Those who walked the aisles of the recent North American International Auto Show got to see what some say is the future of automotive technology. There, automakers that included Ford, GM, Daimler AG's smart car operation, Nissan, along with a host of others all displayed vehicles powered mainly by batteries ready to hit the streets.
But don't hyperventilate over electric vehicles. The demand for hybrid and all-electric vehicles (EVs) was actually down a few percent last year. One report from J.D. Power & Associates says that just 7.3% of the anticipated 70.9M passenger vehicles sold in 2020 will be hybrid and electric. Most importantly, the well-chronicled cost, size, and/ or capacity limitations of batteries, along with lack of a systematic way to charge them from a cleaner power grid, need a decade or two to overcome.
Nevertheless, interest in EV-related technology is growing, and not just among traditional automakers. The “holy trinity” of EV technology consists of electric motors, batteries, and power controllers. It isn't just carmakers who have expertise in these areas. A lot of firms with IP in related fields have looked at entering the EV supply chain. And the modular nature of EV powertrains may provide an opportunity for the birth of new vehicle companies that simply buy the main powertrain systems and put vehicles around them.
Changing economics of EVs is the reason such a scenario begins to look attractive. “The total cost of systems running off an internal combustion engine (ICE) is going up; the cost of hybrids is coming down,” said Jeff Defrank, Chief Technology Officer for ALTe in Auburn Hills, Mich. ALTe is a start-up in the so-far uncommon business of converting traditional low/mid-age vehicles to range-extended electric vehicles in light-truck fleets.
ALTe, which plans to be in full operating mode by the end of the year, has a business plan that could become part of the blue-print for companies in the EV business: It will provide integrated motor-drive kits to nationwide installers. The installers then make the conversions. The cost of a conversion will likely be several tens of thousands of dollars; ALTe estimates the time to recoup costs will be perhaps three years for the typical fleet vehicle's high-mileage profile.
The EV market also looks interesting to firms with smarts in building electric motors. “The challenge for vehicles is that industrial motors are often too large and too heavy,” says Jon Lutz, VP of Technology at UQM Technologies in Longmont, Colo. “The effort to package electric motors within vehicles requires something custom, so it's not as easy as pulling motors out of the catalog and incorporating them into vehicles, because of the size and weight constraints. That led the majority of the auto industry toward permanent magnet technology.”
UQM knows a thing or two about building electric motors for vehicles. In the 1970s, it developed an environmentally friendly, battery electric passenger car tradenamed the Electrek and sold over 75 of them. While developing the Electrek, it perfected designs for electric vehicle motors optimized for delivery of high torque and high speed with exceptional energy efficiency. These efforts led to the invention of a patented core-permanent magnet motor technology. Other IPs to come of this work include methods of manufacture, switching circuit architecture, the packaging of an electromechanical brake inside a motor, and a low-cost way of accurately sensing rotor position.
To get a high power density in three-phase ac permanent-magnet motors for EVs today, it looks as though Neodymium is the most practical magnetic material. “The magnet has to be a high energy product,” says Lutz. “The two basic rare-earth materials are samarium-cobalt and neodymium-iron-boron. Neodymium is a bit stronger, and it's typically cheaper. What we've done is develop ways to minimize how much magnet we require in our motors while getting the most power out. That alleviates the cost issues.” UQM's products include a so-called standard line from 50 kW (67 hp) to 200 kW (268 hp) in two frame sizes: 11 and 16-in-diameter. Products up to about 145 kW are suited to automobiles; products for trucks and buses cover the range from 150 to 200 kW.
Continue on next page
In practice, the magnet makers — including Sumitomo Special Metals and the Magnequench Div. of Neo Material Technologies (Toronto) — take raw materials and deliver a magnetizable piece to the motor manufacturer. The motor maker, in choosing the appropriate grade of material, basically trades off magnet strength for temperature qualities. Most magnetic pieces are magnetized in the latter stages of motor drive manufacture (working with pre-magnetized magnets creates several handling issues).
The motor maker or system house integrates the control and power inverter electronics (generally including a DSP with discrete drivers as delivered by such semiconductor houses as IR, TI, and STMicroelectronics) — to coordinate operation among the motor, high-voltage battery system, and the drive train.
