(student course project on the topic "Tube sound")
In recent years, little attention has been paid in the technical literature to tube circuitry and the design of equipment using radio tubes. However, the sound of tube amplifiers is still considered unsurpassed and “tube sound” is of great interest among music lovers, musicians and sound engineers. Leading companies producing professional audio equipment include tube microphones, tube preamplifiers, tube equalizers and even tube sound processors in their range of products. And this technique has the highest price category. Tube final amplifiers currently occupy a strong place among the highest quality sound reproduction devices, identifying such a concept among music lovers as Hi-End. The few remaining industrial production facilities of radio tubes in the world have already been acquired by the world's leading manufacturers, modernized and expanding the output of their products. Radio tubes in the field of sound reproduction showed that they began to be forgotten prematurely, and that there are areas of radio engineering where they did not lose their ground with the advent of semiconductors.
Proposed terminal tube amplifier does not pretend to provide listeners with Hi-End quality sound, however, it has a typical pleasant tube sound, contains exclusively purchased radio elements and can be assembled in just a week by an averagely qualified radio amateur.
This circuit turned out to be so successful in its simplicity and low cost of components, while simultaneously combining warmth and pleasant sound even when evaluated by professional sound engineers, that for the second year now I have been using the assembly of this amplifier as a course project for first-year students of the Faculty of Sound Engineering of the Humanitarian Institute of Television and Radio Broadcasting, in where I teach a survey course in radio engineering.
Application. The amplifier (monoblock) is designed for home listening to compact discs (CDs) on bookshelf speaker systems. Can be used by guitarists for home rehearsals, vocalists to listen to their recordings, or for home concerts. It is especially good to use this amplifier when playing jazz and blues compositions, as well as when scoring chamber concerts of bard songs, where it is necessary not only to convey the soulfulness of the sound of the voice and classical guitar, but also to add the warmth of sound inherent exclusively in tube terminal amplifiers.
Options. Maximum output power amplifier on a sine wave signal, measured at a load of 8 or 4 ohms reaches 15 watts. Frequency range at half power (0.707 voltage) from 40 Hz to 25 KHz. Sensitivity at 1 kHz, at maximum output power, is 1.55 volts effective on an unbalanced input.
Schematic diagram. The amplifier is made using three radio tubes according to a classic push-pull transformer circuit and contains two stages - a preliminary stage, based on a 6N23P double triode, and a final stage, based on two 6P43P beam tetrodes. Moreover, not only the output amplification stage is symmetrical, but also the preliminary one, made according to the circuit of a paraphase differential amplifier with cathode coupling.
The anode current of each 6N23P triode is 5.8 mA, which is set by an auto-bias resistor (330 ohms) in the common cathode circuit. The gain of such a circuit from the input to each of the outputs is 14. The preliminary stage is powered by increased voltage+ 360 volts from the bridge rectifier to provide high gain linearity and better circuit symmetry with an unbalanced input signal due to the large value of the cathode coupling resistor (5.1 Kom) and, accordingly, the large voltage drop across it (+63 volts). Also, based on the requirement of high linearity of amplification, the voltage distribution between the load resistances - 140 volts and triodes - 160 volts was selected.
Since, when the amplifier is turned on, while the lamps are warming up, the voltage after the bridge rectifier at idle reaches 500 volts, which exceeds the maximum operating voltage of electrolytic capacitors, the circuit uses a chain of quenching (2.7 Kom) and ballast (150 Kom) resistors, protecting the circuit from overvoltage.
If you want to supply a paraphase input signal to the amplifier, you need to apply an inverse signal to the grid of the second triode through the capacitor present in the circuit (0.47 μF) by disconnecting its lower terminal from the common bus. In this case, the sensitivity of the amplifier for each input will be 2 × 0.775 volts.
If you want to introduce feedback into the amplifier, it should also be connected to the grid of the second triode, and the signal feedback can be taken from winding 7 - 8 of the output transformer through a resistive or frequency-dependent voltage divider, depending on the desired functions of the OOS circuit. In the author's layout, to improve damping, a voltage divider consisting of two resistors 10 and 1 Kohm with a transmission coefficient of 0.091 was used. Of course, the sensitivity of the amplifier decreased.
The output stage of the amplifier operates in class AB 1 mode. Tube mode parameters:
Ea = 185 V, Eg2 = 185 V, Rk = 130 Ω (Eg1 = minus 16 V), Iо = 2 x 60 mA, Ig2 = 2 × 1.5 mA,
Raa = 3250 Ω, Uin max = 2 × 11.3 V eff. Pout = 14.4 W.
The output stage is powered by a full-wave rectifier formed by two bridge diodes with grounded anodes, and the +210 volt potential is drawn from the midpoint of the anode winding.
Design. The amplifier is made on a metal chassis using wall-mounted mounting, classic for tube circuits. The amplifier chassis can be made either from construction duralumin corners 30 × 30 × 2 mm and 30 × 60 × 2 mm, both 300 mm long, forming a U-shaped structure (as done in the author’s layout), or can be bent from sheet materials ( AMC or AMG alloys) 2 mm thick. It is also possible to use 1.5 mm thick structural steel sheet for the chassis. The use of sheet materials is recommended only if you have a bending device at your disposal. It is almost impossible to bend sheet aluminum or steel exactly over a length of 300 mm at home.
The two halves of the chassis, if it is made from duralumin corners, are fastened when installing transformers, a choke and lamp panels of output lamps on it with their mounting screws. In this case, no additional fastening is required.
