Battery management units must perform cell balancing and coulomb counting while ensuring that the entire pack meets functional safety requirements...
The battery pack is among the most expensive components found in battery-operated products, such as power tools, scooters, and electric vehicles (EVs). Battery pack performance greatly influences vehicle-level care-abouts in EVs, including vehicle range, battery pack service life, and charge time, not to mention vehicle safety and reliability. Therefore, it is not surprising that battery management is the subject of intense research and ongoing development effort.
From a vehicle systems point of view, key performance indicators (KPIs) for the battery pack include parameters such as DC link voltage, energy density, specific power, and cell life expectancy. Thus far, lithium-ion (Li-ion) batteries provide good results; however, Li-ion chemistry places considerable burden on vehicle electronics for the ‘care and feeding’ of the battery pack.
The use of Li-ion requires that the battery management units (BMUs) ‘push the limits’ of measurement accuracy in a noisy electrical environment in the presence of common-mode voltages exceeding several hundred volts. Aside from monitoring the voltage and temperature of battery cells, the BMU must perform critical functions like cell balancing and coulomb counting, while ensuring that the entire pack works within a safe operating envelope that complies with stringent ISO 26262 functional safety requirements.
Energy density (W-h/l) and specific power (Energy/kg) are two primary figures of merit for EV battery cell design. These figures of merit are driven by several vehicle-level performance parameters; perhaps the most important is range per charge. To optimize range per charge, energy storage must be compact and light.
The higher the energy density, the more energy capacity can be transported in the vehicle; this coupled with a lighter payload due to higher specific power can result in greater vehicle range. In addition to influencing vehicle range, battery pack compactness leaves space for other critical EV systems like the on-board charger and the traction drive that convert electrical energy into motion. This enables the so-called ‘skateboard’ configuration popular on several EV platforms.
Figure 1 These graphs show the chemistry characteristics of a battery cell. Source: STMicroelectronics
Figure 1 compares a few common battery cell technologies. Presently, Li-ion is the clear choice and its use is prevalent in vehicle electrification today. Nonetheless, Li-ion has drawbacks. It is finicky to charge and it is difficult to gauge a Li-ion pack’s state of charge.
Li-ion cells can be tricky, as thermal runaway issues with consumer products like hoverboards have demonstrated. Finally, Li-ion is an expensive technology, not just because of the exotic materials comprising the cells, but also because of the complexity of the battery and thermal management systems that must be present to optimize performance and safety.
Figure 2 depicts the charge and discharge characteristics for a typical Li-ion cell. Once the cell reaches saturation during charging—or even when discharging—the cell voltage remains nearly constant for most of the operational envelope. The flat discharge curve makes it an attractive energy source for an EV because the battery provides nearly-constant energy over a wide operational envelope.
However, this characteristic, along with other intrinsic qualities, presents battery-management challenges. More importantly, battery characteristics largely determine the range that a vehicle can travel, battery service life, safety, and the usability of the vehicle. For instance, knowing how much further the user can travel before charging is required.
Different materials comprise the anode/cathode, which impacts cell characteristics. For example, Li-ion cells charge to 3.8 V to 4.2 V with a tolerance of around ±50 mV, depending upon the anode/cathode materials employed. The cell is considered fully charged when charge current falls below 3% of the cell’s A-h rating. While raising the charge current does not affect the total charge time, it can accelerate the time to reach a plateau of around 70% capacity.
In fact, charging the cell to something less than 100% is desirable to extend service life, as Li-ion cells cannot accept an overcharge without some cell damage and/or a compromise to safety. Thus, the system designer must trade off parameters like range/charge, battery service life, safety, and charge time.
There are other challenges and nuances to consider. Cells are connected in an array—series and parallel combinations to increase voltage and capacity—which complicates the issue of managing over- or under-charging. BMUs implement ‘cell balancing’ to ensure that all cells in the stack—multiple cells connected in series—are virtually at the same level of charge.
For several reasons, it is also important to monitor cell temperature. A significant temperature rise during charging is an indication of a fault. In addition, Li-ion cells do not charge well at cold temperatures, for instance, freezing. In this case, the BMU can heat the battery to compensate.
Finally, even if charging and discharging are tightly controlled, the capacity of the battery decreases over time as it undergoes many charge and discharge cycles. An EV can compensate for this by conveying the range a vehicle has left instead of the battery capacity or state of charge. A brand-new vehicle may charge to 70% and discharge to 30%. As the pack ages and capacity diminishes, the BMU can expand the charge and discharge window to allow the vehicle to maintain its “full-charge” travel range over the life of the vehicle.
