Rectifiers with thyristor voltage regulator. Thyristor power regulator: circuit, operating principle and application For the “power amplifier for SV radio station” circuit

A selection of circuits and a description of the operation of a power regulator using triacs and more. Triac power regulator circuits are well suited for extending the life of incandescent lamps and for adjusting their brightness. Or for powering non-standard equipment, for example, 110 volts.

The figure shows a circuit of a triac power regulator, which can be changed by changing the total number of network half-cycles passed by the triac over a certain time interval. The elements of the DD1.1.DD1.3 microcircuit are made with an oscillation period of about 15-25 network half-cycles.

The duty cycle of the pulses is regulated by resistor R3. Transistor VT1 together with diodes VD5-VD8 is designed to bind the moment the triac is turned on during the transition of the mains voltage through zero. Basically, this transistor is open, respectively, a “1” is sent to the input DD1.4 and transistor VT2 with triac VS1 are closed. At the moment of crossing zero, transistor VT1 closes and opens almost immediately. In this case, if the output DD1.3 was 1, then the state of the elements DD1.1.DD1.6 will not change, and if the output DD1.3 was “zero”, then the elements DD1.4.DD1.6 will generate a short pulse, which will be amplified by transistor VT2 and open the triac.

As long as there is a logical zero at the output of the generator, the process will proceed cyclically after each transition of the mains voltage through the zero point.

The basis of the circuit is a foreign triac mac97a8, which allows you to switch high-power connected loads, and to regulate it I used an old Soviet variable resistor, and used a regular LED as an indication.

The triac power regulator uses the principle of phase control. The operation of the power regulator circuit is based on changing the moment the triac is turned on relative to the transition of the mains voltage through zero. At the initial moment of the positive half-cycle, the triac is in the closed state. As the mains voltage increases, capacitor C1 is charged through a divider.

The increasing voltage on the capacitor is shifted in phase from the mains voltage by an amount depending on the total resistance of both resistors and the capacitance of the capacitor. The capacitor is charged until the voltage across it reaches the “breakdown” level of the dinistor, approximately 32 V.

At the moment the dinistor opens, the triac will also open, and a current will flow through the load connected to the output, depending on the total resistance of the open triac and the load. The triac will be open until the end of the half-cycle. With resistor VR1 we set the opening voltage of the dinistor and triac, thereby regulating the power. At the time of the negative half-cycle, the circuit operation algorithm is similar.

Option of the circuit with minor modifications for 3.5 kW

The controller circuit is simple, the load power at the output of the device is 3.5 kW. With this homemade amateur radio you can adjust lighting, heating elements and much more. The only significant drawback of this circuit is that you cannot connect an inductive load to it under any circumstances, because the triac will burn out!


Radio components used in the design: Triac T1 - BTB16-600BW or similar (KU 208 or VTA, VT). Dinistor T - type DB3 or DB4. Capacitor 0.1 µF ceramic.

Resistance R2 510 Ohm limits the maximum volts on the capacitor to 0.1 μF; if you put the regulator slider in the 0 Ohm position, the circuit resistance will be about 510 Ohms. The capacitance is charged through resistors R2 510 Ohm and variable resistance R1 420 kOhm, after U on the capacitor reaches the opening level of dinistor DB3, the latter will generate a pulse that unlocks the triac, after which, with further passage of the sinusoid, the triac is locked. The opening and closing frequency of T1 depends on the level of U on the 0.1 μF capacitor, which depends on the resistance of the variable resistor. That is, by interrupting the current (at a high frequency) the circuit thereby regulates the output power.

With each positive half-wave of the input alternating voltage, capacitance C1 is charged through a chain of resistors R3, R4, when the voltage on capacitor C1 becomes equal to the opening voltage of dinistor VD7, its breakdown will occur and the capacitance will be discharged through the diode bridge VD1-VD4, as well as resistance R1 and control electrode VS1. To open the triac, an electrical chain of diodes VD5, VD6, capacitor C2 and resistance R5 is used.

It is necessary to select the value of resistor R2 so that at both half-waves of the mains voltage, the regulator triac operates reliably, and it is also necessary to select the values ​​of resistances R3 and R4 so that when the variable resistance knob R4 is rotated, the voltage on the load smoothly changes from minimum to maximum values. Instead of the TC 2-80 triac, you can use TC2-50 or TC2-25, although there will be a slight loss in the permissible power in the load.

KU208G, TS106-10-4, TS 112-10-4 and their analogs were used as a triac. At the moment when the triac is closed, capacitor C1 is charged through the connected load and resistors R1 and R2. The charging speed is changed by resistor R2, resistor R1 is designed to limit the maximum value of the charge current

When the threshold voltage value is reached on the capacitor plates, the switch opens, capacitor C1 is quickly discharged to the control electrode and switches the triac from the closed state to the open state; in the open state, the triac bypasses the circuit R1, R2, C1. At the moment the mains voltage passes through zero, the triac closes, then capacitor C1 is charged again, but with a negative voltage.

Capacitor C1 from 0.1...1.0 µF. Resistor R2 1.0...0.1 MOhm. The triac is switched on by a positive current pulse to the control electrode with a positive voltage at the conventional anode terminal and by a negative current pulse to the control electrode with a negative voltage at the conventional cathode. Thus, the key element for the regulator must be bidirectional. You can use a bidirectional dinistor as a key.

Diodes D5-D6 are used to protect the thyristor from possible breakdown by reverse voltage. The transistor operates in avalanche breakdown mode. Its breakdown voltage is about 18-25 volts. If you don’t find P416B, then you can try to find a replacement for it.

The pulse transformer is wound on a ferrite ring with a diameter of 15 mm, grade N2000. The thyristor can be replaced with KU201

The circuit of this power regulator is similar to the circuits described above, only the interference suppression circuit C2, R3 is introduced, and the switch SW makes it possible to break the charging circuit of the control capacitor, which leads to instant locking of the triac and disconnecting the load.

C1, C2 - 0.1 MKF, R1-4k7, R2-2 mOhm, R3-220 Ohm, VR1-500 kOhm, DB3 - dinistor, BTA26-600B - triac, 1N4148/16 V - diode, any LED.

The regulator is used to regulate load power in circuits up to 2000 W, incandescent lamps, heating devices, soldering iron, asynchronous motors, car charger, and if you replace the triac with a more powerful one, it can be used in the current regulation circuit in welding transformers.

The principle of operation of this power regulator circuit is that the load receives a half-cycle of the mains voltage after a selected number of skipped half-cycles.


