Push-pull pulse voltage converter. Push-pull converters. Voltage converter with PWM control

What does an embedder do when he has nothing to do? Of course, he studies push-pull self-generating converters! In fact, there is something to do, and a lot, but something is too lazy. Therefore, today I will still explore a push-pull self-generating converter. Like this: Just like in the picture above, they are drawn in books, but I don’t like this drawing; Not only does the converter look like a multivibrator in this design (which is far from the true principle of its operation), but the output is also located on top (I slightly corrected this in the first picture). Therefore, I offer my option:
The picture gets ahead of itself a little - I’ll explain where all these numbers come from as the article progresses. First, let's look at the general operating principle of the circuit. When power is applied, the first transistor to open will be the one whose base-emitter voltage is lower or whose current transfer coefficient is greater (there are no exactly identical transistors in nature). Let it be T2. Then an increasing current will begin to flow through winding B. In this case, windings A and B work together as an autotransformer, as a result of which a voltage even greater than the supply voltage will be applied to the base of T2 through resistor R2. This guarantees saturation of the transistor (since both junctions, collector and emitter, are open). T1 is closed in this case, because the voltage on the collector of saturated T2 is low. T2 is open, the current through winding B is increasing, everything is cool. However, this will continue until the magnetic circuit of the transformer reaches saturation. As soon as this happens, the inductance of the windings will drop sharply, and, consequently, the current through them will begin to tend to infinity, limited almost only by the winding resistance. In fact, after all

UPD: I analyzed the operation of this circuit in more detail and correctly.

Like everything on earth, such a converter has pros and cons. The first and most obvious advantage is fantastic simplicity. Only four parts are required, not including the transformer. Another advantage is that the transformer in such a converter will never go too far into saturation, which limits losses. In addition, this is a true push-pull circuit, so the transformer does not need a gap, which means that you can use, for example, rings from savings (which is what I'm going to do next). With all the pros, this scheme also has a lot of cons. Firstly, the magnetic circuit will still enter saturation, so there will be losses that could be avoided. Secondly, it is clear that the ability to operate such a converter is closely tied to the real properties of the transformer’s magnetic core (the error in indicating which in datasheets reaches 30%) and slightly to the imperfection of the transistors. That is, calculate Such a converter is impossible - its parameters can only be roughly estimated, or measured on a real circuit. The operating frequency will be determined by how quickly the magnetic circuit enters saturation, that is, it will depend on the input voltage. Above I talked about savings rings. For a toroidal core, the expression for induction in the magnetic circuit is as follows: where μ is the magnetic permeability of the ring, μ 0 is the magnetic constant, N is the number of turns of the winding, I is the current in the winding, R is the radius of the ring. The rate of increase of the current in the winding is proportional to the applied voltage (see the very first formula), that is, the rate of increase of the magnetic flux will also be proportional to it, that is, the operating frequency will depend on the input voltage. In this case, the absolute value of induction will be proportional to the product of the number of turns and the current, therefore the no-load current will be determined by the number of turns in windings A and B (the more turns, the less current saturation will be achieved). This leads to another drawback - in order to obtain a low no-load current, you need to wind a lot of wire, which is especially tedious in the case of a toroidal core. Well, the no-load current will also depend on the applied voltage. From all that has been said, we can conclude that such a scheme is suitable when the simplicity of the converter outweighs the need for accurate predictability and quality of its characteristics. For example, when the goal is to have a little fun on a spring evening.