On top, on the chassis, there are power and output transformers, a smoothing filter choke, radio tubes and the axis of a trimming resistor balancing the output stage with a lock nut. All three connectors (input, output, network) are located on one side wall (rear). On the other side wall (front) there is only a mains power switch. All other radio elements of the amplifier and mounting conductors are located in the chassis basement, protected by side walls.
Electrolytic capacitors are secured using clamps 12 - 15 mm wide made of thin (0.4 mm) white tinned tin, which can be taken from empty cans of stewed meat or from used cans of gasoline for refilling ZIPPO lighters. To prevent the clamps from scratching the capacitors when fastening them, and also to soften the pressure, before installation it is advisable to wrap the electrolytic capacitors with three to four layers of thin (0.15 mm) varnished cloth.
As an installation conductor for “ground” and incandescent circuits, a tinned copper single-core wire with a diameter of 0.8 mm is used, on which cambric - a cotton thread woven tube impregnated and varnished, with an internal diameter of 1 mm. The use of such historical insulation as cambric (it appeared in the 19th century before last, as wire insulation in electrical appliances), which currently smacks of archaism, but, nevertheless, has been preserved and is still produced (!), has its basis quite reasonable technical solution. Well, what other insulation will withstand and work properly for a long time in proximity to the sharp edges of the holes in the metal chassis, will provide mechanical strength and good insulation at high voltages (hundreds of volts) at temperatures up to 60-70 degrees, during long-term operation of the amplifier, where radio tubes are and transformers give off a lot of heat? After all, all modern insulating materials developed for transistor or microcircuit structures do not withstand “lamp” operating conditions, mounted installation and fairly frequent amplifier adjustments with each lamp change!
Long connections to transformers and the choke are made with flexible mounting wire MGShV-0.2. When the MGShV wire passes through the holes in the chassis, pieces of cambric with an internal diameter of 2 mm and a length of 20-25 mm are put on it in these places.
Double-leaf mounting posts are used for additional contact support points. In the absence of industrially manufactured mounting racks (cast from carbolite), it is possible to use homemade ones - machined from rod PCB according to the attached drawing. In this case, tinned copper wire with a diameter of 1 mm is used as mounting petals, the ends of which are bent into rings with an outer diameter of 3.5 mm for easy fastening of mounting conductors.
The design, made according to the basic scheme, uses 4 mounting posts. If feedback is introduced into the amplifier, its elements are mounted on the fifth mounting rack, as shown in the photograph of the chassis basement.
Details. The amplifier mainly uses widely used radio components, however, some of them need to be used only those indicated on the schematic diagram, since not all modern radio components are capable of high-quality operation in tube circuits.
Fixed resistors types MLT, S2-23. Variable resistor PP2-11, PP3-43 with locking of the adjusting axis with a locknut. It is also possible to use resistors SPO, SP2-2, SP3-30, SP4-2m and others, the overall dimensions of which allow them to be installed in this design.
A very important addition! The 10 Ohm resistors that are located in the anode circuits of the output lamps must be selected identical with an accuracy of 1 percent or better. To do this you need to buy about ten to fifteen of them and digital tester(with a multimeter) select a pair of identical ones (or negotiate with the seller in the store and make such a selection before purchasing). At the same time, the exact value of their denomination is not so important and can differ within 10 percent, the main thing is that they are the same with high accuracy. And somehow mark the selected pair of resistors so that they are installed in the amplifier during installation.
Electrolytic capacitors from JAMICON operating temperature not lower than 85 degrees, and better - 105 degrees. Capacitors from this company are durable and work well in tube circuits at high voltages and high temperatures at a very affordable price.
It is highly undesirable to use cheap electrolytic capacitors in an amplifier. Saving on radio components leads to poor sound quality of the amplifier and instability of its parameters over time. In addition, poor (cheap) electrolytic capacitors tend to leak and sometimes explode, contaminating the surrounding installation with electrolyte. So, saving is more expensive.
It is better to use input and interstage coupling capacitors of the types indicated in the diagram. Input - K78-2, interstage K73-P2. It is also possible to replace interstage capacitors with K78-2, with an operating voltage of at least 400 volts, since when the amplifier tubes are turned on and warmed up on these capacitors, the voltage briefly reaches 400 volts. The circuit proximity of interstage capacitors with powerful (1 watt) anode load resistors of the preliminary stage leads to the fact that during operation of the amplifier they heat up to a temperature of about 50 degrees. For sealed capacitors in a metal case with glass soldered insulators, K73-P2 is not a problem. This will not cause them to dry out or boil. How modern types of capacitors (as a rule, not sealed, but open-frame, filled with compound) will behave under such conditions can only be guessed at. In addition, the sound quality of the amplifier greatly depends on the coupling capacitors.
The bridge rectifier type KBPC606 can be replaced with the BR606 or the domestic KTs402A.
6P43P radio tubes can be replaced with 6P18P, while the 130 Ohm resistor in the common cathode circuit will need to be reduced to 75 Ohms. 6N23P can be replaced with E88SS, for example, from TESLA. In this case, no changes are required in the amplifier circuit, however, its sound will change in character and become more gentle and intimate. And the domestic 6N23P lamp in this circuit gives a very clear, transparent and musical sound.
Lamp sockets PLC-9-D-60 - for output lamps and PLC-9-D-35 for the preliminary stage lamp. They differ only in the length of the fixing springs that wrap around the lamp cylinder.
The mains connector is a three-pin “computer” instrument plug, which allows you to connect a standard power cord with a European mains plug and a grounding pin to the amplifier.