No reluctance to switch
Though there are a lot of suppliers with electric motor expertise, it seems as though major carmakers want to keep the holy trinity systems under their own control. As such they would be following Toyota Motor Corp's lead. Toyota fabricates its own motors and semiconductors. General Motors also has a pilot program to make its own motors.
But it seems that automakers and electric motor makers alike are looking for alternatives to permanent magnet motors, to get away from dependence on rare earths. “Permanent magnet motors are the best choice if you're concerned strictly about maximizing torque for a given size and weight,” says Mike Turner, Technical Director for Switched Reluctance Drivers at Nidec Motor Corp. (St. Louis and the U.K.) “In terms of torque and power density they are the best, because you're paying for the magnetic field up front by buying the magnets, instead of paying for it by putting current in the windings. The downside is that it's difficult to make a PM motor work efficiently and with good stability over a wide constant power speed range.”
Myriad considerations and tradeoffs are drawing Nidec into the switched reluctance camp. “In recent years, people have designed passenger hybrid cars on the assumption that a permanent magnet motor would be used,” says Nidec Managing Director for SRDL Roy Blake. “In the last year or two, they've said ‘wait a minute’ in terms of size and cost. People are finding the answer may not be only PM motors. People who are worried about the rare-earth material monopoly in China drive some of that. And so switched reluctance and induction motors will find their way into the market.”
Switched reluctance types will not be without their disadvantages, often cited as high-torque ripple when operating at low speed, more noise, and the need for a more complex drive. And induction motors, often cited as just not having enough basic performance, may have a longer climb. (Though Toyota has been developing an EV motor that does not use rare earths and has indicated it is a kind of induction motor. But as of this writing, it will not release details.)
Beyond materials and cost, the PM motor has its performance issues, says Mike Turner. “In PM motors, light-load efficiency is poor at high speeds. There's good efficiency at full load. But at light load, efficiency is impaired by speed-related losses that arise regardless of load torque. The alternating magnetic field from the rotor induces significant eddy currents at high speeds. Furthermore, if you have to put current in the windings to weaken the magnetic field (i.e., field-weakening operation), you have to do that pretty much regardless of load torque. So on the freeway, motor speed is high and you're in field weakening for sure. As you take your foot off the gas pedal and settle to cruising speed, the motor goes from 100% to, say, 15% load. And the advantages of a PM motor suddenly aren't at all obvious with respect to load and speed.”
In making the case for switched reluctance, he also cites another factor. “In EV apps, you might typically want to operate at constant torque at up to about 1,200 rpm, and then a constant power speed range from 1,200 to 6,000 rpm,” he says. (The constant power/speed ratio is the ratio of the top speed of a motor to its base speed, and represents its ability to deliver both high torque and high speed.) “Induction motors do this rather poorly: they're presently confined to a 2:1 or 3:1 ratio. PM can do between 3 and 5. Switched reluctance, because it has no magnets and doesn't suffer from the effects of parasitic inductances in the same way as an induction motor, has no such limitations, and furthermore can maintain high system efficiency across a wide range of speed and loads. We've demonstrated systems with constant power speed ratios as high as 20:1.”
Continue on next page
Far from discarding the induction motor, manufacturers are revisiting them. “Induction motors are a possibility, but so far they tend to be larger, heavier, and less efficient,” said Jon Lutz. “A PM motor system can be 94%, where an induction motor may be 90%.” Gaining an edge on the major motor technologies thus seems a daunting task.
Semiconductor makers put the pedal to the metal
From the standpoint of semiconductor makers, EV work is just starting to get juiced up. “Hybrid development is mostly in phase one right now; car makers want to prove out the systems,” says STMicroelectronics Director of Automotive Marketing Joseph Notaro. “Phase two is optimizing the system; once you understand where the market is going, whether you need parallel hybrid, series, series parallel, that's not defined today. It's about implementing lessons learned. And phase three will be the breakthrough phase: the right battery technology, a new generation of power technology… That's about three to five years off, in most cases, for true production-ready parts. For qualifying to automotive, another three years, so we're talking beyond 2016.”
