Features of the circuitry of power supplies for tube amplifiers. Features of the circuitry of power supplies for tube amplifiers Power supplies for anode circuits and screen grids

When building any low-power design on lamps, the question of anode power is one of the first to arise.

The power supply is so - in principle - the most important part of any electronic device, but why in this article I mention the power supply of low-power and precisely lamp devices? And in general - what do I mean by these very devices?

Well, first of all, according to the theme of the blog, these are sound reinforcement devices. And it can be - first of all - pre-amplifiers for sound recording, which have recently been very popular on tubes. Well, devices based on them - tube phono stage, tube timbre blocks, tube guitar effects.

The specificity of supplying low-power lamps is a low current, but at the same time it is quite high voltage... And - for this type of device - constant voltage with very good filtering, i.e. maximally smoothed, with a minimum (no?) ripple.

In classic power amplifiers with linear power supplies, the problem of ripple is solved, as a rule, by using high-capacity capacitors (often connected in large quantities in parallel) and even chokes. But for a reason at the very beginning I emphasized that we are talking about a power supply unit specifically for a micropower (pre) amplifier. In this case, large capacitors will be

take up too much space if the design is compact

cost, possibly more than the entire structure as a whole

overload a low-power anode transformer at the time of charging

To ensure good signal filtering while saving space / money, a popular design called "electronic choke" helps.

This scheme has been known for a very long time and has a huge number of repetitions and modifications; hundreds of radio amateurs-designers have used it. Therefore, I will not describe the principle of the operation (we are against copy-paste!), Although I still recommend reading the most successful, in my opinion, article about this scheme from Oleg Ivanov.

We do not claim authorship of this circuit, and, in turn, took as a basis the circuit described in the article at the link above and slightly modified it, as, at one time, Oleg Ivanov modified one of the first stabilizer circuits.

This diagram is below.

At the beginning - as usual - there is a diode bridge, which can be made either from four separate diodes, or as a structure in one case. We recommend using diodes for a current of at least 2A. Despite the fact that the operating currents of the circuits that will be powered by this design are tens, or even units of milliamperes, the current is relatively high and abrupt at the moment the capacitor is charged. It can damage low-power diodes even with an intact and externally functional structure.

Then there are two or more high-voltage capacitors connected in parallel, the capacitance of which is relatively small (maybe 22μF, 33μF, 47μF). The decision in favor of just several capacitors connected in parallel, instead of one large one, was made in favor of lowering the cost of the structure and reducing its size.
Then, through a resistor of 0.47 - 1 kOhm, in order to ensure the second order in filtering, one or more connected capacitors are switched on in parallel, with a total capacity comparable to the total capacity of the capacitors in front of the resistor.

Next is a diagram using field-effect transistor, the principle of operation of which is described in detail in the article, one of the key parts of which is a set of metal-film or other non-electrolytic capacitors connected in parallel. However, some other authors consider it permissible to use oxide capacitors in this design, while observing the polarity.
After the stabilizer itself, we provided a voltage divider, from which, if necessary, you can apply a bias voltage to the lamp filament, as recommended by the designers of lamp technology, especially in the SRPP cascade, in order to reduce the background and the likelihood of breakdown through the filament.

Resistor R8 is needed if a milliammeter or an indicator of the appearance of a load will be introduced into the circuit. Its resistance is selected in such a way that the voltage drop across it at operating current corresponds to the required voltage for deflecting the indicator arrow or LED glow. So, R = U / I, where U is the required voltage, I is the operating current. For example, in order for an LED with an operating voltage of 2.2V to light up at a current of 10mA, a 22Ω resistance with a power of at least 0.25W is required.
If there is no need for indication, the resistor should be replaced with a shunt.

Now let's look at the design that we have developed and now serial production for use by fellow radio amateurs in their products.

On one 170x40mm printed circuit board, we, in addition to an electronic choke, placed a rectifier and a voltage regulator. Its operating current, however, is small and this part of the circuit can only be used in the case of operation on one lamp with a filament current of 150mA and an input voltage of no more than 12V. To work with lamps with a higher filament current, but not more than 1A, you will need a more massive radiator.
When the heater is supplied with alternating voltage or from a separate rectifier, this (lower) part of the circuit (the left side of the board) part of the circuit is not assembled.

As you can see in the layout image, there is space on the board for diodes of different sizes, as well as for a diode bridge. AC high voltage from the anode transformer is applied to 250V AC in points.

Two capacitors in parallel to the second part of the filter can be replaced with one of a larger capacity, there is a place for OR for two small ones OR for one large one. On the very right side of the board, there is a place for connecting several capacitors in parallel. It is made in the form of a breadboard specifically so that you can install a different number of capacitors of different sizes (suppose, 3 capacitors of 3.3μF 400V or 4 capacitors of 2.2μF 400V).
It is also possible to place a fuse-fuse or a reusable thermostatic fuse on the board. Rectified and filtered voltage output - HV DC out + -, divider output for filament bias - heat DC shift.

There are several modifications of this design. You can download the do-it-yourself layout files from the links below. You can also order from us high-quality (factory) ready-made boards for this project .

To do this, use the contact form on the left.

Continuation of the article on the materials of the electronic network Internet with reflections from " Notebook"Yuri Ignatenko and my comments and corrections

Accessories for the selected scheme. Resistors

Put any resistors, Soviet or Chinese, there is no difference. The main thing is that their power corresponds to the required one and slightly exceeds it.

Question... I would like to know about PTMN and MLT resistors? Can they be used in ULF?

Answer. Standard, mass-produced resistors of all types can be used in ULF, for this they were manufactured by the industry. Any good resistor is fine. It should be remembered that resistors of one particular type do not introduce distortion noticeable compared to resistors of another particular type. At par, as a rule, it does not matter whether they "float" or not "float". The question was asked about the use of resistors in the ULF. In ULF resistors with typical drift are applicable. It's not scary that the nominal value will float away from heating, let's say 100 kOhm, as it was at 20 degrees. and will become 100.1 kOhm at 80 degrees. So what? Highly accurate resistors with a low thermal coefficient are needed for instruments, oscilloscopes, space, etc. with wild ranges of temperature change and a thousandfold margin. And having put all the PTMN resistors in the ULF, no listener will distinguish the sound of the amplifier from the filling with MLT. In addition, the difference in the used nominal value by 5-10% from that specified in the circuit, as a rule, is easily digested by any tube amplifier. Moreover, when adjusting the instrument mode, the denomination may turn out to be even farther from the original in the picture. If we evaluate the noise characteristics of different types of resistors, then for tube circuits with a gain of the order of 100, the difference will be negligible even for evaluating by instruments.

