Over the past 10-20 years, the number consumer electronics has grown many times. A huge variety of electronic components and ready-made modules appeared. Power requirements have also increased, many require a stabilized voltage or a stable current.

The driver is most often used as a current stabilizer for LEDs and charging car batteries. Such a source is now in every LED spotlight, lamp or lamp. Consider all options for stabilization, ranging from old and simple to the most effective and modern. They are also called led driver.

• 1. Types of stabilizers
• 2. Popular models
• 3. Stabilizer for LEDs
• 4. Driver for 220V
• 5. Current stabilizer, circuit
• 6. LM317
• 7. Adjustable current stabilizer
• 8. Prices in China

Types of stabilizers

Pulse adjustable DC

15 years ago, in my first year, I took tests in the subject "Power Sources" for electronic equipment. From then until today, the LM317 chip and its analogues, which belongs to the class of linear stabilizers, remain the most popular and popular.

At the moment, there are several types of voltage and current stabilizers:

1. linear up to 10A and input voltage up to 40V;
2. pulse with a high input voltage, lowering;
3. pulse with low input voltage, increasing.

On a pulse PWM controller, usually from 3 to 7 amperes according to the characteristics. In reality, it depends on the cooling system and efficiency in a particular mode. Boosting from a low input voltage makes a higher output voltage. This option is used for power supplies with a small number of volts. For example, in a car, when you need to make 19V or 45V out of 12V. With a buck, it's easier, the high is reduced to the desired level.

Read about all the ways to power LEDs in the article "to 12 and 220V". Connection schemes are described separately from the simplest ones for 20 rubles to full-fledged blocks with good functionality.

By functionality, they are divided into specialized and universal. Universal modules usually have 2 variable resistances to adjust the Volts and Amps output. Specialized ones most often do not have building elements and the output values ​​\u200b\u200bare fixed. Among the specialized, current stabilizers for LEDs are common, circuits in in large numbers is on the Internet.

Popular Models

Lm2596

Among the impulse ones, the LM2596 has become popular, but by modern standards it has a low efficiency. If more than 1 amp, then a heatsink is required. A small list of similar ones:

1. LM317
2. LM2576
3. LM2577
4. LM2596
5. MC34063

I will supplement with a modern Chinese assortment, which is good in terms of characteristics, but is much less common. On Aliexpress, the search for the marking helps. The list is compiled by online stores:

• MP2307DN
• XL4015
• MP1584EN
• XL6009
• XL6019
• XL4016
• XL4005
• L7986A

Also suitable for Chinese DRL daytime running lights. Due to the low cost, LEDs are connected through a resistor to a car battery or car network. But the voltage jumps up to 30 volts in pulses. Low-quality LEDs cannot withstand such surges and begin to die. Most likely you have seen flashing DRLs or running lights, for which some LEDs do not work.

Do-it-yourself circuit assembly on these elements will be simple. Mostly these are voltage stabilizers, which are switched on in the current stabilization mode.

Do not confuse the maximum voltage of the entire unit and the maximum voltage of the PWM controller. 20V low voltage capacitors can be installed on the unit when pulse chip has an input up to 35V.

LED Stabilizer

It is easiest to make a current stabilizer for LEDs with your own hands on the LM317, you only need to calculate the resistor for the LED on online calculator. Food can be used at hand, for example:

1. 19V laptop power supply;
2. from the printer for 24V and 32V;
3. from consumer electronics at 12 volts, 9V.

The advantages of such a converter are low price, easy to buy, minimum parts, high reliability. If the current stabilizer circuit is more complicated, then it becomes not rational to assemble it with your own hands. If you are not a radio amateur, then a switching current stabilizer is easier and faster to buy. In the future, it can be modified to the required parameters. You can find out more in the section "Ready-made modules".

Driver for 220 V

If you are interested in a driver for a 220v LED, then it is better to order or buy it. They are of medium difficulty to manufacture, but setup will take more time and setup experience will be required.