The diode bridge rectifies alternating voltage. Resistor R1 and zener diode VD2, together with the filter capacitor, form a 10 V power source to power the K561IE8 microcircuit and the KT315 transistor. The rectified positive half-cycles of the voltage passing through capacitor C1 are stabilized by the zener diode VD3 at a level of 10 V. Thus, pulses with a frequency of 100 Hz follow to the counting input C of the K561IE8 counter. If switch SA1 is connected to output 2, then a logical one level will be constantly present at the base of the transistor. Because the microcircuit reset pulse is very short and the counter manages to restart from the same pulse.

Pin 3 will be set to a logical one level. The thyristor will be open. All power will be released at the load. In all subsequent positions of SA1 at pin 3 of the counter, one pulse will pass through 2-9 pulses.

The K561IE8 chip is a decimal counter with a positional decoder at the output, so the logical one level will be periodic at all outputs. However, if the switch is installed on output 5 (pin 1), then counting will only occur up to 5. When the pulse passes through output 5, the microcircuit will be reset to zero. Counting will begin from zero, and a logical one level will appear at pin 3 for the duration of one half-cycle. During this time, the transistor and thyristor open, one half-cycle passes to the load. To make it clearer, I present vector diagrams of the circuit operation.

If you need to reduce the load power, you can add another counter chip by connecting pin 12 of the previous chip to pin 14 of the next one. By installing another switch, you can adjust the power up to 99 missed pulses. Those. you can get about a hundredth of the total power.

The KR1182PM1 microcircuit has two thyristors and a control unit for them. The maximum input voltage of the KR1182PM1 microcircuit is about 270 Volts, and the maximum load can reach 150 Watts without the use of an external triac and up to 2000 W with the use, and also taking into account the fact that the triac will be installed on the radiator.


To reduce the level of external interference, capacitor C1 and inductor L1 are used, and capacitance C4 is required for smooth switching on of the load. The adjustment is carried out using resistance R3.

A selection of fairly simple regulator circuits for a soldering iron will make life easier for a radio amateur.

Combination consists in combining the ease of use of a digital regulator and the flexibility of adjusting a simple one.


The considered power regulator circuit works on the principle of changing the number of periods of the input alternating voltage going to the load. This means that the device cannot be used to adjust the brightness of incandescent lamps due to visible blinking. The circuit makes it possible to regulate power within eight preset values.

There are a huge number of classic thyristor and triac regulator circuits, but this regulator is made on a modern element base and, in addition, was phase-based, i.e. does not transmit the entire half-wave of the mains voltage, but only a certain part of it, thereby limiting the power, since the triac opens only at the required phase angle.

The article describes how a thyristor power regulator works, the diagram of which will be presented below

In everyday life, very often there is a need to regulate the power of household appliances, such as electric stoves, soldering irons, boilers and heating elements, in transport - engine speed, etc. The simplest amateur radio design comes to the rescue - a power regulator on a thyristor. Assembling such a device will not be difficult; it can become the very first home-made device that will perform the function of adjusting the temperature of the soldering iron tip of a novice radio amateur. It is worth noting that ready-made soldering stations with temperature control and other nice functions are an order of magnitude more expensive than a simple soldering iron. A minimal set of parts allows you to assemble a simple thyristor power regulator for wall mounting.

For your information, surface mounting is a method of assembling radio-electronic components without using a printed circuit board, and with good skill it allows you to quickly assemble electronic devices of medium complexity.

You can also order a thyristor regulator, and for those who want to figure it out on their own, a diagram will be presented below and the principle of operation will be explained.

By the way, this is a single-phase thyristor power regulator. Such a device can be used to control power or speed. However, first we need to understand this because this will allow us to understand for what load it is better to use such a regulator.

How does a thyristor work?

A thyristor is a controlled semiconductor device capable of conducting current in one direction. The word “controlled” was used for a reason, because with its help, unlike a diode, which also conducts current only to one pole, you can select the moment when the thyristor begins to conduct current. The thyristor has three outputs:

  • Anode.
  • Cathode.
  • Control electrode.

In order for current to begin flowing through the thyristor, the following conditions must be met: the part must be in a circuit that is energized, and a short-term pulse must be applied to the control electrode. Unlike a transistor, controlling a thyristor does not require holding the control signal. The nuances do not end there: the thyristor can be closed only by interrupting the current in the circuit, or by generating a reverse anode-cathode voltage. This means that the use of a thyristor in DC circuits is very specific and often unwise, but in AC circuits, for example in a device such as a thyristor power regulator, the circuit is constructed in such a way that a condition for closing is ensured. Each half-wave will close the corresponding thyristor.

Most likely, you don’t understand everything? Do not despair - below we will describe in detail the process of operation of the finished device.

Scope of application of thyristor regulators

In what circuits is it effective to use a thyristor power regulator? The circuit allows you to perfectly regulate the power of heating devices, that is, influence the active load. When working with a highly inductive load, the thyristors may simply not close, which can lead to failure of the regulator.

Is it possible to have an engine?

I think many of the readers have seen or used drills, angle grinders, which are popularly called “grinders,” and other power tools. You may have noticed that the number of revolutions depends on the depth of pressing the trigger button of the device. It is in this element that a thyristor power regulator is built in (the diagram of which is shown below), with the help of which the number of revolutions is changed.

Note! The thyristor regulator cannot change the speed of asynchronous motors. Thus, the voltage is regulated on commutator motors equipped with a brush assembly.

Scheme of one and two thyristors

A typical circuit for assembling a thyristor power regulator with your own hands is shown in the figure below.

The output voltage of this circuit is from 15 to 215 volts; in the case of using the indicated thyristors installed on heat sinks, the power is about 1 kW. By the way, the switch with the light brightness control is made according to a similar scheme.

If you don't need to fully regulate the voltage and just want an output of 110 to 220 volts, use this diagram, which shows a half-wave thyristor power regulator.

How it works?

The information described below is valid for most schemes. Letter designations will be taken in accordance with the first circuit of the thyristor regulator

A thyristor power regulator, the operating principle of which is based on phase control of the voltage value, also changes the power. This principle lies in the fact that under normal conditions the load is affected by the alternating voltage of the household network, changing according to a sinusoidal law. Above, when describing the operating principle of a thyristor, it was said that each thyristor operates in one direction, that is, it controls its own half-wave from a sine wave. What does it mean?

If you periodically connect a load using a thyristor at a strictly defined moment, the value of the effective voltage will be lower, since part of the voltage (the effective value that “falls” on the load) will be less than the mains voltage. This phenomenon is illustrated in the graph.