“The major difference between industrial and automotive is not so much the electrical, environment conditions and temperature ranges, but different mission profiles,” says Notaro. “You turn on industrial equipment once a day, maximum. You might turn on a vehicle 15 to 30 times a day. As for temperature, with a hybrid vehicle using an internal combustion engine, you'll have lots of heat that brings temperatures higher than the power electronics. With an electrical vehicle, you won't have that heat source to deal with anymore. The mission profile will change. That's phase two optimization.”
The basic DSP control devices and IC drivers for today's hybrid motor drives (some with control/drive units the size of just a 12-V battery) are good enough to tackle most jobs. Most major semiconductor makers have an automotive portfolio, with the device families often industrial ICs and components qualified to automotive applications.
“The architecture of the electronic control unit (ECU) network of an HEV is often similar to that of cars with a standard combustion engine,” says Dr. Marc Bachmann, Texas Instruments' worldwide strategic marketing director for Transportation and Safety MCU. “The control of the electrical drive is usually just added to the control ECU for the combustion engine. Architecture and system setup of the remaining ECU-network can remain more or less unchanged. In the case of an EV, there's more freedom in the system architecture, especially when the car is not derived from an existing car platform. Still, the need for cost efficiency will lead to the re-use of existing and proven ECUs wherever possible.”
Better IC technology cuts costs, boosts safety requirements, and improves the power-density. “On the IC level, one way to cut cost is functional integration, which reduces cost on the silicon level and, at the same time, helps footprint optimization on the PCB level,” says Bachmann. “In addition, both ECUs for electric motor control and battery management are critical for the functional safety of the vehicle.”
In that context, the National Electrical Manufacturers Association's (NEMA, Arlington, VA) new Electrical Vehicle Supply Equipment/Systems Section held its first technical meeting in January to address safety in electric vehicles.
“As for power density, smaller electrical drives enable new options for design, packaging and styling of HEVs and EVs. We can enhance power density by increasing the number of poles of a BLDC motor,” says Bachmann. “To the same degree, the performance requirements grow for the components in a field-oriented control-loop, especially affecting the MCU and bridge driver.”
The typical motor drive itself isn't too complicated at first glance. “It consists of the main 10 to 20 kHz, dc/ac inverter block to drive the motor, pumps, and cooling fans. There are various other motors for door locks, windshield wipers, electrical power steering, and so on,” said Marzak Li, International Rectifier product marketing manager for the Automotive Products Business Unit. “Operating voltages can range from 12 to above 400 V.” In that connection, some system designers are resurrecting the intermediate system bus idea (at about 48 V), depending on the degree of electrification, to maximize overall system efficiency.
Continue on next page
New devices provide the expected improvements in switching and conduction losses. One goal is to make IGBTs in the 600 to 1,200-V range more similar to MOSFETs at a lower price while also improving thermal management. “It's getting away from wire bonding on power devices for IGBTs as much as possible to improve reliability and power density,” says Li. It means dual-sided cooling and using a power sandwich to cool the part and to provide electrical contact (as with the company's DirectFET devices).
Currently IR has more than 50 products in its electric vehicle portfolio, led by its newest WARP2 devices, which signify very fast, high-power switching. “We already have better microcontrollers, mixed signal processes (e.g., BCD, 20 to 800 V for industrial apps), and discrete power MOSFETs and IGBTs,” adds STMicroelectronics' Joseph Notaro. “Silicon technology is key, but we need to get the power outside the chip. We need ribbon bonding for higher currents and multiple die packages.”
Beyond the need for more tweaked power components, however, most present product efforts focus on controls, where microprocessors and microcontrollers rule. These include TI's TMS570 microcontroller family, which targets safety-related automotive applications. Lithium-ion battery management and safety systems for HEVs/EVs are also a priority; recent arrivals include chips from STMicroelectronics and Intersil, among others.
The ultimate key to success in HEVs and EVs rests on improving the battery and finding better ways to charge it. “The batteries are very expensive,” admits Paul Williamson, national manager of Lexus College at Toyota. “But with each subsequent generation, NiMH to Li-ion, we have seen a similar decreasing trend line in cost-per-power. We're optimistic for more potential even after the power-density curve flattens out.”
