The article describes promising methods for increasing the efficiency of switching power supplies. In particular, the quasi-resonant method of controlling power transistors and the synchronous rectification method. The features of using these methods are described, and practical implementation on the Renesas HA16163 controller is shown.
Modern element base makes it possible to obtain quite high efficiency in classic PWM solutions - up to ~95%. In budget designs, where weight and size characteristics are not important, they are content with more modest characteristics. But there are application areas where size and efficiency come first - power supplies for the defense industry, for aircraft, server power (passive cooling), small-sized sources for laptops, telecommunications, etc. The main losses in a classic PWM switching power supply are distributed approximately as follows: 50% power switches, 40% output rectifier, 10% transformer and snubbers. As you can see, the main losses in the form of heat are dissipated in the key elements and the output rectifier. Dynamic losses in key transistors are significantly reduced due to the soft switching mode (resonant and quasi-resonant control method). This allows you to use “slower” transistors at higher conversion frequencies or using standard mass transistors to obtain conversion frequencies several times higher than in the standard PWM topology. When switching switches at zero current (PNT) or at zero voltage (ZVN), losses on snubber elements are significantly reduced, in some cases it is even possible to abandon snubbers.
The use of quasi-resonant circuitry provides the following advantages - higher efficiency than classic PWM circuits, a wide range of loads (unlike a frequency-controlled resonant circuit). In a quasi-resonant circuit, unlike a resonant circuit, the oscillatory circuit does not accumulate energy, but only participates in the transfer of energy to the load. This eliminates the need to use large resonant circuit components. However, the quasi-resonant circuit has its drawback - when the load decreases, the circuit goes into hard switching mode and the efficiency drops. In the load range in which soft switching occurs, the circuit emits a narrow spectrum of noise that is easier to suppress.
Losses in the output rectifier in the range from units to hundreds of watts, with output voltages of 1.8-80V, can be significantly reduced due to synchronous rectification.
Let's consider the circuit of a quasi-resonant converter with synchronous rectification. Figure 7 shows a timing diagram explaining the operation of the converter.
Time point 1 – Fig. 1
At moment t0, transistors S3 and S6 are open, the supply voltage Vin is connected through inductor Lr to the primary winding of the transformer, and a voltage proportional to the voltage in the primary winding appears on the secondary winding. The synchronous rectifier switches S14S15 are turned off, S16S17 are turned on. The voltage from the primary winding is supplied to the load through inductor L1.
Time point 2 – Fig. 2
At moment t1, transistor S3 is turned on. When transistor S6 is turned off, a co-induction voltage surge occurs at the resonant inductor Lr. Switches S14S15 and S16S17 of the output rectifier are turned on, thereby shunting the output winding, the energy stored in the resonant inductor Lr goes into the output capacitance of the transistor S6 - C12, C12 is charged at a speed where N = N1/N2 is the transformation ratio Iload is the load current, C12 is the output transistor capacity 
Fig.1
Fig.2
Time point 3 – Fig. 3
At moment t2, transistor S4 turns on. By this time, the output capacitance of transistor S6-C12 is charged to the supply voltage Vin (so that switching occurs at zero voltage with minimal losses). Turn-on delay of transistor S4 - 
Time point 4 – Fig.4
At time t3, transistor S3 turns off, and the output capacitance of transistor S5-C11 is discharged. There is a transfer of energy from capacitor C11 to the resonant choke Lr. Free harmonic oscillations occur in the circuit. Natural resonant frequency of the circuit 
Fig.3
Fig.4
Time point 6 – Fig.6
At time t5, when the voltage on capacitor C11 reaches zero, transistor S5 turns on. The current in the output winding changes its direction, the voltage of the secondary winding is connected to the load through inductor L2.
Rice. 5
Fig.6
Fig.7
Electronic components manufacturer Renesas produces a quasi-resonant ZVS controller HA16163, which has 4 low-current outputs for controlling the bridge circuit of the converter and 2 outputs for controlling the synchronous rectifier switches. The microcircuit allows you to build converters on its basis with a switching frequency of 1 MHz (2 MHz oscillator frequency)!