The shaded area is the area of ​​stress that is under load. The letter “a” on the horizontal axis indicates the opening moment of the thyristor. When the positive half-wave ends and the period with the negative half-wave begins, one of the thyristors closes, and at the same moment the second thyristor opens.

Let's figure out how our specific thyristor power regulator works

Scheme one

Let us stipulate in advance that instead of the words “positive” and “negative”, “first” and “second” (half-wave) will be used.

So, when the first half-wave begins to act on our circuit, capacitors C1 and C2 begin to charge. Their charging speed is limited by potentiometer R5. this element is variable, and with its help the output voltage is set. When the voltage necessary to open dinistor VS3 appears on capacitor C1, the dinistor opens and current flows through it, with the help of which thyristor VS1 will be opened. The moment of breakdown of the dinistor is point “a” on the graph presented in the previous section of the article. When the voltage value passes through zero and the circuit is under the second half-wave, the thyristor VS1 closes, and the process is repeated again, only for the second dinistor, thyristor and capacitor. Resistors R3 and R3 are used for control, and R1 and R2 are used for thermal stabilization of the circuit.

The principle of operation of the second circuit is similar, but it controls only one of the half-waves of alternating voltage. Now, knowing the principle of operation and the circuit, you can assemble or repair a thyristor power regulator with your own hands.

Using the regulator in everyday life and safety precautions

It must be said that this circuit does not provide galvanic isolation from the network, so there is a danger of electric shock. This means that you should not touch the regulator elements with your hands. An insulated enclosure must be used. You should design the design of your device so that, if possible, you can hide it in an adjustable device and find free space in the case. If the adjustable device is located permanently, then in general it makes sense to connect it through a switch with a dimmer. This solution will partially protect against electric shock, eliminate the need to find a suitable housing, has an attractive appearance and is manufactured using an industrial method.

The device shown in Fig. 1 is designed for smooth regulation in low-power loads. With its help, you can power a second additional radio device from one power source that has power reserves. For example, a 15...20 V power supply powers the required circuit, but you need to additionally power a transistor receiver from it, which has a lower supply voltage (3...9 V). Scheme made on a field-effect epitaxial-planar transistor with a p-n junction and an n-channel KP903. When operating the device, the property of the current-voltage characteristics of this transistor at different voltages between the gate and source is used. The KP903A...B family of characteristics is given in. The input supply voltage of this device is 15...20 V. Resistor R2 type PPB-ZA with a nominal value of 150 Ohms. With its help you can set the required voltage in the load. Disadvantage regulator is the increase in the internal resistance of the device when the operating voltage decreases. T160 current regulator circuit Figure 2 shows scheme indicator voltage the above-described regulator assembled on a field-effect transistor KP103. The device is designed to control voltage under load. Connecting this indicator to the device regulator is carried out according to the given diagram. Depending on the letter index KP103 of the indicator installed in the circuit (Fig. 2), we will record (by the moment the HL1 LED lights up when the output voltage increases) the operating voltage in the load. The effect of fixing different voltages in the load is obtained as a result of the fact that the KP103 channel transistors have different voltage cutoff depending on the letter index, for example, for the KP103E transistor it is 0.4-1.5 V, for the KP103Zh - 0.5-2.2 V, for the KP103I - 0.8-3 V, etc. Having installed the transistor...

For the "Simple power regulator" circuit

The load of this simple power can include incandescent lamps, heating devices of various types, etc., the power corresponding to the thyristors used. The method for setting up the regulator is contained in the selection of a variable control resistor. However, it is best to select such a potentiometer in series with a constant resistor so that the voltage at the power output varies within the widest possible range. A. ANDRIENKO, Kostroma....

For the "Universal low voltage power supply" circuit

In practice, very often 3 to 12 V are required to power various devices. The described power supply allows you to obtain the following series: 3; 4.5(5); 9; 12 V at load current up to 300 mA. It is possible to quickly change the polarity of the output voltage. ...

For the "VOLTAGE CONVERTER" circuit

Power supply CONVERTER S. Sych225876, Brest region, Kobrin district, Orekhovsky village, Lenin st., 17 - 1. I propose a simple and reliable converter circuit voltage for managing varicaps in various designs, which produces 20 V when supplied from 9 V. The converter option with a voltage multiplier was chosen, since it is considered the most economical. In addition, it does not interfere with radio reception. A pulse generator close to rectangular is assembled on transistors VT1 and VT2. A voltage multiplier is assembled using diodes VD1...VD4 and capacitors C2...C5. Resistor R5 and zener diodes VD5, VD6 form a parametric voltage stabilizer. Capacitor C6 at the output is a high-pass filter. The current consumption of the converter depends on voltage power supply and number of varicaps, as well as their type. It is advisable to enclose the device in a screen to reduce interference from the generator. A correctly assembled device works immediately and is not critical to the ratings of the parts....

For the circuit "Voltage converter 5 -> 230V"

Power supply Converter 5 -> 230 V Chips: DD1 - K155LA3 DD2 - K1554TM2 Transistors: VT1 - VT3 - KT698G, VT2 - VT4 - KT827B, VT5 - KT863Resistors: R1 - 910, R2 - 1k, R3 - 1k, R4 -120 0.25 W , R5 - 120 0.25 W, R6 - 500 0.25 W, R7 - R8 - 56 Ohm 2W, R9 - 1.5 kOm2W Diode VD5 - KC620A two in series Capacitors: C1 - 10H5 C2 - 22 μF x450V Transformer: T1 - two windings of 10 volts t connected series 16A; one winding at 220 volts, current 1A, frequency 25 kHz = Converter voltage 5 - 230V...