The input connector is a “tulip” with an external conductor isolated from the housing, connector type is an instrument socket with an RCA flange - RJ-RU CANARE. Output connector - XLR instrument plug - NC3MD-L-1 NEUTRIK. The amplifier connectors are clearly visible in the photographs.
Power transformer for powering the TAN17-127/220-50 or TAN17-220-50 monoblock. It is possible to replace it with TAN31, but in this case you will have to move the installation holes 2 mm in each direction along the chassis, since the center-to-center installation size for TAN17 is 46 mm, and for TAN31 it is 50 mm. In all other respects, everything is identical for the 17th and 31st TANs, and even the size of the chassis allows it to accommodate a more powerful transformer. If you want to power a stereo amplifier from one power transformer (two monoblocks combined into one design), then you need to take a more powerful standard, having the same voltages and designed for high currents: TAN45-127/220-50 or TAN45-220- 50.
TN39-127/220-50 or TN38-127/220-50 is used as an output transformer. Please note that power incandescent transformers of the TN series, which have one network (unsplit) winding of only 220 volts, are not suitable for use in this circuit as an output transformer.
The smoothing filter choke D40-5-0.18 can be replaced with D31-5-0.14, but since it has smaller dimensions and different installation dimensions, minor changes to the amplifier design will be required.
The amplifier uses a fuse with solder terminals of type VP1-2 for a current of 1 ampere. It is possible to use a different type of fuse close overall dimensions and even without leads, by soldering to it wire leads from tinned copper single-core wire with a diameter of 0.8 mm. Using a fuse with solder pins saves the plumbing work of installing a holder for a replaceable fuse, which is not required in this circuit, since the amplifier contains exclusively industrial-made radio components of sufficiently high reliability. However, the presence in the design mains fuse mandatory according to electrical safety conditions.
Power switch - toggle switch TV1-2 or TV1-4. When changing the diameter of the installation hole, you can use toggle switches TP1-2 or MT-3.
Chassis manufacturing. The chassis drawing is made in such a way that it is convenient to mark each of the corners separately (30 x 30 x 2 and 30 x 60 x 2), that is, the dimensions are based on two bases. To mark the chassis you will need a marking caliper. Actually, for the manufacture itself you will need the following tools: a hacksaw for metal (to cut off corners of the required length - 300 mm), a flat file with a personal notch (to process the edges of the cut and remove burrs), a drill and drills of the appropriate diameters (indicated in the drawing ), a semicircular file with a brute notch to cut large-diameter holes after drilling them along the contour with a small drill (for example, 2 mm), a benchtop bench vice with a jaw width of at least 60 mm (in order to clamp workpieces and chassis parts during processing ), a drill with a diameter of 7 - 8 mm, sharpened at an angle of 90 degrees, in order to remove burrs and chamfer all holes in the chassis after drilling them, a sharpener and a diamond file to sharpen the drills while working.
Chassis drawing.
Amplifier assembly. To properly assemble the amplifier, you must follow the assembly order. First of all, transformers are installed on the chassis. The power transformer must be installed on the chassis so that the terminals of its network winding 1 - 6 face the short edge of the chassis. When installing the output transformer, it is necessary to rotate it so that the terminals of its filament windings 7 - 16 are facing the edge of the chassis.
Transformers are secured with M4x12 screws and M4 nuts. Washers are placed under the screw head and under the nut. In addition, a grower is placed directly under the nut. The screws are inserted so that their heads are in the basement of the chassis, and the nuts are facing the transformers. An “earth” petal is installed under the lower left mounting screw of the power transformer (see the chassis basement installation diagram), under the washer. It is at this one point that the chassis connects to the amplifier circuit common.
The throttle is then installed on the chassis. Its conclusions must face the power transformer. Since there are threads in the mounting holes of its legs, it is fastened only with M4x12 screws without nuts. It is necessary to place washers under the screw heads.
Next, lamp sockets are installed on the chassis and secured with M3x6 screws with grommets and M3 nuts. Washers are not used when fastening the panels. After this, a variable amplifier balance resistor of type PP2-11 is installed on the chassis basement side and secured on the top side of the chassis using standard bushings and nuts. After tightening the fastening nut, screw on the lock nut, which fixes the position of the resistor axis after adjustment. Attention! The thread on the variable resistor PP2-11 is plastic. Therefore, there is no need to tighten the fastening nut too much to avoid breaking the resistor.
Network, input and output connectors and a power switch are installed on the side walls of the chassis. Before installation, the connectors are unfolded so that their terminals are in the position as shown in the wiring diagram. The switch is secured with a standard nut, and the connectors are secured with M3x10 screws, grommets and M3 nuts. Washers are not used when attaching these connectors.
Then electrolytic capacitors are installed in the chassis basement in accordance with the wiring diagram and with mandatory observance of the polarity of the terminals. To do this, you will need to unscrew some of the mounting screws of the transformers and the inductor. The fastening clamps are cut from sheet metal in place.
After this, mounting racks are installed in the chassis basement. Before installation, it is necessary to mold their terminals, and when fastening them, unfold them as shown in the wiring diagram. The stands are secured with M3x12 screws, washers, screws and M3 nuts. In this case, the washer is placed on the mounting ear of the mounting stand, then the groover is placed and the nut is tightened. There is no washer placed under the screw head on the top side of the chassis.
Lastly, a bridge rectifier is installed in the chassis basement. It must be positioned as shown in the wiring diagram. The orientation is given by the beveled corner and the marked “+” terminal. The bridge is fastened with an M3x12 screw, a washer, a screw and an M3 nut, similar to the fastening of the mounting posts.
