Design of battery monitoring module in communication power supply monitoring system

1 Introduction

In the local communication power supply monitoring system, the battery monitoring module is a relatively independent unit with its own processor unit and data acquisition unit. Therefore, it can be used as a part of the local communication power supply monitoring system, and at the same time, it can be simply expanded to become a stand-alone battery online tester. This article introduces in detail the design of the battery monitoring module in a local communication power supply centralized monitoring system with a two-level distributed system structure.

2 Overall Implementation of Battery Monitoring Unit

Battery monitoring has always been a hot and difficult issue in domestic and foreign research. In this system, the battery monitoring unit mainly completes the following functions: online detection of remaining capacity, switching between equalization and floating charge modes, single terminal voltage test and lagging battery detection, battery body temperature test, etc. Its overall implementation is shown in Figure 1.

Figure 1 Overall hardware structure of the battery monitoring unit

The processor module is the core of the battery monitoring unit. Here we use ATMEL's latest RISC high-performance microcontroller AT90S8515 and large-capacity 8KB FLASHROM, which not only ensures high-speed analysis and processing of large amounts of data, but also realizes the storage and query of data.

In the data acquisition module, since the data that needs to be processed in the battery monitoring unit has special requirements on accuracy (for example, the measurement of the battery internal resistance is usually at the mΩ level, and there must be enough bits), and since the battery internal resistance and voltage are both slowly changing low time-varying signals, we use a 16-bit Σ-Δ A/D converter AD7715, which has the functions of automatic zeroing and automatic range calibration, thereby ensuring high measurement accuracy, and has an SPI interface, which can be easily interfaced with the microcontroller.

The battery monitoring unit is equipped with an RS485 communication interface, which exchanges data with the main processor of the front-end machine in the form of communication. Therefore, in this system, the battery monitoring module is actually connected to the main monitoring module as an intelligent device. The internal resistance detection module, the single cell voltage test module, and the single cell temperature test module are introduced in detail below. Since the DC data collection and processing of the current test module are similar to those of the main processing unit, they will not be repeated here.

3 Online detection of remaining battery capacity

The remaining capacity of the battery is a problem that users are most concerned about. It is closely related to the reliability of the entire power supply system. The higher the remaining capacity of the battery, the higher the system reliability, otherwise it is the opposite. Therefore, how to monitor the remaining capacity of the battery in real time online without consuming the battery energy and affecting the normal operation of the electrical equipment will be of great practical significance.

The battery is a complex electrochemical system. When it is operated under different load conditions, the amount of electricity that can be released by the battery is also different. As the battery is used for a longer time, the amount of electricity that can be released will also decrease. In the past, the quality of the battery and the amount of its remaining power were often judged based on the terminal voltage of the battery, but this method has great limitations. As the battery ages, its terminal voltage does not change significantly. Therefore, it is difficult to use the change in terminal voltage to estimate its remaining power, and the error is large.

3.1 Several Commonly Used Remaining Power Prediction Methods

At present, the most representative solutions for predicting the remaining battery power are as follows:

(1) Density method: The remaining capacity of a battery is closely related to the density of its internal electrolyte, which is determined by lead sulfate, lead oxide and lead. By measuring the density of the electrolyte, its remaining capacity can be indirectly estimated. However, in the later stages of battery use, as the positive and negative plates corrode and break, the ratio of the above three substances differs greatly from the ratio formulated during battery manufacturing, resulting in the inaccuracy of using the density value to estimate the remaining capacity. At the same time, since most of the current communication power systems use valve-regulated lead-acid batteries, this method is difficult to apply.

(2) Open circuit voltage method: As mentioned above, the degree of charge of a battery is closely related to the density of the battery electrolyte, and the N.RST equation describes the relationship between the electrolyte and the battery electromotive force. Therefore, by measuring the open circuit voltage of the battery, the remaining capacity of the battery can be calculated. Its disadvantage is that as the battery ages and the remaining capacity decreases, the open circuit voltage does not change significantly, so it is impossible to accurately predict the remaining capacity. In addition, the open circuit voltage is the steady-state voltage of the battery when it is unloaded, so it can only be measured when the battery is stationary, which is not suitable for real-time online measurement.

(3) Timed discharge method: by applying a load to the battery, the rate of change of the battery terminal voltage per unit time is calculated, and the remaining power is estimated based on the size of the rate of change. A small change means a large remaining power, otherwise it is the opposite. In order to achieve online measurement and shorten the measurement time, it is necessary to discharge the battery with a large current, which will cause serious damage to the battery and seriously affect the battery life.

