AVX Corp., Fountain Inn, S.C., www.avx.com
The typical goal in applications incorporating supercaps is to substitute these devices for some of the energy capacity that would otherwise be provided by a battery. The resulting battery/supercap combo can typically take up less space (and weigh less) than an equivalent battery operating alone. Supercaps also charge up more quickly than batteries with equivalent energy storage.
By most accounts, supercap makers are trying to both advance their storage technology with such exotic methods as carbon nanotubes, and to keep down the cost of their products. Industry observers say the latter tactic will dominate the industry for the near future. Several vendors are mechanically redesigning their bread-and-butter components. The price of some large supercap systems for industrial applications has also fallen to one-tenth that of a decade ago. And though supercap technology is still considered to be in its infancy, the ultracapacitor community now has its own industry organization, KiloFarad International, Arlington, Va.
“Right now it’s all about cost reduction, not energy-density,” says Steve Minnihan, Power Analyst for energy storage at Lux Research (Boston). “There are some companies that have tried to improve energy-density, but that’s poor technological strategy. It doesn’t make economic sense, when you can just couple the cap with a battery, which has an energy density that the supercap will never be able to achieve.”
Lux pegs the global supercapacitor market at $122M in 2008 for the consumer areas for the 20 or so companies in the business worldwide. Minnihan sees the consumer market growing to $550M by 2014. “As a full fledged commercial product, supercaps are in mid adolescence,” says Mike Sund, VP of Communications and Investor Relations for Maxwell Technology (San Diego). “The market research is tricky to get right because the supercap market isn’t reported as such in some of the bigger companies.” These include players in Japan and Korea (e.g., Panasonic, Nesscap, Nippon Chemi-Con) that mainly concentrate on their own country’s industrial applications. “A loose estimate for the overall market is about $200M,” he says, noting that Maxwell’s portion was $5M in 2004 and $44M in 2009 (of which a good part was in automotive), and may rise by 50% this year.
Power for portables
The first battery-with-capacitor systems introduced almost a decade ago typically went into board-mounted backup applications. Devised by such companies as Tadiran (Lake Success, N.Y.), they delivered peak pulses measured in amps and a charge-discharge cycle life measured in the thousands. Tadiran’s PulsesPlus product included a lithium-based “hybrid layer capacitor” (a supercapacitor, although not so named), which had an equivalent series resistance (ESR) of a few hundred milliohms.
In contrast, today’s supercap products are more typically called on for a burst of energy that lets a BlackBerry send email or a smart utility meter transmit a few packets of data. Their life cycle has risen at least one hundredfold and their ESR has been cut by a factor of between 10 and 100. “Today you can generally get a half-million cycles of charge/discharge before you see noteworthy degradation,” says Steve Minnihan. “With the rate at which consumers are upgrading their cell phones and cameras, the supercap will not break down before the cell phone does.”
Continue on next page
The improved cycling qualities and ESR profile makes life much easier for the battery and, in so doing, improves system efficiency. “Supercaps actually save power,” says Minnihan. “Batteries have a limited power-density. If you try to exceed that power metric the battery’s efficiency and energy capacity degrade with time. The supercap is much more capable of delivering a high-power flash in a camera, for example. It’s also used in camera auto-focus mechanisms. You can take 50% more pictures with a digital camera if you incorporate the supercap to power auto-focus rather than use double-A batteries to power the entire camera.”
Ultimately, the technical issue is in finding ways to cut ESR. In a handheld device drawing up to a few tens of milliamps, the additional voltage drop from a little excess ESR isn’t too significant. Above a few hundred milliohms, though, a five-amp surge can drop the available battery voltage by almost half.
Supercaps are also finding use in portable applications as a way of reducing battery chatter – a short loss of mechanical contact between the battery and its connectors that potentially causes a loss of data. Says Bharat Rawal, Group Leader for BestCap capacitors, AVX (Myrtle Beach, S.C.), “You may get chatter if you throw your handheld device on the table or you drop it on the ground. It’s particularly bad in the emergency room where you’re using a scanner to read blood flow or urine flow.”
A supercap wired across the battery pack can prevent momentary power losses. Doing so typically buys two or three hundred milliseconds, enough time to back up the data settings on a flash drive or similar device. For similar reasons, solid-state disk drives often integrate supercaps, which function as a sort of micro-UPS.