Note: This is comparable to the removal of the brain to the seller for 1 kopeck when buying a Lexus in a car dealership. Any reasoning about the advantages of "non-inductive" resistors in the ULF should be regarded as doggy delirium (or paranoia). We can recommend the following attitude to this topic: A thief has come to your house, allegedly profitable product... And he rubs cotton wool into your ear for the sole purpose of robbing you. The goal is simple - to legally take your hard-earned money in exchange for sweet speeches. This is pink marketing nonsense that managers in crimson jackets need to beat their faces hard for. Evgeny Bortnik

Volume control

For a stereo amplifier, a dual volume control is needed, preferably with an inverse logarithmic characteristic. Pay attention to the absence of dust, dirt and rust. The resistor, before its use, should simply be stored normally and not creak. Chinese resistor RG 50kOhm. Take class A, this is their inverse logarithmic. Our class B is inverse logarithmic, and their B class is linear. An example of a resistor is shown in the picture.

The volume control should be no more than 50 kOhm. Now there are no piezo cartridges, as before, the sources are all low-impedance, so a variable resistor of 500 kOhm or 1MOhm is not needed at the input. An increase in resistance by 10-20 times reduces the input currents by the same amount. Therefore, for small input currents, background noise will be more noticeable. By doing high quality amplifier with good sound Do not put redundant RC circuits in the signal path. It is impossible to put a resistor with a high resistance in series in the signal path, because with the Miller capacitance and the input capacitance of the lamp and the installation itself, the same RC-chain is obtained, which sits all the "sound transparency". Serial-parallel chains of accelerating and decelerating harmonics of different frequencies appear on the signal path. Therefore, you cannot use volume controls over 50 kOhm.

Question. Is there a benefit to installing an Alps volume control?

Answer. Not much benefit because there is no difference. Is that in the ambition of the client, since the Alps volume control to put this $ 35 or Chinese is 4 hryvnia, and the used USSR is free. There is a large, very arrogant and aggressive bazaar. This is an economic war, like an ordinary big business in which big money is spinning. The man in the street is shit in the ear, using his uncertainty, due to his poor technical readiness and sensitivity to flattery. Verified reliably.

Tone controls

This is also an RC-chain, which sits all the "transparency of sound", so no shielded wires and no tone controls. Listen to the recordings as the director recorded them. In this he is more competent than you. Get rid of arrogance, show culture. Sound engineer (previously they were professionals high class) recorded the sound the way it should, and not the way you want it. You will listen to a tube amplifier tuned according to the instruments for a month without tone controls on the linear path and think to yourself: Was I not sick?

Electrolytic capacitors

One channel in the PSU requires three electrolytic capacitors of at least 100 mkF, 100 mkF and 50 mkF, a voltage margin of 400-450 volts determines the ultimate strength. For the reliability of the UMZCH, the age of the capacitors can be limited to 20 years, although the real state of affairs must be looked at in fact. It is better not to use dried electrolytes from TV 150 + 30x350 volts. Imported parts are optional. Although you can do on them. There is no difference in sound. To reduce the background, the first electrolytic capacitor for power supply must be at least 100 μF, the second at least 100-150 μF. The capacities in the filter of the power supply do not need to be spared. However, pay close attention to the nature of the oscillation of the transient. At high currents of consumption, the wires are chosen thicker. Consequently, their resistance is less and tricks are possible without load. In the presence of filter chokes, the transient process must be considered even more carefully.

Question... How critical is it if you reduce the capacitance in the power filter? What level of ripple is allowed on the output? And in the power supply circuit of the anode 6g2? Do they need to be stowed away in the basement, or can they be positioned over the chassis?

Answer. It doesn't matter where the electrolytic capacitors are. Most importantly, they must be isolated from the chassis. The capacitor case should only be connected to the ground bus. The larger the container, the better the filtration. And we can install any serviceable containers. For low-voltage circuits 150 + 150X250 volts from the TV. Here you have 300 microfarads or 150 + 30 X 350volts already 180 microfarads. Most of the Sovdepov electrolytic capacitors have a positive capacity of up to 30%. It is possible to use sequential electrolyte switching. One plus with one minus together. In this case, it is desirable to shunt each electrolyte with a 100-150kΩ resistor. And a film capacitor with a large voltage in parallel will not hurt each shunt. The voltage limit for the serial pair will double. It should be remembered about the increase in the rectified DC voltage by 1.4 times of the effective AC voltage during the no-load of the source. For 6p3s lamps, it is easy to jump out to XX voltages of 500-600 volts. Push-pull circuits are less sensitive to power quality than single-ended circuits. In a high-quality tube UMZCH, the ripple of the power supply of the output stage is less than 20-50mV. The pre-stage power supply is more demanding. It can be recommended to reduce the ripple by an order of magnitude.

Question... Can you tell me more about these green hats - tantalum electrolytes?

Answer. Tantalum is the best electrolytes in the USSR. Feel free to put in the cathodes of the lamps.

Question. The network is now 267 volts, during the day it was 240 volts, now electrolytes have 365 volts, they are designed for 350 volts - is it dangerous?

Answer. Serviceable sovdep capacitors have a fairly large voltage margin. After turning off the amplifier, you need to feel with your hand whether the electrolytes are warming up or not. If the hot is 50-80 degrees, then there is a chance that they will pshikat. If the temperature is normal, then they will work more. If 350 volts is written on our capacitors, then up to 450 volts will not explode. Soviet ones are not imported electrolytic capacitors, on which if 350 volts is written, then at a voltage of 360 volts, breakdown is inevitable. The sovdepov electrolytes have a supply voltage of 1.5-2 times. The increased voltage in the power supply of the amplifier will be only when turned on. In a minute, the lamps will warm up and there will be 310-320 volts.

Note. Keep in mind the following. 1. The fact of an increased likelihood of an explosion with a cold start is indisputable. 2. The fact of the presence of the effect of "poisoning" of the cathodes is indisputable. 3. The fact of increased wear of lamps when high voltages are switched on to the cold cathode also exists, regardless of the smart ones. Therefore, it is possible to recommend the use of starting automatics with anode supply delay. And if the source is started at XX, then the voltages will be high. Youthful bravado with increased stress is not needed. Use capacitors with a voltage rating equal to or greater than the voltage specified in the amplifier circuit. There are circuits with starting damping resistors. The circuitry is varied. Line voltage dash can be more dangerous for fixed-bias triode circuits. This is no longer typical for electrolytes, but for bulbs capable of self-heating, for example 6s33s. There are organizational and schematic ways of dealing with an accident. From auto displacement to progressive, adaptive and tracking displacement. Evgeny Bortnik

Answer. This recommendation was for kenotrons. For modern silicon diodes, it is quite acceptable to set 220 μF, however, the diodes must withstand large peak currents (tenfold) when switched on to discharged capacitors. The first two capacitors can be supplied at 100 μF each, and as the last one, use one of the first. It will turn out to be 100, 100 and 50 uF, respectively. And put the electrolyte to ground from a divider of 20-50 microfarads per 25 volts.