The chip has the following functions:
- soft start;
- the ability to turn on/off the converter through an additional input of the microcircuit;
- external synchronization input;
- cycle-by-cycle load current limitation;
- complete shutdown of the microcircuit during a short circuit;
- the microcircuit has a built-in error amplifier;
- the microcircuit contains 3 pins that allow you to program delays at outputs A and B, C and D, E and F.
Figure 8 shows a typical connection diagram. Integrated half-bridge drivers can be used as drivers for outputs A, B, C, D; International Rectifier offers a wide range. You can also use drivers on discrete elements using transformer isolation. As drivers for outputs E and F, it is necessary to use either optical drivers or use isolation transformers (based on the isolation requirements of the primary - secondary side of the converter).
Fig.8
Let us note important points that must be taken into account when designing a quasi-resonant converter
65 nanometers is the next goal of the Zelenograd plant Angstrem-T, which will cost 300-350 million euros. The company has already submitted an application for a preferential loan for the modernization of production technologies to Vnesheconombank (VEB), Vedomosti reported this week with reference to the chairman of the board of directors of the plant, Leonid Reiman. Now Angstrem-T is preparing to launch a production line for microcircuits with a 90nm topology. Payments on the previous VEB loan, for which it was purchased, will begin in mid-2017.
Beijing crashes Wall Street
Key American indices marked the first days of the New Year with a record drop; billionaire George Soros has already warned that the world is facing a repeat of the 2008 crisis.
The first Russian consumer processor Baikal-T1, priced at $60, is being launched into mass production
The Baikal Electronics company promises to launch into industrial production the Russian Baikal-T1 processor costing about $60 at the beginning of 2016. The devices will be in demand if the government creates this demand, market participants say.
MTS and Ericsson will jointly develop and implement 5G in Russia
Mobile TeleSystems PJSC and Ericsson have entered into cooperation agreements in the development and implementation of 5G technology in Russia. In pilot projects, including during the 2018 World Cup, MTS intends to test the developments of the Swedish vendor. At the beginning of next year, the operator will begin a dialogue with the Ministry of Telecom and Mass Communications on the formation of technical requirements for the fifth generation of mobile communications.
Sergey Chemezov: Rostec is already one of the ten largest engineering corporations in the world
The head of Rostec, Sergei Chemezov, in an interview with RBC, answered pressing questions: about the Platon system, the problems and prospects of AVTOVAZ, the interests of the State Corporation in the pharmaceutical business, spoke about international cooperation in the context of sanctions pressure, import substitution, reorganization, development strategy and new opportunities in difficult times.
Rostec is “fencing itself” and encroaching on the laurels of Samsung and General Electric
The Supervisory Board of Rostec approved the “Development Strategy until 2025”. The main objectives are to increase the share of high-tech civilian products and catch up with General Electric and Samsung in key financial indicators.
The principle is a device with an efficiency above 100%, you will say that this is a fake and everything is not real, but this is not true. The device was assembled using domestic parts. The design of the transformer has one feature: the transformer is W-shaped with a gap in the middle, but in the gap there is a neodymium magnet, which sets the initial impulse to the feedback coil. The pickup coils can be wound in any direction, but at the same time, pinpoint precision is required in their winding; they must have the same inductance. If this is not observed, then there will be no resonance; a voltmeter connected in parallel to the battery will inform you about this. I have not found any particular application in this design, but you can connect a light source in the form of incandescent lamps.
