Boost converter for solar battery. Stabilized voltage converter on the YX8018 chip. Network type converter

Poonam Deshpande

Electronic Design

A simple combination of a solar battery, several LEDs and a small DC/DC regulator will allow you to illuminate dark corners of the room during the daytime and at the same time provide stabilized power to low-power loads

A solar-powered lamp that only runs during the daytime may seem virtually useless, but there are many areas of homes and offices that remain relatively dark even during the day. This “daylight” glows from a nearby solar panel, and in addition has an additional stabilized 0.5 W source capable of powering small loads such as a VHF receiver.

A photovoltaic panel with a nominal power of 10 W is used to power the fluorescent lamp (Figure 1). Its voltage, at the point of maximum power equal to 17.3 V, powers two identical LED chains (LED1... LED5 and LED6... LED10). Each chain consists of five white LEDs with a power of 1 W each. Series resistors R1 and R2 with a resistance of 22 Ohms with a permissible dissipation power of 2 W set the currents of the circuits.

The output of the photovoltaic panel is connected through a switch to the input of a switching voltage stabilizer (PVS) (Figure 2). The capacitor at the input of the converter chip reduces the dependence of the brightness of the LEDs on changes in the load current, which depends on the level of the audio signal at the output of the VHF receiver.

There are quite a few cheap switching voltage converter ICs that are well suited for this application, and three of them are very similar in prevalence, switching frequency, output voltage, L and C values, and load resistance. These are LM3524, MC34063 and LM2575. All other things being equal, an IC-based converter loses less battery voltage due to lower current consumption and lower power switch saturation voltage. It is clear that this particular microcircuit was chosen for the power source.

The input supply voltage (V IN) is supplied to pin 6 of the MC34063 DC/DC converter through the SW switch (Figure 3). The 2200 µF smoothing capacitor C1, located after the switch, is designed to minimize voltage fluctuations caused by changes in light intensity. Capacitor C2 with a capacity of 100 pF at pin 5 sets the converter switching frequency to 33 kHz.

The output voltage is filtered by elements L1 and C3. The 220 μH inductance is made independently by winding 48 turns of wire on a toroidal core, for which it is quite possible to use a core with a diameter of 10 mm and a height of 20 mm, extracted from an old computer cable. The resistances of resistors R1 and R2 are selected so that the output voltage is 5 V. If the output should have a different voltage, the resistance of resistor R1 should be changed. For example, for an output voltage of 6 V, the resistance of R1 should be 27 kOhm, and for 4.5 V - about 39 kOhm. The assembled circuit is shown in Figure 4, and the complete system is shown in Figure 5.

To get more light, you can make a day lamp with two solar panels connected in series (Figure 6). However, in this case, the maximum output voltage of the photovoltaic source may exceed 40 V, which is the limit value set for the MC34063 chip. To solve this problem, the DC/DC converter is not connected directly to the output of the solar panel, but to one of two LED strings. Each chain consists of ten LEDs with a maximum forward voltage of 3.5 V. Thus, the voltage on the chain does not exceed 35 V.

Links

Related materials

Switching DC/DC converters DC DC CONVERTER CONTROL CIRCUITS

• Super!!! Light up during the day, darken at night!!! Everything is simply ingenious!!! Now I finally understand what a “fluorescent lamp” is!!!
• The above is not our way! Our people are much more economical! Ours, a domestic young technician, a 5th grade student. buys a dynamo flashlight for 19 UAH. (RUB 40-45) and... just puts it in his pocket. Savings - $20 on purchasing a solar panel and all sorts of resistor diodes from foreign capitalists. http://www.leroymerlin.ua/p/%D0%9B%D...4-307ee51a3035. Would you say it's inconvenient? Under the guidance of a retired former physics teacher from the “Crazy Hands” school club, the student, having learned the multiplication table by the 5th grade, calculates the work his grandmother does when opening the door to a dark pantry: he multiplies 2 kgf of effort by 1 meter of edge movement doors and receives 20 joules. Looking into the school physics room, the student learns that 2 LEDs of the mentioned flashlight at a voltage of 2 volts and a current of 10 milliamps have a power consumption of only 20 mW! By opening the door just once, you can illuminate the pantry for as much as 50 seconds - the energy in the flashlight does not disappear, but charges the battery built into the Chinese flashlight! Now the whole family of the young talent opens and closes the door to the pantry during morning exercises - the student’s father, during a break in a football match, attached a dynamo flashlight to the door to the pantry! And our student’s younger brother attached a switch to the same door from the door of the old refrigerator - when the pantry is closed, there is no light in the pantry - the flashlight battery does not discharge. They are already collecting signatures for petitions to the Government. If each of the 100 million residents saved just 100 watts of electricity, it would be possible to close all the country's power plants forever! Details and further actions - https://www.youtube.com/watch?v=WVMolYlx-h8.
• A. Raikin wanted to tie a dynamo to the ballerina...
• What for a goat accordion and an accordion for the ass? the receiver can be powered by free energy and what the hell with that solar panel
• Give a working example...don’t suggest a detector receiver.