Note. DFor a steeper budget and a high-quality amplifier, the electrolyte capacity can be increased by an order of magnitude. However, the power supply must first be modeled or mocked up. In complex sources, the problem arises not only of limiting the charge current, but also the question of its balanced duration, the absence of oscillation, acceptable quality factor, the absence of local overvoltages and resonances, as well as the need for an accelerated discharge when turned off. A block-modular amplifier design can be recommended. The power supply is the main module. It is a monolithic plug-in unit, functionally complete and fully pre-tuned and rehearsed independently of the amplifier. Evgeny Bortnik.

Question. In general, increasing the capacity above a certain threshold gives anything? Some viewers put capacities in the filters in thousands of microfarads, or even tens of thousands.

Answer. There is a reasonable limit in everything. Shmelev's devices show how the anode power is filtered. You should set such a capacity so that -70 -80dB is a peak at a frequency of 100 Hz. Such suppression of pulsations is practically inaudible in acoustics. According to the picture, there is a 50 Hz line noise to the input and the input cable. The 150 Hz peak is the harmonic of the 50 Hz pickup. The 100Hz peak shows what the flattening of the plate voltage is. Acceptable anti-aliasing. The fact is that the use of more powerful electrolytes is not only an increase in the cost of an amplifier, but also a fight against the consequences of this very increase in capacity.

Question... How do Soviet electrolytes differ from modern imported ones?

Answer. Specially spending a day on measuring the parameters of Soviet electrolytic capacitors and foreign remakes "ala China" managed to obtain reliable information. Sovdep turned out to be better both in capacity and in reliability and in instantaneous energy output. In terms of size, the Sovdep now significantly outperforms foreign ones. It is curious that in foreign it is written or in the datasheet it costs 100mkF -20 + 20%, and the capacity there is always less, i.e. 80-85 uF. The bourgeoisie work to the minimum of tolerances. In the Council of Deputies, 100 μF -20 + 80% is always a valid capacitance of 130-140 μF. In the USSR electrolytic capacitors, high-quality plates made of thick aluminum tape are used, which can give off a lot of energy instantly. They have a layer of thin foil sprayed on, which does not allow the removal of such energy as from our K50 series. Of course they also have good electrolytic capacitors. But in our sale their price tag will go off scale. The cost of the $ 50 capacitor is too high. Variations are possible depending on capacity and voltage. Merchants bring in cheaper capacitors for $ 0.3-2 and sell them for $ 0.6-4, welding 100% of the margin. This is disgusting. The photo shows that nothing happened to the plates of the capacitor from the times of the USSR for 40 years of proper storage.

The electrolyte of the plate has not corroded. The capacitor - as soon as it came off the assembly line. This was done reliably in the USSR. And I will not say anything about the details with the VP stamp or OS.

Question... Well, what everyone calls an electrolyte and supposes that it dries out…. is it not dry? Is it wet to the touch?

Answer. And where can the electrolyte go from the sealed capacitor? I have electrolytic capacitors and 1953. And all workers and capacity are not lost. Disassembled capacitors of the USSR to show their advantage over imported waste. As you can see, there is no inductance in the co-electrolytic capacitor, because the entire plate, on one side, goes out with each of its turns and all the turns are connected together. Therefore, there is no inductive component (the effect of winding turns) and the capacitor operates in a very wide frequency range, without requiring shunting with film and other capacitors.

This fact also shows that it is permissible to remove instantaneous power from the sovdep capacitor, much greater than from imported ones. The design feature of cheap foreign capacitors is shown in the figure below. Two wire leads are visible. They go from one single point of the lining, therefore, access to the rest of the surface occurs through linear inductance. In addition to significant inductance, this design is characterized by a small instantaneous current return.

Question... How to check an electrolytic capacitor?

Answer... You can try methods of different degrees of stiffness. First check- A defective electrolytic capacitor, prone to gurgling and explosion, is always hot. You need to turn on the amplifier. It will work for 15 minutes. It is necessary to turn off and touch after one to three minutes (so that the electrolytes are discharged) all electrolytic capacitors for heating, the temperature of the faulty one will be increased to 60 - 70 degrees. In practice, verification can be unsafe. I checked this method - I connected the assembled power supply unit to the network and waited. At the fourteenth minute, one of the six capacitors exploded. Conclusion: the temperature should be checked every 5 minutes for 15 minutes. And if the temperature does not rise, then give the capacitors another hour to train to restore capacity. Another check- diode D226 is connected in series with an electrolytic capacitor. They are plugged into a 220 V network (without confusing the polarity, otherwise it will explode). Format the hour. Then they turn it off and after 1 - 2 minutes measure the voltage across it with a multimeter. If it is 0 volts, they still try to format it. If not less than 150 volts, then this is an excellent capacitor with low losses and good capacity. Then you can short-circuit. If a spark fires, it gives excellent energy. Another way- check the capacity by comparison. To do this, use a 500 Ohm resistor for 2 W + diode. The electrolyte is charged through this chain for 30 seconds from a 220 volt network. A 220 V 60-watt light bulb is connected to the electrolyte through the button. Press the button and evaluate with what brightness the light bulb flashed. Next, replace the electrolyte with the next one and again evaluate with what brightness the light bulb flashed.

Question... Do I need to bypass electrolytic capacitors with paper capacitors to better work in the HF range?

Answer... Serviceable electrolytic capacitors (especially Soviet ones) work perfectly up to 30 kHz without blockage. Therefore, they do not need to be shunted with film. If there is a Spectralab, the Shmelev complex, then you can check it yourself. If there are doubts about the serviceability and time is more expensive than money, then shunting with a good film will not hurt.

Interstage capacitors

There is no tangible difference to the viewer in domestic and imported serviceable capacitors. There are only two interstage capacitors in a simple circuit. We put any, it is better to first ring them with the device. K78-2, K-72, K78-19, etc. The voltage is allowed at least 300 volts. You can buy imported film. Set from 0.1 to 0.5 μF. Not essential. With a large input impedance of the subsequent stage, low frequencies go without blockage. Sovdep capacitors BMT and MBM are designed non-inductively, made quite high quality, it is only important to maintain tightness. If you look at the photo, for example, where a small capacitor with an electrolyte is shown, as in Fig. 31, then everything will become clear. The covers are also connected on one side with the output of all turns, and not like imported "audiophile" leads at one point with the cover are contacted and then rolled up like a roll. That is why serviceable domestic capacitors have an advantage. If in doubt, try opening the capacitor yourself.

Question... Are old capacitors of the BM series similar to imported ones or not?

Answer... All known serviceable sovdep capacitors are good, use it boldly. The inductance of the interstage capacitors has practically no effect on the sound quality, because the input impedance of the lamp of the next stage is 200 - 400 kOhm. Input capacity 30-200 pF. The inductance of the capacitor is simply scanty, the effect will be at hundreds of kHz and MHz. Look at the diagrams of tube oscilloscopes with a bandwidth of 5 - 40 MHz. The usual stages, the usual USSR-interstage capacitors and the normal bandwidth is obtained. All measuring technology The USSR was made on resistors MLT, VS on its own capacitors and lamps. And everything worked, the resistors did not make noise, the capacitors did not affect and the lamps were correctly amplified. The marketing hysteria on the websites was inflated by dealers according to the plans of the owners of foreign factories. Bourgeois need to sell their "audiophile" capacitors and resistors. The average viewer should only observe the selected voltage limits. Particularly demanding ones should remember that different capacitors give different tail and amplitude of harmonics. “Audiophiles” may continue to rush about, picking up capacitors to their taste, and not on the fidelity of reproduction.