Technical characteristics at resonance:
Efficiency is above 100%
Reverse current is 163-167 milliamps (I don’t know how this happens, but the battery is charging)
Current consumption is 141 milliamps (it turns out that 20 milliamps is free energy and goes to charge the battery)
Red wire coil L1
Green wire coil L2
The black wire is the pickup coil
Settings
From my own experience, I was convinced that coil L1, wound with the same wire, is more easily tuned to resonance with L2, creating more current than is consumed. As I understand it, ferromagnetic resonance is created, which powers the load and charges the battery with a high current. To adjust the resonance, there must be two identical coils or one; when the device is turned on, they move under the load of an incandescent lamp (in my case, a 12 Volt 5 Watt lamp). To set up, connect a voltmeter in parallel to the battery and start moving the coil(s). At resonance, the voltage on the battery should begin to increase. Having reached a certain threshold, the battery will stop charging and discharging. You need to install a large heatsink on the transistor. In the case of two coils, everything is more complicated, since you need to wind them so that the inductances are practically the same; with different loads, the location of the right and left coils will change. If these tuning rules are not followed, then resonance may not occur, but we will get a simple boost converter with high efficiency. My coil parameters are 1:3, that is, L1 8 turns, L2 24 turns, both with the same wire cross-section. L1 dangles on top of L2. Removable coils, no matter what kind of wire, but I have 1.5mm.
Photo
The finished device is in a non-resonant state (coils connected in series)
Test of self-powering from a removable coil through a diode. (Result: failure, runs for 14 seconds with decay)
The state of resonance on one coil without self-powering through a diode. The experiment was successful, with the battery connected, the converter worked for 37 hours 40 minutes, without losing voltage on the battery. At the beginning of the experiment, the battery voltage was 7.15 volts, by the end it was 7.60 volts. This experience has proven that the converter is capable of delivering efficiency above 100%. For the load I used a 12 Volt 5 Watt incandescent lamp. I refused to try to use other devices, since the magnetic field around the device is very strong and creates interference within a radius of one and a half meters, the radio stops working within a radius of 10 meters.
List of radioelements
| Designation | Type | Denomination | Quantity | Note | Shop | My notepad |
|---|---|---|---|---|---|---|
| VT1 | Bipolar transistor | KT819A | 1 | KT805 | To notepad | |
| C1 | Capacitor | 0.1 µF | 1 | To notepad | ||
| C2 | Electrolytic capacitor | 50 µF 25 V | 1 | To notepad | ||
| R1 | Resistor | 2.2 kOhm | 1 | To notepad | ||
| R2 | Resistor | 62 Ohm | 1 | To notepad | ||
| Bat1 | Battery | 12 Volt | 1 |
The described device provides exceptionally high conversion efficiency, allows regulation of the output voltage and its stabilization, and operates stably when the load power varies. This type of converter is interesting and undeservedly little widespread - quasi-resonant, which is largely free from the disadvantages of other popular circuits. The idea of creating such a converter is not new, but practical implementation became feasible relatively recently, after the advent of powerful high-voltage transistors that allow significant pulse collector current at a saturation voltage of about 1.5 V. The main distinctive feature and main advantage of this type of power source is the high efficiency of the voltage converter , reaching 97...98% without taking into account losses on the secondary circuit rectifier, which are mainly determined by the load current.
The quasi-resonant converter differs from a conventional pulse converter, in which by the moment the switching transistors are closed, the current flowing through them is maximum, the quasi-resonant one differs in that by the moment the transistors are closed, their collector current is close to zero. Moreover, the reduction in current at the moment of closing is ensured by the reactive elements of the device. It differs from resonant in that the conversion frequency is not determined by the resonant frequency of the collector load. Thanks to this, it is possible to regulate the output voltage by changing the conversion frequency and realize stabilization of this voltage. Since by the time the transistor closes, the reactive elements reduce the collector current to a minimum, the base current will also be minimal and, therefore, the closing time of the transistor is reduced to the value of its opening time. Thus, the problem of through current that occurs during switching is completely eliminated. In Fig. Figure 4.22 shows a schematic diagram of a self-oscillating unstabilized power supply.
Main technical characteristics:
Overall efficiency of the unit, %................................................... ....................92;
Output voltage, V, with a load resistance of 8 Ohms....... 18;
Operating frequency of the converter, kHz....................................20;
Maximum output power, W...................................................55;
Maximum amplitude of output voltage ripple with operating frequency, V
The main share of power losses in the unit falls on the heating of the rectifier diodes of the secondary circuit, and the efficiency of the converter itself is such that there is no need for heat sinks for transistors. The power loss on each of them does not exceed 0.4 W. Special selection of transistors according to any parameters also not required. When the output is shorted or the maximum output power is exceeded, generation is interrupted, protecting the transistors from overheating and breakdown.