The YX8018 chip is widely used in inexpensive LED lawn lights, where an unstabilized step-up voltage converter is built on it. It powers the lighting LED(s) from a Ni-Cd battery. The current through the LED (from fractions to several milliamps) is set by the inductance of the storage choke in the converter. Therefore there is no need to stabilize the voltage. A special feature of the YX8018 and similar microcircuits is the presence of a control input, with which you can also turn on the voltage converter switch. It is this input that is used in LED lawn lights to automatically turn them on after dark. The same input can be used to build a stabilized boost voltage converter.

The circuit of such a converter on the YX8018 chip is shown in Fig. 1. It can be used to power from one Ni-Cd, Ni-Mh battery or galvanic cell of various radio-electronic devices requiring a supply voltage of 2 to 5 V. In the initial state, there is a voltage close to the voltage at the CE input (pin 3) of the microcircuit nutrition. This is due to the presence of a built-in resistor connecting this pin to the power supply positive. Therefore, the converter turns on, the voltage pulses at its output L (pin 1) are rectified by the diode VD1, and the smoothing capacitors C2 and C3 are charged - the output voltage increases. When the gate voltage of transistor VT1 reaches a threshold value (about 2 V), the resistance of the transistor channel will decrease and the voltage at its source (and the CE input of the microcircuit) will also decrease - the converter will turn off. The output voltage will begin to decrease, which will lead to the closing of the field-effect transistor and turning on the converter.

Thus, the converter periodically turns on and off, maintaining the output voltage set by trimming resistor R1. The operating frequency of the converter is about 200 kHz, and the on/off frequency depends on the output current and the capacitance of capacitor C2 (the higher the current and the smaller the capacitor capacitance, the higher the frequency) and can range from several hertz to tens of kilohertz. The dependences of the output voltage of the converter (2.7 V) on the input voltage for different values ​​of the load current and the limit values ​​of the load current are presented in Fig. 2. The ripple amplitude is about 10 mV, remains almost unchanged and depends within small limits on the output voltage and parameters of the field-effect transistor. The ripple frequency depends on the operating frequency of the converter and the frequency of switching on/off the converter and can vary within wide limits. Thermal stability is determined primarily by the parameters of the field-effect transistor. In this case, the temperature coefficient of voltage is negative and amounts to several millivolts per degree Celsius.

All elements can be mounted on a single-sided printed circuit board made of foil fiberglass, its drawing is shown in Fig. 3. A tuning resistor SP3-19 was used, the oxide capacitor was imported, the rest were K10-17. Instead of the 1N5817 diode, low-power pulsed or detector germanium diodes or Schottky diodes can be used. The inductor is wound on a ferrite ring with a diameter of 6...9 mm from the electronic ballast transformer of a compact fluorescent lamp and contains 5 turns of PEV-2 0.4 wire. The output voltage in the range of 2.2.5 V is set with a trimming resistor; it can be replaced with a resistive divider with a total resistance of at least 1 MOhm. To reduce ripple with a frequency of 200 kHz between capacitors C2 and C3, you need to install a choke, for example EC24, with an inductance of 470...1000 μH in the positive power line.