To be continued.

Evgeny Bortnik, August 2015, Russia, Krasnoyarsk

Making a good power supply for a power amplifier (ULF) or other electronic device is a very demanding task. The quality and stability of the entire device depends on what the power source will be.

In this publication I will talk about making it not difficult transformer unit food for my homemade amplifier low frequency power "Phoenix P-400".

Such a simple power supply can be used to power various low frequency power amplifier circuits.

Foreword

For the future power supply unit (PSU) to the amplifier, I already had a toroidal core with a wound primary winding at ~ 220V, so the task of choosing a "pulsed PSU or based on a network transformer" was not a problem.

Switching power supplies have small dimensions and weight, high output power and high efficiency. A power supply based on a mains transformer is heavy, easy to manufacture and set up, and also does not have to deal with dangerous voltages when setting up a circuit, which is especially important for beginners like me.

Toroidal transformer

Toroidal transformers, in comparison with transformers on armored cores made of W-shaped plates, have several advantages:

less volume and weight;
higher efficiency;
better cooling for windings.

The primary winding already contained about 800 turns of 0.8mm PELSHO wire, it was embedded in paraffin and insulated with a layer of a thin tape made of fluoroplastic.

Having measured the approximate dimensions of the iron of the transformer, you can calculate its overall power, thus you can estimate whether the core is suitable for obtaining the required power or not.

Rice. 1. Dimensions of the iron core for the toroidal transformer.

Overall power (W) = Window area (cm 2) * Sectional area (cm 2)
Window area = 3.14 * (d / 2) 2
Sectional area = h * ((D-d) / 2)

For example, let's calculate a transformer with iron dimensions: D = 14cm, d = 5cm, h = 5cm.

Window area = 3.14 * (5cm / 2) * (5cm / 2) = 19.625 cm 2
Sectional area = 5cm * ((14cm-5cm) / 2) = 22.5cm 2
Overall power = 19.625 * 22.5 = 441 W.

The overall power of the transformer I used turned out to be clearly less than I expected - somewhere around 250 watts.

Selection of voltages for secondary windings

Knowing the required voltage at the output of the rectifier after the electrolytic capacitors, it is possible to approximately calculate the required voltage at the output of the secondary winding of the transformer.

The numerical value of the DC voltage after the diode bridge and smoothing capacitors will increase by about 1.3..1.4 times, compared with the alternating voltage supplied to the input of such a rectifier.

In my case, to power the UMZCH, you need a bipolar constant voltage - 35 Volts on each shoulder. Accordingly, an alternating voltage must be present on each secondary winding: 35 Volts / 1.4 = ~ 25 Volts.

Following the same principle, I performed an approximate calculation of the voltage values for the other secondary windings of the transformer.

Calculation of the number of turns and winding

To power the remaining electronic units of the amplifier, it was decided to wind several separate secondary windings. A wooden shuttle was made for winding the coils with enameled copper wire. It can also be made of fiberglass or plastic.

Rice. 2. Shuttle for winding the toroidal transformer.

Winding was carried out with enameled copper wire, which was available:

for 4 power windings UMZCH - wire with a diameter of 1.5 mm;
for other windings - 0.6 mm.

I selected the number of turns for the secondary windings experimentally, since I did not know the exact number of turns of the primary winding.

The essence of the method:

We carry out winding of 20 turns of any wire;
We connect the primary winding of the transformer to the ~ 220V network and measure the voltage on the wound 20 turns;
We divide the required voltage by the one obtained from 20 turns - we find out how many times 20 turns are needed for winding.

For example: we need 25V, and from 20 turns it turned out 5V, 25V / 5V = 5 - we need to wind 5 times 20 turns, that is, 100 turns.

The calculation of the length of the required wire was made as follows: I wound 20 turns of wire, made a mark on it with a marker, unwound and measured its length. Divided the required number of turns by 20, multiplied the resulting value by the length of 20 turns of wire - I got approximately the required length of the wire for winding. By adding 1-2 meters of stock to the total length, you can wind the wire on the shuttle and safely cut it off.

For example: you need 100 turns of wire, the length of 20 coiled turns is 1.3 meters, we find out how many times 1.3 meters need to be wound to get 100 turns - 100/20 = 5, we find out the total length of the wire (5 pieces of 1, 3m) - 1.3 * 5 = 6.5m. Add 1.5m for the stock and get the length - 8m.

For each subsequent winding, the measurement should be repeated, since with each new winding, the wire length required for one turn will increase.

To wind each pair of 25 Volt windings on the shuttle, two wires were laid in parallel at once (for 2 windings). After winding, the end of the first winding is connected to the beginning of the second - two secondary windings are obtained for a bipolar rectifier with a connection in the middle.

After winding each of the pairs of secondary windings to power the UMZCH circuits, they were insulated with a thin fluoroplastic tape.

Thus, 6 secondary windings were wound: four for powering the UMZCH and two more for the power supplies of the rest of the electronics.

Rectifier and voltage stabilizer circuit

Below is a schematic diagram of the power supply for my homemade power amplifier.

Rice. 2. Schematic diagram of a power supply for a homemade LF power amplifier.

To power the LF power amplifier circuits, two bipolar rectifiers are used - A1.1 and A1.2. Rest electronic components amplifiers will be powered by voltage stabilizers A2.1 and A2.2.

Resistors R1 and R2 are needed to discharge electrolytic capacitors when the power lines are disconnected from the power amplifier circuits.

In my UMZCH there are 4 amplification channels, they can be turned on and off in pairs using switches that switch the power lines of the UMZCH handkerchief using electromagnetic relays.

Resistors R1 and R2 can be excluded from the circuit if the power supply is permanently connected to the UMZCH boards, in which case the electrolytic capacities will be discharged through the UMZCH circuit.

Diodes KD213 are designed for a maximum forward current of 10A, in my case this is enough. The diode bridge D5 is designed for a current of at least 2-3A, assembled it from 4 diodes. C5 and C6 are capacitors, each of which consists of two 10,000 uF 63V capacitors.

Rice. 3. Schematic diagrams constant voltage stabilizers on microcircuits L7805, L7812, LM317.

Explanation of names on the diagram:

STAB - voltage stabilizer without regulation, current no more than 1A;
STAB + REG - adjustable voltage stabilizer, current no more than 1A;
STAB + POW - adjustable voltage stabilizer, current approx. 2-3A.