220 LED driver can be removed from faulty LED lamps, lamps and spotlights, in which the circuit with LEDs is faulty. In addition, almost any existing driver can be modified. To do this, find out the model of the PWM controller on which the converter is assembled. Typically, the output parameters are set by a resistor or several. Look at the datasheet to see what resistance should be in order to get the required amps.

If you put an adjustable resistor of the calculated value, then the number of amperes at the output will be configurable. Just do not exceed the rated power that was indicated.

Current stabilizer, circuit

I often have to look through the assortment on Aliexpress in search of inexpensive but high-quality modules. The difference in cost can be 2-3 times, it takes time to find the minimum price. But thanks to this I make an order for 2-3 pieces for tests. I buy for reviews and consultations of manufacturers who buy components in China.

June 2016 the best choice became a universal module on XL4015, the price of which is 110 rubles with free shipping. Its characteristics are suitable for connecting powerful LEDs up to 100 watts.

Schematic in driver mode.

V standard version the XL4015 case is soldered to the board, which serves as a heatsink. To improve cooling on the XL4015 case, you need to put a radiator. Most put it on top, but the effectiveness of such an installation is low. Better system cooling should be placed at the bottom of the board, opposite the place where the microcircuit is soldered. Ideally, it is better to unsolder it and put it on a full-fledged radiator through thermal paste. The legs will most likely need to be lengthened with wires. If such serious cooling is required for the controller, then the Schottky diode will also need it. It will also have to be put on a radiator. Such a refinement will significantly increase the reliability of the entire circuit.

In general, the modules do not have protection against incorrect power supply. This instantly disables them, be careful.

LM317

Application (roll) does not even require any skills and knowledge of electronics. The number of external elements in the circuits is minimal, so this affordable option for anyone. Its price is very low, its possibilities and application have been repeatedly tested and verified. Only it requires good cooling, this is its main drawback. The only thing to be wary of is low-quality Chinese LM317 microcircuits, which have worse parameters.

Due to the absence of unnecessary noise at the output, linear stabilization microcircuits were used to power high-quality Hi-Fi and Hi-End DACs. For DACs, power cleanliness plays a huge role, so some use batteries for this.

The maximum power for the LM317 is 1.5 Amps. To increase the number of amperes, you can add a field effect transistor or a regular one to the circuit. It will be possible to get up to 10A at the output, it is set by low-resistance. In this scheme, the KT825 transistor takes on the main load.

Another way is to put an analogue with higher technical specifications on the big system cooling.

Adjustable current stabilizer

As a radio amateur with 20 years of experience, I am pleased with the range of ready-made blocks and modules for sale. Now you can assemble any device from ready-made blocks in a minimum time.

I began to lose confidence in Chinese products after I saw in the "Tank Biathlon" how the best Chinese tank had a wheel fall off.

Chinese online stores have become the leader in the range of power supplies, DC-DC current converters, drivers. In their free sale, you can find almost any modules, if you look better, then very highly specialized ones. For example, for 10,000 thousand rubles, you can assemble a spectrometer worth 100,000 rubles. Where 90% of the price is a markup for a brand and slightly modified Chinese software.

The price starts from 35 rubles. for a DC-DC voltage converter, the driver is more expensive and has two three trimming resistors instead of one.

For more universal use an adjustable driver is better. The main difference is the installation variable resistor in the circuit that sets the amperes at the output. These characteristics can be specified in typical schemes inclusions in the specifications for the microcircuit, datasheet, datasheet.

The weak points of such drivers are the heating of the inductor and the Schottky diode. Depending on the PWM controller model, they can withstand 1A to 3A without additional cooling of the microcircuit. If above 3A, then cooling of the PWM and a powerful Schottky diode is required. The inductor is rewound with a thicker wire or replaced with a suitable one.

The efficiency depends on the operating mode, the voltage difference between the input and output. The higher the efficiency, the lower the heating of the stabilizer.

Prices in China

The cost is very low considering that shipping is included in the price. I used to think that because of the goods for 30-50 rubles, the Chinese will not even get dirty, a lot of work with a low income. But as practice has shown, I was wrong. Any penny nonsense they pack and send. It comes in 98% of cases, and I have been buying on Aliexpress for more than 7 years and for large amounts, probably already about 1 million rubles.