The filter, consisting of capacitors C1...SZ and inductor LI, L2, is designed to protect the supply network from high-frequency interference from the converter. The autogenerator is started by circuit R4, C6 and capacitor C5. The generation of oscillations occurs as a result of the action of positive feedback through transformer T1, and their frequency is determined by the inductance of the primary winding of this transformer and the resistance of resistor R3 (as the resistance increases, the frequency increases).
Chokes LI, L2 and transformer T1 are wound on identical ring magnetic cores K12x8x3 made of 2000NM ferrite. The inductor windings are performed simultaneously, “in two wires,” using PELSHO-0.25 wire; number of turns - 20. Winding I of the TI transformer contains 200 turns of PEV-2-0.1 wire, wound in bulk, evenly around the entire ring. Windings II and III are wound “in two wires” - 4 turns of PELSHO-0.25 wire; winding IV is a turn of the same wire. For the T2 transformer, a K28x16x9 ring magnetic core made of 3000NN ferrite was used. Winding I contains 130 turns of PELI10-0.25 wire, laid turn to turn. Windings II and III - 25 turns of PELSHO-0.56 wire each; winding - “in two wires”, evenly around the ring.
Choke L3 contains 20 turns of PELI10-0.25 wire, wound on two folded together ring magnetic cores K12x8x3 made of 2000NM ferrite. Diodes VD7, VD8 must be installed on heat sinks with a dissipation area of at least 2 cm2 each.
The described device was designed for use in conjunction with analog stabilizers for various voltage values, so there was no need for deep ripple suppression at the output of the unit. Ripple can be reduced to the required level by using LC filters that are common in such cases, such as, for example, in another version of this converter with the following basic technical characteristics:
Rated output voltage, V................................................... 5,
Maximum output current, A................................................... ......... 2;
Maximum pulsation amplitude, mV............................................50;
Change in output voltage, mV, no more, when the load current changes
from 0.5 to 2 A and mains voltage from 190 to 250 V........................150;
Maximum conversion frequency, kHz.................................... 20.
The circuit of a stabilized power supply based on a quasi-resonant converter is shown in Fig. 4.23.
The output voltage is stabilized by a corresponding change in the operating frequency of the converter. As in the previous block, powerful transistors VT1 and VT2 do not need heat sinks. Symmetrical control of these transistors is implemented using a separate master pulse generator assembled on a DDI chip. Trigger DD1.1 operates in the generator itself.
The pulses have a constant duration specified by the circuit R7, C12. The period is changed by the OS circuit, which includes optocoupler U1, so that the voltage at the output of the unit is maintained constant. The minimum period is set by circuit R8, C13. Trigger DDI.2 divides the repetition frequency of these pulses by two, and the square wave voltage is supplied from the direct output to the transistor current amplifier VT4, VT5. Next, the current-amplified control pulses are differentiated by the circuit R2, C7, and then, already shortened to a duration of approximately 1 μs, they enter through the transformer T1 into the base circuit of transistors VT1, VT2 of the converter. These short pulses serve only to switch transistors - closing one of them and opening the other.
In addition, the main power from the excitation generator is consumed only when switching powerful transistors, so the average current consumed by it is small and does not exceed 3 mA, taking into account the current of the zener diode VD5. This allows it to be powered directly from the primary network through the quenching resistor R1. Transistor VT3 is a control signal voltage amplifier, as in a compensation stabilizer. The stabilization coefficient of the block's output voltage is directly proportional to the static current transfer coefficient of this transistor.