For the diagram "Repairing a charger for an MPEG4 player"

After two months of use, the “nameless” charger for a pocket MPEG4/MP3/WMA player failed. Of course, there was no schematic for it, so I had to draw it up from the circuit board. The numbering of the active elements on it (Fig. 1) is conditional, the rest correspond to the inscriptions on the printed circuit board. Converter assembly voltage implemented on a low-power high-voltage transistor VT1 type MJE13001, output stabilization unit voltage produced on transistor VT2 and optocoupler VU1. In addition, transistor VT2 protects VT1 from overload. Transistor VT3 is intended to indicate the end of battery charging. Upon inspection of the product, it turned out that transistor VT1 “went to a break”, and VT2 was broken. Resistor R1 also burned out. Troubleshooting took no more than 15 minutes. But with proper repair of any radio-electronic product, it is usually not enough to just eliminate the malfunctions; you also need to find out the reasons for their occurrence so that this does not happen again. Radomkrofon circuits As it turned out, during the operating hour of the charger, moreover, with the load turned off and the case open, transistor VT1, made in the TO-92 case, heated up to a temperature of approximately 90 ° C. Since there were no more powerful transistors nearby that could replace the MJE13001, I decided to glue a small heat sink to it. A photograph of the charger is shown in Fig. 2. A duralumin radiator with dimensions of 37x15x1 mm is glued to the transistor body using Radial teleconductive glue. The same glue can be used to glue the radiator to the circuit board. With the heat sink, the temperature of the transistor body dropped to 45...50°C. The reason for the initially strong heating of transistor VT1. Perhaps it lies in the “simplification” when assembling its damper circuit. The design and topology of the printed circuit board give reason to believe that...

For the diagram "Power regulator on three parts"

Recently, resistor and transistor power regulators have experienced a real renaissance. They are the most uneconomical. You can increase the efficiency in the same way as by turning on a diode (see figure). In this case, a more convenient control limit is achieved (50-100%). Semiconductor devices can be placed on one heatsink. Yu.I.Borodaty, Ivano-Frankivsk region. Literature 1. Danilchuk A.A.

Power regulator for soldering iron / /Radioamator-Electric. -2000. -No. 9. -P.23. 2.Rishtun A Tension regulator on six parts //Radioamator-Electric. -2000. -No. 11. -P.15....

For the circuit "Converter DC 12 V to AC 220 V" scheme Power supplyConverter DC 12 V to AC 220 V Anton Stoilov Offered voltage 12V AC 220V, which when connected to a 44Ah car battery can power a 100W load for 2-3 hours. It consists of a master oscillator on a symmetrical multivibrator VT1, VT2, loaded on powerful paraphase switches VT3-VT8, which switch the current in the primary winding of the step-up transformer TV. VD3 and VD4 protect powerful transistors VT7 and VT8 from overvoltages when operating without load. The transformer is made on a magnetic core Ш36х36, windings W1 and W1" each have 28 turns of PEL 2.1, and W2 - 600 turns of PEL 0.59, and W2 is wound first, and W1 is wound on top of it with a double wire (with the goal of achieving symmetry of the half-windings). When adjusting with trimmer RP1, minimal distortion of the output shape is achieved voltage"Radio Television Electronics" N6/98, p. 12,13....

For the "LED voltage indicator" circuit

In the practice of a radio amateur, a situation often arises when it is necessary to monitor the readings of one or another parameter. I propose a diagram of an LED indicator “ruler”. Depending on the input, more or less LEDs are lit, arranged in a line (one after the other). Range of permissible voltage- 4...12V, i.e. at an input voltage of 4 V, only one (first) LED will glow, and at 12 V, the entire line will glow. The capabilities of the circuit can be easily expanded. To monitor alternating voltage, it is enough to install a diode bridge of low-power diodes before resistor R1. The supply voltage can be varied from 5 to 15 V by selecting resistors R2...R8 accordingly. The brightness of the LEDs mainly depends on the power supply of the circuit, while the input characteristics of the circuit practically do not change. In order for the brightness of the LEDs to be the same, resistors should be selected as follows: where Iк max is the collector current VT1, mA; R3=2R2; R4=3R2; R5=4R2; R6=5R2; R7=6R2; R8 = 7R2. Thus, when using the KT312A transistor (lK max = 30 mA) R2 = 33 Ohm. Resistor R1 is included in the divider voltage and regulates the operating mode of transistor VT1. Diodes VD1...VD7 can be replaced with KD103A, KD105, D220, LEDs HL1...HL8 - with AL102. Resistor R9 limits the base current of transistor VT1 and prevents the latter from failing when high voltage is applied to the input of the circuit. A. KASHKAROV, St. Petersburg....

For the diagram "Universal voltage regulator and charger-starter for"

Quite often in amateur radio practice there is a need to adjust the AC within 0...220 V. LATRs (autotransformers) are widely used for this purpose. But their age has already passed and these bulky devices have been replaced by modern thyristor regulators, which have one drawback: the voltage in such devices is regulated by changing the duration of alternating voltage pulses. Because of this, it is impossible to connect a highly inductive load to them (for example, a transformer or inductor, as well as any other radio device containing the elements listed above). The voltage regulator shown in the figure is free from this drawback. It combines: overcurrent protection device, thyristor regulator voltage with bridge regulator, high efficiency (92...98%). In addition, the regulator is a simple thermostat based on a triac and works in conjunction with a powerful transformer and rectifier, which can be used to charge car batteries and as a starting device when the battery is discharged. Main parameters regulator voltage: Rated supply voltage, V 220 ± 10%; AC output voltage, V 0...215; Efficiency, no less, percent(s) 92; Maximum load power, kW 2. Main parameters of the charging and starting device: DC output voltage, V 0...40; Direct current consumed by the load, A 0...20; Starting current (with starting duration 10 s), A 100. Switch...

When developing an adjustable power supply without a high-frequency converter, the developer is faced with the problem that with a minimum output voltage and a large load current, a large amount of power is dissipated by the stabilizer on the regulating element. Until now, in most cases, this problem was solved this way: they made several taps at the secondary winding of the power transformer and divided the entire output voltage adjustment range into several subranges. This principle is used in many serial power supplies, for example, UIP-2 and more modern ones. It is clear that the use of a power source with several subranges becomes more complicated, and remote control of such a power source, for example, from a computer, also becomes more complicated.

It seemed to me that the solution was to use a controlled rectifier on a thyristor, since it becomes possible to create a power source controlled by one knob for setting the output voltage or by one control signal with an output voltage adjustment range from zero (or almost from zero) to the maximum value. Such a power source could be made from commercially available parts.

To date, controlled rectifiers with thyristors have been described in great detail in books on power supplies, but in practice they are rarely used in laboratory power supplies. They are also rarely found in amateur designs (except, of course, for chargers for car batteries). I hope that this work will help change this state of affairs.

In principle, the circuits described here can be used to stabilize the input voltage of a high-frequency converter, for example, as is done in the “Electronics Ts432” TVs. The circuits shown here can also be used to make laboratory power supplies or chargers.

I give a description of my work not in the order in which I carried it out, but in a more or less orderly manner. Let's look at general issues first, then “low-voltage” designs such as power supplies for transistor circuits or charging batteries, and then “high-voltage” rectifiers for powering vacuum tube circuits.