Publication date: 07.05.2014

Readers' opinions

• Sergey (other) / 04/14/2019 - 14:49
  And garden lamps do not need to “shine all night.” They need it to “shine all evening and part of the night.” They are also a “decorative element”. For lighting and other beauty. And not at all for illuminating anything with “bright light”. They don't have to keep the light on all night.
• Sergey / 08/13/2018 - 12:12
  The problem with garden lamps is that the sun is weak; it doesn’t feed the battery enough, and therefore it’s not enough even for the night. I paralleled two solar ones - now after a day there is 18 hours of sunshine.
• clim / 06/09/2018 - 07:25
  in the datasheet there are just 2 options - from 1 and from 2 batteries
• clim / 06/09/2018 - 07:24
  I checked the lawn lamp, the solar battery is 4*4 cm, in the bright sun it gives up to 10 mA, not microamperes, so everything is ok, it can be fully charged in a day (solar)
• badgers / 01/05/2018 - 08:18
  I looked through all the “data-seets” - nowhere is the MAXIMUM input voltage for the YX8018 specified, specifically is it possible to give 3.2 V (when powering the flashlight from two elements), in practice it seems to work, but I would like to act according to legal specifications, I am trained as a designer ...
• z123 / 12/10/2017 - 00:36
  The solar cell provides a microampere current and cannot in any way charge a battery that requires at least tens of MILLIAMPS. Support (so that she lives longer) - maybe. But don't charge. Therefore, circuits where only this YX8018 + battery, resistor, switch, LED and solar element = this is a circuit for a short time, then the battery dies and that’s it. Either dispose of it (for spare parts) or convert it into something completely different. Those who make and sell this are cheaters. Counting on fools to fool and swindle. And then it doesn’t matter anymore.
• Grandfather Sergey / 10/07/2017 - 00:04
  No, for some this topic is really relevant, there is no need to laugh in vain. I also have this problem - there are a lot of batteries left with a resource of 10-30%. They are no longer suitable for a flashlight; for other devices, it is better to buy new ones. But the YX1808 for night lighting of my apartment, as long as it doesn’t fit into the door with its forehead in the dark, is just IT! And, if the LED in THIS device has already gone out, then THIS battery is truly dead. No other device will suck anything out of it! You can safely say thank you to her for her cooperation and, saying goodbye, dispose of it.
• Danil / 05/30/2017 - 14:28
  How to charge a phone using this chip? What would be powered by the sun and charge your phone?
• Dmitry / 05/16/2017 - 23:36
  Yuri, the end of the wire that comes from the middle of the resistor should continue to the transistor at control input 3. In the picture it is cut off. According to the logic of work, it should be like this. I bought a lamp with such a converter and immediately disassembled it. The plus of the solar cell is soldered to input 3. It is not for charging, but just a light sensor. You need to charge the AAA battery yourself by removing it from the lamp.
• Andrey / 05.25.2016 - 16:32
  Fixed prices sell garden night lights. Inside there is a 4-pin YX8018 microcircuit, an LED, a nickel tablet, a solar panel, a switch and, like, a choke for a resistor type. It charges during the day, and if you cover the diesel fuel (or in the evening), the diode lights up. Googled it a bit. 8018 is a DC-DC converter for solar panel
• Yuri / 03/22/2015 - 18:05
  Is the author mistaken about the internal resistor at pin 3? Most likely it is connected to ground.
• TL494 / 12/16/2014 - 13:10
  And if you calculate how much a kW/hour stored in HIT costs? Everything is quite natural. Although at home I recycle old batteries in batches of 2-3, to zero, without any diagrams.
• Vladislav / 06.12.2014 - 15:25
  Dear I Nechaev, Thank you for your publication, it is relevant for me, since I am looking for a low-cost circuit for recycling voltage of about 1 volt at XX, there is something to recycle in large quantities. In garden lanterns, a similar circuit, such as JD 1803B, probably works most likely. THESE CHARACTERISTICS CAN NOT BE FOUND ON IT, on some of these flashlight controllers there are no markings at all, THERE IS ANALOGUE ANA 608-6, ANA 618 BUT there are Chinese symbols, there are other controllers like max 1724 or 1722 and others that work from 0.7 - 0.8 volts with an output voltage of up to 5.5 volts at a current of 150 to 300 mA, since I am not a strong electronics engineer, I need additional. discussing the circuit design, my skype vladislav14211 mail [email protected] I will be glad to cooperate and discuss the technical solution I need based on your scheme
• Sergey / 05/10/2014 - 07:18
  Get several ma at 9...15 volts from one element a larger capacity is sufficient - this is understandable. For example, to power a multimeter. I assembled similar circuits myself if necessary. But from the voltage that 1 element gives you get 2 volts, this is strong, guys!!! This is more likely due to an excess of time. I understand a man who finds himself in the heat of the “promised homeland” (look at this site) But in the imperial capital, when you spit, you end up in a store or kiosk where there is a heap of batteries.