The use of transistor optocoupler U1 ensures reliable galvanic isolation of the secondary circuit from the network and high noise immunity at the control input of the master oscillator. After the next switching of transistors VT1, VT2, the capacitor SY begins to recharge and the voltage at the base of the transistor VT3 begins to increase, the collector current also increases. As a result, the optocoupler transistor opens, maintaining the master oscillator capacitor C13 in a discharged state. After the rectifier diodes VD8, VD9 are closed, the capacitor SY begins to discharge to the load and the voltage across it drops. Transistor VT3 closes, as a result of which capacitor C13 begins charging through resistor R8. As soon as the capacitor is charged to the switching voltage of the trigger DD1.1, a high voltage level will be established at its direct output. At this moment, the next switching of transistors VT1, VT2 occurs, as well as the discharge of the SI capacitor through the opened optocoupler transistor.
The next process of recharging the capacitor SY begins, and the trigger DD1.1 after 3...4 μs will return to the zero state again due to the small time constant of the circuit R7, C12, after which the entire control cycle is repeated, regardless of which of the transistors is VT1 or VT2 - open during the current half-term. When the source is turned on, at the initial moment, when the capacitor SY is completely discharged, there is no current through the optocoupler LED, the generation frequency is maximum and is determined mainly by the time constant of the circuit R8, C13 (the time constant of the circuit R7, C12 is several times smaller). With the ratings of these elements indicated in the diagram, this frequency will be about 40 kHz, and after it is divided by the DDI.2 trigger - 20 kHz. After charging the capacitor SY to the operating voltage, the OS stabilizing loop on the elements VD10, VT3, U1 comes into operation, after which the conversion frequency will already depend on the input voltage and load current. Voltage fluctuations on the capacitor SY are smoothed out by filter L4, C9. Chokes LI, L2 and L3 are the same as in the previous block.
Transformer T1 is made on two ring magnetic cores K12x8x3 folded together from 2000NM ferrite. The primary winding is wound in bulk evenly throughout the entire ring and contains 320 turns of PEV-2-0.08 wire. Windings II and III each contain 40 turns of wire PEL1110-0.15; they are wound “in two wires”. Winding IV consists of 8 turns of PELSHO-0.25 wire. Transformer T2 is made on a ring magnetic core K28x16x9 made of 3000NN ferrite. Winding I - 120 turns of PELSHO-0.15 wire, and II and III - 6 turns of PEL1110-0.56 wire, wound “in two wires”. Instead of PELSHO wire, you can use PEV-2 wire of the appropriate diameter, but in this case it is necessary to lay two or three layers of varnished cloth between the windings.
Choke L4 contains 25 turns of wire PEV-2-0.56, wound on a ring magnetic core K12x6x4.5 made of 100NNH1 ferrite. Any ready-made inductor with an inductance of 30...60 μH for a saturation current of at least 3 A and an operating frequency of 20 kHz is also suitable. All fixed resistors are MJIT. Resistor R4 - adjusted, of any type. Capacitors C1...C4, C8 - K73-17, C5, C6, C9, SY - K50-24, the rest - KM-6. The KS212K zener diode can be replaced with KS212Zh or KS512A. Diodes VD8, VD9 must be installed on radiators with a dissipation area of at least 20 cm2 each. The efficiency of both blocks can be increased if, instead of KD213A diodes, Schottky diodes are used, for example, any of the KD2997 series. In this case, heat sinks for diodes will not be required.
Half-bridge quasi-resonant power supply
To improve the characteristics of switching power supplies assembled on the basis of bridge and half-bridge converters, in particular, to reduce the likelihood of through current and increase efficiency, the authors propose to transfer such sources to a quasi-resonant operating mode. The described article provides a practical example of such a power supply.
Often, to reduce size and weight, power supplies (PS) with a network transformer are replaced with pulse voltage converters. The benefit from this is obvious: lower weight and dimensions, significantly lower copper consumption for winding products, high efficiency of the power supply. However, pulsed power supplies also have disadvantages: poor electromagnetic compatibility, the possibility of through current appearing through transistors in push-pull converters, the need to introduce overcurrent protection circuits, and the difficulty of starting a capacitive load without taking special measures to limit the charging current.
Let us consider, using the example of a push-pull half-bridge self-oscillating voltage converter, how to a certain extent these disadvantages can be eliminated or reduced by changing its operating mode. Let's switch the converter to a quasi-resonant operating mode by introducing a resonant circuit. The shape of the current through the primary winding of the pulse transformer in this case is shown in Fig. 1.