Let's move from theory to practice. In my bins lay an unidentified ring, taken from a savings account. Its diameter is 10 mm, height - 3.5 mm, thickness - 2 mm. That is, it looks like an EPCOS R 10 x 6 x 4 ring.
I wound 10 turns of wire around it and measured the inductance of the resulting coil. The result was 286 μH, which corresponds to a permeability of about 8000. That is, according to the datasheet above, the ring material is either T37 or T38. Their saturation induction is something like 400 mT. I figured that I wouldn’t be too lazy to wind no more than 15 turns. Using the second formula, we can calculate that the saturation current will be something around 65 mA. Fine; fits well within the capabilities of the main “just transistors” - BC547/847/817. After that, I wound the windings - the primary, 15 turns in two wires, and the secondary, 63 turns (as many as I could). The transformation ratio turned out to be 4.2, that is, from 1.5 V we get approximately 6.3 V.
I put together a diagram. I installed 510 Ohm resistors in the bases of the transistors (as I found). At the same time, at a minimum input voltage (I took a minimum of 0.9 V with an eye on the battery as a source), the base current will be sufficient to provide a collector current sufficient to saturate the transformer (we calculated above about 65 mA). Collected:
Gave 1.5 V. It worked!
The output is 6.3 V RMS, exactly as designed. You can install a doubling rectification circuit and get 12 V. Voltage at the collectors:
It can be seen that the pulse amplitude is 3 V, that is, twice the supply voltage. So practice really coincides with theory - the primary winding works like an autotransformer. Voltage at the bases (do not trust the frequency measurement, the oscilloscope is glitchy due to surges; the time grid is the same as above):
Current consumption. I measured the voltage across a 10 Ohm resistor connected in series with the converter:
About 76 mA peak. Using the second formula, you can calculate the saturation induction - it turns out to be about 457 mT, that is, the ferrite is apparently still T38. The average idle current at a voltage of 1.5 V was about 30 mA. The converter starts at an input voltage of 0.5 V. As for me, such a circuit is an excellent way to use rings from savings in simple converters 1.5 - 5 V / 3.3 V. Of course, it would be nice to also install a stabilizer at the output (with a diode bridge, of course), in the simplest case is linear, the same L78L33. The efficiency of such a solution will not be particularly high, but in terms of cost and simplicity it will probably outperform even Chinese products.

1. Computer architecture computer systems telecommunications networks

   Document

   Dependencies in order to simplification. 5. Presentation of the received... single-cycle (a) and two-stroke(b) RS flip-flops. ..., specialized converters information, ... index ( index search) allows... for scientific and technical calculations, mathematical problems...

2. Management

   LPC 1.0 is given calculation simplified Index I 8 Index 8 Index two-stroke scheme converter with transformerless...

3. archive

   LPC 1.0 is given calculation interface bandwidth... . Meet and simplified options, without... DS3) 6 FDEDIN (DS3) 8 Index I 8 Index 8 Index 10 Motor On A o 10 ... applies here two-stroke scheme converter with transformerless...

4. Management

   LPC 1.0 is given calculation interface bandwidth... . Meet and simplified options, without... DS3) 6 FDEDIN (DS3) 8 Index I 8 Index 8 Index 10 Motor On A o 10 ... applies here two-stroke scheme converter with transformerless...

The most widespread are push-pull secondary power sources, although they have a more complex electrical circuit compared to single-cycle ones. They allow you to obtain significantly higher output power with high efficiency.
Circuits of push-pull converter-inverters have three types of connection of key transistors and the primary winding of the output transformer: half-bridge, bridge and with a primary winding tapped from the middle.

Half bridge diagram of the key cascade construction.
Its feature is the inclusion of the primary winding of the output transformer at the midpoint of the capacitive divider C1 - C2.

The amplitude of the voltage pulses at the emitter-collector transistor transitions T1 and T2 does not exceed Upit the value of the supply voltage. This allows the use of transistors with a maximum voltage Uek up to 400 volts.
At the same time, the voltage on the primary winding of transformer T2 does not exceed the value Upit/2, because it is removed from the divider C1 - C2 (Upit/2).
A control voltage of opposite polarity is supplied to the bases of key transistors T1 and T2 through transformer Tr1.


IN pavement In the converter, the capacitive divider (C1 and C2) is replaced by transistors T3 and T4. Transistors in each half-cycle open in pairs diagonally (T1, T4) and (T2, T3).

The voltage at the transitions Uec of closed transistors does not exceed the supply voltage Upit. But the voltage on the primary winding of transformer Tr3 will increase and will be equal to the value of Upit, which increases the efficiency of the converter. The current through the primary winding of transformer Tr3 at the same power, compared to a half-bridge circuit, will be less.
Due to the difficulty in setting up the control circuits of transistors T1 - T4, a bridge switching circuit is rarely used.

Inverter circuit with so-called push-pull output is most preferable in powerful converters-inverters. A distinctive feature in this circuit is that the primary winding of the output transformer Tr2 has a terminal from the middle. For each half-cycle of voltage, one transistor and one half-winding of the transformer alternately operate.

