history and present condition of the battery.

New Legislation

A bill in Congress could require many household appliances to operate on standby power of 1 W or less. While this does not specifically address battery-operated devices, the implications are tremendous. If the legislation is passed, electronics manufacturers will be forced to produce designs that consume far less power than those being sold today. This will certainly result in designs that consume less battery power.

H.R.4, Securing America’s Future Energy Act of 2001, passed the U.S. House of Representatives on August 1 and is now before the Senate. The bill is intended “to enhance energy conservation and research and development and to provide for security and diversity in the energy supply for the American people.”

 

Smart battery technology is the science of knowing a battery’s precise condition. The ultimate goal is extremely smart batteries at extremely low cost””batteries that you never have to think or worry about. Another desirable characteristic is fast charge time. In practice, though, this is difficult because each halving of charge time more than doubles charge current, which directly impacts the cost of charging circuits and power supplies.

Optimum charging and discharging of batteries generally results in the following:

  • Service life is doubled in comparison to crude methods that use static models of battery behavior.
  • Batteries that have been mistreated work fine for a while but then deteriorate rapidly once the aging process kicks in; the life of optimally charged batteries tapers off gradually.
  • Optimally charged batteries hold more energy, and the difference becomes more pronounced with age.

Managing Current Flow.

A battery consists of a metal case, a plastic separator, metal plates, rubber seals, and liquid electrolyte. As the battery heats up, the materials expand, and the victim of the expansion is the separator, which is made of porous plastic. Dendrites grow from one side to the other, and the battery becomes self-shorting. Constant flow of current in one direction accelerates the aging process, but occasional current reversals break up the flow and clear the electrodes of gas bubbles, preventing crystal migration into the separator.

Current reversal is also used in the plating of fine metal objects, a process similar to the chemical activity inside a battery. Constant current in only one direction can result in corrosion, but reverse pulses break it up. “If you want to corrode somüthing, put it in a caustic environment and add DC current,” remarks Dave Whitmer, product applications manager at Galaxy Power. “That’s exactly the condition you have inside a rechargeable battery””metals immersed in potassium hydroxide, with current flowing constantly,” referring to NiCd and NiMH batteries.

Major Types of Rechargeable Batteries
PROS CONS
Lead Acid
This type of battery charges fast, supports heavy loads (e.g., car starters), withstands harsh environments, and has long life. Size and weight can be a problem. There are environmental issues with lead and sulfuric acid.
Nickel Cadmium (NiCD)
Overall, these can’t be beat. They’re reliable and fast, handle heavy loads, have good power density (the ratio of energy to size and weight), use an endothermic chemical reaction (i.e., they absorb heat when charging; all others create heat when charging””exothermic). There are environmental problems with cadmium. Modern units don’t suffer from “memory effect,” but they do exhibit a voltage depression if improperly charged.
Nickel Metal Hydride (NiMH)
Similar to NiCD, these batteries can store a bit more energy, and they’re less environmentally hostile. These also exhibit some voltage depression.
Lithium Ion
These offer long life, light weight, and high power density; one cell is 3.6 V, compared with typical 1.2 V of most other cell types (e.g., NiCD or NiMH). This type is expensive and requires charge/discharge protection, active/passive overtemperature precautions, and overcurrent and over-/undervoltage safeguards. These batteries are volatile under stress and can explode. Lithium is reactive, and electrolyte is flammable.
Lithium Polymer
The electrolyte is nonliquid, resulting in a solid-state battery, which can be made in any size and shape you want. These are the most expensive type of battery. Their charging characteristics are similar to lithium ion.
Rechargeable Alkaline
These are low cost and offer good power density. This type has a short service life””it can handle far fewer charge cycles than the other battery types.

