1 Introduction
A battery is a power source that outputs electrical energy through discharge and absorbs and recovers electrical energy through charging. Low-voltage power supplies composed of lithium-ion batteries are critical components in underwater robot systems. Improper maintenance and management of lithium-ion batteries directly affect their efficiency and lifespan, and can even damage them, thus impacting the overall performance of the underwater robot and, in severe cases, leading to safety accidents. By measuring the parameters of lithium-ion battery packs online, it is possible to understand their operating status, characteristics, and maintenance requirements in a timely manner; therefore, the development of online monitoring systems for lithium-ion batteries is imperative.
To monitor the parameters of lithium-ion power batteries, the first step is to design a parameter acquisition module to collect parameters such as voltage, current, and temperature of the lithium-ion power battery and upload them to a microcontroller with an A/D conversion module for recording and display.
Overview of the monitoring system for 2 lithium-ion power battery packs
This system employs distributed data acquisition and centralized data processing. Separate voltage, current, and temperature acquisition circuits are designed, and the data is then sent to a microcontroller for centralized processing. The system structure diagram is shown in Figure 2-1.
Figure 2-1 System structure diagram.
This system monitors the lithium-ion battery pack of an underwater robot system under the National 863 Program. The battery pack uses TS-LFP160AHA model lithium-ion batteries manufactured by Shenzhen Leitian Technology Co., Ltd., and consists of eight individual cells. The system needs to monitor the terminal voltage of each individual cell and make over-voltage and under-voltage judgments; it also needs to measure temperature at multiple points, monitoring the temperature of each battery and the temperature and humidity of the environment in which the battery pack is located; since the eight individual cells are connected in series, only the series current needs to be measured to make overcurrent judgments.
This paper uses the TMS320LF2407A chip. The use of this chip as the CPU in the battery monitoring system is also reflected in the following aspects:
1. Energy saving has become a hot topic in modern electronic device design. This issue becomes even more prominent and important when devices are powered by rechargeable batteries. The DSP used in this design is powered by a 3.3V power supply, reducing controller losses. The chip's power management includes a low-power mode, which can independently switch peripheral devices into low-power mode.
2.16-channel input A/D converter. This is particularly useful for multi-channel acquisition sub-circuits. The output of the acquisition circuit can be directly connected to the DSP's A/D conversion channel, eliminating the need for a separate A/D conversion circuit outside the DSP.
3.40 individually programmable or multiplexed input/output pins. Can be used to control safety switches and other peripheral circuits.
4. The Serial Communication Interface (SCI) and the 16-bit Serial Peripheral Interface (SPI) module can be connected to the display section of the monitoring system.
3. System Hardware Design
The system's hardware design mainly includes voltage acquisition circuits, current acquisition circuits, and temperature acquisition circuits. The acquisition circuits use the TMS320LF2407A as the CPU. The TMS320LF2407A is a high-performance 16-bit fixed-point DSP device designed by TI specifically for real-time control, with an instruction cycle of 33ns. It integrates a front-end sampling A/D converter and a back-end PWM output hardware, simplifying hardware circuit design while meeting the system's real-time requirements.
3.1 Voltage Acquisition Circuit Design
This design focuses on managing lithium-ion power batteries. The battery pack consists of eight 3.6V lithium batteries. Each battery cell has a rated voltage of 3.6V and a fully charged terminal voltage of 4.25V. The voltage acquisition accuracy must be controlled within 1.5%. The minimum sampling frequency required by the battery management system is 20ms.
The system uses linear optocouplers as isolation and signal transmission sampling devices for the data acquisition system, thus isolating the voltage of each battery cell at the front end. The high voltage of the batteries is scaled down proportionally to accurately reflect the changing voltage values to the DSP. The voltage then passes through a multiplexer before entering the microprocessor for calculation. The advantages of optocoupler isolation are high speed (optocoupler speed is in the microsecond range, far less than the millisecond range of relays) and good real-time performance. Furthermore, the signals at both ends of the optocoupler are completely isolated electrically, so even a short circuit at the optocoupler output will not affect battery operation. The optocoupler converts the voltage signal into a current signal for acquisition, solving the common ground problem. Compared to voltage sensors, optocouplers offer better cost-effectiveness.
When selecting components, we considered both cost-effectiveness and practicality. For the optocoupler, we chose the TLP521 manufactured by Toshiba Corporation of Japan, and for the operational amplifier, we chose the dual operational amplifier TL082.
The voltage measurement circuit for a single battery cell is shown in Figure 3-1 below.
Figure 3-1 Single cell voltage acquisition circuit.
VIN, the voltage of a single battery cell, forms a circuit with the LED in the optocoupler via R1, converting the voltage signal (VIN) into a current signal (I11). I11 and I21 have a certain proportional relationship: I11∝I21. UU1 is used as a comparator here. When the voltage Va at point A is greater than the voltage Vb at point B, UU1 outputs a higher voltage value; when Va at point A is lower than Vb at point B, UU1 outputs a lower voltage value. In the entire voltage sampling circuit, the comparator forms a feedback loop, keeping the voltage values at points A and B consistent. The purpose of this is that the voltage at point B is obviously 15/2 = 7.5V, Va = Vb = 7.5V, indicating that the transistors in the upper and lower optocouplers are conducting in the same way. Thus, the conduction of the transistors is controlled by the LED. It can be seen that when I21 = I22, I11 = I22. Therefore, VIN/ = I11 = I22 = Vout/R4. It can be seen that Vout is proportional to VIN.
