Design of NiMH battery fast charger based on ATmega16

Introduction: Nickel-metal hydride battery is a widely used rechargeable battery, widely used in various electronic devices, such as mobile phones, cameras, walkmans and MP3s.

Nickel-metal hydride batteries are widely used rechargeable batteries in various electronic devices, such as mobile phones, cameras, walkmans and MP3 players. There are also many types of chargers based on nickel-metal hydride batteries on the market. According to research, most of the existing charging systems on the market use a fixed mode to charge the battery, and the charging time is too long (slow charging is about 10 hours, and fast charging is about 5 hours). Fast charging must increase the charging current, but due to the lack of necessary detection methods, it is easy to cause the battery to overheat or overcharge, affecting the battery life [1]. Based on the above reasons, the author designed an intelligent fast charger. The charger is based on the ATmega16 microcontroller, uses a switching power supply and fuzzy control technology, adopts a pulse charging method, and detects the charging status online in real time, achieving good results.

1) Working principle

The system consists of a 5V switching power supply circuit, a DC-DC conversion circuit, a voltage, current, and temperature detection circuit, and a discharge circuit. The charging principle is shown in Figure 1. The 5V switching power supply circuit provides a stable 5V DC power for the microcontroller and the main charging circuit. Its working principle is as follows: The microcontroller collects the current, voltage, and temperature signals during charging in real time through the internal integrated A/D converter, adjusts the duty cycle of the PWM control signal, controls the opening and closing of the MOS tube, and thus adjusts the amplitude of the pulse voltage and the size of the charging current. The pulse voltage output by the MOS tube is filtered by DC and becomes a stable DC power added to the battery to charge the battery. The discharge control depolarizes and discharges the battery in a timely manner.

Figure 1 Charging principle diagram

II) System hardware composition

The main charging circuit and detection circuit are shown in Figure 2. Due to space constraints, the 5V switching power supply circuit is not drawn, and only one circuit (this charger has two circuits and can charge two AA/AAA batteries at the same time) of the main charging circuit and detection circuit are drawn.

2.1AVR MCU

AVR microcontroller is an enhanced high-speed 8-bit microcontroller with a reduced instruction set and built-in Flash developed by ATMEL in 1997. Atmega16 microcontroller is a high-performance microcontroller in the AVR microcontroller series. It has 16kB in-system programmable Flash program memory, 512b EEPROM, 1kB on-chip SRAM; 32 programmable general-purpose I/O ports; 32 general-purpose working registers; real-time clock (RTC); two 8-bit timers. Counters, one 16-bit timer. Counter, with channel PWM function; 8-channel 10-bit ADC; optional programmable gain; on-chip oscillator and programmable watchdog timer. It is a microcontroller with high reliability, strong functions, high speed, low power consumption and low price.

2.2 Charging main circuit

The main charging circuit is actually a buck converter, which is driven by the AVR's PWM through the bipolar NPN transistor 8050 to drive the P-channel MOS tube IRF9540. The switch tube is connected to the inductor, diode and capacitor, and the diode D2 is used to prevent the battery from supplying power to the microprocessor when the power is off.

Figure 2 Charging main circuit and detection circuit

2.3 Detection Circuit

2.3.1 Voltage Detection

In order to monitor the charging voltage between the positive and negative poles of the battery, an op amp is used in the design. The op amp circuit for measuring the battery voltage is a common differential op amp circuit. The measurement range of the analog-to-digital converter ADC (10-bit ADC, 1024 quantization units) integrated in the microcontroller is AGND~AREF, where AGND is the ground voltage and AREF is the reference voltage. Taking R8=R10=10k8, R9=R11=11k8, the voltage value corresponding to a certain quantization unit can be calculated: Umeasured=N×2193(mV), where N is the number collected and converted by the microcontroller ADC.

2.3.2 Current Detection

The charging current is obtained through the 0118 precision resistor R13. In order to improve the measurement accuracy, this voltage is amplified by the operational amplifier and then sent to the ADC of the microcontroller. Taking R12=1k8, R14=10k8, the current value corresponding to a certain quantitative unit can be calculated: Imeasured=N×2193(mA).

