Marelli Says New BMS Offers Deeper Insight into State of EV Battery Cells

The system uses electrochemical impedance spectroscopy to monitor changes in state of charge, state of health and internal temperature.

January 6, 2025

Chris Clonts

Marelli says its new EIS-based BMS will eventually be able to work with frequencies above 1 kHz once it moves the analysis onto a dedicated IC. (Marelli)

Battery cell management is a never-ending quest for deeper and faster insight into a cell’s energy production and health. Supplier Marelli says its new battery management system (BMS) will raise the bar for battery efficiency and performance.

There are two basic forms of wired battery management systems, centralized and distributed. Marelli is developing a wireless solution that it says will debut in 2027. (Marelli)

Speaking at a Car Training Institute symposium in Berlin in December, Marelli BMS product manager Davide Cavaliere said the system is based on electrochemical impedance spectroscopy (EIS) and provides real-time insights in lithium-ion batteries’ degradation.

EIS was already popular in labs for its detailed data, but it was too expensive for mass production. In a news release, Marelli said the new solution is cost-effective and scalable for large-scale production.

David Cavaliere’s presentation included this overview of EV battery management systems. (Marelli)

Giovanni Mastrangelo, CTO of Marelli’s propulsion solutions group, said the new system offers automakers a technological edge. “The adoption of EIS and AI in BMS technology is anticipated to…enhance safety, reliability and performance standards in the management of electric vehicles.”

The company calls its current generation of systems “EIS ready.” Its next-gen “Full EIS” BMS will be released later this year and improves real-time estimation of the state of charge, state of power, state of health and internal temperature of cells. Cavaliere said in his presentation that the result is longer battery life and extended range as well as prevention of thermal runaway situations.

When internal resistance in a cell increases, it is a sign of the degradation of the cell’s performance potential. (Marelli)

“The EIS principle relies on applying a periodic stimulus to the cell [current] at multiple frequencies,” he said. “We sweep this frequency from high to low and see the voltage response of the cell. Then we can calculate the internal impedance in modulus and phase.” It’s observing the changes in impedance that allows the deep insight into the state of the cells.

The analysis is done by running the results of the sweep through a fast Fourier transform (FFT) algorithm.

Cavaliere said the potential bottleneck is the “speed of the communication protocol between the monitoring IC and the microcontroller.” Still, Marelli’s new BMS can transfer samples once every 10 milliseconds. That means the system can’t reach higher frequencies than 10 Hz.

The fast Fourier transform is currently completed before data is communicated to the integrated controller. Once Marelli moves it to the IC, it will be able to work with much higher frequencies. (Marelli)

“We are already working in the next generation of BMS that will push beyond this limit,” he said. “By switching the calculation of the FFT to the monitoring IC, we can increase the frequency to higher than one kilohertz.”

Cavaliere said Marelli now makes three kinds of BMS: A centralized solution in which one board monitors all cells in a pack, a distributed daisy-chain architecture that helps when the number of cells is too cumbersome, and a wireless solution that is projected to be available in 2027.