Electrical installation. First, the necessary jumpers are installed on the transformer terminals. Adjacent petals are connected by tinned copper single-core wire with a diameter of 0.7 - 0.8 mm. Longer jumpers are made with MGShV 0.2 wire.
Then, the installation of incandescent circuits is carried out. A winding wire is taken, for example PEV-2-0.8, pieces of the required length are measured with a small margin, the insulation at the ends is stripped to a length of 5 - 7 mm and tinned, a cambric 1 mm thick is put on each conductor and the conductors in the cambric are twisted together . After this, filament circuits are laid with such a double wire to each lamp, as shown in the wiring diagram and can be seen in the photograph. The filament circuit conductors must be laid directly on the bottom of the chassis.
Next, all network connections are installed. After this, you can turn on the amplifier and make sure that the lamps glow. Really, it’s a magical sight when a structure that has not yet been completely assembled already begins to show the first signs of life! The lights of the lamp cathodes are beautiful!
The next step will be laying the “ground”, that is, the common wire. It is made with tinned copper single-core wire with a diameter of 0.7 - 0.8 mm. You can again take a winding wire with a diameter of 0.8 mm, strip it of insulation with a knife, carefully tin it along its entire length and use it for installation.
Then, we install and solder interstage separating capacitors K73-P2. They are large, take up a lot of space, and should be one of the first radio components installed. They should hang on their terminals at a distance of 4 - 5 mm above the bottom of the chassis basement. Following them, we install input coupling capacitors K78-2; they should lie on the side wall and at the bottom of the chassis basement.
After this, install the cathode jumper of the output lamps. It is performed with bare copper tinned single-core wire with a diameter of 0.7 - 0.8 mm. Then, in place, all the amplifier resistors are mounted in random order. What and where is more convenient to install. Don't forget about the common cathode resistor of the output tubes. It is located far from the main circuit of the amplifier, but turning on the amplifier without it will damage the cathode electrolytic capacitor. It will simply explode, because without a cathode resistor it will have about 200 volts, and its maximum operating voltage is only 50. And before installing, check its rating again. It should be 130 ohms.
Then the output transformer is connected with pieces of wire MGShV 0.2. The conductors must be measured to the required length locally with a small margin. The wires should not be tight, but they should not dangle or form unnecessary loops. In places where conductors pass through holes in the chassis, it is necessary to put on them pieces of cambric with an internal diameter of 2 mm and a length of 20 - 25 mm. The output windings of the output transformer and their connections to the output connector are made with a thicker wire MGShV 1.0 or 1.5.
Lastly, the anode rectifier and smoothing filter circuit is installed. Well, so that there is no temptation to turn on the rectifier without load and, thereby, disable the electrolytic capacitors, or even explode them.
After completing the installation of the entire amplifier, it is necessary to check it according to the circuit diagram and eliminate errors or install missing connections and parts. This happens quite often.
And the last but very important thing. Measure and compare the resistance values of the resistors in the anode circuits of the output lamps (nominal 10 Ohms) - whether the resistors you installed are the same. They must be identical and matched to within 1 percent or better!
Warnings and safety rules. Before turning on the amplifier for the first time, it is necessary to once again and very carefully check the correctness of the installation in principle and wiring diagrams. It is advisable not to do this yourself, but to ask your friend, colleague or a more experienced radio amateur. Other people's mistakes, from the outside, are more visible than your own. It is especially worth paying attention to the polarity of the bridge rectifier and electrolytic capacitors. Tube designs, unlike transistor ones, contain high voltages and sufficient power to cause significant damage or injury if installed incorrectly electric shock.
1. There must be another adult in the room where you are doing any work with the custom lamp structure. So that there is someone to give you first aid. Please note that voltages up to 1000 volts do not produce fatal destruction in the human body. But the heart can stop from electric shock, while being completely healthy. Therefore, in case of electric shock, if a person has lost consciousness, it is necessary first of all to tear him away from the current-carrying wires, lay him on his back on a flat surface, unbutton his clothes, and as quickly as possible make sure whether his heart is beating. If it stops, start doing chest compressions and artificial respiration. At the same time, for one blowing of air into the victim’s lungs through the mouth, you need to make 5-6 sharp pressures on the lower third of the chest, displacing it by 4-5 centimeters. After each pressure, you should quickly remove your hands to freely straighten your chest. In a minute it is necessary to carry out 48-50 compressions on the chest and 10-12 blows of air into the lungs. And without interruption, as quickly as possible, call ambulance, be sure to say that the person has cardiac arrest as a result of electric shock, and you are giving him artificial respiration and chest compressions. At the same time, you cannot stop for a minute. And even after the doctors arrived! For three to five seconds to quickly open the door for them, it’s possible. And immediately after this, continue both cardiac massage and artificial respiration with greater intensity. You can stop performing cardiac massage and artificial respiration only after the person affected by the current asks for it.
The second rule allows you to avoid the above-described unpleasant consequences even in case of electric shock. It was introduced in the century before last by physicist and electrical engineer Nikola Tesla. And this rule bears his name. So.
2. If you need to touch live circuit elements, you should do this with one hand, and put the other hand in your trouser pocket. As a last resort, if you are not wearing trousers and there are no pockets on your dress, you need to put your hand behind your back. In this case, even if you simultaneously touch two elements of the circuit with one hand under a large potential difference, you will be strongly “jerked”, but such an electric shock will not lead to sad consequences. It’s better to put your left hand in your pocket and work with your right hand in the pattern.