Operation of a thyristor rectifier with a capacitive load

The literature describes a large number of thyristor power regulators operating on alternating or pulsating current with a resistive (for example, incandescent lamps) or inductive (for example, an electric motor) load. The rectifier load is usually a filter in which capacitors are used to smooth out ripples, so the rectifier load can be capacitive in nature.

Let's consider the operation of a rectifier with a thyristor regulator for a resistive-capacitive load. A diagram of such a regulator is shown in Fig. 1.

Rice. 1.

Here, as an example, a full-wave rectifier with a midpoint is shown, but it can also be made using another circuit, for example, a bridge. Sometimes thyristors, in addition to regulating the voltage at the load U n They also perform the function of rectifier elements (valves), however, this mode is not allowed for all thyristors (KU202 thyristors with some letters allow operation as valves). For clarity of presentation, we assume that thyristors are used only to regulate the voltage across the load U n , and straightening is performed by other devices.

The operating principle of a thyristor voltage regulator is illustrated in Fig. 2. At the output of the rectifier (the connection point of the cathodes of the diodes in Fig. 1), voltage pulses are obtained (the lower half-wave of the sine wave is “turned” up), designated U rect . Ripple frequency f p at the output of the full-wave rectifier is equal to twice the network frequency, i.e. 100 Hz when powered from mains 50 Hz . The control circuit supplies current pulses (or light if an optothyristor is used) with a certain delay to the thyristor control electrode t z relative to the beginning of the pulsation period, i.e. the moment when the rectifier voltage U rect becomes equal to zero.

Rice. 2.

Figure 2 is for the case where the delay t z exceeds half the pulsation period. In this case, the circuit operates on the incident section of a sine wave. The longer the delay when the thyristor is turned on, the lower the rectified voltage will be. U n on load. Load voltage ripple U n smoothed by filter capacitor C f . Here and below, some simplifications are made when considering the operation of the circuits: the output resistance of the power transformer is considered equal to zero, the voltage drop across the rectifier diodes is not taken into account, and the thyristor turn-on time is not taken into account. It turns out that recharging the filter capacity C f happens as if instantly. In reality, after applying a trigger pulse to the control electrode of the thyristor, charging the filter capacitor takes some time, which, however, is usually much less than the pulsation period T p.

Now imagine that the delay in turning on the thyristor t z equal to half the pulsation period (see Fig. 3). Then the thyristor will turn on when the voltage at the rectifier output passes through the maximum.


Rice. 3.

In this case, the load voltage U n will also be the largest, approximately the same as if there were no thyristor regulator in the circuit (we neglect the voltage drop across the open thyristor).

This is where we run into a problem. Let's assume that we want to regulate the load voltage from almost zero to the highest value that can be obtained from the existing power transformer. To do this, taking into account the assumptions made earlier, it will be necessary to apply trigger pulses to the thyristor EXACTLY at the moment when U rect passes through a maximum, i.e. t z = T p /2. Taking into account the fact that the thyristor does not open instantly, but recharging the filter capacitor C f also requires some time, the triggering pulse must be submitted somewhat EARLIER than half the pulsation period, i.e. t z< T п /2. The problem is that, firstly, it is difficult to say how much earlier, since it depends on factors that are difficult to accurately take into account when calculating, for example, the turn-on time of a given thyristor instance or the total (taking into account inductances) output resistance of the power transformer. Secondly, even if the circuit is calculated and adjusted absolutely accurately, the turn-on delay time t z , network frequency, and therefore frequency and period T p ripples, thyristor turn-on time and other parameters may change over time. Therefore, in order to obtain the highest voltage at the load U n there is a desire to turn on the thyristor much earlier than half the pulsation period.

Let's assume that we did just that, i.e. we set the delay time t z much less T p /2. Graphs characterizing the operation of the circuit in this case are shown in Fig. 4. Note that if the thyristor opens before half the half cycle, it will remain in the open state until the process of charging the filter capacitor is completed C f (see the first pulse in Fig. 4).


Rice. 4.

It turns out that for a short delay time t z fluctuations in the output voltage of the regulator may occur. They occur if, at the moment the trigger pulse is applied to the thyristor, the voltage on the load U n there is more voltage at the output of the rectifier U rect . In this case, the thyristor is under reverse voltage and cannot open under the influence of a trigger pulse. One or more trigger pulses may be missed (see second pulse in Figure 4). The next turn on of the thyristor will occur when the filter capacitor is discharged and at the moment the control pulse is applied, the thyristor will be under direct voltage.

Probably the most dangerous case is when every second pulse is missed. In this case, a direct current will pass through the winding of the power transformer, under the influence of which the transformer may fail.

In order to avoid the appearance of an oscillatory process in the thyristor regulator circuit, it is probably possible to abandon pulse control of the thyristor, but in this case the control circuit becomes more complicated or becomes uneconomical. Therefore, the author developed a thyristor regulator circuit in which the thyristor is normally triggered by control pulses and no oscillatory process occurs. Such a diagram is shown in Fig. 5.


Rice. 5.

Here the thyristor is loaded onto the starting resistance R p , and the filter capacitor C R n connected via starting diode VD p . In such a circuit, the thyristor starts up regardless of the voltage on the filter capacitor C f .After applying a trigger pulse to the thyristor, its anode current first begins to pass through the trigger resistance R p and then when the voltage is on R p will exceed the load voltage U n , the starting diode opens VD p and the anode current of the thyristor recharges the filter capacitor C f . Resistance R p such a value is selected to ensure stable startup of the thyristor with a minimum delay time of the trigger pulse t z . It is clear that some power is uselessly lost at the starting resistance. Therefore, in the above circuit, it is preferable to use thyristors with a low holding current, then it will be possible to use a large starting resistance and reduce power losses.

Scheme in Fig. 5 has the disadvantage that the load current passes through an additional diode VD p , at which part of the rectified voltage is uselessly lost. This drawback can be eliminated by connecting a starting resistor R p to a separate rectifier. Circuit with a separate control rectifier, from which the starting circuit and starting resistance are powered R p shown in Fig. 6. In this circuit, the control rectifier diodes can be low-power since the load current flows only through the power rectifier.


Rice. 6.

Low voltage power supplies with thyristor regulator

Below is a description of several designs of low-voltage rectifiers with a thyristor regulator. When making them, I took as a basis the circuit of a thyristor regulator used in devices for charging car batteries (see Fig. 7). This scheme was successfully used by my late comrade A.G. Spiridonov.


Rice. 7.