Therefore, I place an order in advance, usually 2-3 pieces of the same name. Unnecessary sell on a local forum or Avito, everything sells like hot cakes.

So, the core and main component of the LED light bulb is the LED. From the point of view of circuitry, light-emitting diodes are no different from any others, except that in the sense of using them as diodes themselves, they have terrible parameters - a very small allowable reverse voltage, a relatively large junction capacitance, a huge operating voltage drop (about 3.5 V for white LEDs - e.g. for rectifier diode it would be a nightmare), etc.

However, we understand that the main value of LEDs for mankind is that they glow, and sometimes quite brightly. In order for an LED to glow happily ever after, it needs two conditions: a stable current through it and good heat dissipation from it. The quality of the heat sink is ensured by various design methods, so now we will not dwell on this issue. Let's talk about why and how modern humanity achieves the first goal - a stable current.

Speaking of white LEDs

It is clear that white LEDs are most interesting for lighting. They are made on the basis of a crystal that emits blue light, filled with a phosphor that re-radiates part of the energy in the yellow-green region. The title picture clearly shows that the current-carrying wires go into something yellow - this is the phosphor; the crystal is located underneath. On a typical spectrum white LED the blue peak is clearly visible:

Spectra of LEDs with different color temperatures: 5000K (blue), 3700K (green), 2600K (red). Read more.

We have already figured out that in the circuitry sense, the LED differs from any other diode only in the parameter values. Here it must be said that the device is fundamentally nonlinear; that is, he does not obey the Ohm's law familiar from school at all. The dependence of the current on the applied voltage on such devices is described by the so-called. current-voltage characteristic (CVC), and for the diode it is exponential. It follows from this that the smallest change in the applied voltage leads to a huge change in current, but that's not all - with a change in temperature (as well as aging), the I–V characteristic shifts. In addition, the position of the IV characteristics is slightly different for different diodes. I will specify separately - not only for each type, but for each instance, even from the same batch. For this reason, the distribution of current through diodes connected in parallel will necessarily be uneven, which cannot have a good effect on the durability of the structure. In the manufacture of matrices, they try to either use series connection, which solves the problem at the root, or choose diodes with approximately the same forward voltage drop. To facilitate the task, manufacturers usually indicate the so-called "bin" - the code for the selection by parameters (including voltage), into which a particular instance falls.

VAC of a white LED.

Accordingly, for everything to work well, the LED must be connected to a device that, regardless of external factors will be with high precision automatically select a voltage at which a given current flows in the circuit (for example, 350 mA for one-watt LEDs), and continuously monitor the process. In general, such a device is called a current source, but in the case of LEDs, it is fashionable these days to use the overseas word "driver". In general, a driver is often referred to as solutions that are primarily designed to work in a specific application - for example, "MOSFET driver" - a microcircuit designed to drive specifically powerful field-effect transistors, "seven-segment indicator driver" - a solution to drive specifically seven-segment devices, etc. . That is, by calling a current source an LED driver, people are hinting that this current source is designed specifically to work with LEDs. For example, it may have specific functions - something in the spirit of having a DMX-512 light interface, detecting an open and short circuit at the output (and a conventional current source, in general, should work without problems on short circuit), etc. However, concepts are often confused, and, for example, they call the most common adapter (voltage source!) For LED strips a driver.

In addition, devices designed to set the mode lighting device often referred to as ballast.

So, current sources. The simplest current source can be a resistor in series with the LED. This is done at low powers (somewhere up to half a watt), for example, in the same LED strips. As the power increases, the losses on the resistor become too high, and the requirements for current stability increase, and therefore there is a need for more advanced devices, the poetic image of which I drew above. All of them are built according to the same ideology - they have a regulatory element controlled by current feedback.

Current stabilizers are divided into two types - linear and pulse. Linear circuits are relatives of the resistor (the resistor itself and its analogues also belong to this class). They usually do not give a special gain in efficiency, but they increase the quality of current stabilization. Impulse circuits are best solution however, they are more complex and expensive.