2.3.3 Temperature measurement

The battery temperature is measured by the negative temperature coefficient resistor R15, whose resistance is approximately 10k8 at 25℃. The NTC is part of the voltage divider and is powered by the reference voltage (AREF, 3.3V). From Figure 2, we can get:

Since the resistance of NTC is not linear, in practical applications, the table lookup method can be used to find the corresponding temperature.

2.4 Discharge control

The battery discharge is mainly carried out through the high-power resistor R5. The transistor D880 mainly acts as a switch. Usually, D880 is in the off state. When discharge is required, the microcontroller gives a discharge signal, D880 is turned on, and the battery starts to discharge.

3) Fuzzy control and software design

Research has shown that the charging current characteristic curve is related to the ambient temperature, the age of the battery, the depth of discharge of the battery, the battery charging capacity and the change of the load current. It is a nonlinear, time-varying system. Therefore, it is difficult to find an accurate mathematical model to describe the characteristic curve. The most effective way is to apply fuzzy mathematics theory to fuzzy control, establish fuzzy control rules based on the knowledge and experience of experts, complete the control decision-making process through fuzzy logic reasoning, and finally achieve regulation and control of the controlled object. From the charging characteristics of nickel-hydrogen batteries, it can be seen that the charging current that nickel-hydrogen batteries can accept is different in different charging stages. After the charging voltage reaches the peak value, the battery terminal voltage changes little and is relatively flat, with only a small increment. However, if the charging current is too large, the battery temperature will rise rapidly, seriously affecting the battery life and even causing the battery to explode [2]. Therefore, the end of charging becomes the key point of control. Based on the above reasons, the charging rules are formulated as follows: In the first stage, a preset small current (0.1C) is used for constant current charging (where C represents the capacity of the battery, such as 1800mAh, 0.1C represents charging with a current of 180mA), and when the battery voltage exceeds a certain threshold (1.0V), the second stage is reached, which can prevent the over-discharged battery from being damaged when charged with a large current; in the second stage, a preset large current (1C) is used for constant current pulse charging, and depolarization discharge is performed in a timely manner; in the third stage, when the battery voltage exceeds the second threshold (1.45V), the fuzzy control method is used to adaptively control the charging current, so that the charging current gradually decreases as the battery voltage increases, so as to achieve the best charging effect. When charging is terminated, the highest voltage detection and the temperature change rate △T/△t detection are used together, and the maximum charging time and the highest battery temperature are set at the same time.

3.1 Fuzzy Controller

The fuzzy controller takes the difference between the ideal voltage and the actual voltage △U and its change rate △Un-△Un-1 as input, and the charging current I as output, forming a two-dimensional fuzzy control system. During the charging process, the battery is depolarized and discharged in time. Both △U and △Un-△Un-1 have 5 language values, namely "NB(-2)", "NS(-1)", "Z(0)", "PS(+1)", and "PB(+2)". The fuzzy control table is shown in Table 1.

Table 1 Fuzzy control table (I)

In Table 1, "I" is 0, indicating that the charging current remains unchanged; "+1" indicates that the current increases by one level; "-1" indicates that the current decreases by one level; "+2", "-2" and so on. The rule table is pre-stored in the EEPROM of the microcontroller, and the program selects the corresponding control rule by looking up the table.

3.2 System Software Process

The system software flow chart is shown in Figure 3.

Figure 3 System software flow chart

IV) Conclusion

(1) This system was used to conduct multiple fast charging experiments on AA Super NiMH batteries. By combining pulse charging with fuzzy control, the battery temperature rise was reduced by an average of about 5°C compared to constant current (1C) charging.

(2) Due to the use of advanced charging control strategies, the charging voltage and current are intelligently controlled, which greatly shortens the charging time (it can be fully charged in 1h to 2h), and the charging speed is increased by 3 to 5 times compared with conventional chargers.

(3) The charging process is safe and reliable and will not damage the battery or shorten the battery life, thus achieving better results.

references:

[1] Zhang Xing, Yang Weimin, Li Gangyuan. Research on intelligent battery charging device[J]. Journal of University of Shanghai for Science and Technology, 2004(4):3812384.

[2] Chang Jiang, Jing Zhanrong, Gao Tian. Design and application of fuzzy neural controller based on HGA[J]. Computer Engineering and Applications, 2006(14):2222224.