The elements circled in the diagram (Fig. 7) were installed on a small printed circuit board. Several similar schemes are described in the literature; the differences between them are minimal, mainly in the types and ratings of parts. The main differences are:

1. Timing capacitors of different capacities are used, i.e. instead of 0.5m F put 1 m F , and, accordingly, a variable resistance of a different value. To reliably start the thyristor in my circuits, I used a 1 capacitorm F.

2. In parallel with the timing capacitor, you do not need to install a resistance (3 k Win Fig. 7). It is clear that in this case a variable resistance may not be required by 15 k W, but of a different magnitude. I have not yet found out the influence of the resistance parallel to the timing capacitor on the stability of the circuit.

3. Most of the circuits described in the literature use transistors of the KT315 and KT361 types. Sometimes they fail, so in my circuits I used more powerful transistors of the KT816 and KT817 types.

4. To base connection point pnp and npn collector transistors, a divider of resistances of a different value can be connected (10 k W and 12 k W in Fig. 7).

5. A diode can be installed in the thyristor control electrode circuit (see the diagrams below). This diode eliminates the influence of the thyristor on the control circuit.

The diagram (Fig. 7) is given as an example; several similar diagrams with descriptions can be found in the book “Chargers and Start-Chargers: Information Review for Car Enthusiasts / Comp. A. G. Khodasevich, T. I. Khodasevich -M.:NT Press, 2005.” The book consists of three parts, it contains almost all chargers in the history of mankind.

The simplest circuit of a rectifier with a thyristor voltage regulator is shown in Fig. 8.


Rice. 8.

This circuit uses a full-wave midpoint rectifier because it contains fewer diodes, so fewer heatsinks are needed and higher efficiency. The power transformer has two secondary windings for alternating voltage 15 V . The thyristor control circuit here consists of capacitor C1, resistances R 1- R 6, transistors VT 1 and VT 2, diode VD 3.

Let's consider the operation of the circuit. Capacitor C1 is charged through a variable resistance R 2 and constant R 1. When the voltage on the capacitor C 1 will exceed the voltage at the point of resistance connection R 4 and R 5, transistor opens VT 1. Transistor collector current VT 1 opens VT 2. In turn, the collector current VT 2 opens VT 1. Thus, the transistors open like an avalanche and the capacitor discharges C 1 V thyristor control electrode VS 1. This creates a triggering impulse. Changing by variable resistance R 2 trigger pulse delay time, the output voltage of the circuit can be adjusted. The greater this resistance, the slower the capacitor charges. C 1, the trigger pulse delay time is longer and the output voltage at the load is lower.

Constant resistance R 1, connected in series with variable R 2 limits the minimum pulse delay time. If it is greatly reduced, then at the minimum position of the variable resistance R 2, the output voltage will disappear abruptly. That's why R 1 is selected in such a way that the circuit operates stably at R 2 in the minimum resistance position (corresponds to the highest output voltage).

The circuit uses resistance R 5 power 1 W just because it came to hand. It will probably be enough to install R 5 power 0.5 W.

Resistance R 3 is installed to eliminate the influence of interference on the operation of the control circuit. Without it, the circuit works, but is sensitive, for example, to touching the terminals of the transistors.

Diode VD 3 eliminates the influence of the thyristor on the control circuit. I tested it through experience and was convinced that with a diode the circuit works more stable. In short, there is no need to skimp, it’s easier to install D226, of which there are inexhaustible reserves, and make a reliably working device.

Resistance R 6 in the thyristor control electrode circuit VS 1 increases the reliability of its operation. Sometimes this resistance is set to a larger value or not at all. The circuit usually works without it, but the thyristor can spontaneously open due to interference and leaks in the control electrode circuit. I have installed R 6 size 51 Was recommended in the reference data for thyristors KU202.

Resistance R 7 and diode VD 4 provide reliable starting of the thyristor with a short delay time of the trigger pulse (see Fig. 5 and explanations thereto).

Capacitor C 2 smoothes out voltage ripples at the output of the circuit.

A lamp from a car headlight was used as a load during the experiments with the regulator.

A circuit with a separate rectifier for powering the control circuits and starting the thyristor is shown in Fig. 9.


Rice. 9.

The advantage of this scheme is the smaller number of power diodes that require installation on radiators. Note that the diodes D242 of the power rectifier are connected by cathodes and can be installed on a common radiator. The anode of the thyristor connected to its body is connected to the “minus” of the load.

The wiring diagram of this version of the controlled rectifier is shown in Fig. 10.


Rice. 10.

To smooth out output voltage ripples, it can be used L.C. -filter. The diagram of a controlled rectifier with such a filter is shown in Fig. eleven.


Rice. eleven.

I applied exactly L.C. -filter for the following reasons:

1. It is more resistant to overloads. I was developing a circuit for a laboratory power supply, so overloading it is quite possible. I note that even if you make some kind of protection circuit, it will have some response time. During this time, the power source should not fail.

2. If you make a transistor filter, then some voltage will definitely drop across the transistor, so the efficiency will be low, and the transistor may require a heatsink.

The filter uses a serial choke D255V.

Let's consider possible modifications of the thyristor control circuit. The first of them is shown in Fig. 12.


Rice. 12.

Typically, the timing circuit of a thyristor regulator is made of a timing capacitor and a variable resistance connected in series. Sometimes it is convenient to construct a circuit so that one of the terminals of the variable resistance is connected to the “minus” of the rectifier. Then you can turn on a variable resistance in parallel with the capacitor, as done in Figure 12. When the motor is in the lower position in the circuit, the main part of the current passing through the resistance is 1.1 k Wenters timing capacitor 1mF and charges it quickly. In this case, the thyristor starts at the “tops” of the rectified voltage pulsations or a little earlier and the output voltage of the regulator is the highest. If the engine is in the upper position according to the circuit, then the timing capacitor is short-circuited and the voltage on it will never open the transistors. In this case, the output voltage will be zero. By changing the position of the variable resistance motor, you can change the strength of the current charging the timing capacitor and, thus, the delay time of the trigger pulses.

Sometimes it is necessary to control a thyristor regulator not using a variable resistance, but from some other circuit (remote control, control from a computer). It happens that the parts of the thyristor regulator are under high voltage and direct connection to them is dangerous. In these cases, an optocoupler can be used instead of a variable resistance.


Rice. 13.