(4) Internal resistance method: Studies have shown that there is a high correlation between the internal resistance of the battery and the degree of charge. The American GNB company has tested nearly 500 VRLA batteries with capacities ranging from 200 to 1000A. When the battery pack voltage ranges from 18 to 360V, the experimental results show that the correlation between internal resistance and battery capacity is very good, and the correlation coefficient can reach 88%. Therefore, by measuring the internal resistance of the battery, its remaining power can be predicted more accurately. When the battery is fully charged (full) and fully discharged (discharged), its internal resistance differs by about 2 to 4 times. As the battery charging process proceeds, the internal resistance gradually decreases; as the discharge process proceeds, the internal resistance gradually increases. In addition, as the battery ages, its internal resistance gradually increases, and its remaining power also decreases. The relationship curve between the internal resistance of the battery and the remaining power is shown in Figure 2.

Figure 2 Relationship curve between battery internal resistance and remaining capacity

Since the change rate of the internal resistance of the battery when it is fully charged and fully discharged is much greater than the change rate of the battery terminal voltage (the terminal voltage change rate is about 30%~40%), it is much more accurate to use the internal resistance of the battery to predict its remaining power than the open circuit voltage method. The advantage of the internal resistance method is that for batteries used online, this method has the least impact on the system and can be accurately measured throughout the battery's entire service life.

Through the introduction and comparison of the above measurement methods, it is not difficult to see that the internal resistance method is most suitable for online measurement of the remaining power of sealed batteries. Therefore, this system uses the internal resistance method to measure the remaining capacity.

3.2 Implementation plan for predicting remaining power using the internal resistance method

The specific implementation method of predicting the remaining power by the internal resistance method is: first fully charge the battery (taking a 2V battery as an example, charge it to 2.35V and the floating current to 10mA), then discharge the battery at a discharge rate of 0.1C, and record the size of the internal resistance and power during the discharge process. When the battery is discharged (the 2V battery is discharged to 1.75V), a complete discharge curve can be obtained, that is, the relationship between the remaining power and the internal resistance of the battery. Store this curve in the EPROM. When testing batteries of the same model and specification in the future, the microcontroller calculates the remaining power value based on the internal resistance of the battery measured online by looking up the table. Therefore, the key to this method lies in how to measure the internal resistance of the battery online. The measurement principle is as follows: apply a constant AC audio current source Is to both ends of the battery, and then detect the battery terminal voltage Vo and the angle θ between Is and Vo. Obviously, the relationship between the three is

as well as

R is the internal resistance of the battery that we want to obtain. The specific implementation scheme is shown in Figure 3:

Figure 3 Implementation of the internal resistance method to predict the remaining power

The 300Hz signal generation circuit is composed of a 14-bit binary serial counter/frequency divider CD4060 and a low-pass filter circuit, and the specific circuit is shown in Figure 4. The constant current power amplifier part uses an audio power amplifier with a power of up to 4W.

Figure 4 300Hz signal generation circuit

4 Measurement of battery cell voltage

The Technical Requirements for Centralized Monitoring Systems of Communication Power and Air Conditioning stipulates that the battery detection device must measure the single cell voltage of each battery. Since the batteries are connected in series to power the communication equipment, the potential of each battery to the ground is different, and the highest common mode voltage can reach 60V, which is difficult for general multi-channel analog switches and A/D converters to bear. Therefore, to test it, the floating ground signal must first be processed with a common ground or isolation measures must be taken. The traditional and more mature testing method is to use relays and large electrolytic capacitors for isolation processing. The basic principle is shown in Figure 5.

Figure 5 Traditional single cell voltage test method

The basic test principle is: first close the relay to area A to charge the electrolytic capacitor; when the voltage of the battery needs to be measured, close the relay to area B to isolate the electrolytic capacitor from the battery. Since the electrolytic capacitor retains the voltage signal of the battery, the test part only needs to measure the voltage on the electrolytic capacitor to obtain the corresponding battery voltage. This method does not require the use of relatively expensive devices such as linear optical isolation, and has the advantages of simple principle and low cost. However, due to the defects of slow mechanical action and low service life of the relay, practice has proved that the detection device implemented according to this principle is difficult to be satisfactory in terms of speed, service life, and working reliability.

4.1 Implementation of the hardware direct subtraction method

The idea of ​​hardware direct subtraction comes from the concept of mathematical subtraction. Imagine that if a high differential mode gain op amp is used to compress the high potential on the battery proportionally, that is, first compress the high-end potential of the nth battery according to the ratio of Rn1/Rn2 to the extent that the analog electronic switch can withstand, measure the compressed voltage value, and then multiply the compression factor back by the software to get the high-end potential of the nth battery. Similarly, the low-end potential of the nth battery can be obtained, and then the two are subtracted by the software to get the single cell voltage of the nth battery. This method is feasible in theory, but it is difficult to achieve in practice. For example, the absolute error of a 40V potential is ±40mv through a test system with a test accuracy of 0.1%, while the absolute error of a 38V potential is ±38mv through a system with the same test accuracy. The absolute error between the two is ±78mv. Obviously, the relative error can reach 8%, which is far from meeting the requirements of the communication power supply monitoring system. Therefore, this subtractor method is impossible to implement in engineering, but its idea is very valuable for reference: if the problem of continuous accumulation of errors can be solved, it is possible to obtain satisfactory measurement results. For this reason, we use two high differential mode gain amplifiers to design a hardware direct subtraction circuit, and its principle circuit is shown in Figure 6.

