Low Voltage Battery Monitoring for High Voltage Electric Vehicles

If you don’t already drive an electric vehicle (EV) – a hybrid electric vehicle (HEV), plug-in hybrid electric vehicle (PHEV), or all-electric vehicle – then there’s a good chance that you may soon. Range anxiety is a thing of the past. You can now help the environment without having to worry about being stuck in one. Governments around the world are offering generous financial incentives to offset the premium of EVs, hoping to steer you away from buying internal combustion engine (ICE) cars. Some governments have already taken steps to require automakers to build and sell EVs, hoping that the market will eventually be dominated by them, while others have drawn a more defined line in the sand; Germany, for example, is already pushing to ban ICE cars by 2030.

Electric car mover

If you’re not already driving an electric vehicle (EV) — a hybrid electric vehicle (HEV), plug-in hybrid electric vehicle (PHEV), or all-electric vehicle — then there’s a good chance that you may be soon. Range anxiety is a thing of the past. You can now help the environment without having to worry about being stuck in one. Governments around the world are offering generous financial incentives to offset the premium of EVs, hoping to steer you away from buying internal combustion engine (ICE) cars. Some governments have already taken steps to require automakers to build and sell EVs, hoping that the market will eventually be dominated by them, while others have drawn a more defined line in the sand; Germany, for example, is already pushing to ban ICE cars by 2030.

For most of the history of the automobile, innovation has focused on improving the fuel efficiency of the internal combustion engine and cleaning up emissions while providing a comfortable user experience. However, the vast majority of recent innovations in internal combustion vehicles are a direct result of advances in electronics—improvements in chassis systems, powertrains, autonomous and advanced driver assistance systems (ADAS), infotainment, and safety systems. Electric vehicles have many of the same electronic systems as internal combustion vehicles, and of course the drivetrain itself. According to Micron Technology, the electronics portion of an electric vehicle’s value is as much as 75%, and this portion is increasing as advances in semiconductor technology continue to reduce the cost of various electronic modules and subsystems. Even non-traditional automotive players, such as Intel, are looking to get a piece of the action.

Not surprisingly, of all the electronic subsystems in an electric vehicle, manufacturers and consumers focus on the heart of the electric vehicle, the battery system. The battery system consists of the rechargeable battery cells themselves (lithium-ion is the current standard), as well as the battery management system (BMS), which maximizes battery usage and safety by monitoring the cells.

Bare Metal Server Monitoring

The primary function of a BMS is to monitor the state of a battery, or in the case of an electric vehicle, a very large battery pack or battery bank. A BMS typically monitors individual cell and battery pack voltage, current, temperature, state of charge (SOC), state of health (SOH) and other related functions, such as coolant flow. In addition to the obvious safety and performance benefits provided by a BMS, accurately monitoring these parameters can often translate into a better driving experience, with the driver fully informed of real-time battery conditions.

To be effective, BMS measurement circuits must be accurate and fast, have high common-mode voltage rejection, consume low power, and communicate securely with other devices. Other responsibilities of an EV BMS include recovering energy back into the battery pack (i.e., regenerative braking), balancing cells, protecting the battery pack from dangerous levels of voltage, current, and temperature, and communicating with other subsystems such as chargers, loads, thermal management, and emergency shutdown.

Automakers use a variety of BMS monitoring topologies to meet their needs for accuracy, reliability, ease of manufacturing, cost, and power requirements. For example, the distributed topology shown in Figure 1 emphasizes high accuracy of local intelligence, high manufacturability of series-connected battery packs, and low power and high reliability of inter-IC communication through low-power SPI and isoSPI interfaces.

The topology in Figure 1 includes an EV battery stack monitor (in this case, the Analog Devices LTC2949) in a low-side current sensing configuration with isoSPI communication lines connected in parallel with the bottom cell monitor (LTC6811-1). For enhanced reliability, a dual communication scheme can be implemented by connecting a second isoSPI transceiver to the top of the battery stack and creating a ring topology that can communicate in both directions. Isolated communication with the SPI master controller is achieved through an LTC6820isoSPI-to-SPI signal converter. Analog Devices’ stackable LTC681x family of multicell battery monitors can be used to measure individual voltages of up to 6, 12, 15, or 18 series-connected cells, while a single LTC2949 is used to measure total stack parameters. Together, the LTC681x and LTC2949 form a comprehensive EV BMS monitoring solution—a circuit that may be better known to some as the analog front end (AFE) of a BMS.