3.2 Current Acquisition Circuit Design
All the individual cells of a lithium-ion power battery pack are connected in series to form the entire power supply system, requiring only one current sampling point.
This paper uses a Hall current sensor to collect data.
The schematic diagram of the Hall current sensor is shown in Figure 3-2. The magnetic field generated by the current In flowing through the conductor is compensated by the magnetic field generated by the compensation current Im flowing through the secondary coil, which is controlled by the output signal of the Hall element. When the magnetic fields of the primary and secondary sides are balanced, the compensation current Im can accurately reflect the value of the primary current In.
Figure 3-2 Schematic diagram of Hall current sensor.
This system uses the Yusen CBH100SF closed-loop Hall current sensor. The measurement frequency is 0-100kHz, rated current is 100A, measurement range is 0-±150A, turns ratio is 1:1000, accuracy is 0.2%-1%, and response time is <1µs. The structure is shown in Figure 3-3.
Figure 3-3 shows the external shape and connection diagram of CHB100.
The sampling resistor Rm should be a precision resistor for sampling; a high-precision metal film resistor with low temperature drift (not exceeding 2ppm) is recommended. Due to its relatively large parasitic inductance, precision wire-wound resistors should be avoided in high-frequency sampling applications. The rated secondary output current of the sampling resistor should be less than the power supply voltage, with a difference greater than 4V. The power rating of the sampling resistor must be sufficient, Rm = 30Ω.
3.3 Temperature Acquisition Circuit Design
Battery operating temperature is a crucial factor in calculating remaining battery capacity. Furthermore, real-time temperature data acquisition is necessary for assessing battery safety and thermal management. This design incorporates temperature signal acquisition for eight individual battery cells, as well as real-time acquisition of ambient temperature.
This system uses a thermistor to detect the temperature of the battery itself. Combined with a bridge circuit, the temperature signal is converted into a voltage signal. The circuit diagram is shown in Figure 3-4.
Figure 3-4 Single cell temperature sampling circuit
The RMDZ1 is a thermistor, chosen primarily for its cost-effectiveness, small size, and long connecting wires, allowing it to be directly attached to the battery cell casing. Its drawback is poor linearity. Battery temperature detection mainly involves reporting the temperature difference between upper and lower limits, calculating the temperature difference between batteries, and identifying abnormal batteries. This doesn't involve complex functions or calculations, and the linearity requirement is not high, so a thermistor is sufficient.
A novel temperature sensor, the LM35, was selected for ambient temperature measurement. Its key features include an output voltage proportional to the ambient temperature in Celsius, internal calibration within the integrated circuit (IC), and no external calibration required. It boasts a sensitivity of 10.0 mV/℃, an accuracy of 0.5℃, an operating voltage range of 4V-30V, extremely low power consumption, and low output impedance. The LM35's full-scale range [55℃, 150℃] connection method was adopted. To prevent negative output voltage at sub-zero temperatures, which would hinder sampling into the DSP, a subtractor circuit was designed. The output voltage was adjusted to [0, 4.5V] within the ambient temperature range of [-45℃, 75℃].
4. System Software Design
The software design of this system adopts C language programming on a DSP (TMS320LF2407A) and implements a modular design, which increases the readability and portability of the program. This design mainly focuses on lithium-ion power batteries used in underwater robots, while striving for better compatibility; that is, replacing other batteries should not require hardware modifications, only software modifications, or even minimal software modifications. For this system, the control software should meet the following requirements:
It collects signals such as current, voltage, and temperature to determine battery fault signals, processes them, takes corresponding protective measures, and displays fault information.
The acquisition of analog data includes individual battery cell voltage, current, individual battery cell temperature, and ambient temperature. Voltage acquisition requires controlling an analog multiplexer, with the voltage values of each individual battery cell entering the DSP at different times. The requirement is to acquire voltage and current at the same moment. By fully utilizing the TMS320F2407 A/D module, four quantities—voltage, current, battery temperature, and ambient temperature—are acquired simultaneously, and the analog sampling of multiple batteries in the battery pack is completed using a loop.
5. Summary
This paper designs a monitoring system based on TMS320LF2407A to address the characteristics and testing requirements of lithium-ion power battery packs. It proposes a scheme of distributed data acquisition and centralized data processing, presents the hardware and software schemes for voltage, current and temperature acquisition of the battery monitoring system, and builds the underlying acquisition module framework of the battery monitoring system with up to 8 individual batteries.
Based on this, battery information can be easily collected into the DSP for recording and battery status estimation, and communicated with the central controller via the CAN network to form a complete battery monitoring system.
The main research content of this project lies in the design of the overall battery monitoring system and its hardware circuitry. The core of this project is a scheme combining distributed data acquisition with centralized data processing. Voltage, circuitry, and temperature of individual batteries are collected separately, and this basic information is sent to a DSP for centralized and comprehensive analysis and processing. The hardware design focuses on the design of several acquisition circuits and the application of the DSP subsystem within the monitoring system. The voltage acquisition circuit, while ensuring performance, offers flexibility and a significant price advantage. Inter-channel interference and acquisition speed are improved, meeting the system's real-time performance and measurement accuracy requirements. By adding peripheral sample-and-hold devices, voltage and current can be acquired simultaneously. The current and temperature acquisition of the battery management system utilizes Hall effect high-current sensors , thermistors, and Hall effect temperature sensors for measurement, respectively.
For more information, please visit the Power Electronics channel.