An example of connecting an optocoupler to a thyristor regulator circuit is shown in Fig. 13. Type 4 transistor optocoupler is used here N 35. The base of its phototransistor (pin 6) is connected through a resistance to the emitter (pin 4). This resistance determines the transmission coefficient of the optocoupler, its speed and resistance to temperature changes. The author tested the regulator with a resistance of 100 indicated in the diagram k W, while the dependence of the output voltage on temperature turned out to be NEGATIVE, i.e., when the optocoupler was very heated (the polyvinyl chloride insulation of the wires melted), the output voltage decreased. This is probably due to a decrease in LED output when heated. The author thanks S. Balashov for advice on the use of transistor optocouplers.


Rice. 14.

When adjusting the thyristor control circuit, it is sometimes useful to adjust the operating threshold of the transistors. An example of such adjustment is shown in Fig. 14.

Let's also consider an example of a circuit with a thyristor regulator for a higher voltage (see Fig. 15). The circuit is powered from the secondary winding of the TSA-270-1 power transformer, providing an alternating voltage of 32 V . The part ratings indicated in the diagram are selected for this voltage.


Rice. 15.

Scheme in Fig. 15 allows you to smoothly adjust the output voltage from 5 V to 40 V , which is sufficient for most semiconductor devices, so this circuit can be used as a basis for the manufacture of a laboratory power supply.

The disadvantage of this circuit is the need to dissipate quite a lot of power at the starting resistance R 7. It is clear that the lower the thyristor holding current, the greater the value and the lower the power of the starting resistance R 7. Therefore, it is preferable to use thyristors with low holding current here.

In addition to conventional thyristors, an optothyristor can be used in the thyristor regulator circuit. In Fig. 16. shows a diagram with an optothyristor TO125-10.


Rice. 16.

Here the optothyristor is simply turned on instead of the usual one, but since its photothyristor and LED are isolated from each other; the circuits for its use in thyristor regulators may be different. Note that due to the low holding current of the TO125 thyristors, the starting resistance R 7 requires less power than in the circuit in Fig. 15. Since the author was afraid of damaging the optothyristor LED with large pulse currents, resistance R6 was included in the circuit. As it turned out, the circuit works without this resistance, and without it the circuit works better at low output voltages.

High voltage power supplies with thyristor regulator

When developing high-voltage power supplies with a thyristor regulator, the optothyristor control circuit developed by V.P. Burenkov (PRZ) for welding machines was taken as a basis. Printed circuit boards were developed and produced for this circuit. The author expresses gratitude to V.P. Burenkov for a sample of such a board. A diagram of one of the prototypes of an adjustable rectifier using a board designed by Burenkov is shown in Fig. 17.


Rice. 17.

The parts installed on the printed circuit board are circled in the diagram with a dotted line. As can be seen from Fig. 16, damping resistors are installed on the board R 1 and R 2, rectifier bridge VD 1 and zener diodes VD 2 and VD 3. These parts are designed for 220V power supply V . To test the thyristor regulator circuit without alterations in the printed circuit board, a TBS3-0.25U3 power transformer was used, the secondary winding of which is connected in such a way that the alternating voltage 200 is removed from it V , i.e. close to the normal supply voltage of the board. The control circuit works similarly to those described above, i.e. capacitor C1 is charged through a trimmer resistance R 5 and a variable resistance (installed outside the board) until the voltage across it exceeds the voltage at the base of the transistor VT 2, after which the transistors VT 1 and VT2 open and capacitor C1 is discharged through the opened transistors and the LED of the optocoupler thyristor.

The advantage of this circuit is the ability to adjust the voltage at which the transistors open (using R 4), as well as the minimum resistance in the timing circuit (using R 5). As practice shows, having the ability to make such adjustments is very useful, especially if the circuit is assembled amateurishly from random parts. Using trimmers R4 and R5, you can achieve voltage regulation within a wide range and stable operation of the regulator.

I started my R&D work on developing a thyristor regulator with this circuit. In it, the missing trigger pulses were discovered when the thyristor was operating with a capacitive load (see Fig. 4). The desire to increase the stability of the regulator led to the appearance of the circuit in Fig. 18. In it, the author tested the operation of a thyristor with a starting resistance (see Fig. 5.


Rice. 18.

In the diagram of Fig. 18. The same board is used as in the circuit in Fig. 17, only the diode bridge has been removed from it, because Here, one rectifier common to the load and control circuit is used. Note that in the diagram in Fig. 17 starting resistance was selected from several connected in parallel to determine the maximum possible value of this resistance at which the circuit begins to operate stably. A wire resistance 10 is connected between the cathode of the optothyristor and the filter capacitorW. It is needed to limit current surges through the optoristor. Until this resistance was established, after turning the variable resistance knob, the optothyristor passed one or more whole half-waves of rectified voltage into the load.

Based on the experiments carried out, a rectifier circuit with a thyristor regulator was developed, suitable for practical use. It is shown in Fig. 19.


Rice. 19.


Rice. 20.

PCB SCR 1 M 0 (Fig. 20) is designed for installation of modern small-sized electrolytic capacitors and wire resistors in ceramic housings of the type SQP . The author expresses gratitude to R. Peplov for his help with the manufacture and testing of this printed circuit board.

Since the author developed a rectifier with the highest output voltage of 500 V , it was necessary to have some reserve in the output voltage in case of a decrease in the network voltage. It turned out to be possible to increase the output voltage by reconnecting the windings of the power transformer, as shown in Fig. 21.

Rice. 21.

I also note that the diagram in Fig. 19 and board fig. 20 are designed taking into account the possibility of their further development. To do this on the board SCR 1 M 0 there are additional leads from the common wire GND 1 and GND 2, from the rectifier DC 1

Development and installation of a rectifier with a thyristor regulator SCR 1 M 0 were conducted jointly with student R. Pelov at PSU. C with his help photographs of the module were taken SCR 1 M 0 and oscillograms.


Rice. 22. View of the SCR 1 M module 0 from the parts side


Rice. 23. Module view SCR 1 M 0 solder side


Rice. 24. Module view SCR 1 M 0 side

Table 1. Oscillograms at low voltage

No.

Minimum voltage regulator position

According to the scheme

Notes

At the VD5 cathode

5 V/div

2 ms/div

On capacitor C1

2 V/div

2 ms/div

i.e. connections R2 and R3

2 V/div

2 ms/div

At the anode of the thyristor

100 V/div

2 ms/div

At the thyristor cathode

50 V/div

2 ms/de


Table 2. Oscillograms at average voltage

No.