Smart Battery Sensors

You might imagine that smart batteries would have all sorts of expensive chemical sensors, but typically only temperature sensors are used. The most common type is the thermistor””a resistor that has a large, precise, and predictable change in resistance when subjected to a corresponding change in temperature. This device must be biased with a voltage and pullup resistor. Most often, negative-temperature-coefficient (NTC) units are used. Their resistance decreases as temperature increases. The most popular types are glass-encapsulated units with axial leads.

Battery packs often use strap-type polymeric positive-temperature-coefficient (PTC) resettable fuses. These are easily installed in series with the cells, and they increase the size of the pack by only 1 mm or so. The fuses limit dangerously high currenz flow during fault conditions (e.g., accidental short circuits). Sometimes bimetallic circuit breakers are used instead.

What Standards Apply to Smart Batteries?

Industry standards help create agreement among battery manufacturers and device vendors and advance the adoption of smart technology. One such standard is the Smart Battery System Specifications, which was created by Intel and nine other companies in 1995. The specification defines message formats for smart batteries, smart-battery chargers, hosts, and multiple-battery system managers/selectors. The goal of the standard is to provide adequate information for power management and charge control regardless of the particular battery’s chemistry. The following are three primary components of the specification.

  • The Smart Battery Data Specification covers the data communicated by a smart battery, including message formats and communication protocols
  • The Smart Battery Charger Specification covers data communicated to and from a smart charger, including messaging formats and the characteristics of the charger itself.
  • The Smart Battery System Manager Specification describes the interface for a component or system that manages multiple smart batteries. It connects one or more batteries to power the system, charges multiple batteries, and reports the properties of the batteries. The specification implements safety measures, and it details the requirements for the selector and message formats.
The Basics of Battery Power Management

Charge the battery precisely to its maximum storage point and no further. (After that it becomes a resistor.) The current flow causes heat that damages the battery and shortens its life.

Condition the battery as you charge it with brief discharge pulses and rest times. Physically, this reduces the buildup of gases inside the battery. Monitoring the temperature of the battery is an important aspect of this; the faster a battery is charged, the more heat and gases are created, which have a negative impact on battery life.

Be careful about how completely you discharge the battery. Beyond a certain point, damage can occur. The battery may be drained to a minimum point (called 100% depth of discharge””1 V per cell for NiCd and NiMH) and then should be recharged. Batteries are less efficient as the discharge current is increased.

An important component of the specification is the System Management Bus (SMBus), which transfers digital data among batteries and devices. SMBus is a derivative of the venerable I2C (Inter-IC) bus, which is popular in mass-produced items, such as televisions, VCRs, and audio equipment. The I2C bus is a bidirectional, two-wire serial network that provides a communications link between integrated circuits. Phillips introduced the I2C bus 20 years ago. The bus transfers data at a rate of 100 Kb to 3.4 Mb per second and supports 7- and 10-bit addresses.

Resistance to Standardization

In contrast to some industries, major companies who work with smart battery technology have been relatively cooperative in creating these standards. The Smart Battery Forum was created in 1995 by 10 companies, including Duracell, Fujitsu, Intel, Linear Technology, Mitsubishi, and Toshiba. Nevertheless, large companies sometimes have reason to shy away from standards.

 
What Do You Measure in a
Smart Battery?

Although chemical readings can be taken in nonsealed batteries, all the data in a typical smart battery are extrapolated from voltage, current, and temperature measurements. Many batteries store historical information (e.g., the number of charges/discharges) and identification information (e.g., manufacturer and serial number). The Smart Battery Forum (www.sbs-forum.org) defines a list of important battery variables.

 

For example, notebook computers have notoriously low profit margins, so the manufacturers make much of their money from accessories. A replacement battery for a notebook typically costs well over $100. But standard communications for smart batteries encourage lower cost knockoff products and gray market. The SMBus specs are available without joining the organization, so it’s pretty easy to figure out how it works. Although it’s possible to define security passwords in the SMBus specification, this is somewhat of a concern. Many manufacturers therefore use their own proprietary communications standards.