When using LM317, 7805 and 7812 microcircuits, the output voltage of the stabilizer can be calculated using a simplified formula:

Uout = Vxx * (1 + R2 / R1)

Vxx for microcircuits has the following meanings:

LM317 1.25;
7805 - 5;
7812 - 12.

Calculation example for LM317: R1 = 240R, R2 = 1200R, Uout = 1.25 * (1 + 1200/240) = 7.5V.

Design

Here's how it was planned to use the voltages from the power supply:

+ 36V, -36V - power amplifiers on TDA7250
12V - electronic volume controls, stereo processors, output power indicators, thermal control circuits, fans, backlight;
5V - temperature indicators, microcontroller, digital control panel.

ICs and voltage regulator transistors were attached to small heat sinks that I removed from non-working computer power supplies. The housings were attached to the radiators through insulating gaskets.

The printed circuit board was made of two parts, each of which contains a bipolar rectifier for the UMZCH circuit and the required set of voltage stabilizers.

Rice. 4. One half of the power supply board.

Rice. 5. The other half of the power supply board.

Rice. 6. Ready-made power supply components for a homemade power amplifier.

Later, during debugging, I came to the conclusion that it would be much more convenient to make voltage stabilizers on separate boards. Nevertheless, the option "all on one board" is also not bad and convenient in its own way.

Also, the rectifier for the UMZCH (diagram in Figure 2) can be assembled by surface mounting, and the stabilizer circuits (Figure 3) in the required quantity - on separate printed circuit boards.

The connection of the rectifier electronic components is shown in Figure 7.

Rice. 7. Wiring diagram for assembling a bipolar rectifier -36V + 36V using surface mounting.

The connections must be made using thick insulated copper conductors.

The diode bridge with 1000pF capacitors can be placed separately on the heatsink. The installation of powerful KD213 diodes (tablets) on one common radiator must be carried out through insulating thermo-gaskets (thermoresin or mica), since one of the diode's terminals is in contact with its metal lining!

For the filtering circuit (electrolytic capacitors of 10000 μF, resistors and ceramic capacitors 0.1-0.33 μF), you can use hastily assemble a small panel - a printed circuit board (Figure 8).

Rice. 8. An example of a panel with cutouts made of fiberglass for mounting rectifier smoothing filters.

To make such a panel, you will need a rectangular piece of fiberglass. Using a homemade cutter (Figure 9), made of a hacksaw blade for metal, we cut the copper foil along the entire length, then cut one of the resulting parts in half perpendicularly.

Rice. 9. Homemade hacksaw blade cutter made on a grinder.

After that, we outline and drill holes for parts and fasteners, clean the copper surface with thin sandpaper and tin it with flux and solder. We solder the details and connect to the circuit.

Conclusion

Here is such a simple power supply unit was made for the future homemade audio power amplifier. It remains to supplement it with a soft start and standby mode.

UPD: Yuri Glushnev sent a printed circuit board for assembling two stabilizers with voltages + 22V and + 12V. It contains two STAB + POW circuits (Fig. 3) on LM317, 7812 microcircuits and TIP42 transistors.

Rice. 10. Printed circuit board of voltage stabilizers for + 22V and + 12V.

Download - (63 KB).

Another printed circuit board designed for the circuit adjustable stabilizer voltage STAB + REG based on LM317:

Rice. 11. Printed circuit board for an adjustable voltage regulator based on the LM317 microcircuit.

In the article, you will learn how to make do-it-yourself tube amplifiers from scrap materials. It's no secret that tube sound- the most beautiful, its fans will exist at all times, despite the fact that the market is replete with a large number of offers of small-sized equipment on transistors and microcircuits. Consider in more detail what you should consider when making a tube amplifier.

Nutrition is the main difficulty

Yes, it is with the power supply that problems can arise, since you need two values of alternating voltage: 6.3 V for powering the filaments and from 150 V for the anodes of the lamps. The very first thing you need to find out for yourself is the power of the future structure. The power of the transformer for the power supply depends on this. Please note that the transformer must have three windings. Without such a power supply, you cannot make tube

In addition to the aforementioned secondary ones, there must also be a network (primary) one. It must contain so many turns for the transformer to work in normal mode. And even with a significant load (and voltage surges in the network up to 250 V), the winding should not overheat. Of course, the dimensions of the power supply will be rather big due to the large size of the transformer.

Rectifier

You will need to make a rectifier to get at least +150 Volts DC at the output. To do this, you need to use a bridge circuit for connecting diodes. D226 diodes can be used in the design of the power supply. If you need to make high reliability, then use D219 (they have a maximum operating current of 10 Amperes). If you are making tube amplifiers with your own hands, then follow the safety rules.

Work well in power supplies diode assemblies... You only need to choose those that are capable of functioning normally at voltages up to 300 volts. Pay particular attention to filtering the DC output voltage - install 3-4 electrolytic capacitors connected in parallel. The capacity of each must be at least 50 μF, the supply voltage is over 300 V.

Lamp circuit

So, now closer to the scheme itself. If you are making a tube Guitar Amplifier do it yourself, or to play music, you need to understand that the most important thing is safety and reliability. The most common circuits contain one or two preamplifier stages and one power amplifier stage. The preliminary ones are built on triodes. Since there are radio tubes that have two triodes in one base, you can save a little space during installation.

And now about what elements the tube amplifiers contain. You will have to assemble everything with your own hands into a single structure. For lamp in pre-amplifier it is best to use 6N2P, 6N23P, 6N1P. Moreover, despite the fact that all these lamps are analogous to each other, the 6N23P sounds much nicer. This lamp can be found in the PTC block (switch television channels) old black-and-white TV sets such as "Record", "Spring-308", etc.

Final amplifier stage

As an output lamp, 6P14P, 6P3S, G-807 are usually used. And the first will be the smallest, but the last two are very impressive in size. And the G-807 has an anode in the upper part of the cylinder. Please note that in tube ULFs it is imperative to use a transformer to connect acoustics. Without such a matching transformer, you cannot make a tube amplifier with your own hands.

Perfectly work as output transformers TVK used in vertical scanning. His primary winding turns on between the plus of the power supply and the anode of the output lamp. A capacitor is connected in parallel to the windings. Moreover, it is very important to choose the right one! First, it must be paper (such as MBM). Secondly, its capacitance must be at least 3300 pF. Do not use electrolytic or ceramic.

Adjustments and stereo sound

It will be very easy to make stereo sound. It is enough just to make two identical amplifiers. You can find a stereophonic tube amplifier in the old Soviet technique. You can repeat the design with your own hands. But you need to take into account some features:

connects directly to the amplifier input. which is used for it, you need to choose such that there are two elements on the axis in one case. In other words, when you rotate the knob, the resistance of two resistors changes at once.
Similar requirements for the frequency regulator. It is included in the anode circuit of the first triode of the preamplifier.