This circuit is characterized by the highest efficiency, low ripple level and low noise emission. This is achieved by reducing the current in the primary winding and reducing the power dissipation in the key transistors.
The voltage amplitude of the pulses in half of the primary winding Tr2 increases to the value Upit, and the voltage Uek on each transistor reaches the value 2 Upit (self-induction emf + Upit).
It is necessary to use transistors with a high value of Ucat, equal to 600 - 700 volts.
The average current through each transistor is equal to half the current consumption from the supply network.

Current or voltage feedback.

A feature of push-pull self-excited circuits is the presence of feedback (Feedback) from the output to the input, in terms of current or voltage.

In the scheme current feedback communication winding w3 of transformer Tr1 is connected in series with the primary winding w1 of output transformer Tr2. The greater the load at the inverter output, the greater the current in the primary winding Tr2, the greater the feedback and the greater the base current of transistors T1 and T2.
If the load is less than the minimum permissible, the feedback current in winding w3 of transformer Tr1 is insufficient to control the transistors and the generation of alternating voltage is disrupted.
In other words, when the load is lost, the generator does not work.

In the scheme voltage feedback The feedback winding w3 of transformer Tr2 is connected through a resistor R to the communication winding w3 of transformer Tr1. This circuit provides feedback from the output transformer to the input of the control transformer Tr1 and then to the base circuits of transistors T1 and T2.
Voltage feedback is weakly dependent on load. If there is a very large load at the output (short circuit), the voltage on winding w3 of transformer Tr2 decreases and a moment may come when the voltage on the base windings w1 and w2 of transformer Tr1 will not be enough to control the transistors. The generator will stop working.
Under certain circumstances, this phenomenon can be used as protection against output short circuit.
In practice, both circuits with feedback in both current and voltage are widely used.

Push-pull inverter circuit with voltage feedback

For example, let's consider the operation of the most common converter-inverter circuit - a half-bridge circuit.
The circuit consists of several independent blocks:

• — rectifier unit – converts alternating voltage 220 volts 50 Hz into direct voltage 310 volts;
• — triggering pulse device – generates short voltage pulses to start the autogenerator;
• — alternating voltage generator – converts a direct voltage of 310 volts into a rectangular alternating voltage of high frequency 20 – 100 kHz;
• - rectifier - converts alternating voltage 20 -100 kHz into direct voltage.

Immediately after turning on the 220 volt power supply, the triggering pulse device, which is a sawtooth voltage generator (R2, C2, D7), begins to operate. From it, triggering pulses arrive at the base of transistor T2. The autogenerator starts.
The key transistors open one by one and in the primary winding of the output transformer Tr2, connected to the diagonal of the bridge (T1, T2 - C3, C4), a rectangular alternating voltage is formed.
The output voltage is removed from the secondary winding of transformer Tr2, rectified by diodes D9 - D12 (full-wave rectification) and smoothed by capacitor C5.
The output produces a constant voltage of a given value.
Transformer T1 is used to transmit feedback pulses from the output transformer Tr2 to the bases of key transistors T1 and T2.


The push-pull UPS circuit has a number of advantages over the single-cycle circuit:

• — the ferrite core of the output transformer Tr2 operates with active magnetization reversal (the magnetic core is most fully used in terms of power);
• — the collector-emitter voltage Uek on each transistor does not exceed the DC source voltage of 310 volts;
• — when the load current changes from I = 0 to Imax, the output voltage changes slightly;
• — high voltage surges in the primary winding of transformer Tr2 are very small, and the level of radiated interference is correspondingly lower.

And one more note in favor of the push-pull circuit!!

Let's compare the operation of two-stroke and single-cycle self-generators with the same load.
Each key transistor T1 and T2 is used only half the time (one half-wave) during one clock cycle of the generator; the second half of the cycle is “resting”. That is, the entire generated power of the generator is divided in half between both transistors and the transfer of energy to the load occurs continuously (from one transistor, then from the other), during the entire cycle. Transistors operate in a gentle mode.
In a single-cycle generator, the accumulation of energy in the ferrite core occurs during half the cycle, and in the second half of the cycle it is released to the load.

The key transistor in a single-cycle circuit operates four times more intensely than the key transistor in a push-pull circuit.

One of the most popular topologies of pulse voltage converters is a push-pull or push-pull converter (literally translated - push-pull).