Still, few designers envision a practical battery architecture that can take the user on a 150-mile round trip to work and back on a single charge. On the other hand, the battery systems are getting lighter: The first battery pack for the Prius back in the 90s reportedly weighed 87 lb; the second-generation 2003 version weighed 78 lb and could deliver 30% more power. And the battery pack for the 2010 Prius weighs 68 lb and added 25% more in total power.
Battery lifetimes are ramping up, too. “We are now seeing battery life spans equal to the life of the vehicle,” said Williamson. “We shoot for 300,000 km. The key is battery management — how high you want to charge it, and to what degree you discharge it. If you run in depletion mode, you'll have challenges. In general terms, we tend to charge our batteries never more than 75 or 80% SOC (state of charge), and never discharge them below 25 or 30% SOC.” Toyota batteries come from Panasonic Electric Vehicle Energy (PEVE), of which Toyota is a half-owner. But neither the performance nor environmental cycle is complete without a better charging source.
On the grid
Most participants say the jump to the big time rests on plug-in hybrids (PHEVs) and charging the battery from a cleaner power grid. It'll be important to charge up at the right time of day, too, if it's going to work environmentally. “There are two challenges,” said Paul Williamson. “The first is known: electric storage cells have physical constraints as to the rate you can put energy in, and that's a function of chemistry and temperature considerations. Any EV is going to take two to three hours to get a decent charge into it at 220 V. And it'll take maybe six to eight at best for 110 V. That doesn't lend itself to convenience when we run out of gas on the road and we need two minutes to get gas at the local station to make it to an appointment.”
The more important long-term issue is where the charging power comes from, and when. The problem reduces to taking power from the grid when both the price and the carbon footprint are at a low. “That's when you want to charge the EV; because that's essentially carbon-neutral energy,” says Williamson. “As long as the companies are using smart power meters to charge that vehicle when the power plant is idle, and the percentage of EVs isn't a majority of all vehicles, you can get a relatively clean EV. But that's at some point in the future.”
This scenario implies that energy producers and suppliers will need to modify their business models to discourage customers from charging their EVs during the day. But will they? “The vehicle and the power grid should be smart enough to talk to each other to start charging when the rates go down,” said Williamson. “If people charge on the road during the day, we'll need to build new power plants. Unless those plants will generate power in a different way than they are now, there won't be a net improvement in air quality.”
Continue on next page
Indeed, the pollution coming out of a power plant is as much an issue as is how to derive hydrogen supplies. “The EV has no tailpipe at all, but smokestack emissions are pretty significant,” said Williamson. “We are still in the state where a new Toyota or Lexus hybrid, when you look at total life cycle emissions from mining to engineering to manufacturing and end-of-life recycling, is cleaner than an electric vehicle. That's unless you're in a few particular countries (or states) that have developed low carbon footprint ways of making electricity. So, for instance, in China today (which doesn't yet have low-carbon footprint power plants coming on line) an EV creates worse smokestack emissions than all of the tailpipe emissions of a good clean hybrid.”
Most consumers are probably interested in the time line for when a hybrid vehicle and a traditional gasoline-powered vehicle with roughly the same performance capabilities will cost about the same. Even now, the differential can be as much as $10,000, although various hybrid-vs.-traditional analyses reveal a significantly lower margin. “Just a few years ago, the hybrid option cost consumers an extra three to five thousand. Today, you can find some cars where the consumer is given the choice of either a large gas engine or a smaller gas engine with hybrid capability that gets better MPG all for the same price,” said Ken Farrington, TI Technology Consultant, Worldwide Strategic Marketing Group. “And the consumer really isn't sacrificing performance with the smaller engine, because the electric motor does a good job assisting with acceleration. Hopefully this parity pricing approach will become more common, as it should help hesitant potential hybrid buyers make their first hybrid vehicle purchases.”
There are also other ways to look at the gains. “We are now on our sixth generation of different hybrid vehicle technologies; most automakers haven't reached their second generation yet,” says Paul Williamson. “The price tag for our current Prius is about $22k. It compares favorably to our Corolla, which is about $17K. Plus you get twice the fuel economy with the Prius.”
Managing the gains
Electric hybrids ultimately use the gas in the tank to charge the battery. How does that argument square with saving on fossil fuels? “We're really not gaining in the mode where the gas tank charges the battery,” says Jon Lutz. “Where you gain is operating the engine in a narrower band of power, and you can downsize the engine if you have an electric motor helping out. That can keep your engine operating at a higher overall efficiency.”
