Detailed explanation of the technical principle of battery capacity meter

As environmental awareness gradually increases, countries around the world are competing to develop environmentally friendly cars. my country is also investing in the development of battery-powered electric cars. An indispensable instrument for electric cars is the battery capacity meter, which, like the fuel gauge of an ordinary car, tells the user how much capacity is left in the battery and how many kilometers it can travel. In fact, not only electric cars need a battery capacity meter, but many occasions where batteries are used have an urgent need for it. The traditional means of monitoring batteries is just a voltmeter, but the voltage cannot accurately reflect the capacity of the battery. The voltage is normal but the capacity is often missing. As a user, what often confuses you is that you don't know how long the battery can be used, which affects the use of many key occasions and is prone to misjudgment and accidents. Therefore, it is very necessary to develop an instrument that reflects the battery capacity. At present, similar products have been launched abroad, but perhaps due to technical confidentiality, no introduction has been made to their implementation methods.

This article uses electric vehicles as the object of use and proposes a battery capacity meter that uses a power measurement method to measure the battery capacity under certain conditions. It is based on the principle that the energy charged into and released from the battery is calculated and multiplied by the corresponding loss coefficient to indicate the battery capacity (the coefficient should take into account the charging efficiency and the battery discharge current size and other factors that affect the battery capacity).

Battery Capacity Meter Basic Principle

In addition to some factors of the battery itself, the capacity of the battery mainly depends on the charge and discharge amount. Obviously, if the charge and discharge conditions of the battery can be recorded at all times, the capacity can be measured. We imagine equipping a traditional single battery with an instrument called a battery capacity meter to achieve the purpose of displaying the capacity. The capacity meter dynamically monitors the total amount of electricity charged into the battery and the total amount of electricity discharged, and displays it intuitively after calculation. Other factors that affect the battery capacity are combined into a loss coefficient, and the coefficient multiplied by the arithmetic sum of the charge and discharge amount is the remaining capacity of the battery. Since the types, sizes, and performances of batteries are different, the loss coefficients are different and are mainly obtained through experiments. Therefore, the coefficient problem is not discussed here, and only the circuit that completes the function of measuring electricity is studied.

There are many ways to charge and discharge batteries, such as constant current, voltage limiting, pulse, negative pulse, etc. Therefore, the method of simply multiplying current by time to measure capacity cannot adapt to other methods except constant current, and the integral method cannot adapt to the needs of negative pulse charging. At the same time, it requires time parameters and is not very suitable. Obviously, the design of the battery capacity meter should meet most charging and discharging methods. Regardless of the charging method, the key parameters that affect the battery capacity are current and time. In the case of negative pulse charging, there is only a negative current at the same time. For this reason, we designed a battery capacity meter circuit with the following working mode, and the principle block diagram is shown in Figure 1.

First, the battery's charge and discharge current is monitored, converted into a voltage signal, amplified, and sent to a voltage-frequency converter to convert it into a frequency signal. Finally, it is sent to a counter to record the number of pulses, and the count value is displayed in a certain way, which constitutes a battery capacity meter. In fact, the frequency represents the magnitude of the current. The larger the current, the higher the frequency, and the more pulses are recorded in the same time, and vice versa. The charge and discharge time is also reflected in the counting of pulses. The longer the time, the more counts. In this way, the calculation of the battery charge and discharge capacity is completed by counting.

Figure 1 Schematic diagram of battery capacity meter

The combination of the absolute value amplifier and the reversible counter realizes the measurement of the discharge gap during charging (i.e. negative pulse charging), and at the same time uses a set of circuits to complete the calculation of both charging and discharging directions. The counting direction of the reversible counter is controlled by the direction of current flow, with forward counting during charging and reverse counting (subtraction) during discharging.

Solution demonstration and key technical solutions

1 Current sampling

The purpose of current sampling is to convert the current signal into a voltage signal. There are generally three ways:

(1) Sampling resistor;

(2) Shunt;

(3) Hall devices (including transformers).

From the perspective of electric vehicle battery use, the current is large, so it is obviously not appropriate to use a sampling resistor, and the shunt is too heavy and large in size, so it is not very suitable. Therefore, the Hall device is more suitable. Its advantages are that the linearity is better than 0.1%, which is suitable for tracking a large range, good dynamic performance, and a response time of less than 1μs, so that the instantaneous current of the car starting can be tracked instantly. In addition, its small size and light weight are suitable for installation in the car. Its disadvantage is that the price is slightly expensive, but it can be completely ignored for the price of batteries used in cars. Since mature products that can be purchased are selected, the circuit is relatively simple and is no longer listed.