Unlike a single-ended flyback converter, energy is not stored in the core of the push-pool, because in this case it is the core of the transformer, and not, it serves as a conductor for the alternating magnetic flux created in turn by the two halves of the primary winding.

However, despite the fact that this is a pulse transformer with a fixed transformation ratio, the stabilization voltage of the push-pull output can still be changed by varying the width of the operating pulses (using).

Due to their high efficiency (efficiency up to 95%) and the presence of galvanic isolation of the primary and secondary circuits, push-pull pulse converters are widely used in stabilizers and inverters with a power of 200 to 500 W (power supplies, automotive inverters, UPS, etc.)

The figure below shows the general circuit of a typical push-pull converter. Both the primary and secondary windings have taps from the middle, so that in each of the two working half-cycles, when only one of the transistors is active, its half of the primary winding and the corresponding half of the secondary winding would be used, where the voltage would drop on only one of the two diodes.

The use of a full-wave rectifier with Schottky diodes at the output of a push-pull converter makes it possible to reduce active losses and increase efficiency, because it is economically much more expedient to wind two halves of the secondary winding than to incur losses (financial and active) with a diode bridge of four diodes.

The switches in the primary circuit of a push-pull converter (MOSFET or IGBT) must be designed for twice the supply voltage in order to withstand not only the source EMF, but also the additional effect of the EMF induced during each other's operation.

The features of the device and mode of operation of a push-pull circuit distinguish it favorably from half-bridge, forward and flyback circuits. Unlike the half-bridge, there is no need to decouple the key control circuit from the input voltage. A push-pull converter operates like two single-ended forward converters in one device.

In addition, unlike a forward converter, a spirit-cycle converter does not need a limiting winding, since one of the output diodes continues to conduct current even when the transistors are closed. Finally, unlike a flyback converter, in a push-pull converter the switches and magnetic circuit are used more sparingly, and the effective pulse duration is longer.

Current-controlled push-pull circuits are becoming increasingly popular in integrated power supplies for electronic devices. With this approach, the problem of increased voltage on the keys is completely eliminated. A shunt resistor is connected to the common source circuit of the switches, from which the feedback voltage is removed for current protection. Each cycle of operation of the switches is limited in duration by the moment the current reaches a given value. Under load, the output voltage is typically limited by PWM.

When designing a push-pull converter, special attention is paid to the selection of switches so that the open channel resistance and gate capacitance are as small as possible. To control the gates of field-effect transistors in a push-pull converter, gate driver microcircuits are most often used, which easily cope with their task even at frequencies of hundreds of kilohertz, typical of switching power supplies of any topology.

Push-pull converters use the magnetic core of a pulse transformer more efficiently. In such circuits there is no need to combat the magnetization of the core, which makes it possible to reduce its dimensions. The output voltage is symmetrical. In addition, the transistors of the converter operate in a lighter mode.

Sometimes, for low power (up to 15 W), the simplest converter is used, made according to the circuit of a self-oscillator (Fig. 4.16, a). This circuit is not critical to the parts used, but selecting the operating point of the transistor operating mode using resistor R2 can improve the characteristics of the device (sometimes a capacitor is installed in parallel with R2). A divider of resistors R1-R2 provides the necessary initial current to start the autogenerator.

Rice. 4.16. Schemes of push-pull self-generators

The 2N3055 universal transistors used are replaced by similar domestic ones KT818GM, KT8150A, and if you change the polarity of the supplied power, then pnp transistors can also be used. The supply voltage of the circuit can be from 12 to 24 V. For long-term operation of the device, transistors must be installed on radiators.

The transformer can be made on a ferrite M2000NM1 ring magic conductor, its working cross-section depends. on the power in the load. For a simplified choice, you can use the recommendations, see table. 4.5.

Table 4.5. Permissible maximum power for ring ferrite magnetic cores of the M2000NM1 brand

When manufacturing transformer T1, windings 1 and 2 are wound simultaneously, but the phasing of their connection must correspond to that shown in the diagram. For a cross-section of a ring magnetic core of standard size K32x20x6, windings 1 and 2 each contain 8 turns (PEL wire with a diameter of 1.2...0.81 mm); 3 and 4, 2 turns each (0.23 mm); 5 - the number of turns of the secondary winding depends on the required voltage (0.1...0.23 mm).

Using this circuit, you can obtain voltages of up to 30 kV if you use a magnetic circuit from transformers used in modern TVs.