Let's now take a quick look at what you can see inside or around LED bulbs these days.

1. Capacitor ballast

The capacitor ballast is an extension of the idea of ​​putting a resistor in series with an LED. In principle, the LED can be connected to the outlet directly like this:

The back-to-back diode is necessary in order to prevent the breakdown of the LED at the moment when the mains voltage changes polarity - I have already mentioned that there are no LEDs with a permissible reverse voltage of hundreds of volts. In principle, instead of a reverse diode, you can put another LED.

The resistor value in the circuit above is calculated for the LED current of about 10 - 15 mA. Since the mains voltage is much greater than the drop across the diodes, the latter can be ignored and calculated directly according to Ohm's law: 220/20000 ~ 11 mA. You can substitute the peak value (311 V) and make sure that even in the limit case, the diode current will not exceed 20 mA. Everything turns out great, except that the resistor will dissipate about 2.5 watts of power, and about 40 mW on the LED. Thus, the efficiency of the system is about 1.5% (in the case of a single LED, it will be even less).

The idea of ​​the method under consideration is to replace the resistor with a capacitor, because it is known that in circuits alternating current reactive elements have the ability to limit current. By the way, you can also use a choke, moreover, they do it in classic electromagnetic ballasts for fluorescent lamps.

Counting according to the formula from the textbook, it is easy to get that in our case a 0.2 μF capacitor is required, or an inductance coil of about 60 H. Here it becomes clear why chokes are never found in such ballasts of LED lamps - a coil of such an inductance is a serious and expensive structure, but a 0.2 uF capacitor is much easier to get. Of course, it must be designed for the peak mains voltage, and better with a margin. In practice, capacitors with an operating voltage of at least 400 V are used. Having slightly supplemented the circuit, we get what we have already seen in the previous article.

Lyrical digression

"Microfarad" is abbreviated exactly as "uF". I dwell on this because I often see people writing "mF" in this context, while the latter is short for "millifarad", that is, 1000 microfarads. In English, "microfarad", again, is not written as "mkF", but, on the contrary, "uF". This is because the letter "u" resembles the letter "μ" with its tail torn off.

So, 1 F/F = 1000 mF/mF = 1000000 uF/uF/μF, and nothing else!

In addition, "Farad" is masculine, as it is named after the great male physicist. So, "four microfarads", but not "four microfarads"!

As I have already said, such a ballast has only one advantage - simplicity and cheapness. Like a ballast with a resistor, current stabilization is not very good here, and, even worse, there is a significant reactive component, which is not very good for the network (especially at noticeable powers). In addition, as the desired current increases, the required capacitance of the capacitor will increase. For example, if we want to turn on a one-watt LED operating at 350 mA, we need a capacitor with a capacity of about 5 microfarads, designed for a voltage of 400 V. This is already more expensive, larger and more complex in terms of design. With the suppression of ripples, everything is also not easy here. In general, we can say that the capacitor ballast is only forgivable for small beacon lamps, nothing more.

2. Transformerless buck topology

This circuit solution belongs to the family of transformerless converters, which includes step-down, step-up and inverting topologies. In addition, transformerless converters also include SEPIC, Chuck converter and other exotics, such as switched capacitors. In principle, an LED driver can be built on the basis of any of them, but in practice they are much less common in this capacity (although boost topology is used, for example, in many flashlights).

One example of a driver based on a transformerless buck topology is shown in the figure below.

In wildlife, such inclusion can be observed on the example of the ZXLD1474 or the ZXSC310 inclusion option (which, by the way, is just a boost converter in the original switching circuit).

Here the LED is connected in series with the coil. The control circuit monitors the current through the measuring resistor R1 and controls the switch T1. If the current through the LED drops below a predetermined minimum, the transistor turns on and the coil with the LED connected in series with it is connected to the power source. The current in the coil begins to increase linearly (red area on the graph), diode D1 is locked at this time. As soon as the control circuit registers that the current has reached a predetermined maximum, the key closes. In accordance with the first switching law, the coil tends to maintain the current in the circuit due to the energy stored in the magnetic field. At this point, current flows through diode D1. The energy of the coil field is consumed, the current decreases linearly (green area on the graph). When the current drops below a predetermined minimum, the control circuit registers this and opens the transistor again, pumping power into the system - the process repeats. Thus, the current is maintained within the specified limits.