2 Absolute value amplifier

Since the directions of charging and discharging currents are different, an absolute value amplifier is used to amplify the positive and negative signals output by the Hall device into a positive signal, which is then sent to the voltage-frequency converter.

There are many ways to design an absolute value amplifier. From the perspective of power supply, there are two methods: single power supply and dual power supply. The number of op amps used is one or more. Since this machine uses Hall devices and is a bidirectional current, a single power supply has no advantages. A single op amp amplifier has too many resistance values, high precision requirements, and the load should also be considered, so it is not very suitable.

This machine uses an absolute value amplifier composed of two operational amplifiers. The low offset and low drift operational amplifier 0P-07 is selected. It has high precision and its performance is not affected by the load. The accuracy of the absolute value amplifier is required here, not to contribute to the system accuracy, but from another point of view. This is what was mentioned earlier. As far as the battery capacity meter is concerned, the best way to monitor the battery should be to be integrated with the battery and always monitor the battery status. This requires that when the battery has no charge or discharge current, the output of the amplifier is zero. Otherwise, after a long period of shelving, the capacity meter indicates full or empty due to the error of the amplifier, resulting in misjudgment. The prototype designed with high precision, low offset and low drift has a full-scale error of 1mv and a zero-scale error of less than 1mv. See Figure 2.

Figure 2 Schematic diagram of absolute value amplifier

3 Voltage-to-frequency converter

The voltage-frequency converter is the core part of the battery capacity meter, which is responsible for converting the amplified signal into a frequency signal. Its linearity and accuracy directly affect the whole machine. There are many ways to achieve voltage-frequency conversion. In terms of form, there are two forms: discrete components and dedicated integrated chips. The general discrete components have higher accuracy, volume, and adjustment complexity than integrated chips, but their prices are lower. The dedicated chips have lower linearity, voltage stability, accuracy and other indicators at an acceptable price. Considering the issues of volume, full-range tracking of charge and discharge, and performance-price ratio, we selected VFC32 as the voltage-frequency conversion device. The good linearity of the device provides a guarantee for the full-range tracking accuracy, and reduces the volume with fewer components. The circuit principle is shown in Figure 3.

Figure 3 Schematic diagram of voltage-to-frequency converter

4 Reversible counter

The counter part is all CMOS circuit, firstly, the power consumption is low, which is very important for battery power supply; secondly, its level matches the operational amplifier level and increases the display range. See Figure 4.

Figure 4 Schematic diagram of reversible counter

A 14-level pulse carry binary counter 4020 and two 4-bit reversible binary counters 4516 are used to form a 21-level counter. The high 7-bit counter value is valid as the count value and output, while the low 14 bits are only used for counting and not for output. In addition, 4020 is a unidirectional counter without a subtraction function.

This design has two major advantages:

(1) 4020 is a highly integrated counter that can replace three and a half 4516s, thus greatly reducing the size.

(2) When used for addition, 4020 can be accurate to the lowest digit; when used for subtraction, the error is the lowest 14 digits, but this 14 digits is also the maximum error at one time and has no accumulation, because the circuit uses a mixed method of asynchronous and synchronous counting. When subtracting 14 numbers (although 4020 is adding), 4020 outputs an asynchronous pulse 4516 minus "1", just like a real subtraction, and the value of 4020 cannot be output, which makes the result very accurate.

5 Control circuit

This part includes preset circuit, anti-overflow circuit and counting direction control circuit.

In order to have a wide range of applications, this prototype has added a toggle switch to the preset and control circuits of the counter, so that the initial value and final value of the counting part can be set by the toggle switch, which can achieve the purpose of detecting the use of known battery capacitance, which is more convenient.

At the same time, in order to prevent the counter from overflowing in both directions, anti-overflow circuits are set up respectively so that the counter will stop counting when it reaches zero or full value to prevent errors.

By comparing the current flow direction, the output pulse controls the counting of the reversible counter to form a direction control circuit.