A similar circuit of a self-oscillator, made using field-effect transistors, is shown in Fig. 4.16, b. It allows the use of a simpler transformer that does not require feedback windings. Zener diodes VD1, VD2 prevent the appearance of dangerous voltages on the gates of transistors.

The operating frequency of such circuits is set by the parameters of the transformer magnetic circuit and the inductance of the windings, since the delay of the feedback signal depends on this (it is better if the frequency is in the range of 20...50 kHz).

The disadvantage of these circuits is their low efficiency, which makes it difficult to use them at high power, as well as the unstabilized output voltage, which can vary greatly depending on changes in the supply voltage. A more successful push-pull converter circuit, made using a specialized microcircuit (Fig. 4.17), is characterized by high efficiency and can maintain a stable voltage across the load.

Rice. 4.17. Push-pull pulse converter circuit

The converter is made on the widely used T114EU4 PWM controller chip (a complete imported analogue of the TL494), which makes the circuit quite simple. In the normal state (at zero gate voltage), transistors VT1, VT2 are closed and open by pulses from the corresponding outputs of the microcircuit. Resistors R7-R9 and R8-R10 limit the output current of the microcircuit, as well as the voltage at the gate of the switches. The circuit of elements C1-R2 ensures a smooth transition to operating mode when the power is turned on (a gradual increase in the pulse width at the outputs of the microcircuit). Diode VD1 prevents damage to circuit elements when the power polarity is incorrectly connected.

Stress diagrams explaining the operation are shown in Fig. 4.18. As can be seen in Figure (a), the trailing edge of the pulse has a longer duration than the leading edge. This is explained by the presence of a gate capacitance of the field-effect transistor, the charge of which is absorbed through resistor R9 (R10) during the time when the output transistor of the microcircuit is closed. This increases the time it takes to close the key. Since in the open state the voltage drops on the field-effect transistor is no more than 0.1 V, power losses in the form of slight heating of VT1 and VT2 occur mainly due to the slow closing of the transistors (this is what limits the maximum permissible load power).

Rice. 4.18. Stress diagrams

The parameters of this circuit when operating on a 100 W lamp are given in Table. 4.6. At idle, the current consumption is 0.11 A (9 V) and 0.07 A (15 V). The operating frequency of the converter is about 20 kHz.

Table 4.6. Basic parameters of the scheme

Transformer T1 is made on two ring cores made of ferrite grade M2000NM1, size K32x20x6, folded together. The parameters of the windings are indicated in the table. 4.7.

Table 4.7. Parameters of windings of transformer T1

Before winding, the sharp edges of the core must be rounded off with a file or coarse sandpaper. When making a transformer, the secondary winding is first wound. Winding is carried out turn to turn, in one layer, followed by insulation with varnished cloth or fluoroplastic tape. Primary windings 1 and 2 are wound with two wires simultaneously, as shown in Fig. 4.19 (evenly distributing the turns on the magnetic circuit). This winding can significantly reduce voltage surges at the fronts when closing the field switches. The transistors are installed on a heat sink, which is made from a duralumin profile (Fig. 4.20).

Rice. 4.19 Type of design of a pulse transformer

Rice. 4.20. Radiator design

Heatsinks are fixed to the edges of the printed circuit board. A single-sided printed circuit board made of fiberglass with a thickness of 1.5...2 mm has dimensions of 110x90 mm (see Fig. 4.21 and 4.22).

Rice. 4.21. PCB topology

Rice. 4.22. Arrangement of elements

This circuit can be used to power a load that constantly consumes power up to 100 W. For more power, it is necessary to reduce the switching time of the field switches. This can be done by specially designed microcircuits that have a complementary output stage designed to control powerful field-effect transistors, for example, K1156EU2, UC3825.

In the above circuit, N-type transistors with static induction KP958A (BCIT-Bipolar Static Induction Transistor) can also be used as power switches for power up to 60 W. They are designed specifically to operate in high frequency power supplies. The physics of operation of such a transistor is close to the operation of a conventional bipolar one, but due to its design features it has a number of advantages:

1) low source-drain voltage drop in the open state;
2) increased gain;
3) high speed when switching;
4) increased resistance to thermal breakdown.

In this case, it is better to select transistors with the same parameters, and reduce resistors R9 and R10 to 100...150 Ohms.