In Fig. Figure 2 shows the voltage and current waveforms for one of the switching transistors. From the figures it can be seen that the converter operates in a quasi-resonant mode - there is no through current in this case.
The voltage at the base of the switching transistor decreases and becomes zero at the end of the pulse. Thus, the transition to a quasi-resonant operating mode completely eliminates dynamic losses in switching transistors and problems associated with the electromagnetic compatibility of sensitive devices with pulsed power supply, since the spectrum of generated oscillations is sharply narrowed.
A half-bridge converter differs from a push-pull bridge converter in the smaller number of transistors used; from a push-pull with a middle output - half the voltage on the transistors. A self-generating converter differs from converters with a master oscillator, first of all, in the minimum number of elements, the maximum possible efficiency, and the use of a saturable auxiliary transformer is guaranteed to exclude the possibility of through current.
The circuit of a half-bridge quasi-resonant power supply, devoid of the listed disadvantages, is shown in Fig. 3.
(click to enlarge)
Main technical characteristics
- Supply voltage change interval, V....198...264
- Maximum efficiency, %......92
- Output voltage, V, with a load resistance of 36 Ohms......36
- Operating frequency conversion interval, kHz......12...57
- Maximum output power, W......70
- Maximum amplitude of output voltage ripple with operating frequency, V......2.2
The IP contains the following components: noise suppression filter C1C2L1, which prevents the penetration of high-frequency ripples created by the converter into the supply network; network rectifier VD1 with filter capacitor C3; protection circuits against overload and short circuits in the load R1R2VD2K1U1VD3VD4R6R7C7. The protection circuit consumes insignificant current, therefore it has little effect on the overall efficiency of the source, but if necessary, the efficiency can be increased slightly by replacing the zener diode VD2 with a higher voltage one. Resistors R6 and R7 form a voltage divider necessary to turn on the emitting diode of the thyristor optocoupler. If these fixed resistors are replaced with one variable resistor, the protection threshold can be adjusted within very wide limits. If you plan to power a load with a large capacitance (more than 5000 μF), to eliminate false protection triggers, you should increase the capacitance of capacitor C7, however, the waiting time before turning on the source will increase in this case.
Elements R3, R4, C4, C5 form a voltage divider. Resistors R3, R4 are necessary to discharge the capacitors of filter C3 and divider C4C5 after turning off the power supply. Capacitor C6 and inductor L2 are a resonant circuit. The triggering circuit is exactly the same as in the device described in the article. It consists of transistor VT3, resistors R10-R12 and capacitor C10. Transistor VT3 operates in avalanche mode. The triggering pulse opens transistor VT2, providing initial asymmetry.
Diodes VD5-VD8 - output rectifier with filter capacitors C8, C9. LED HL1 indicates the presence of voltage at the output of the IP. Self-generation of oscillations occurs as a result of positive feedback from winding III of transformer T1 to winding III of transformer T2 through current-limiting resistor R9. As its resistance decreases, the conversion frequency decreases, which leads to a shift in the maximum efficiency of the source towards a higher load power.
The device uses capacitors K73-17 (C1, C2, C6, C9, C10), K73-11 (C4, C5), K50-32 (C3), K50-24 (C7, C8). All resistors are C2-23. Instead of the specified capacitors and resistors, it is possible to use other components, however, capacitors should be selected with a minimum dielectric loss tangent in the operating frequency range of the power supply conversion.
Diode bridge VD1 - any with a permissible forward current of more than 1 A and a permissible reverse voltage of at least 400 V, for example BR310. It is also possible to use discrete diodes, for example KD202R, connected via a bridge circuit. It is best to use the KT315G (VT3) transistor in the device - the triggering circuit will work with it immediately, the KT315B transistor will have to be selected, and it is better not to use the KT315A, KT315V transistors. Transistors KT826V (VT1, VT2) are interchangeable with any of the KT826 or KT812A, KT812B series. Due to low losses, transistors can not be installed on heat sinks. The diodes of the output rectifier KD213A (VD5-VD8) can be replaced with KD213B, KD213V or the KD2997, KD2999 series. They should be installed on a heat sink with a cooling surface area of at least 10 cm2.