Figure 6 Circuit for measuring single cell voltage using hardware direct subtraction method

In Figure 6, ICL7650 is an operational amplifier with a differential mode gain of up to 105/mV, which can ensure that the potentials of the operational amplifier's non-inverting input and inverting input are equal, both equal to the ground potential. Rnp is a balancing resistor to ensure the operation of the operational amplifier. Vna is the high-end potential of battery n, and Vnb is the low-end potential of battery n.

The basic principle is as follows: Operational amplifier A forms an inverting amplifier, that is:

Operational amplifier B forms an adder, that is:

It can be seen from formula (2) that as long as the resistance values ​​of Rn1, Rn2, Rn3, Rn4 and Rn5 are reasonably selected to meet the conditions:

Then formula (2) can be transformed into:

This enables direct hardware subtraction and avoids the accumulation of errors.

4.2 Selection of component parameters

The battery for communication is usually composed of 24 cells with a single voltage of 2V. Its highest common mode voltage can reach about 60V. It needs to be moved to a ground voltage of about 2V and ensure the working safety of the operational amplifier. Therefore, it is more appropriate to choose between 25 and 35. Considering other factors such as the thermal stability of the resistor, we choose the resistance value of Rn2 and Rn3 as 1.5kΩ, and Rn1, Rn4 and Rn5 as 50kΩ. At the same time, since it is difficult to ensure high accuracy at this level of resistance, a 5kΩ potentiometer should be added for adjustment.

5. Measurement of battery cell temperature

The temperature of the battery body is an important parameter of VRLA batteries, which has a great influence on the remaining capacity and service life of the battery. We use the digital temperature sensor DS1620 of Dallas Company to measure the battery body temperature. It has the advantages of wide temperature measurement range, stable reading, convenient interface with single chip microcomputer, etc. Its temperature measurement resolution can reach 0.50C. If adjusted by software, it can reach a higher accuracy of 0.10C, which is very suitable for measuring the temperature of battery cells. Here, we only explain the method of achieving 0.10C accuracy by software.

5.1 Further analysis of temperature measurement principle

To obtain a higher temperature measurement solution, in addition to knowing the temperature value directly read by the DS1620, it is also necessary to know the value of the counter at that temperature and the count value for each 10C increase in that temperature, which can be read from the nonlinear accumulator. The nonlinear accumulator circuit is used to compensate for the nonlinear effect of the temperature oscillator, which helps to obtain higher temperature measurement accuracy.

Use the microcontroller to control DS1620, convert the corrected temperature direct reading value into a decimal number (in units of 0.50C), recorded as temp_read. At the same time, read the value stored in the counter after the counting gate is closed, recorded as count_remain. Then read the value in the nonlinear accumulator as the count value per degree Celsius at this temperature, recorded as count_per_c. After the above parameters are determined, the actual temperature T with an accuracy of 0.10C can be calculated using the following formula, that is:

5.2 Software method to achieve 0.10C temperature measurement resolution

According to the above analysis, through software programming, the MCU can be used to control the DS1620 to achieve a temperature measurement resolution of 0.10C. The software flow chart is shown in Figure 7.

Figure 7 Program flow to achieve a temperature measurement resolution of 0.10C

The specific implementation process is as follows:

(1) Send the "write configuration" command to initialize the DS1620, setting it to single temperature conversion mode and processor control state. The command is 0CH, 03H;

(2) Send the "start conversion" command (EEH);

(3) Send the "read configuration" command to read the status register data. Repeat this command until the DONE bit is "1", which means that the temperature conversion has been completed;

(4) Send the "read temperature" command to read data from the temperature register and convert it into an integer temp_read;

(5) Send a "read counter" command to read a 9-bit value from the counter, i.e., count_remain;

(6) The value in the nonlinear accumulator is read into the counter. At this time, there is no data exchange between the external unit and the DS1620.

(7) Resend the "read counter" command to read the value of the counter at this time, that is, count_per_c;

(8) The accurate temperature value is calculated by formula (5-10).

6 Experimental Results

In the following experimental results, the standard source and calibration device used for the tests of various electrical quantities are ST-9020 electric energy meter field detector (0.01) level; the test battery is GFM200 of Nandu Company, and the capacity obtained by discharging it at a constant current of 0.1C discharge rate under rated load is used as the standard capacity; the test environment temperature is 240C. The test results of the battery remaining capacity are shown in Table 1; the test results of the battery single cell voltage are shown in Table 2.

From the above test results, it can be seen that the system has high measurement accuracy and can fully meet the requirements of the Technical Requirements for Centralized Monitoring System of Communication Power Supply and Air Conditioning. This proves the feasibility of the battery monitoring module design proposed in this paper in the centralized monitoring system of local communication power supply, and has engineering practical value.