The device is a simple boost converter and voltage limiter that charges 12V batteries from a 6V solar panel. The device also has MPPT (Maximum Power Point Tracking) function. When we think of MPPT, we usually think of microcontrollers and complex power computing algorithms. However, such algorithms are not really needed.

The article presents two schematic solutions. The first circuit simply illustrates a boost switching converter, while the second shows a homemade working circuit of the device. It is recommended for more advanced experimenters who have an oscilloscope at their disposal. The circuit may also be of interest to students and those who simply want to expand their knowledge of electronics.

Boost converter topology diagrams and homemade solar converter circuit diagram

TheoreticalintelligenceOincreasingconverter

In the boost converter topology diagram, coil L1 is charged when transistor Q1 is on. When transistor Q1 is turned off, coil L1 discharges to the battery through zener diode D1. Performing this operation several thousand times per second will result in a significant output current. This process is also called inductive discharge. For it to function, the input voltage must be lower than the output voltage. Also, if you have a solar panel, you must use an energy storage element - a capacitor (C1), which will allow the solar panel to continuously output current between cycles.

Description of the boost converter circuit diagram

The circuit consists of three main blocks, including a 555 MOS gate generator, a 555 PWM modulator, and an operational amplifier with a voltage limiter. The 555 series with cascaded output can provide a current of about 200mA and makes an excellent low power pulse generator. The 555 PWM modulator is a classic oscillator circuit based on the 555 series. To adjust the discharge time of capacitor C3 (coil charging time), a voltage of 5V is applied to pin 5.

Limitationvoltage

Operational amplifier U1A calculates the battery voltage signal when the divided voltage setpoint is compared with the 5V reference voltage. When the voltage exceeds the set value, the output switches in the negative direction, thereby reducing the frequency of the generator's PWM pulses and limiting any subsequent charge. This effectively prevents overcharging.

Powering the circuit from a solar panel

To prevent unnecessary battery drain when the sun is not shining, all circuits are powered through the solar panel, with the exception of the closed-loop voltage divider, which draws about 280uA.

MOSFET logiclevel

Since the circuit must operate at low voltage levels (this circuit operates from an input voltage of at least 4V), it is necessary to install a logic level MOSFET. It will open at a voltage of 4.5V. For this purpose I used a power MOSFET transistor MTP3055.

Voltage clamping using a zener diodeD2

In this circuit, DO NOT DISCONNECT the battery, otherwise the MOSFET transistor will burn out. Therefore, to protect it, I installed a 24V zener diode D2. Without this zener diode, I myself have burned out many MOS transistors.

MPPT function

When the solar panel voltage/current increases, the PWM generator increases the pulse frequency, which in turn causes the output current to increase. At the same time, additional voltage is applied to the coil, thus increasing its charging current. The result is that the boost converter actually "goes hard" when the voltage goes up, or "goes hard" when the voltage goes down. To maximize energy transfer in bright sunlight, potentiometer R8 is adjusted so that the battery charging current is maximum - this will be the point of maximum power. If the circuit is working correctly, there will be a very flat peak when R2 is rotated. Diode D3 performs automatic MPPT regulation more accurately by subtracting a fixed voltage from the voltage difference between the battery and the average voltage through capacitor C3. In low light conditions you will find that resistor R3 is not optimal, however it will not be completely removed from the chain. Note that smart MPPT controllers can also perform better at full range, but this improvement is extremely ineffective.