Supercaps also play a big part in wind turbines and potentially could find their way into electric vehicles. However, “We feel supercapacitors are a rather poor fit for the transportation market; the big growth area will be in consumer electronics; specifically mobile phones and digital cameras,” stresses Steve Minnihan. “But while they aren’t particularly a good fit for propulsion power, there are some applications in the automotive area where, for instance, they’ll be useful in bringing a heavy vehicle up to speed from a dead stop and for auxiliary components such as headlights.
The market for those transportation applications will be about $314M in 2014.” The main use for supercaps in hybrids and EVs has been as a way to store energy reclaimed from regenerative braking. Mike Sund explains the reason why. “Think about how long it takes your cell phone or digital camera or laptop to charge -- one to three hours. Think about a braking event -- five seconds. How much energy can a battery absorb as the result of a brief braking event (during which time the system converts kinetic energy into electrical energy)? Some, but not very much. The battery absorbs energy at a given rate (i.e., charge acceptance). Ultracaps can absorb energy much faster than a battery (higher power density). Also when it gets colder, the battery’s ability to charge and discharge is diminished because it’s a function of temperature for a given chemical reaction. Ultracaps will work to 40 below. Batteries begin to lose their ability charge and discharge energy at about zero.”
Supercap systems can also serve as a way to reduce the carbon footprint, particularly where an electric motor gives diesel vehicles a boost as they start up from a dead stop. “That’s where the particulate emissions are highest,” says Sund. “A battery-capacitor system will not only reduce fuel consumption by perhaps 25% under these conditions but will also reduce total pollution by as much as 90%.”
Maxwell’s latest advances include a new stamp-sized, prismatic form factor for its PC-10 – the company’s oldest device still in production – for boardmounted backup (cost is about $2 each). These caps are suited for keeping power to a solid-state drive, and can mount in the drive itself. The company’s largest device, at 3,000 F and costing about $40, is the size of a soda can. Its hybrid transit modules incorporate about 300 of these supercaps. They’re arranged in 48-V modules (each contain 18 supercaps) measuring about 9×9 in, with 20 such modules required for a 750-V transit bus application. These modules usually mount on the roof of a bus. The firm’s 125-V modules, containing 48 supercaps, are about the size of a two-drawer file cabinet.
Continue on next page
Maxwell has also introduced a radial version of its axial “D Cell” supercap modules. The older axial version needed a two-circuit board hookup; the new design, with contacts on the same side, requires only one. Future work is scheduled for 3-V products (and consequently higher energy storage capability). Similarly, Nichicon recently developed a new EVerCAP UW Series of 1-82 F devices (2.7-V radial lead devices, a case size smaller than its UM series).
AVX, which claims to make the lowest-leakage supercap today at 1 to 5 μA (one to two orders below that of a typical capacitor for battery backup found in clock radios and the like), recently developed the BZ01 for energy harvesting. This device joins other products for energy harvesting that include Microd Gen Systems’ (Ithaca, N.Y.) Bolt120C ultracapacitor. The BZ01 is used in Powercast’s P2110 energy harvesting device. Also, AVX has future BZ05 and BZ09 products under development at PowerCast. These devices will feature extremely low leakage currents, on the order of 0.6 to 0.8 μA, a necessary ingredient for collecting energy without losing most of it. In addition, its BZ02 is a 120-mF, 16-V device that lets users deliver a pulse of 10 A for several milliseconds (and 100 A for 0.5 msec).
AVX delivers high-voltage performance through its use of aqueous (versus organic) electrolytes, and can deliver multiple cells in series in the same package. Devices using this technology, designed for handheld apps, can provide virtually any value between 3 and 18 V. The aqueous arrangement uses a so-called bipolar construction in which the cathode can also operate as an anode. With the same current collector used as an anode and cathode, no balancing is required. And AVX says the total package will be smaller.
Nippon Chemi-Con (Japan) has begun work on massproducing its new nano-hybrid capacitor, with samples scheduled for April 2011. Nippon says the new supercaps will have three times the energy density of existing devices, while retaining their power-density capabilities. In addition, LS Mtron (Korea) developed a line of large cylindrical ultracapacitors (up to 3000 F) that are suited to hybrid autos, wind turbines, and UPSes.
In other developments, CAP-XX (Australia) earlier demonstrated prototypes of a surface-mounted supercap suited for portables that will withstand solder reflow processes to 260°C with minimal effect on total capacitance and ESR (less than 10%; initial capacitances were 1 and 0.5 F, ESR of 60 and 100 mΩ, and had rated voltages of 2.75 and 5.5 V, respectively).