6 Display

There are two display modes: digital and pointer. To ensure intuitive display, the original voltmeter indicating battery voltage is still used in the car, while the instrument of ordinary car is used as much as possible. A switch is set on the voltmeter to switch the indication of voltage and capacity, which is more convenient.

This requires converting the binary number of the counter into voltage. Obviously, D/A conversion is possible, but the circuit complexity increases and the cost also increases. Therefore, in order to simplify the circuit, we only borrow the idea of ​​D/A conversion network and use the weighted resistor T-shaped network to convert the 7-bit value of 4516 into analog output to drive the voltage meter indication, as shown in Figure 5.

Figure 5 shows the circuit schematic

7 Working power supply part

The battery capacity meter is different from other instruments in that it can only use batteries as power source. However, due to the changes and fluctuations in battery voltage, direct use is obviously inappropriate. Therefore, a secondary power source must be generated by the battery.

First of all, the Hall device needs a power supply of ±12V, and the circuit control counting and other parts also use ±12V. In addition, we consider that in order to make the capacity indication more intuitive and clear, its maximum voltage range should be larger, and at the same time, it can make full use of its voltmeter effective indication. Its voltmeter range is 40V, and the maximum battery voltage is 30V, so the maximum capacity indication is set to 28V, which requires a power supply voltage of 30V.

Since the battery discharges a large current when it starts, the voltage fluctuates greatly, about 15 to 30V. In order to adapt to its changes and reduce the power consumption of the capacity meter itself and improve efficiency, the design uses switching power supplies.

First, +12V is obtained by using the LM2575 step-down regulator. The chip has an input voltage of up to 40V, a fixed oscillation frequency of 52kHz, and good voltage and current regulation rates, which meets the requirements of the capacity meter.

-12V is converted by using +12V as input through 34063DC/DC converter. This loses some power. Our original design used M2575HV (input voltage 60V) to directly introduce the battery voltage, but because the 60V LM2575HV could not be purchased, we had to give up. If there is a batch in the future, you can order it. Fortunately, the power of -12V is limited and the loss is small. A 30V power supply has high voltage and low current. If ordinary DC converters such as 2575 or other devices are used, the volume is too large, and the magnetic core components are greatly wasted, which is not worth the loss. Therefore, we have been looking for a simple method in the design. Finally, after experiments, we decided to use the 555 oscillator to boost the voltage and use the voltage doubler rectification method to increase 12V to 30V, and the effect is very good, as shown in Figure 6.

Product design and calculation

1 Setting the voltage/frequency relationship

Voltage 0～10V corresponds to frequency 0～10kHz

Figure 6 30V power supply schematic

Current 0～1000A corresponds to voltage 0～10V

The selection of these values ​​takes into account the optimal values ​​of the Hall element, amplifier, F/V conversion design and the needs of the test prototype.

2 counting digits

4020-14-bit 4516 has two pieces with 8 bits in total, which adds up to 22 bits. Only 21 bits are used, and the number of counts is:

221=2.097152×106.

Counting time for 10kHz

T=(221×1/104) seconds=3.49 minutes.

When 10kHz corresponds to 1000A, for a 45Ah battery

T=C/I=45/1000=0.045h=2.7 points<3.49 points,

It can be seen that the timing is enough, and the full-time timing ampere-hour is

(221×1/104)×1000/3600=58.25Ah.

3. Error calculation

The counting time for the first 14 levels is △T=214, the total counting time is T=221, and the relative error is △T/T=214/221=0.78%.

It can be seen that the error of the first 14 levels is very small, less than 1%, and it only appears once when doing subtraction, so it can be ignored. Therefore, it is reasonable to use one 4020 instead of three 4516.

Performance Test Results

The whole machine was tested under the condition of charging and discharging current of 15A. The voltage (representing capacity) indicated full capacity as 28.002V. After the battery capacity was fully discharged, the voltage (representing capacity) indicated 0V. The error between the indicated capacity and the actual capacity was 3%, which met the design requirements.

Under the mode that the output capacity is equal to the input capacity multiplied by the loss coefficient, this article takes electric vehicles as the use object and makes an in-depth and detailed explanation of the input sampling, absolute value amplification, voltage-frequency conversion, display and working power supply. It has carried out a very beneficial exploration and is one of the effective methods for measuring battery capacity at present. It is suitable for batteries with no memory effect and relatively stable performance.