Amplifier housing

If you are making a tube guitar amplifier with your own hands, then it makes sense to use a metal case. He will not be afraid of blows and other minor shocks. But if you are making an amplifier for use at home, for example, to connect to a player, a computer, then it is wiser to use a wooden case. But it must be taken into account that it is advisable to fix the power transformer to the case with rubber gaskets. With their help, vibrations are reduced.

Much depends on what the tube amplifier will be like. With their own hands, many craftsmen make cases from sheet aluminum. If even small vibrations are applied to the lamp, its mesh will begin to vibrate. And these vibrations will begin to intensify, and the result is a buzz in the speakers. You also need to make a common bus, which should pass near all the lamps that make up the structure. All wires carrying a signal must be shielded as much as possible - this will get rid of various kinds of interference.

Circuits with transistors

And another interesting design is tube-transistor amplifiers. You can do these with your own hands literally in the evening. But lamp structures, as a rule, are made by hinged installation. It turns out to be the most convenient and simple. And in case transistors are used, you need to use printed wiring. In addition, a voltage of 9 or 12 volts is required to power the transistor stages. Moreover, transistors are used only to build a preliminary amplification stage. In other words, you only have one tube left - in the output stage (or two, if it comes about the stereo version).

UMZCH repair technique

Repair of UMZCH is almost the most frequent of the questions asked on amateur radio forums. And besides, it is one of the most difficult. Of course, there are "favorite" faults, but in principle, any of several tens, or even hundreds of components that make up the amplifier can fail. Moreover, there are a great many UMZCH schemes.

Of course, it is not possible to cover all cases encountered in the practice of repair, however, if you follow a certain algorithm, then in the overwhelming majority of cases it is possible to restore the device's performance in a completely acceptable time... This algorithm was developed by me from the experience of repairing about fifty different UMZCH, from the simplest, for a few watts or tens of watts, to concert "monsters" of 1 ... 2 kW per channel, most of which were received for repair without schematic diagrams.

The main task of repairing any UMZCH is to localize a failed element, which entailed the inoperability of both the entire circuit and the failure of other stages. Since there are only 2 types of defects in electrical engineering:

the presence of contact where it should not be;
lack of contact where it should be,

then the "super task" of the repair is to find the broken or broken element. And for this - to find the cascade where it is located. Further - "a matter of technology." As doctors say: "The correct diagnosis is half of the treatment."

The list of equipment and tools required (or at least highly desirable) during the repair:

Screwdrivers, side cutters, pliers, scalpel (knife), tweezers, magnifying glass - that is, the minimum required set of conventional assembly tools.
Tester (multimeter).
Oscilloscope.
A set of incandescent lamps for various voltages - from 220 V to 12 V (2 pcs.).
Low frequency sinusoidal voltage generator (highly desirable).
Bipolar regulated power supply 15 ... 25 (35) V with output current limiting (highly desirable).
Capacitance meter and equivalent series resistance ( ESR ) capacitors (highly desirable).
And, finally, the most important tool is the head on the shoulders (required!).

Consider this algorithm on the example of repairing a hypothetical transistor UMZCH with bipolar transistors in the output stages (Fig. 1), not too primitive, but also not very complicated. This scheme is the most common "classic of the genre". Functionally, it consists of the following blocks and nodes:

a) bipolar power supply (not shown);

b) differential input stage on transistors VT 2, VT 5 with current mirror on transistors VT 1 and VT 4 in their collector loads and a stabilizer of their emitter current on VT 3;

v) voltage amplifier on VT 6 and VT 8 in a cascode connection, with a load in the form of a current generator on VT 7;

G) transistor quiescent current thermal stabilization unit VT 9;

e) unit for protection of output transistors against overcurrent on transistors VT 10 and VT 11;

e) current amplifier based on complementary triplets of Darlington transistors in each arm ( VT 12 VT 14 VT 16 and VT 13 VT 15 VT 17).

Rice. 1.

The first point of any repair is an external examination of the subject and sniffing it (!). This alone sometimes allows us to at least assume the essence of the defect. If it smells like burnt, it means that something is clearly burning.

Checking the presence of mains voltage at the input: the mains fuse is stupidly blown, the power cord wires are loose in the plug, an open in the power cord, etc. The stage is the most commonplace in its essence, but at which the repair ends in about 10% of cases.

We are looking for a circuit for an amplifier. In the instructions, on the Internet, from acquaintances, friends, etc. Unfortunately, more and more often in recent years - unsuccessfully. Didn't find it - we sigh heavily, sprinkle ashes on our head and start drawing the circuit on the board. You can skip this stage. If the result is not important. But it’s better not to miss it. It's dreary, long, disgusting, but - "It is necessary, Fedya, it is necessary ..." ((C) "Operation" Y "...).

We open the subject and make an external examination of its "giblets". Use a magnifying glass if needed. You can see the destroyed cases of p / n devices, darkened, charred or destroyed resistors, swollen electrolytic capacitors or electrolyte drips from them, broken conductors, tracks printed circuit board etc. If one is found, this is not yet a reason for joy: the destroyed parts may be the result of the failure of some "flea", which is visually intact.

We check the power supply. We unsolder the wires going from the PSU to the circuit (or disconnect the connector, if any)... We take out the mains fuse and solder a 220 V lamp (60 ... 100 W) to the contacts of its holder. It will limit the current in the primary winding of the transformer, as well as the currents in the secondary windings.

We turn on the amplifier. The lamp should blink (while the filter capacitors are charging) and go out (a weak glow of the filament is allowed). This means that K.Z. on the primary winding of the network transformer there is no obvious short circuit. in its secondary windings. With a tester in the alternating voltage mode, we measure the voltage on the primary winding of the transformer and on the lamp. Their sum must be equal to the network. We measure the voltages on the secondary windings. They should be proportional to what is actually measured on the primary winding (relative to the nominal). You can turn off the lamp, put the fuse in place and turn on the amplifier directly to the network. We repeat the voltage check on the primary and secondary windings. The ratio (proportion) between them should be the same as when measured with a lamp.

The lamp burns constantly at full incandescence - which means we have a short circuit. in the primary circuit: check the integrity of the insulation of the wires coming from the mains connector, power switch, fuse holder. We unsolder one of the reasons going to the primary winding of the transformer. The lamp went out - most likely the primary winding (or turn-to-turn circuit) is out of order.

The lamp burns constantly in incomplete incandescence - most likely, a defect in the secondary windings or in the circuits connected to them. We unsolder one wire from the secondary windings to the rectifier (m). Do not confuse, Kulibin! So that later it would not be excruciatingly painful from improper soldering back (mark, for example, with pieces of adhesive masking tape). The lamp went out - it means that everything is in order with the transformer. On - we sigh again heavily and either look for a replacement for it, or rewind.