First start. The first switching on and adjustment of the amplifier is carried out without connecting the speaker systems and signal source to the amplifier.
Before plugging in the amplifier, make sure that the outlet is actually 220 volts. Otherwise, taking measurements with any accuracy makes no sense. Turn on the tester to measure alternating voltage with a measurement limit of at least 300 volts and with one hand insert the tester probes into the sockets of the electrical outlet. The tester can show voltage values ranging from 198 to 242 volts. That is, 220 volts, plus or minus 10 percent, is normal. However, all further considerations and measurement techniques are given for the case when there is exactly 220 volts in the outlet. And, of course, When starting to work with high voltages, ensure that rule number one is followed! High voltages, potentially life-threatening, in electrical engineering are considered to be voltages above 40 volts.
Set the variable balance resistor of the amplifier to the middle position, insert the radio tubes. Set the power switch to the “off” position, connect the power cord and plug it into an AC outlet of 220 volts, 50 Hz.
Connect a tester with a measurement limit of at least 500 volts of direct voltage with the negative probe to the chassis or to the common wire, and the positive probe to the “+360 volt” point of the circuit, that is, to the “hot” end of the 150 Kom resistor. Secure the tester probes with crocodile clips. Turn on the power switch. Observe how the measured voltage quickly increases to a value of + 450 ... 470 volts and then, within one minute, as the lamps warm up, decreases to an operating value of + 350 ... 360 volts. If everything happens this way, it means that the anode power supply of the preliminary stage is working correctly.
Then you need to turn off the amplifier with the power switch and wait until the cathodes of the lamps completely go out. Leaving the negative probe of the tester on the common wire, connect the positive probe using an alligator clip to pin 4 of the output transformer. At the same time, remember and follow rule number two!!! On electrolytic capacitors and after turning off the power, the charge can be maintained for a long time.
Turn on the amplifier. Observe how the voltage quickly increases to a value of 250...270 volts and then, as the lamps warm up, decreases to a working 185...190 volts. After the voltage is established around the operating value (plus or minus 5 percent), the amplifier can be turned off and considered that the first turn-on was successful.
If the nature of the voltage change is different or other values occur, then it is necessary to turn off the amplifier and check the correct installation, serviceability and ratings of the radio components.
Adjustment. To adjust the amplifier, you will need a screwdriver, a tester with alligator clips on both probes, a pencil and a sheet of paper.
Turn on the amplifier and let it warm up for 3 to 4 minutes. Set the tester's measurement limit to 2 volts DC. Remember rule number two!
Using a tester, measure the voltage drop across the 10 Ohm resistor in the anode of the top output lamp in the circuit. Moreover, connect the negative probe to the anode side of the lamp, and the positive probe to the terminal 6 side of the output transformer. The voltmeter reading should be within 0.5 - 0.6 volts. If the voltage is less or more, then by turning the axis of the “balance” trimmer resistor with a screwdriver, set it within the specified limits. Write down the voltage value you set to three decimal places.
Be careful when measuring the voltage drop across anode resistors! Careless touching of one tester probe to any point in the circuit, including the chassis, when the second tester probe is connected to the anode circuit, will lead to failure of the tester. After all, it is switched on to a measurement limit of only 2 volts!
After this, measure the voltage drop across the 10 Ohm resistor at the anode of the lower output lamp circuit. Moreover, connect the negative probe to the anode side of the lower lamp, and the positive probe to the output side of output transformer 1. Read the voltmeter reading to three significant figures and write it down next to the first reading. Add these two values and divide the amount in half. Using a screwdriver and a variable “balance” resistor, set the calculated voltage value you obtained.
Now measure the voltage again at the anode resistor of the upper lamp. Make sure that there is exactly the same voltage value there too. If the voltages on the anode resistors are different, then repeat the measurements, calculations and adjustments of the “balance” resistor until you set the same voltages on the anode resistors of both output lamps. Once you have succeeded, the adjustment is complete. Turn off the amplifier, wait until the tubes cool (and the capacitors discharge) and carefully tighten the trimmer retaining nut.
If the resistor adjustment range is not enough to set the same voltages, then it is necessary to replace one of the output lamps. This pair had too much variation in parameters. Not all lamps you buy will be suitable for high-quality work in push-pull amplifier. The spread of radio tube parameters for this circuit should not exceed 10 percent. Otherwise, when adjusting, you will not be able to balance the circuit, and the amplifier will distort loud sounds. Therefore, it is necessary to either buy selected pairs of radio tubes, or select them yourself, purchasing a obviously larger number of tubes and selecting from them those pairs, when turned on, the amplifier circuit will be balanced most accurately (at the center of the “balance” resistor). To some approximation, ideal pairs can be considered lamps that, when swapped, do not require balancing and provide the same anode currents.
Well, remember that when changing the lamps of the output stage, its adjustment should be repeated again.
Exploitation. The tube amplifier heats up during operation. Radio tubes are heating up, power transformer and throttle. Almost all resistors located in the basement of the chassis get hot. They get quite hot. But at the same time, all radio components for lamp structures are designed to operate at elevated temperatures. Therefore, a tube amplifier does not require fans during its operation; however, natural convection cooling is necessary. The main thing to ensure the correct temperature regime of a tube amplifier is to give it natural contact with the surrounding air. That is, either the amplifier must be operated openly as it is shown in the photograph, or a mesh or lattice protective housing must be made for it, which will freely pass air to all elements of its design. Moreover, air access is also necessary to the elements in the chassis basement. Therefore, you cannot place the amplifier on a solid surface. Either you need to place transverse slats under it, with a cross-section of 15 - 15 mm, made of any solid material, or, covering the basement of the chassis from below with a metal grid, install legs on it, which will provide the required distance from the supporting surface to the side edges of the chassis. A tube amplifier must not be placed in a closed housing in which there is no flow of external air. In this case, the radio components will overheat and quickly fail.
