Designers are adopting new battery chemistries and experimenting with new architectures to get more miles out of a single charge.
EV high-voltage battery management system (BMS) technologies are evolving rapidly. To get more miles out of a single charge, reduce charging times, and minimize the overall cost of EV battery packs, designers are adopting new battery chemistries and experimenting with new architectures. Intelligent battery junction boxes (BJBs) and domain-controlled BMS are the next evolutionary steps for EV BMS architecture, offering increased design flexibility, reduced software overhead, and higher battery pack performance.
A traditional BMS has three main subsystems: the battery management unit (BMU), BJB, and cell supervision unit (CSU). The BMU contains the main BMS MCU, which is responsible for state-of-charge (SoC) and state-of-health (SoH) calculations of the battery pack. Precise measurement of SoC and SoH is key to reducing cost and providing an accurate representation of battery life and driving range. Additionally, most of the electronics needed for pack voltage and current monitoring, insulation resistance measurement, and contactor and pyro fuse drivers are on the BMU. The CSU contains the electronics for cell voltage and temperature monitoring, and the BJB primarily functions as an electromechanical box where shunts, contactors, and pyro fuses are located.
TI’s newest battery monitors and balancers, such as the BQ79616-Q1, support a broad spectrum of battery chemistries, including LiFePO4, to improve cell-voltage–monitoring accuracy and enable precise SoC and SoH measurements.
The traditional BMS architecture requires many cables running between the BMU and BJB that consume precious real estate in the battery pack and add weight to the car. In response, manufacturers increasingly are moving electronics such as the UIR sensor (for pack voltage, current, and insulation resistance measurement), contactor drivers, and pyro fuse drivers to the BJB. These intelligent BJBs significantly reduce the amount of cabling between the BMU and BJB while providing greater flexibility for locating the BMU and BJB in the battery pack. The BMU becomes a low-voltage–only board, reducing its complexity and cost.
This new design architecture presents a new challenge: The intelligent BJB and BMU still need to communicate with each other through a Controller Area Network (CAN) bus or an isolated daisy chain, among other options. The CAN approach requires that you place a safety MCU, CAN transceiver, and associated power circuits on the BJB. On the other hand, an isolated daisy chain offers a simple, two-wire twisted-pair protocol that does not require any MCU or associated software, serving OEMs’ requirements for reduced complexity and lower bill-of-materials costs.
In intelligent BJBs, the UIR sensor communicates with the BMU as well as all other circuitry on the BJB, without the need for an MCU. The UIR sensor uses the same isolated daisy chain as the cell monitors sitting on the CSU, thus enabling the BJB and all of the CSUs to sit on the same isolated daisy-chain bus and communicate with the BMU.
The next step in the evolution of BMS architectures is to take the BMU out of the battery pack and integrate it into the powertrain domain controller to create a domain-controlled BMS. The key enabler of this architecture is the intelligent BJB and the BMU’s low-voltage board. Such an architecture greatly reduces the complexity and cost of BMS design for both hardware and software.
A domain-controlled BMS architecture requires a communication interface between the domain control unit (DCU) and the battery pack. Adopting a standardized communication interface enables OEMs to use off-the-shelf DCUs and reduce their dependency on any specific semiconductor supplier.
TI’s automotive BMS portfolio of high-performance, fully scalable solutions enables OEMs and Tier 1 manufacturers to meet their performance and cost targets as they electrify their fleets.
This article was originally published on EEWeb.
Ankush Gupta is a product line manager at Texas Instruments Inc.