Middle position of voltage regulator

According to the scheme

Notes

At the VD5 cathode

5 V/div

2 ms/div

On capacitor C1

2 V/div

2 ms/div

i.e. connections R2 and R3

2 V/div

2 ms/div

At the anode of the thyristor

100 V/div

2 ms/div

At the thyristor cathode

100 V/div

2 ms/div

Table 3. Oscillograms at maximum voltage

No.

Maximum voltage regulator position

According to the scheme

Notes

At the VD5 cathode

5 V/div

2 ms/div

On capacitor C1

1 V/div

2 ms/div

i.e. connections R2 and R3

2 V/div

2 ms/div

At the anode of the thyristor

100 V/div

2 ms/div

At the thyristor cathode

100 V/div

2 ms/div

To get rid of this drawback, the regulator circuit was changed. Two thyristors were installed - each for its own half-cycle. With these changes, the circuit was tested for several hours and no “emissions” were noticed.

Rice. 25. SCR 1 M 0 circuit with modifications

An article about various ways to connect a load to a microcontroller control unit using relays and thyristors.

All modern equipment, both industrial and domestic, is powered by electricity. At the same time, its entire electrical circuit can be divided into two large parts: control devices (controllers from the English word CONTROL - to control) and actuators.

About twenty years ago, control units were made on microcircuits with a low and medium degree of integration. These were the series of microcircuits K155, K561, K133, K176 and the like. They are called because they perform logical operations on signals, and the signals themselves are digital (discrete).

Exactly the same as ordinary contacts: “closed - open”. Only in this case these states are called “logical one” and “logical zero”, respectively. The logical one voltage at the output of microcircuits ranges from half the supply voltage to its full value, and the logical zero voltage of such microcircuits is usually 0...0.4V.

The operation algorithm of such control units was carried out through the appropriate connection of microcircuits, and their number was quite large.

Currently, all control units are developed based on . In this case, the operating algorithm is laid down not by the circuit connection of individual elements, but by a program “stitched” into the microcontroller.

In this regard, instead of several dozen or even hundreds of microcircuits, the control unit contains a microcontroller and a number of microcircuits for interacting with the “outside world”. But, despite this improvement, the signals of the microcontroller control unit are still the same digital as those of the old microcircuits.

It is clear that the power of such signals is not enough to turn on a powerful lamp, motor, or even just a relay. In this article we will look at, what ways can you connect powerful loads to microcircuits?.

The most. In Figure 1, the relay is turned on using transistor VT1; for this, a logical unit is applied to its base through resistor R1 from the microcircuit, the transistor opens and turns on the relay, which turns on the load with its contacts (not shown in the figure).

Cascade shown in Figure 2 works differently: to turn on the relay, a logical 0 must appear at the output of the microcircuit, which closes transistor VT3. in this case, transistor VT4 will open and turn on the relay. Using the SB3 button you can turn on the relay manually.

In both figures you can see that diodes are connected parallel to the relay windings, and in relation to the supply voltage in the opposite (non-conducting) direction. Their purpose is to extinguish the self-induction EMF (can be ten or more times the supply voltage) when the relay is turned off and to protect the circuit elements.

If the circuit contains not one or two relays, but much more, then for connecting them a specialized chip ULN2003A, allowing connection of up to seven relays. This connection diagram is shown in Figure 3, and in Figure 4 the appearance of a modern small-sized relay.

Figure 5 shows (instead of which, without changing anything in the circuit, you can connect a relay). In this diagram, you should pay attention to the transistor switch, made on two transistors VT3, VT4. This complication is caused by the fact that some microcontrollers, for example AT89C51, AT89C2051, hold the logical 1 level on all pins for several milliseconds during the reset time when turned on. If the load is connected according to the diagram shown in Figure 1, then the load will operate immediately when the power is turned on, which can be a very undesirable phenomenon.

In order to turn on the load (in this case, the LEDs of optocoupler thyristors V1, V2) a logical 0 should be applied to the base of transistor VT3 through resistor R12, which will lead to the opening of VT3 and VT4. The latter will light the LEDs of the optothyristors, which will open and turn on the mains load. Optocoupler thyristors provide galvanic isolation from the network of the control circuit itself, which increases the electrical safety and reliability of the circuit.

A few words about thyristors. Without going into technical details and current-voltage characteristics, we can say that this is a simple diode, they even have similar designations. But the thyristor also has a control electrode. If a pulse positive relative to the cathode is applied to it, even a short-term one, the thyristor will open.

The thyristor will remain in the open state as long as current flows through it in the forward direction. This current must be no less than a certain value called the holding current. Otherwise, the thyristor simply will not turn on. You can turn off the thyristor only by breaking the circuit or applying a voltage of reverse polarity. Therefore, in order to pass both half-waves of alternating voltage, a counter-parallel connection of two thyristors is used (see Fig. 5).

In order not to make such an inclusion, triacs are also produced in bourgeois language. They already have two thyristors in one housing, connected back-to-back - in parallel. They have a common control electrode.

Figure 6 shows the appearance and pinout of thyristors, and Figure 7 shows the same for triacs.

Figure 8 shows diagram for connecting the triac to the microcontroller (chip output) using a special low-power optotriac type MOC3041.

This driver inside contains an LED connected to pins 1 and 2 (the figure shows a top view of the microcircuit) and the optotriac itself, which, when illuminated by the LED, opens (pins 6 and 4) and, through resistor R1, connects the control electrode to the anode , due to which a powerful triac is opened.

Resistor R2 is designed to prevent the triac from opening in the absence of a control signal at the moment the power is turned on, and the chain C1, R3 is designed to suppress interference at the time of switching. True, the MOC3041 does not create any special interference, since it has a CROSS ZERO circuit (voltage transition through 0), and switching on occurs at the moment when the mains voltage has just passed through 0.

All considered circuits are galvanically isolated from the supply network, which ensures reliable operation even with significant switching power.

If the power is insignificant and galvanic isolation of the controller from the network is not required, then it is possible to connect thyristors directly to the microcontroller. A similar diagram is shown in Figure 9.

This is the diagram Christmas tree garland produced, of course, in China. The control electrodes of the MCR 100-6 thyristors are connected directly to the microcontroller (located on the board under a drop of black compound). The power of the control signals is so low that the current consumption for all four at once is less than 1 milliamp. In this case, the reverse voltage is up to 800V and the current is up to 0.8A. The overall dimensions are the same as those of KT209 transistors.

Of course, it is impossible to describe all the schemes at once in one short article, but it seems that we were able to describe the basic principles of their operation. There are no particular difficulties here, the schemes have all been tested in practice and, as a rule, do not cause any problems during repairs or self-production.

Boris Aladyshkin