Unlike transistor amplifiers, which can be left on for days, weeks and months (this means professional equipment), simply forgetting about them, with tube amplifiers you need to be careful and careful. Turned it on, warmed it up for half an hour, listened; I finished listening and turned it off! However, you should not turn the amplifier on and off every hour either. Frequent switching on and off is harmful to lamps, perhaps more than long-term operation. Alas, radio tubes do not last forever, and require respect! And they have their own, albeit rather long, but limited resource. Moreover, during operation, the parameters of radio tubes gradually degrade. Therefore, do not forget about the included tube amplifiers and waste the resource of radio tubes. Well, sometimes, during operation, you need to check the balancing of the amplifier.
Now connect the signal source to the amplifier, sound system, turn it on, let the tubes warm up for 20 - 30 minutes, and enjoy the soft, gentle and velvety tube sound!
Literature.
1. S. Komarov, “UMZCH on “television” lamps with TN transformers.” “Radio” No. 12 for 2005 and No. 1 for 2006.
2. http://www..html
Set of parts for assembling this amplifier can be purchased at the Radio magazine store at the address: Moscow, Seliverstov lane, building 10, building 1. (1st floor, first door on the left in the hall): kits.radio.ru
If you want this set of parts to be sent to you by mail, then you need to contact the Chip-Set company: www.chipnabor.ru, phone: +7 916 080 2446, E-mail: [email protected]
Good afternoon.
It all started with a discussion of a hybrid headphone amplifier circuit. Her distinctive feature is to use a 6N23P lamp in low supply voltage mode (and with low anode currents). An input tube stage similar in modes is shown in the figure.
Note: We did not have exact mode values. Therefore, I determined the modes of the original circuit by calculation, taking into account the original circuit of the averaged I-V characteristics of the lamp and the cascade supply voltage (60V). The test cascade is set to close mode. It is quite possible that in original scheme the measurement results would differ slightly from those obtained for our cascade.
The measurements were carried out as usual: for several levels of the output signal and different frequencies. I will give graphs only for a frequency of 1 KHz. Because for frequencies of 100 Hz and 10 kHz the results differed slightly (within the measurement error).
First of all, I present graphs for output signals with an amplitude of 1 V and 2 V. These are probably the levels that were in the hybrid amplifier from which it all began.
The distortion level is high and increases rapidly with increasing output signal amplitude. Already at a level of 2 V, the second harmonic reaches 1%, the third - 0.03%... Is this good or bad? It seems to me that honest and clear sound such an amplifier should not be called. Rather, it is “heavily colored.”
Note: When using this stage in a real circuit, the distortion level will certainly be higher. The influence of the next cascade will be felt.
Let's increase the output signal level:
Distortion increased (more than 3%), harmonics up to the 7th appeared in the spectrum. But the lamp is not to blame for this. We just want too much from her. Conclusion: no need to torture the lamps :).
General conclusion: 6N23P operates acceptably with a supply voltage of 60V (at the anode 40V) with an output signal level of up to 1-2 V.
Now let's see what this lamp can do in a high-voltage cascade.
So: 6N23P lamp in “high-voltage” mode:
We will carry out measurements for a conventional resistive cascade and for SRPP. Let's start with resistive. You need to choose a mode. Anode voltage, anode current, supply voltage and load resistance can greatly influence the parameters of the cascade. So which mode should you choose?
I admit: I didn’t want to optimize or try all the modes that are claimed to be “the best” for this lamp. Therefore, I chose at my own discretion. I “adjusted” the setting by conducting test measurements of the spectrum. I did not notice any significant changes in the results. I settled on the mode that seemed best.
Note: I admit that I could have made a mistake with the choice of the cascade operating mode. Maybe there is a more “correct” one?
Measurements were carried out for four output signal levels. Here are the graphs:
The result is as you would expect. Linearity has increased: harmonic levels have decreased and spectra have become shorter. The cascade can be used in an “ear” amplifier with an output signal level of up to 5-8 V (for high-impedance headphones).
Probably, the sound will become cleaner and more reliable (compared to the “low-voltage” mode).
Let's look at 6N23P in SRPP:
Measurement results:
Linearity has increased a little more. If you use this lamp in a hybrid amplifier for low-impedance headphones (Uout< 2 В), то уровень искажений каскада не превысит 0.1 %. Так как спектр в этом режиме представлен только второй гармоникой, то можно предполагать, что искажения будут совершенно не заметны на слух и не дадут окраски.
Let everyone decide for themselves whether to use this tube in the pre-stage of a hybrid headphone amplifier or look for something else.
By changing the type of cascade and operating mode of the lamp, its sound can be changed in a wide range: from strongly “colored” to almost “honest”. And the choice of what and how to listen is a personal matter :). And any choice will be absolutely correct if you like the sound...
It's time to move on to the second participant in today's measurements.
I did not plan to use the 6N6P lamp in a voltage amplifier and did not intend to take measurements for it. I noticed it after a recent discussion on Facebook.
Some time ago, one active connoisseur of tube sound tried to convey to my colleague Nikita the idea that semiconductors versus tubes are the same as “a carpenter versus a joiner” (c).