A distinctive feature of buck topology is the ability to make ripples luminous flux arbitrarily small, since in such a connection the current through the LED is never interrupted. The way to approach the ideal lies through an increase in inductance and an increase in the switching frequency (today there are converters with operating frequencies up to several megahertz).

Based on such a topology, a driver was made Gauss lamps discussed in the previous article.

The disadvantage of the method is the lack of galvanic isolation - when the transistor is open, the circuit is directly connected to the voltage source, in the case of network LED lamps - to the network, which can be unsafe.

3. Flyback converter

Although the flyback converter contains something that looks like a transformer, in this case it is more correct to call this part a two-winding choke, since current never flows through both windings at the same time. In fact, flyback converters are similar in principle to transformerless topologies. When T1 is open, the current in the primary rises, energy is stored in the magnetic field; at the same time, the polarity of switching on the secondary winding is deliberately chosen so that the diode D3 is closed at this stage and no current flows on the secondary side. The load current at this moment supports the capacitor C1. When T1 closes, the polarity of the voltage on the secondary is reversed (because the derivative of the current in the primary reverses sign), D3 opens and the stored energy is transferred to the secondary. In terms of current stabilization, everything is the same - the control circuit analyzes the voltage drop across the resistor R1 and adjusts the time s e parameters so that the current through the LEDs remains constant. Most often, the flyback converter is used at powers not exceeding 50 W; further, it ceases to be appropriate due to increasing losses and the necessary dimensions of the inductor transformer.

I must say that there are options for flyback drivers without an opto-isolator (for example). They rely on the fact that the currents of the primary and secondary windings are coupled, and with certain reservations, one can limit oneself to analyzing the current of the primary winding (or, more often, a separate auxiliary winding) - this saves on details and, accordingly, reduces the cost of the solution.

The flyback converter is good in that, firstly, it provides isolation of the secondary part from the mains (higher safety), and, secondly, it makes it relatively easy and cheap to manufacture lamps compatible with standard dimmers for incandescent lamps, as well as to arrange coefficient correction power.

Lyrical digression

The flyback converter is so called because a similar method was originally used to obtain high voltage in televisions based on cathode ray tubes. The high voltage source was circuitically combined with the horizontal sweep circuit, and a high voltage pulse was obtained at the time reverse electron beam.

A little about pulsations

As already mentioned, impulse sources work for enough high frequencies(in practice - from 30 kHz, more often about 100 kHz). Therefore, it is clear that a serviceable driver in itself cannot be a source of a large ripple factor - primarily because this parameter is simply not normalized at frequencies above 300 Hz, and, besides, high-frequency ripples are quite easy to filter out anyway. The problem is the mains voltage.

The fact is that, of course, all the circuits listed above (except for the circuit with a quenching capacitor) operate on direct voltage. Therefore, at the input of any electronic ballast, first of all, there is a rectifier and a storage capacitor. The purpose of the latter is to feed the ballast at those moments when the mains voltage goes below the threshold of the circuit. And here, alas, a compromise is needed - high-voltage high-capacity electrolytic capacitors, firstly, cost money, and, secondly, they take up precious space in the lamp housing. This is also the root cause of power factor problems. The described circuit with a rectifier has an uneven current consumption. This leads to the appearance of higher harmonics of it, which is the reason for the deterioration of the parameter of interest to us. Moreover, the better we try to filter the voltage at the ballast input, the lower the power factor we will get if we do not make separate efforts. This explains the fact that almost all the low ripple lamps that we have seen show a very mediocre power factor, and vice versa (of course, the introduction of an active power factor correction will affect the price, so they prefer to save on it for now). Add tags

Educational article on LED current stabilizers and more. The circuits of linear and pulse current stabilizers are considered.