The larger picture is that a standard communications bus for batteries brings the usual advantages of networking: distributed intelligence, interchangeable parts, enhanced diagnostics, and plug-and-play functionality. Intelligence is useful only if it can be communicated across a distance.

Components of Networked Multibattery System

The figure below shows the hardware architecture of a smart battery system. The SMBus carries messages in the system. Any of the devices can be the master of the bus at any given time. This system has two batteries, a charger, a system manager, and a system host.

The charging circuit polls the smart battery to obtain the desired charging characteristics and matches itself to the requirements. Now the battery controls its charging cycle. If the battery is 100% charged and AC power is connected, the system disconnects the battery. The smart charger receives critical-event notification from the smart battery when it detects a problem. The notifications include alarms for overcharge, overvoltage, overtemperature, and temperature rising too fast.

The system host is powered by the battery and communicates with it. The host collects data for the system’s power-management scheme. It may also provide information about the battery’s present state and capabilities. Like the charger, the host is notified of emergency conditions. It receives alarms for end of discharge and remaining capacity.

A smart battery system manager or selector feature is necessary in a multiple-battery system. Power must be switched without interruption or damage to any component, regardless of the event””the AC power comes and goes or the batteries switch as they deplete.

 

Smart Batteries in Hybrid Electric Vehicles

Hybrid Electric vehicles, such as the Toyota Prius and the Honda Insight, use a gasoline engine-battery combination to power the electric motors. Because rechargeable batteries handle the constant fluctuations in power demand, the gasoline engine can run at optimum conditions all the time. Also, regenerative braking converts kinetic energy into battery power. These advantages double the fuel efficiency of these compact cars, from the typical 30 mpg to around 60 mpg.

Dan Friel of PowerSmart points out a different challenge that had to be worked out in the development of hybrid electric vehicles. Their 300 V NiMH batteries operate at maximum efficiency between 40% and 80% charge, and it’s not desirable to fully deplete or fully charge the batteries. Friel says that you could fully charge the batteries every few months on an “oil change” basis to recalibrate the reference limits for the electronics, but it’s not necessary. His company developed a battery model that can precisely determine the charge state of the battery based on other measurements, without taking it to its charge limits. This reduces vehicle maintenance.

The Economics of the Battery Industry

The world eagerly awaits new, superior energy storage products. At times, it’s been a hot topic on Wall Street, and there is a pent-up demand for anything that represents a genuine improvement.

The volumes in the battery industry are potentially huge. A former Duracell employee indicated that some lines produce upward of a billion batteries per year, and even a battery-powered product such as a camcorder can sell hundreds of thousands of units per year.

Nevertheless, breakthroughs are not an everyday occurrence. The tooling for a battery line costs megabucks, and it’s one thing to produce a lab queen, but it’s another to make a million batteries a month with uniform characteristics.

SIDEBAR:

Hybrid Electric Sport Utility Vehicles

In contrast to Japanese car manufacturers, who have introduced compact vehicles, U.S. manufacturers are implementing hybrid technology in sport utility vehicles. This improves the usual 15-30 mpg. Because SUVs are hot sellers, there’s a lot more value in doubling a low mpg number than in doubling one that’s already good to start with.

In a hybrid vehicle, the battery is quite expensive, so there’s plenty of incentive to keep it running optimally and to avoid having to replace it. But is smart technology suitable for ordinary vehicle batteries? One problem with extending battery life is that it reduces the aftermarket for battery manufacturers. But IQPower, a German-based company, is entering the market, touting the superior ability of a smart battery to start your car on cold mornings.

According to the company, at about 0ºF standard batteries can produce only 40% of their normal current, while the engine requires 110% more torque because of increased oil viscosity. The chasm between low battery current and the sluggish engine grows to the point at which your car simply won’t start anymore. IQPower says that smart technology can more than double the cold starting power of a standard battery and allow it to be charged up to six times as fast. This is accomplished by precise modeling of the battery’s behavior and dynamic charging based on that model.