Note: Nikita is one of the co-authors of this blog and an indispensable member of our small team. Among his many other activities, he also represents us on Facebook... That’s why he had to communicate with a lover of tube_sound.
The connoisseur’s manner of communication was not very correct and was not conducive to a friendly conversation. And he compensated for the lack of argumentation with emotions and peremptory statements. It felt like communication was becoming a completely pointless waste of time. Almost all conversations on the topic “which is better? Tubes or semiconductors?” are like this.
At that moment, when it became completely sad and boring, the interlocutor voiced the statement that a 6N6P lamp with a power supply of 300 V is capable of producing an output signal with an amplitude of 100 V and a distortion level of 0.01%. Numbers and specifics appeared. Communication could finally become constructive, useful and interesting.
In what cascade and in what modes should the lamp be run in order to obtain such results? For some reason, the opponent did not want to answer this question:(. He lost interest in us and stopped answering. We can only guess what scared him off. Maybe Nikita’s inability to develop a total love for lamps, or maybe his own lack of love for exact numbers.. If Nikita's interlocutor appears again, we may find out.
Due to the unexpected end of the conversation, it remained unclear: in what mode to measure... But I couldn’t get rid of the desire to check the lamp. I even went to the store and bought several pieces born in 1986.
In general, the discussion was about simple preliminary stages, so I thought that I could try the same resistive stage and SRPP. Again, I chose the modes at my discretion based on a series of short test measurements.
6N6P. Resistive stage
Measurement results
Much has been written on this topic. I was prompted to study this very common lamp by the experience of failures when setting up single-ended output stages on directly heated triodes - it is very difficult to get rid of the background alternating current without making complex decisions. I paid attention to the GU-50 - quite high power dissipation and indirect heat. Initially I installed a 6N1P for driving, but I didn’t like the sound of this lamp. After several tests, I settled on 6N23P from the Kaluga plant. Moreover, I had foreign analogues for comparison - ECC88 Tesla, Telefunken, Philips - I’ll be honest - their sound is no better. The diagram is shown. Everything is assembled in a custom-made metal case. Choke and output transformers – TW10SE from Audioinstrument. I would like to take this opportunity to thank Sergei Glazunov for his help and advice in choosing transformers.
Now the most important thing is the sound. It’s not for nothing that I titled the article that way. What I heard from the test monitors exceeded all my expectations. I put this amplifier even in my best home system instead of the famous Audio Note Otto Phono (the rest - Lite tube ADC, Magnat 709, Beringer, RCF ART312 speakers) - and the GU-50 sounded BETTER!!! The limitation is only low (about 5 watts) output power. An attentive reader, having calculated the power dissipation at the anode of the output lamp (approximately 20 Watts), will object - how can this be, this lamp can dissipate 2 times more! Yes, it can diffuse, but the sound quality suffers from this - for a number of reasons. In general, in my opinion, often those who design tube amplifiers are in vain interested in “squeezing” the maximum possible power from the output tube - the sound quality, as a rule, deteriorates significantly from this. In addition, the service life of the lamp and lamp socket is reduced, the amplifier loses stability and it is necessary to use very complex techniques to stabilize the operating point, which in the end still help little due to quick change parameters (aging) of the lamp.
Unfortunately, the initially installed interstage capacitor C12 K71-7 - (in my opinion, one of the best Soviet capacitors for interstage use) still does not reveal all the capabilities of the circuit - after installing the German Mudorf, the picture became complete. I would compare the sound of this amplifier with a 6P7S in a triode, but with a significantly higher output power. PT might be suitable, but I haven't tried it.
I highly recommend trying to assemble this circuit - you won’t regret it. Of course, without deteriorating the sound, you can replace the 5U4G with a 5Ts3S, and instead of the network Cantenburry Windings, you can use a voltage-appropriate TAN or TA + TN. If you don't like the sound of 6N23P, install 6N1P without any modifications. If there is an increased background of 50 Hz, the resistance of R12 can be increased to 6.2 Koma.
Thank you for your attention, this concludes the description of project no. 9. This project was carried out in December 2008, the amplifier was presented as a souvenir to my Chinese friend Peter. Unfortunately, there are no photos left.
Updated April 2015. Due to the fact that many questions come in, I want to answer the most frequently asked about this publication - the output transformer in this design has a Ra of about two, and it is very advisable not to increase this value too much, despite advice often found on the Internet that the optimum is 3.5 kilo-ohms. By increasing Ra you will lose in output power and, most importantly, the amplifier’s singing will become dry and analytical.
- Soft, detailed and clear sound
- Excellent transmission of vocals, stage and volume
- Simple design, no configuration required
- A complete set of protections implemented on the chip chip
- High concept - a vacuum double triode acts as a current buffer. The maximum linearity of the phase response and frequency response has been achieved, an inverting connection with T-OOS has been used.
- The basis is the popular LM3886 MC manufactured by National Semiconductors
- Average power – 68 W/4 Ohm. Peak – 135 W.
The LM series amplifier chips have the best sound among analogues. This also applies to flagship models of various levels, such as LM1875, LM3876 and its logical continuation - LM3886. The author's article continues the debate on the topic of circuit design and Thorsten's developments. An amplifier based on LM3875 is being considered. Its best sound, stability and linearity is achieved with inverting switching. However, this connection, when operating on the classic output impedance of the source, has a number of disadvantages. In short: with increasing frequency, the nonlinearity of the frequency response and phase increases. This is due to the fact that with inverting switching, the signal must come from a current source, and CD players and sound cards have an output impedance of about 200 Ohms. Current source on field effect transistors also disappears due to high losses, high input capacitance and pronounced nonlinearity. A current buffer on a triode successfully copes with this task.