It was determined that the transformer is in order, and the defect is in the rectifiers or filter capacitors. We call the diodes (it is advisable to unsolder under one wire going to their terminals, or to solder it out, if it is an integral bridge) with a tester in ohmmeter mode at the minimum limit. Digital testers often lie in this mode, so it is advisable to use dial gauge... Personally, I have been using the “squeaker” dial tone for a long time (Fig. 2, 3). Diodes (bridge) are punctured or cut off - we change. Integers - call the filter capacitors. Before measuring, they must be discharged (!!!) through a 2-watt resistor with a resistance of about 100 ohms. Otherwise, you could burn the tester. If the capacitor is intact, when closed, the arrow first deviates to the maximum, and then rather slowly (as the capacitor is charged) "creeps" to the left. We change the connection of the probes. The arrow first goes off scale to the right (the charge from the previous measurement remains on the capacitor) and then creeps to the left again. If there is a capacitance meter and ESR then it is highly desirable to use it. We replace broken or broken capacitors.

Rice. 2. Fig. 3.

Are the rectifiers and capacitors intact, but is there a voltage regulator at the output of the power supply? No problem. Between the output of the rectifier (s) and the input (s) of the stabilizer (s), we turn on the lamp (s) (chain (s) of lamps) for a total voltage close to that indicated on the filter capacitor case. The lamp lit up - a defect in the stabilizer (if it is integral), or in the reference voltage generation circuit (if it is on discrete elements), or a capacitor at its output is broken. The broken regulating transistor is determined by ringing out its terminals (evaporate it!).

Is everything in order with the power supply (the voltages at its output are symmetrical and nominal)? Let's move on to the most important thing - the amplifier itself. We select a lamp (or chains of lamps) for a total voltage not lower than the nominal voltage from the power supply output and through it (them) we connect the amplifier board. Moreover, it is desirable for each of the channels separately. We include. Both lamps lit up - both arms of the output stages are punctured. Only one - one of the shoulders. Although not a fact.

The lamps are not lit or only one of them is lit. This means that the output stages are most likely intact. We connect a 10 ... 20 Ohm resistor to the output. We include. The lamps should blink (there are usually more power capacitors on the board). We apply a signal from the generator to the input (gain control - to the maximum). The lamps (both!) Went on. This means that the amplifier amplifies something (although it wheezes, phonite, etc.) and further repair consists in finding an element that takes it out of the mode. More on this below.

For further verification, I personally do not use the standard power supply of the amplifier, but use a 2-pole stabilized power supply with a current limitation of 0.5 A. If there is none, you can also use the power supply of the amplifier, connected, as indicated, through incandescent lamps. You just need to carefully insulate their caps so as not to accidentally cause a short circuit and be careful not to break the flasks. But an external power supply is better. At the same time, the consumed current is also visible. A well-designed UMZCH allows fluctuations in supply voltages within fairly large limits. After all, we do not need its super-duper parameters when repairing, just operability is enough.

So, everything is in order with the PSU. We pass to the amplifier board (Fig. 4). First of all, it is necessary to localize the cascade (s) with punctured (s) / dangling (s) component (s). For this extremely desirable have an oscilloscope. Without it, the effectiveness of repairs drops significantly. Although you can also do a lot with the tester. Almost all measurements are taken without load(idling). Let's assume that at the output we have a "skew" of the output voltage from several volts to the full supply voltage.

To begin with, we turn off the protection unit, for which we solder the right terminals of the diodes from the board VD 6 and VD 7 (in my practice it was three the case when the reason for the inoperability was the failure of this particular node). We look at the voltage not output. If it has returned to normal (there may be a residual skew of several millivolts - this is the norm), we call VD 6, VD 7 and VT 10, VT 11. There may be breaks and breakdowns of passive elements. We found a broken element - we change and restore the connection of the diodes. Is the output zero? Is there an output signal (when a signal from a generator is applied to the input)? The renovation is completed.

er = 0 width = 1058 height = 584 src = "amp_repair.files / image004.jpg">

Rice. 4.

Has anything changed with the output signal? Leave the diodes off and move on.

We solder the right terminal of the OOS resistor from the board ( R 12 together with the right terminal C 6), as well as the left conclusions R 23 and R 24, which we connect with a wire jumper (shown in Fig. 4 in red) and through an additional resistor (without numbering, about 10 kOhm) we connect to the common wire. We bridge the collectors with a wire jumper (red) VT 8 and VT 7, excluding the capacitor C8 and the quiescent current thermal stabilization unit. As a result, the amplifier is disconnected into two independent units (an input stage with a voltage amplifier and a stage of output repeaters), which must work independently.

Let's see what we have at the output. Is the voltage imbalance still present? This means that the transistor (s) of the "skewed" arm is (s) broken. We solder, call, replace. At the same time, we also check passive components (resistors). The most common defect, however, I should note that very often it is consequence failure of some element in the previous stages (including the protection unit!). Therefore, the following points are still desirable to fulfill.

Is there no skew? Hence, the output stage is presumably intact. Just in case, we send a signal from the generator with an amplitude of 3 ... 5 V to point "B" (connections of resistors R 23 and R 24). The output should be a sinusoid with a well-defined "step", the upper and lower half-waves of which are symmetrical. If they are not symmetrical, it means that one of the shoulder transistors is “burnt out” (lost parameters), where it is lower. We solder, we call. At the same time, we also check the passive components (resistors).

Is there no output signal at all? This means that the power transistors of both shoulders flew out "through and through". It's sad, but you have to solder everything and ring it with a subsequent replacement.

Component breaks are not excluded. Here you really need to include the "8th instrument". Checking, replacing ...

Have you achieved a symmetrical repetition at the output (with a step) of the input signal? The output stage has been repaired. And now you need to check the performance of the quiescent current thermal stabilization unit (transistor VT nine). Sometimes there is a violation of the contact of the engine of a variable resistor R 22 with resistive track. If it is included in the emitter circuit, as shown in the above diagram, nothing terrible can happen with the output stage, because at the base connection point VT 9 to divider R 20 - R 22 R 21, the voltage simply rises, it opens up more and, accordingly, the voltage drop between its collector and emitter decreases. A pronounced “step” will appear in the idle output signal.

However (very often), a trimmer resistor is placed between the collector and the VT9 base. Extremely "fool-proof" option! Then, when the engine loses contact with the resistive track, the voltage at the base of VT9 decreases, it closes and, accordingly, the voltage drop between its collector and emitter increases, which leads to a sharp increase in the quiescent current of the output transistors, their overheating and, naturally, thermal breakdown. An even more stupid version of this stage is if the VT9 base is connected only to the variable resistor engine. Then, if contact is lost, anything can be on it, with corresponding consequences for the output stages.

If possible, it is worth rearranging R 22 in the base-emitter circuit. True, in this case, the adjustment of the quiescent current will become expressed nonlinear from the angle of rotation of the engine, but IMHO this is not such a big price to pay for reliability. You can just replace the transistor VT 9 to another, with the reverse type of conductivity, if the layout of the tracks on the board allows. This will not affect the operation of the thermal stabilization unit in any way. he is bipolar and does not depend on the type of conductivity of the transistor.