The current stabilizer for the LED is installed in many luminaire designs. LEDs, like all diodes, have a non-linear current-voltage characteristic. This means that as the voltage across the LED changes, the current changes disproportionately. As the voltage increases, at first the current rises very slowly, while the LED does not light up. Then, when the threshold voltage is reached, the LED starts to glow and the current increases very quickly. With a further increase in voltage, the current increases catastrophically and the LED burns out.

The threshold voltage is specified in the LED specifications as a forward voltage at rated current. The current rating for most low power LEDs is 20 mA. For high power lighting LEDs, the current rating can be as high as 350mA or more. By the way, powerful LEDs generate heat and must be installed on a heat sink.

For correct operation LED, it must be powered through a current stabilizer. What for? The fact is that the threshold voltage of the LED has a spread. different types LEDs have a different forward voltage, even the same type of LEDs have a different forward voltage - this is indicated in the LED characteristics as the minimum and maximum values. Therefore, two LEDs connected to the same voltage source parallel circuit will carry different currents. This current can be so different that the LED may fail earlier or burn out immediately. In addition, the voltage regulator also has parameter drift (depending on the primary power level, on load, on temperature, just in time). Therefore, turning on LEDs without current equalization devices is undesirable. Various ways current leveling are considered. This article discusses devices that set a well-defined, given current - current stabilizers.

Types of current stabilizers

The current stabilizer sets the specified current through the LED, regardless of the voltage applied to the circuit. When the voltage on the circuit increases above the threshold level, the current reaches the set value and then does not change. With a further increase in the total voltage, the voltage on the LED stops changing, and the voltage on the current regulator increases.

Since the voltage on the LED is determined by its parameters and is generally unchanged, the current regulator can also be called the LED power regulator. In the simplest case, allocated by the device active power(heat) is distributed between the LED and the stabilizer in proportion to the voltage on them. Such a stabilizer is called linear. There are also more economical devices - current stabilizers based on pulse converter(key converter or converter). They are called pulsed, because they pump power inside themselves in portions - pulses as needed for the consumer. The correct pulse converter consumes power continuously, internally transfers it with pulses from the input circuit to the output circuit and outputs power to the load again continuously.

Linear Current Stabilizer

The linear current regulator heats up the more, the more voltage is applied to it. This is its main drawback. However, it has a number of advantages, for example:

• Linear stabilizer does not create electromagnetic interference
• Simple in design
• Low cost in most applications

Since a switching converter is never completely efficient, there are applications where a linear regulator has comparable or even greater efficiency - when the input voltage is only slightly higher than the LED voltage. By the way, when powered from the mains, a transformer is often used, at the output of which a linear current stabilizer is installed. That is, first the voltage is reduced to a level comparable to the voltage on the LED, and then, using a linear stabilizer, the required current is set.

In another case, you can bring the LED voltage closer to the supply voltage - connect the LEDs in a series chain. The voltage across the string will equal the sum of the voltages across each LED.

Schemes of linear current stabilizers

The most simple circuit current stabilizer - on one transistor (scheme "a"). Since the transistor is a current amplifier, its output current (collector current) is greater than the control current (base current) by h 21 times (gain). The base current can be set using a battery and a resistor, or using a zener diode and a resistor (diagram "b"). However, such a circuit is difficult to tune, the resulting stabilizer will depend on temperature, in addition, transistors have a large spread of parameters and when replacing a transistor, the current will have to be selected again. The circuit with feedback "c" and "d" works much better. The resistor R in the circuit acts as a feedback - as the current increases, the voltage across the resistor increases, thereby locking the transistor and the current decreases. Scheme "g", when using the same type of transistors, has greater temperature stability and the ability to minimize the value of the resistor, which reduces the minimum voltage of the stabilizer and the power dissipation on the resistor R.

