Introduction

Modern electric vehicle (EV) architectures are designed with meticulous attention to detail, incorporating both 400-V and 800-V systems to accommodate different performance and efficiency requirements. At the heart of these systems lies the battery pack, a complex arrangement of electrochemical cells that store and release electrical energy, providing the power source for EV propulsion and range capabilities

Battery basics
At the most fundamental level, deconstructing the battery pack to its battery cells shows how these basic electrochemical units store and release electrical energy. Multiple serially connected battery cells assembled into a
single unit form a module which serves as intermediate level of energy storage. Finally, multiple modules connected, typically in series, form a complete battery pack. The chosen configuration of the pack matches the specific energy (kWh) and power (kW) requirements of the EV. While the overall capacity of the pack (in ampere-hours, Ah) is a function of the number of cells connected in parallel, the total voltage of the pack is determined by the modules connected in series. Figure 1 visually demonstrates this concept.

Figure 1 – Overview cell connections in a modern EV from cell to module to battery pack.

Individual cells can vary in chemistry, capacity, and voltage. By connecting these cells in parallel, as illustrated in the Figure 1, the combined capacity (Ah) increases, providing a higher energy reserve while maintaining the same voltage level. This configuration is beneficial for extending the operational range of the EV by increasing the total available ampere-hours.

Eatron Technologies has adopted the lumped supercell model approach (Figure 2) to efficiently manage battery pack configurations. This approach simplifies the complex structure by treating a group of parallel-connected cells as a single entity – a supercell. Eatron’s use of the supercell model provides detailed insight into each supercell’s state and enables AI features.

Figure 2 – The lumped supercell concept.

Battery management system basics
An advanced battery management system (BMS) oversees the state, temperature, and health of a battery pack to optimize performance and longevity. The BMS ensures that each cell within the pack operates within its safe and efficient operating window, balancing the cells during charge and discharge cycles to maintain the pack’s overall health. Figure 3 below provides a simplified overview of a Battery Management system based on Infineon hardware.

Figure 3 – Simplified block diagram of the BMS solution.

For this proof-of-concept, Neutron Control’s ECU8-based demonstrator was used as the hardware reference platform for development. Some of the components are described below.

The AURIX™ TC4x microcontroller (MCU) with integrated Parallel Processing Unit (PPU) and common automotive interfaces (e.g., CAN, LIN, Ethernet, etc.) possesses the optimal compute, peripherals and safety support for advanced BMS solutions. The AURIX™ TC4x’s integrated Parallel Processing Unit (PPU) hardware accelerator enables edge-based, real-time execution of typical electrochemical PP-p2D battery models and neural networks with an acceleration factor of up ~20x compared to TriCore™ scalar implementations. It addresses the scalability and computational challenges faced by conventional MCUs and enables on-the-fly and efficient parallel processing and execution of complex AI algorithms. With ASIL-D compliance per ISO 26262 and certified ISO 21434 security features, the MCU is a prime solution for both present and future BMS implementations.

The TLE9012DQU is a multi-channel (up to 12 cells in series) battery monitoring and balancing IC fulfilling four main functions: cell voltage measurement, temperature measurement, cell balancing, and isolated communications to the main battery controller. The device achieves best-in-class application reliability and robustness under noisy conditions, high accuracy (±0.2 mV at 25°C including soldering drift), as well as ISO 26262 Safety Element out of Context (SEooC) capability up to ASIL-D level — all ideal features for the BMS environment. Equipped with an independent 16-bit delta sigma analog-to-digital converter (ADC), the TLE9012DQU enables synchronous measurement of each cell to maximize its performance and ensure safe battery operation. Another unique aspect of the IC is its integrated current source which supports direct connection to negative temperature coefficient (NTC) sensors. The TLE9012DQU is just one device in a family of multi-channel battery monitoring ICs enabling scalability for feature upgrades (e.g
TLE9016DQK or TLE9018DQK, 16-channel and 18-channel respectively.)

The TLE9015DQU is a dedicated isolated (iso) UART communication IC developed together with TLE9012DQU to operate either with capacitive or inductive isolation enabling the daisy-chain architecture. The device provides the dual UART ports for serial communication to the host microcontroller and dual iso UART ports for daisy-chain communication inside the battery pack. It supports up to 2 Mbit/s communication speeds and the ring mode/topology requires only 1 device in the daisy-chain. Stable capacitive-isolated communications are confirmed with various tests to provide the lowest cost of isolated communication while offering the functional redundancy via the ring topology.

The PSoC™ 4 High Voltage (HV) Precision Analog (PA) device with high-precision sigma delta ADCs (16-20 bits) provides precise and accurate battery monitoring (e.g., current sensing via shunt) for advanced applications like AIbased BMS solutions. The programmable system on chip (PSoC™) MCU supports synchronization with other BMS devices by communicating in the same daisy chain as the TLE9012DQU battery monitoring and cell balancing ICs. Also, the programmability of the PSoC™ 4 HV PA enables additional functionality to be integrated into a single IC, including pack-monitoring, contact-monitoring, isolation resistance, and pyro-fuse driving.

Eatron’s AI-ISL software solution synergizes with Infineon’s full BMS hardware enabling real-time advanced battery protection and predictive AI functions. Figure 4 shows an example of a high-voltage BMS software architecture that highlights the different components of a BMS stack in an AUTOSAR environment. Within the application software layer, the ISL software, (highlighted in green) integrates with traditional BMS functions, basic software for hardware abstraction (BSW), and runtime environment (RTE).

Figure 4 – Overview of modern BMS software stack.