Verification of this cascade is complicated by the fact that, as a rule, connections to the collectors VT 8 and VT 7 are made with printed conductors. We'll have to lift the legs of the resistors and make connections with wires (Fig. 4 shows the breaks in the conductors). Between the buses of the positive and negative supply voltages and, accordingly, the collector and emitter VT 9, resistors of about 10 kOhm are turned on (without numbering, shown in red) and the voltage drop across the transistor is measured VT 9 when the trimmer slider rotates R 22. Depending on the number of cascades of repeaters, it should vary in the range of about 3 ... 5 V (for "triplets, as in the diagram) or 2.5 ... 3.5 V (for" twos ").

So we got to the most interesting, but also the most difficult one - a differential cascade with a voltage amplifier. They work only together and it is fundamentally impossible to separate them into separate nodes.

We bridge the right terminal of the OOS resistor R 12 with VT 8 and VT manifolds 7 (point " A", Which is now his" exit "). We get a "stripped-down" (without output stages) low-power op-amp, quite efficient at idle (no load). We send a signal with an amplitude of 0.01 to 1 V to the input and see what will happen at the point A... If we observe an amplified signal with a symmetrical shape with respect to the ground, without distortion, then this stage is intact.

The signal is sharply reduced in amplitude (low gain) - first of all, check the capacitance of the capacitor (s) C3 (C4, because manufacturers, to save money, very often put only one polar capacitor for a voltage of 50 V or more, expecting that in reverse polarity it will still work, which is not gut). When it dries up or breaks down, the gain decreases sharply. If there is no capacity meter, we check it simply by replacing it with a known good one.

The signal is skewed - first of all, check the capacitance of the capacitors C5 and C9, which bypass the power supply buses of the preamplifier part after the resistors R17 and R19 (if there are these RC filters at all, since they are often not installed).

The diagram shows two common options for balancing the zero level: with a resistor R 6 or R 7 (there may, of course, be others), if the contact of the engine is broken, the output voltage may also be skewed. Check by rotating the engine (although, if the contact is broken "thoroughly", this may not give a result). Then try to bridge their extreme conclusions with the output of the engine with tweezers.

There is no signal at all - let's see if there is any at all at the input (break in R3 or C1, short circuit in R1, R2, C2, etc.). Only first you need to unsolder the VT2 base, because on it, the signal will be very small and look at the right terminal of the resistor R3. Of course, the input circuits can be very different from those shown in the figure - include the "8th instrument". It helps.

Naturally, it is not realistic to describe all possible causal variants of defects. Therefore, further I will simply outline how to check the nodes and components of this cascade.

Current stabilizers VT 3 and VT 7. Breakdowns or breaks are possible in them. The collectors are soldered from the board and the current between them and the ground is measured. Naturally, you first need to calculate the voltage at their bases and the values of the emitter resistors, what it should be. ( N. B .! In my practice, there was a case of self-excitation of the amplifier due to an excessively large value of the resistor R 10 supplied by the manufacturer. The adjustment of its rating on a fully working amplifier helped - without the above division into cascades).

Similarly, you can check the transistor. VT 8: if you bridge the collector-emitter of the transistor VT 6, it also stupidly turns into a current generator.

Differential transistors VT 2 V 5 T and current mirror VT 1 VT 4 as well as VT 6 are checked by dialing after a tap. It is better to measure the gain (if the tester has this function). It is advisable to select with the same gain.

A few words "not for the record." For some reason, in the overwhelming majority of cases, transistors of more and more power are installed in each subsequent stage. There is one exception to this dependence: on the transistors of the voltage amplification stage ( VT 8 and VT 7) dissipates 3 ... 4 times more power than on the pre-driver VT 12 and VT 23 (!!!). Therefore, if there is such an opportunity, they should be immediately replaced with medium-power transistors. A good option would be KT940 / KT9115 or similar imported ones.

Quite common defects in my practice were non-soldering ("cold" soldering to the tracks / "spots" or poor maintenance of the leads before soldering) of the component legs and breaks in the leads of transistors (especially in a plastic case) directly near the case, which were very difficult to see visually. Wiggle the transistors, carefully observing their terminals. As a last resort, solder and re-solder.

If you have checked all the active components, and the defect persists, you need (again, with a heavy sigh) to remove at least one leg from the board and check the ratings of the passive components with a tester. There are frequent cases of breaks in permanent resistors without any external manifestations. Non-electrolytic capacitors, as a rule, do not break through / break, but anything can happen ...

Again, from the experience of repair: if darkened / charred resistors are visible on the board, and symmetrically in both arms, it is worth recalculating the power allocated to it. In the Zhytomyr amplifier " Dominator »The manufacturer supplied 0.25 W resistors in one of the cascades, which burned regularly (there were 3 repairs before me). When I calculated their required power, I almost fell off my chair: it turned out that they should dissipate 3 (three!) Watts ...

Finally, everything worked ... We restore all the "broken" connections. The advice seems to be the most banal, but how many times forgotten !!! We restore in the reverse order and after each connection we check the amplifier for operability. Quite often, a cascading check seemed to show that everything was working properly, and after the connection was restored, the defect “crawled out” again. The last we solder the diodes of the current protection stage.

We set the quiescent current. Between the power supply unit and the amplifier board, we turn on (if they were turned off earlier) the "garland" of incandescent lamps for the corresponding total voltage. We connect the load equivalent (4 or 8 ohm resistor) to the UMZCH output. Trimmer motor R 22 we set it to the lower position according to the scheme and at the input we supply a signal from a generator with a frequency of 10 ... 20 kHz (!!!) of such an amplitude that a signal of no more than 0.5 ... 1 V is output at the output. step ", which is difficult to notice at a large signal and low frequency. By rotating the R22 engine, we achieve its elimination. In this case, the filaments of the lamps should glow slightly. You can check the current and an ammeter by connecting it in parallel with each string of lamps. You shouldn't be surprised if it will noticeably (but no more than 1.5 ... 2 times upwards) differ from what is indicated in the recommendations for tuning - after all, it is not “following the recommendations” that is important to us, but the sound quality! As a rule, in the "recommendations" the quiescent current is significantly overestimated, in order to guarantee the achievement of the planned parameters ("at the worst"). We bridge the "garlands" with a jumper, increase the output signal level to 0.7 of the maximum (when the amplitude limitation of the output signal begins) and let the amplifier warm up for 20 ... 30 minutes. This mode is the most difficult for the output stage transistors - maximum power is dissipated on them. If the “step” does not appear (at a low signal level), and the quiescent current has increased by no more than 2 times, the setting is considered complete, otherwise we remove the “step” again (as mentioned above).

We remove all temporary connections (do not forget !!!), assemble the amplifier finally, close the case and pour a glass, which we drink with a feeling of deep satisfaction with the work done. Otherwise it won't work!

Of course, within the framework of this article, the nuances of repairing amplifiers with "exotic" cascades, with an op-amp at the input, with output transistors connected with an OE, with "two-level" output stages, and much more are not described ...

Falconist