The current stabilizer can be made on the basis field effect transistor With p-n junction(scheme "d"). The gate-source voltage sets the drain current. At zero gate-source voltage, the current through the transistor is equal to the initial drain current specified in the documentation. The minimum operating voltage of such a current stabilizer depends on the transistor and reaches 3 volts. Some manufacturers of electronic components produce special devices- ready-made stabilizers with a fixed current, assembled according to this scheme - CRD (Current Regulating Devices) or CCR (Constant Current Regulator). Some call it a diode stabilizer, because it works like a diode in reverse.

On Semiconductor produces a linear regulator of the NSIxxx series, for example, which has two leads and, to increase reliability, has a negative temperature coefficient As the temperature rises, the current through the LEDs decreases.

A current stabilizer based on a pulse converter is very similar in design to a voltage regulator based on a pulse converter, but it does not control the voltage at the load, but the current through the load. With a decrease in current in the load, it pumps up power, with an increase, it reduces. The most common circuits of pulse converters include a reactive element - a choke, which, with the help of a switch (key), is pumped up in portions of energy from the input circuit (from the input capacitance) and, in turn, transfers it to the load. In addition to the obvious advantage of saving energy, pulse converters have a number of disadvantages that have to be dealt with by various circuitry and design solutions:

• Pulse converter produces electrical and electromagnetic interference
• It usually has a complex structure
• It does not have absolute efficiency, that is, it spends energy for own work and warms up
• It usually has a higher cost than, for example, transformer plus linear devices

Since energy savings are critical in many applications, component designers and circuit designers try to reduce the impact of these shortcomings, and often succeed.

Schemes of pulse converters

Since the current stabilizer is based on a pulse converter, let's consider the main circuits of pulse converters. Each pulse converter has a key, an element that can only be in two states - on and off. In the off state, the key does not conduct current and, accordingly, no power is generated on it. In the on state, the key conducts current, but has a very low resistance (ideally, zero), respectively, it releases power close to zero. Thus, the key can transfer portions of energy from the input circuit to the output circuit with virtually no power loss. However, instead of a stable current, which can be obtained from a linear power supply, the output of such a switch will be impulse voltage and current. In order to get stable voltage and current again, you can put a filter.

Using a conventional RC filter, you can get the result, however, the efficiency of such a converter will not be better than a linear one, since all the excess power will be released on the active resistance of the resistor. But if you use a filter instead of RC - LC (circuit "b"), then, due to the "specific" properties of inductance, power losses can be avoided. Inductance has a useful reactive property - the current through it increases gradually, the electrical energy supplied to it is converted into magnetic energy and accumulates in the core. After turning off the key, the current in the inductor does not disappear, the voltage on the inductor changes polarity and continues to charge the output capacitor, the inductance becomes a current source through the bypass diode D. Such an inductance, designed to transfer power, is called a choke. The current in the inductor of a properly operating device is constantly present - the so-called continuous mode or continuous current mode (in Western literature, this mode is called Constant Current Mode - CCM). When the load current decreases, the voltage on such a converter increases, the energy accumulated in the inductor decreases and the device can switch to discontinuous operation when the current in the inductor becomes intermittent. With this mode of operation, the level of interference created by the device increases sharply. Some converters operate in border mode, when the current through the inductor approaches zero (in Western literature, this mode is called Border Current Mode - BCM). In any case, significant D.C., which leads to the magnetization of the core, and therefore, the inductor is made of a special design - with a gap or using special magnetic materials.

The stabilizer based on a pulse converter has a device that regulates the operation of the key, depending on the load. The voltage stabilizer registers the voltage at the load and changes the operation of the key (diagram "a"). The current stabilizer measures the current through the load, for example, using a small measuring resistance Ri (circuit "b"), connected in series with the load.

The converter key, depending on the regulator signal, turns on with different duty cycles. There are two common ways to manage a key - pulse width modulation(PWM) and current mode. In PWM mode, the error signal controls the pulse width while maintaining the repetition rate. In current mode, the peak current in the inductor is measured and the interval between pulses is changed.

In modern key converters, a MOSFET transistor is usually used as a key.

Buck Converter

The version of the converter considered above is called a step-down converter, since the voltage at the load is always lower than the voltage of the power source.