Ultracapacitors, or supercapacitors, are generally defined to span the range from perhaps 0.5 to a few thousand farads in packages that range in size from that of a postage stamp to that of a soft drink bottle. Their official name is electrochemical double-layer devices (EDLCs). They store energy electrostatically rather than chemically as with a battery.
It might be helpful to think of a supercap as an electrolyte-based device, with two electrodes (also called plates or collectors) each coated with high porosity carbon, i.e., activated charcoal powder, to significantly increase total plate area. Carbon nanotubes provide even greater surface area per gram of carbon. But their advantages don’t yet outweigh the higher cost of their processing, a significant barrier to mass production.
Traditional capacitors have a comparatively thick physical dielectric (with or without an electrolyte). In contrast, the supercap has what’s called a separator, an interface where the carbon-based layers (i.e., “the electrical double-layer”) approach physical contact. The thickness of the seperator is measured in nanometers.
Keeping in mind that capacitance (Q/V) is proportional to the equivalent area and inversely proportional to the separation between electrodes, it is clear that the supercap’s ability to store charge is several magnitudes higher than either basic twoplate- with-dielectric capacitors or than traditional electrolytic capacitors. Its dielectric breakdown voltage, however, will be on the order of just 2.3 to 2.7 V or so (the present theoretical maximum for these devices is about 3.2 V). There are other drawbacks. As with most capacitors having an electrolyte, supercapacitors are prone to damage from overvoltage situations as well as heat.
The general operation and charge distribution in a supercap resembles what one what might expect in a two-plate capacitor with an electrolyte. When a source voltage is applied to the plus terminal, it pulls electrons from the capacitor anode’s carbon layer. At the other end, electrons deposit on the cathode’s carbon layer (minus terminal).
More specifically, anions (negatively charged ions) flowing through the electrolyte are attracted to the anode and cations (positively charged ions) are attracted to the cathode. The capacitor’s separator prevents charge from flowing directly between the layers. The rate at which anions and cations accumulate on each layer is directly proportional to the physical capacitance of the charge-storage layers (again, a function of the large equivalent area of the porous carbon layers), the separator distance, and the applied voltage.
Continue on next page
“You can’t wire a supercap across a battery”
One can’t simply connect a supercap across a battery. This reality creates some tough challenges for the ICs that are called on to orchestrate the delivery of a power pulse to a portable device. “The caps are pretty finicky to deal with,” says Sam Nork, Director of Linear Technology’s Boston Design Center. “One of the biggest drawbacks is that the maximum voltage across them is actually quite low, 2.7 V is about the highest at which they’re rated. But if you leave them charged at 2.7 V over a long period of time at high temperatures, their lifetime starts to suffer.”
The tradeoff is to charge them somewhere between 1.8 and 2.5 V (thus losing some capability to deliver a high-energy pulse), and to keep temperatures below 60°C or so. “Your expected lifetime at 60°C for a given charge voltage is maybe half what it will be at room temperature,” says Nork. If the capacitor value drops by 20% from its initial value, or the ESR doubles, it’s time to replace the capacitor. Beyond that, the major issue is watching the capacitor leakage current. It’s an issue ultimately associated with voltage balancing in systems with two or more supercaps in series.
“One of the problems is that there’s dielectric absorption (leakage) as the capacitor charges,” says Nork. “The leakage current ultimately decays a few microamps, but it could take 72 hours to do so. And the individual leakage rates of each cap may be different. So if you’re simply regulating voltage at the top of a 4-V stack, the different leakage currents from each cap could well charge one to 3 V (overvoltage condition for the cap), and the other to 1 V. So you need balancing circuits for applications that use two or more caps in series.”
Equalization over time can be an issue as well, he says. Over the span of a week, for example, there could be a 100:1 difference in leakage currents. It’s best to track this parameter electronically. The passive approach of adding fixed-voltage equalizing resistors across the capacitors can eat up a lot of current and can’t react quickly to time- and application-dependent leakage currents.
Thus it is better to equalize supercap currents with intelligent circuitry often built around an op-amp that monitors the midpoint and stack voltages and source or sink currents. Circuitry basically keeps the midpoint voltage of two capacitors in series equal to half the stack voltage. Some examples include Linear Technology’s LTC3225 and LTC3625 buck-boost charge pumps and LTC4425 linear device, each with progressively higher current-delivery capability, to charge the capacitors at programmable currents and simultaneously balance the capacitors while charging.