Figure 1. Distributed EV BMS monitoring topology using a battery monitor (LTC6811-1) and an electric vehicle battery pack (LTC2949).

The EV battery pack monitor is a high-precision current, voltage, temperature, charge, power and energy meter designed specifically for EVs. By measuring these key parameters, system designers have the essential elements to calculate real-time SOC and SOH of the entire battery pack and other quality factors. Figure 2 shows a block diagram of the LTC2949 for a high-side current sensing configuration. The LTC2949 uses an adjustable floating topology, which enables it to monitor very high voltage battery packs without being limited by its own 14.5 V rated voltage. The power supply for the LTC2949 is provided by an LT8301 isolated flyback converter with a VCC connection to the positive terminal of the battery.

At the heart of an EV battery stack monitor are rail-to-rail, low-offset, Σ-Δ ADCs that ensure accurate voltage measurements. Of the five ADCs available in the LTC2949, two 20-bit ADCs can be used to measure the voltage across two sense resistors (as shown in Figure 2) and infer the current flowing through two independent power rails with 0.3% accuracy; with an offset of less than 1 μV, high dynamic range is provided. Similarly, the total battery stack voltage is measured with up to 18 bits and 0.4% accuracy. Two dedicated power ADCs sense the shunt and battery stack voltage inputs, producing a 0.9% accurate power reading. The final 15-bit ADC can be used to measure up to 12 auxiliary voltages, facilitating use with external temperature sensors or resistor dividers. Using the built-in multiplexer, the monitor can perform differential rail-to-rail voltage measurements between any pair of the 12 buffered inputs with 0.4% accuracy.

To simplify setup, the monitor’s five ADCs form three data acquisition channels. Each channel can be configured for one of two speeds, depending on the application, as shown in Table 1. For example, two channels can be used to monitor a single shunt resistor: one channel for slow (100 ms) high-precision current, power, charge, and energy measurements; the other for fast (782 μs) current snapshots, synchronized with battery pack voltage measurements, for impedance tracking or precharge measurements. Alternatively, two shunt resistors of different sizes monitored by two independent channels (as shown in Figure 2) allow the user to balance accuracy and power loss for each shunt. Meanwhile, a third auxiliary channel enables fast measurements of the optional buffered inputs or automatic cycle (RR) measurements of two configurable inputs (stack voltage, die temperature, supply voltage, and reference voltage).

Table 1. Configuration options for the LTC2949's three data acquisition channels

Since SOH is a point in the life cycle of a battery (or battery pack) and a measure of its condition relative to new cells, it is important to use an accurate EV BMS monitor to not only maximize driving range but also minimize unexpected battery failures. Speaking of battery life, the LTC2949 consumes only 16 mA when on and only 8 μA when asleep. When any of the monitor’s three data acquisition channels are configured in fast mode (782 μs conversion time and 15-bit resolution), the monitor can synchronize its stack voltage and current measurements with the cell voltage measurements of any LTC681x multicell battery monitor to infer individual cell impedance, age, and SOH. With this information, stack battery life can be assessed, since the weakest cell ultimately determines the SOH of the entire stack.

Strength in numbers

The EV monitor’s digital features include an oversampled multiplier and accumulator that generates 18-bit power values ​​and 48-bit energy and charge values, reporting minimum and maximum values, as well as alerts based on user-defined limits. This relieves the BMS controller and bus from the task of continuously polling the monitor for voltage and current data, and the additional task of performing calculations based on the results. By taking power samples at the oversampled ADC clock rate (pre-decimation filter) rather than multiplying by an average, the monitor can accurately measure power in situations where current and voltage change far beyond their conversion rates, for signals up to 50 kHz.

Figure 2. Typical connections for the LTC2949 floating EV battery monitor in a high-side current sensing configuration. Power for the monitor is provided via an LT8301 flyback with VCC connected to the positive terminal of the battery.