Since the inductor constantly flows unidirectional current, the requirement for the output capacitor can be reduced, the inductor with the output capacitor plays the role of an effective LC filter. In some circuits of current stabilizers, for example for LEDs, the output capacitor may be absent altogether. In Western literature, a buck converter is called a Buck converter.

Boost Converter

The switching regulator circuit below also works with a choke, but the choke is always connected to the output of the power supply. When the key is open, power is supplied through the inductor and diode to the load. When the key is closed, the inductor accumulates energy; when the key is opened, the EMF that occurs at its terminals is added to the EMF of the power source and the voltage at the load increases.

Unlike the previous circuit, the output capacitor is charged by an intermittent current, so the output capacitor must be large and an additional filter may be needed. In Western literature, a boost-buck converter is called a Boost converter.

inverter converter

Another circuit of the pulse converter works in a similar way - when the key closes, the inductor accumulates energy, when the key opens, the EMF that occurs at its terminals will have the opposite sign and a negative voltage will appear on the load.

As in the previous circuit, the output capacitor is charged by an intermittent current, so the output capacitor must be large, and an additional filter may be needed. In Western literature, the inverting converter is called the Buck-Boost converter.

Forward and flyback converters

Most often, power supplies are made according to a scheme that uses a transformer in its composition. The transformer provides galvanic isolation of the secondary circuit from the power source, in addition, the efficiency of the power supply based on such circuits can reach 98% or more. The forward converter (circuit "a") transfers energy from the source to the load at the moment the key is on. In fact, this is a modified buck converter. flyback converter(circuit "b") transfers energy from the source to the load during the off state.

In a forward converter, the transformer operates normally and the energy is stored in the inductor. In fact, this is a pulse generator with an LC filter at the output. The flyback converter stores energy in the transformer. That is, the transformer combines the properties of a transformer and a choke, which creates certain difficulties when choosing its design.

In Western literature forward converter called forward converter. Flyback - Flyback converter.

Application of a pulse converter as a current stabilizer

Most switching power supplies are available with output voltage stabilization. Typical circuits of such power supplies, especially powerful ones, in addition to output voltage feedback, have a current control circuit key element such as a low resistance resistor. Such control allows you to ensure the mode of operation of the throttle. The simplest current stabilizers use this control element to stabilize the output current. Thus, the current stabilizer turns out to be even simpler than a stabilizer voltage.

Consider a switching current stabilizer circuit for an LED based on a microcircuit from a well-known manufacturer of electronic components On Semiconductor:

The buck converter circuit operates in continuous current mode with an external switch. The circuit is chosen from many others because it shows how simple and effective a switching current regulator circuit with an external switch can be. In the above diagram, the control chip IC1 controls the operation of the MOSFET switch Q1. Since the converter operates in continuous current mode, it is not necessary to install an output capacitor. In many circuits, a current sensor is installed in the source circuit of the switch, however, this reduces the turn-on speed of the transistor. In the above diagram, the R4 current sensor is installed in the primary power circuit, as a result, the circuit turned out to be simple and effective. The key operates at a frequency of 700 kHz, which allows you to install a compact choke. With an output power of 7 watts, an input voltage of 12 volts when operating at 700 mA (3 LEDs), the efficiency of the device is more than 95%. The circuit works stably up to 15 watts of output power without the use of additional heat dissipation measures.

An even simpler circuit is obtained using key stabilizer microcircuits with a built-in key. For example, a diagram of a key LED current stabilizer based on the /CAT4201 chip:

To operate a device with a power of up to 7 watts, only 8 components are needed, including the microcircuit itself. The switching regulator operates in current limit mode and requires a small output ceramic capacitor to operate. Resistor R3 is needed when powered from 24 volts and above to reduce the slew rate of the input voltage, although this somewhat reduces the efficiency of the device. The frequency of operation exceeds 200 kHz and varies depending on the load and input voltage. This is due to the method of regulation - control of the peak current of the inductor. When the current reaches the maximum value, the key opens, when the current drops to zero, it turns on. The efficiency of the device reaches 94%.