Battery cell management
An automotive EV/hybrid electric vehicle (HEV) battery contains several hundred Li-ion cells in both series and parallel connection and it is clear from the challenges already discussed that safe and lifetime-optimized operation can only be maintained with proper cell management. Every cell in the series connection must be diagnosed and balanced individually.
Figure 3 This diagram shows the signal path for battery stack monitoring and cell balancing. Source: STMicroelectronics
The signal path must deliver the requisite precision to estimate state of charge if the objective is to optimize vehicle-level KPIs (Figure 3). Specifically, cell voltage and stack current measurement precision are critical due to the flatness of the charge/discharge curves shown in Figure 2. Also, battery management solutions sometime incorporate coulomb counting—measuring the flow of Amp-seconds in and out of the stack—as a cross check to estimate state of charge of the entire stack.
With such measurement and control complexities, an integrated multi-channel IC comprising cell balancing, as well as voltage and temperature measurement, represents a cost-efficient and optimized solution. An example of such a monitoring and balancing device is the L9963 chip from STMicroelectronics, which supports up to 14 cells per chip with up to 7 NTC temperature sensor inputs.
Figure 4 These diagrams show three BMU architectural approaches. Source: STMicroelectronics
As shown in Figure 4, one L9963 chip provides the functions needed to implement 14 cell management units (CMUs) along with the module management unit (MMU) function. The battery monitoring and protection chip provides a high-precision cell voltage measurement path and it synchronizes cell voltage and stack current readings affording an indication of the state of charge of the entire stack, cell by cell.
A combination of one or more such devices with a suitable microcontroller—to implement the pack management unit (PMU)—provides a complete battery pack solution (Figure 5).
For every connected cell, the CMU acquires the cell voltages and temperatures and communicates this data via a galvanically-isolated interface to the main processing unit. The CMU directly affects the KPI parameters of the whole battery. The more accurately it can determine the cell voltages, the better it can utilize available cell capacity and the more precisely it can derive other higher-level application parameters, such as state of charge.
To achieve effective charge balancing between the cells, a passive balancing method can be applied. A switchable load is placed in parallel to each cell, so that during the charge phase, the charge level of individual cells can be kept constant or slightly decreased in case the switch is conducting. This balances the level of charge throughout the entire battery stack as the cells with the non-conducting “balancing by-pass” continue to raise their charge level.
Here, the L9963 battery protection chip simplifies this passive balancing as it provides integrated balancing MOSFETs in a way that only the external balancing load is needed. Furthermore, the device offers several configuration options that facilitate an autonomous and simplified control of the balancing process.
The acquired sensor data and diagnostic information must then be transferred to the processing unit using a galvanically-isolated interface to properly separate the high voltage battery domain from the conventional vehicle bus system and supply. The L9963 chip supports both a transformer and capacitor-based coupling to create the galvanically-isolated interface.
Fast communication is key and L9963 allows data rates of up to 2.66 Mbps, which translates to an update interval of less than 4 ms for a complete 400-V battery. In this example, the battery consists of 96 cells in series with seven L9963 devices, each managing a stack of 14 cells and all L9963 devices communicating via a single daisy-chained communication interface.
All of these aspects—acquisition of sensor data, the integrity test of measurements, the transfer of sampled data, and the permanent supervision of the cells—are safety critical for both the operation of the vehicle and for the vehicle occupants. With an appropriate battery management device like L9963, developed according to ISO 26262 standard for safety requirements up to ASIL D, safety features are designed in.
The Li-ion battery chemistry delivers exceptional power density and specific power, and these characteristics are key to maximizing vehicle range per charge. This article emphasizes the importance of the BMU to ensure that the battery delivers the performance expected, and also maximizes battery service life, while addressing safety requirements. At the component level, this means that the signal path must deliver high precision over a wide temperature range and there are controls in place to manage the battery as well.
Figure 6 This block diagram shows the layout of an EV energy transfer and storage system. Source: STMicroelectronics
Still, the battery pack and the BMU are only a portion of the overall energy transfer and storage system associated with an EV (Figure 6). Aside from charging equipment installed in owners’ garages, electric vehicle service equipment (EVSE) is becoming prolific as sales of EVs continue to increase. The EVSE interfaces to an on-board charger that converts incoming power from the grid into high-voltage direct current (HVDC). Some chargers provide HVDC at very high current directly and can charge a vehicle to more than 70% in 20 to 30 minutes.
Battery electric vehicles (BEVs) as well as HEVs have the potential to deliver on the promise of reducing the carbon footprint of transportation. And with proper and appropriate battery-charging technology employed, consumers have discovered that they can do so without compromise of vehicle performance as well as convenience.
John Johnson manages the automotive systems marketing group for the Americas region at STMicroelectronics.
Markus Ekler is senior technical marketing engineer for ASSP/ASIC components at STMicroelectronics.