The IP uses an electromagnetic DC relay GBR10.1-11.24 with an operating voltage of 24 V, capable of switching an alternating current of 8 A in circuits with voltages up to 250 V. It can be replaced by any other with a permissible switched alternating current of at least 1 A in circuits with voltage 250 V. However, it is advisable to use a relay with a minimum switching current to increase the efficiency of the power supply, since the lower the switching current, the greater the resistance of resistors R1, R2 and the less power will be dissipated on them.
Chokes L1, L2 and transformer T1 were used ready-made - from an old EC1060 computer: L1 - I5, L2 - 4777026 or 009-01, T1 - 052-02. You can make them yourself. Inductor L1 is wound (two windings at the same time) on a ring magnetic core K28x16x9 made of ferrite (for example, grades M2000NM-A or M2000NM1-17) or alsifer. Its windings contain 315 turns of PEV-2 0.3 wire.
The resonant choke L2 is wound on a ring magnetic core K20x10x5 made of M2000NM-A ferrite. Its winding contains 13 turns of PEV-2 0.6 wire.
Transformer T1 is wound on a ring magnetic core K45x28x8 made of M2000NM1-17 ferrite. Winding I contains 200 turns of PEV-2 0.6 wire, winding II - 35 turns of PEV-2 1 wire, winding III - 5 turns of PEV-2 0.6 wire. The order of winding the windings on the magnetic circuit is arbitrary. Between the windings it is necessary to lay a layer of insulation, for example, fluoroplastic tape. In addition, the transformer should be impregnated, for example, with paraffin from candles or ceresin. This will not only increase the dielectric strength of the insulation, but also reduce the hum created by the source at idle.
Transformer T2 is wound on a ring magnetic core K20x10x5 made of M2000NM-A ferrite. Windings I and II each contain seven turns of PEV-2 0.3 wire (they are wound simultaneously into two wires), and winding III contains nine turns of PEV-2 0.3 wire.
The design of the power supply can be arbitrary; the relative position of the elements on the board is not critical. It is only important to ensure good air flow to the semiconductor devices by natural convection or install the power supply inside the powered device near the fan.
The described IP practically does not require adjustment, although it is worth making sure that the converter operates in a quasi-resonant mode. To do this, an equivalent load is connected to the output of the power supply - a resistor with a power of 100 W and a resistance of 36 Ohms. An additional resistor with a resistance of 0.1...1 Ohm and a power of 1...2 W is connected in series with capacitor C6. The oscilloscope probes are connected to an additional resistor: common - to the midpoint of the voltage divider R3R4C4C5, signal - to capacitor C6. It is necessary to ensure that the oscilloscope is not galvanically connected to the network. If connected, it should be connected to the network through an isolation transformer with a transformation ratio of 1:1. In any case, safety regulations must be followed. By applying power to the IP, make sure that there are bell-shaped current pulses with a pause at zero. If the pulse shape differs from that shown in Fig. 1, it is necessary to select the number of turns of inductor L2 until resonance is obtained.
On an additional resistor with a resistance of 0.1 Ohm, the pulse amplitude should be about 0.1 V. Now you should compare the shape of the current and voltage on the switching transistor VT2 with those shown in Fig. 2 graphs. If they are close in shape, the IP operates in a quasi-resonant mode.
The protection threshold can be changed. To do this, select the resistance of resistor R7 so that the protection operates at the required load current. If it is necessary for the power supply to be turned off when the load power is less than 70 W, the resistance of resistor R7 should be reduced.
To limit the charging current of capacitor C3 at the moment of switching on, we recommend connecting a resistor with a resistance of 5.6 ... 10 Ohms with a power of 2 W to the gap of any network wire.
Literature
- Baraboshkin D. Improved economical power supply. - Radio, 1985, No. 6, p. 51.52.
- Konovalov E. Quasi-resonant voltage converter. - Radio, 1996, No. 2, p. 52-55.
See other articles section.