The usual way manufacturers gauge efficiency is to determine how many miles the vehicle gets on a gallon of gas. Or we might calculate efficiency as a weighted function of the amount of time the electric motor (about 90% efficient) is in use, versus that of the auto's internal combustion engine (about 15% efficient). Toyota calculates the efficiency of its leading front-drive hybrid platforms in the range of 32 to 36%.
While the engine turns the alternator to charge the battery, gasoline's role in the charging process isn't everything. “We need to be clear in how much kinetic energy is available in decelerating a moving vehicle,” said Paul Williamson. “We are able to recover nearly all of that kinetic energy to the point where in our front-drive platforms, i.e., motor generators directly coupled to the front tires, those vehicles almost never use their friction brakes except in a panic stop. So under normal braking, it's all done regeneratively. About 40% of our total efficiency improvement comes from regenerative braking.” Beyond that, the start/stop drive cycle has the greatest opportunity for improving efficiency. On that score, some new U.S. models will minimize emissions during stops by turning the engine off, then restarting it when the driver hits the accelerator.
System designers see the plug-in hybrid (PHEV), where battery charging proceeds from the ac power grid, as making a much bigger dent in the fossil-fuel argument. There are additional environmental issues. The NiMH and Li-ion batteries used in these vehicles (e.g., several hundred stacked 2.7-V cells) consume considerable natural resources.
A magnet's energy product is a function of the magnet's remnant flux, Br (the amount of magnetism left in the material after you've removed the magnetizing current); and the coercivity, Hc (how easy the magnet is to demagnetize once magnetized), explains Mike Turner. If you have a high Br, you'll tend to get a lot of torque for minimal current in the windings. The downside: Current put through the windings tends to demagnetize the magnets. If the magnet demagnetizes quickly as a function of current in the motor, you won't be able to take much advantage of the high remnant field. So the best motor materials will have high remnant flux and high coercivity. The energy product BH, is the product of the demagnetizing force and the remnant magnetic field.
Neodymium and samarium-cobalt have a high BH product. Samarium-cobalt is a better material in many ways: higher temperature capability before there is irreversible demagnetization; far superior corrosion resistance; and much lower temperature coefficient of the magnetic field strength. But it costs more and doesn't have quite the same BH. “If neodymium prices continue to rise, as they inevitably will, the barriers to using samarium-cobalt may disappear because it won't be that much more expensive than neodymium,” says Turner.
International Rectifier, www.irf.com
J.D Power & Associates, www.jdpower.com
Molycorp Minerals, www.molycorp.com
National Electrical Manufacturers Assn, www.nema.org
Nidec Motor Corp., www.nidec.com
Texas Instruments, www.ti.com
Toyota Motor Corp., www.toyota.com
UQM Technologies, www.uqm.com
U.S. Dept. of Energy, www.energy.gov/
Managing the mines
The “rare earths” aren't too rare, but the laws of supply and demand stand to match name with status because the country most able to deliver them in bulk will be scaling back on its exports. Most of the larger working mines for rare earths (which are usually found naturally and in combination) are in China, which intends to maintain a supply for its own industries, as indicated in the Dept. of Energy's recently released report “Critical Materials Strategy.”
Therefore, much less of China's output will be available to the rest of the world as time goes on. China, which generated about 95% of the world's total output of rare earths in 2009 (125,000 tons), now has about 36% of the total global reserves of 99 million tons. There's reportedly a moratorium on new mining permits there.
In response, Molycorp Inc. (Greenwood Village, Colo., with limited stockpiles in Mountain Pass, Calif.) and Hitachi Metals, Ltd. in Japan have agreed to manufacture rare-earth materials and magnets in the U.S., with formal contracts expected by the end of the year.
In the meantime, the industry is nudging towards other solutions. “The marketplace for neodymium and samarium-cobalt is certainly going to change with respect to the availability of materials and the China factor,” says Mike Turner. “Demand is going to be a big thing — even if other suppliers come on stream, the increasing demand as permanent magnet motors are more widely used is going to push the price up, regardless of the increased supply from other non-Chinese sources.” It's one reason why more developers are looking towards switched reluctance and induction motors.