Component ratings

The circuit is configured for a voltage of 9V, the solar panel for a power of 3W. Boost converters are quite finicky and won't work over a wide range of conditions - if your system uses different power rating limits for the solar panel, then expect a problem. The only components that require adjustment are coil L1 and capacitor C3. I was surprised that the repetition rate was very low (about 2kHz). I started with a 100µH coil, but the circuit works better at 390µH - I originally wanted around 20kHz. For best performance, charge the coil 5 to 10 times the solar panel current, then allow a long period of time (3X) to allow the coil to fully discharge. This will ensure acceptable operation when the power supply voltage is close to the battery voltage. Note that low impedance coils provide the best efficiency. The greatest loss really occurs in a Schottky diode, and the least loss is what these diodes are designed for.

High frequency operation is usually preferred. This will minimize the size of the coil. However, for experimentation, use the coil that will work best.

The proposed components are indicated in the diagram. Naturally, the charger can be adapted to suit your requirements.

Oscillograms

List of radioelements

There are different opinions and different numbers about the efficiency of PWM and MPPT controllers. For some, the PWM controller is more effective in cloudy weather, and MPPT works better in sunny weather. For others, the MPPT controller works better in all respects, and there are those who claim that PWM is much better. But you shouldn’t believe everything at once and take an unambiguous point of view; in each case you need to separately understand why and how it works. There are people who don’t even really know how to use their controllers and then say that they are worse or better.

Conventional PWM (PWM) controllers work very simply and the current from the solar panels passes through them almost directly, the power drop on the power transistors is very small. Therefore, as soon as the solar battery voltage exceeds the battery voltage by about 0.5-1 volts, the battery begins charging. But these controllers do not know how to extract all the power from the solar panel. For solar panels, the maximum current cannot exceed its maximum, for example, for a 12 volt solar panel with a power of 100 watts, the load current is no more than 5.7A. And when our battery voltage is about 13-14 volts, then the power going to the battery will be 14 * 5.7 = 79.8 watts, if the battery is discharged to 12 volts, then the power will be even less. In this case, more than 80% of the maximum power of the solar panel cannot be obtained.

But if the battery voltage were not 13-14 volts, but for example 17 volts, then 18*5.7=96.9 watts. In general, in order to extract all the power from a solar panel in the sun, it is enough for it to have 30 elements, and not 36, but then in cloudy weather such a panel will practically not work, which is why they make panels with a standard 36 elements for a 12V battery, and at idle the voltage is about 21-22 volts for such panels. But in the characteristics they write the full power of the panel, and not when operating on a 12 volt battery through a PWM controller.

MPPT controllers work differently, they have a DC-DC converter that converts high voltage to lower voltage, increasing the charge current. The controller scans the voltage and current of the solar panel, and removes power at the point where the maximum voltage of the solar panel is at maximum current, and then converts it to a low voltage to charge the battery. For example, if the panel is 12 volts, then its maximum power will be at 17-18 volts.

But since in MPPT controllers the work occurs through a DC-DC converter, it has its own efficiency, which is usually 90-96%, depending on the operating mode. The DC-DC module itself, in active mode, consumes its energy no matter how much the battery transmits. This is like the inverter has consumption at idle, and DC-DC also has its consumption. This suggests that if in cloudy weather the power from the solar panels is too small, then simply DC-DC operation can consume all this power and nothing will get into the battery, or much less than directly through the PWM controller.

For DC-DC to work, the voltage must be higher than the output by about 1.5-2 volts, this means that when the voltage on the solar panel drops to 15 volts, charging will stop. But now there are different MPPT controllers, some switch to PWM mode when the voltage and current are very small. There are some that stop working at low power and do not charge the battery. Some simply cannot determine the MPPT point at low power and constantly search for it, wasting energy from the battery, that is, they do not charge, but rather discharge it for the useless operation of the DC-DC module.

I now have two controllers, Solar 30 and Photon 100 50, and I compared how they work from dawn until the sun appears. I filmed all this, and this is what I got:

This test showed a clear victory for a specific MPPT controller over a specific PWM controller. Although Solar 30 says that it is MPPT, this is nothing more than a marketing ploy, it is just a PWM controller.

In the end, what can we say about all this? Even in cloudy weather, a good MPPT is not inferior to PWM, and as soon as conditions allow you to take more from the solar panel, the MPPT controller works much better. Well, if the power from a solar panel or array of panels in cloudy weather is even theoretically 1-2% of the nominal, then there is no point in fighting for these drops. It's better to shoot up to 20% more in brighter light.