Hardware Development and Implementation of a BMS Prototype for the Control of Second-Life Batteries in Small-Scale Applications

This master’s thesis presents the development and implementation of a hardware prototype of a battery management system (BMS) for controlling second-life batteries in small, decentralized applications. The work is part of a two-part master’s project, in which this thesis focuses on the complete hardware design and implementation, while a parallel thesis addresses the software development.

The objective is the safe, reliable, and efficient utilization of used lithium-ion cells, for example from electric vehicles, in small-scale energy storage systems such as balcony power plants or mobile backup power units.

The hardware design is based on the ESP32-DevKitC microcontroller as the central control unit. It includes the integration of the LTC6813 battery monitoring IC via the DC2350B evaluation board from Analog Devices for high-precision cell voltage measurement, as well as the connection of the INA228 high-precision current sensor from Texas Instruments via a galvanically isolated I²C interface. Temperature monitoring is implemented using three NTC thermistors (10 kΩ) connected to the GPIO inputs of the LTC6813.

For the MOSFET-based battery protection circuit and the galvanically isolated I²C communication, a custom printed circuit board was designed and manufactured using KiCad. This PCB integrates a logic-level MOSFET (2N7002), an isolated DC/DC converter (DEP1-3-S5-M), two power MOSFETs (IRF540NPBF) in a back-to-back configuration, and a digital I²C isolator (ADM3260ARSZ-RL7). The hardware development focuses on circuit design and dimensioning of the protection circuitry, selection and integration of suitable sensor components, pin mapping and interface configuration (SPI at 1 MHz in VSPI mode, I²C at 400 kHz), as well as PCB design including component placement, routing, and manufacturing output.

SPI communication between the ESP32 and the LTC6813 utilizes the hardware VSPI bus (GPIO5, GPIO18, GPIO19, GPIO23) to ensure deterministic transmission timing, while I²C communication is implemented on freely assigned pins (GPIO25, GPIO26) to keep the default I²C pins available for future extensions.

The results demonstrate that the developed hardware prototype enables reliable monitoring of up to 18 series-connected cells, protective shutdown of the power path in the event of limit violations (temperature: 25–37 °C, cell voltage: 2.50–3.20 V), and galvanically isolated current measurement with a calibrated current LSB of 520 μA.

The outlook outlines potential extensions, including miniaturization through an integrated ESP32 module, scalability to larger battery packs via the LTC6813 isoSPI daisy-chain interface, and the integration of active balancing methods.

The work demonstrates that a customized, cost-effective hardware platform based on open components enables the economically and technically viable reuse of second-life batteries in small-scale renewable energy applications, thereby contributing to resource efficiency and the circular economy.

Software-Based Development of a BMS Prototype for the Use of Second-Life Batteries in Small-Scale Applications

This master’s thesis presents the development of the software for a battery management system (BMS) prototype for the utilization of second-life batteries in small, decentralized applications. The work is part of a two-part master’s project, in which this thesis focuses on the complete software development, while a parallel thesis addresses the hardware implementation. The objective is to ensure the longevity and safety of used lithium-ion cells—e.g., from electric vehicles—through a cost-effective, scalable, and modular BMS.

The software architecture is implemented close to the hardware on an ESP32-DevKitC microcontroller using the Arduino development environment. It utilizes the LTC6813 battery monitoring IC for cell voltage measurement and the INA228 current sensor for high-precision current sensing. The focus lies on the implementation of key BMS functionalities: cell voltage and temperature monitoring using NTC thermistors with table-based interpolation, dual state-of-charge (SOC) estimation via Coulomb counting and a two-dimensional lookup table, automatic passive cell balancing based on threshold logic, and a MOSFET-based battery protection circuit with temperature and voltage supervision. A custom PCB for the protection circuit and galvanically isolated I²C communication was designed and manufactured using KiCad.

A WiFi-based web interface operating in access point mode enables wireless monitoring and control of the BMS via an integrated HTTP server with a REST API.

The results demonstrate that the developed prototype provides reliable cell monitoring with a response time of 100 ms, persistent SOC storage across power interruptions, and robust protection shutdown in the event of limit violations.

Future work includes the integration of Kalman filter-based sensor fusion, impedance spectroscopy for state-of-health (SOH) estimation, and cloud connectivity via MQTT. The thesis demonstrates that advanced BMS functionality can be realized on a cost-effective open-source platform, thereby contributing to resource efficiency and the circular economy in the field of renewable energy systems.