In addition, this kind of buffer has a voltage gain of less than 1. Due to this, the OOS depth of the microcircuit itself is reduced, which also has an extremely beneficial effect on the sound quality. It is known that deep OOS, implemented by a classical divider, coarsens and deadens the sound. In the scheme proposed by Rasmussen ( Fig.1), a T-shaped OOS has been introduced, which increases the input resistance at the inverting input and makes it possible to reduce the grounding resistance at the direct input. The downside of this approach is the increase in noise and interference, but this is the first impression. If the wiring and shielding of the amplifier unit are done properly, interference will be almost invisible.
Now let’s look at what I personally didn’t like about the original scheme.
The author has LM3875 installed as a PA. Its disadvantages are imperfect protection, operation only with an 8-Ohm load, and low power. Instead, the LM3886 MC with a full set of protections and a powerful output stage was chosen, allowing it to deliver long-term power of 68 W and short-term power of 135 W into a 4-Ohm load. In addition, the amplifier is equipped with a full set of protections and a built-in mute mode.
At the exit Fig.1 There is a current limiter - a wirewound SQP resistor. The SPiKe system implemented in the LM3886 allows you to abandon it.
For the convenience of mixing channel parameters and reducing the size of the amplifier, the popular vacuum double triode 6N23P-EV was used as a buffer. It is distinguished low voltage power supply, relevant in this circuit, and at the same time, good sound. Although we have to admit that in this case its application is far from classical.
For our own reasons, the following features were added to the board:
Taking into account all the above considerations, the scheme took the following form ( Fig.2):
Here are the elements C 1 , C 3 , C 4 as well as terminals CN 1.. CN 6 – common for both channels. Each channel also contains half of a double triode 6N23P-EV .
Here, let’s take a break from the circuit design of the PA for a few seconds and consider the power supply, so as not to return to this topic again.
To power the entire circuit, a four-polar power supply with a common ground and an independent heating winding is used, the circuit of which is shown in Fig.3:
Diode bridges are either ready-made or assembled from diodes of the types that appeal to you, everything from D213 to Schottky diodes. For ±36 V 0.2 A – D 1 for a voltage of at least 200 V and a current of at least 4 A. For ±27V 4 A – D 2 for a voltage of at least 100 V and a current of at least 8 A. For incandescent - D 3 for any voltage and current of at least 4 A. This seemingly overestimation of parameters is not accidental. The fact is that, despite the peak reserve of the diodes, the current during charging of the containers exceeds the nominal one several times. But the price of diodes or ready-made bridges does not differ much, so for own peace of mind I don't recommend saving.
Capacities C 1, C 2 (for voltage not less than 50 V), C 5, C 6 (for voltage not less than 35 V), C 9 (for a voltage of at least 16 V) – imported electrolytic type K50-35. C 3, C 4, C 7, C 8, C 10 – type K73-17 at 63 V.
Any power transformer with an overall power of at least 200 W that satisfies the parameters of currents and voltages in the secondary windings indicated in the diagram (incandescent current of at least 0.8 A per lamp) can be used as a transformer.
In addition, it is possible to use two separate transformers. One is powerful for powering the PA, and the other is for powering the lamp. The second can be selected from a number of standardized lamp " T transformers A butno- N Akalnye". I use TAN1.
So, we managed to fit both channels onto one printed circuit board measuring 130x80 mm. Assembled module (without additional blocking containers) C8, C9 ) looked like this ( Fig.4).
Cute, isn't it?
The original layout of the elements is shown in Fig.5:
Now a few words about the details and the intricacies of assembly.
Resistors
Most resistors require pairing across channels with an accuracy of at least 1%. These conditions are fully satisfied by resistors of the C2-23 series. So, selection is required R 1 , R 3.. R 9 . Moreover R 1 , R 3 And R 4 It is better to use metal film type MLT, OMLT or imported analogues.
Resistors R 2 And R 10 no selection required. Can be of the MLT-0.25, S1-4 or S2-23 type at 0.125/0.25 W. R 11 And R 12 – imported at 2 W. The output inductance winds over R 11 , dressed in an insulating cambric, with a wire in enamel or epoxy insulation with a diameter of 0.6-0.8 mm until filled and soldered to the legs of the resistor. Although in this case I am a resistor R 11 didn't install. Instead, a coil was soldered, wound on the handle of a file and containing 15 turns of wire with a diameter of 0.8 mm.
VR 1 , VR2 – double variable resistor. In my case, Taiwan for 44 clicks, selected with an accuracy of 0.5% from 5 pieces.
Capacitors
C 1 , C 3 , C 8 , C 9 , C 10 – polar electrolytic type K50-35, preferably imported well-known brands. However, the circuit does not contain electrolytes in the audio circuit, which significantly improves the sound, reduces the criticality of the elemental base and increases the reliability of the system as a whole.
C1 – 16 V, C3 – 100 V, S8-S10 – 50 V.
C 4 , C 5 , C 7 , C 11 – metal film type K73-17. C 4 - at 250 V, the rest - at 63 V.
C2 – metal film or metal paper of the highest available quality, preferably no worse than polypropylene. The permissible voltage is also not lower than 63 V. Although this circuit sounds great with a K73-17 capacitor.
C6 – ceramics, preferably without piezo effect. KM or disk type. In extreme cases, of course, the K10-17B will do, but it’s hard to imagine a worse option.
Active components
The LM3886 amplification IC can be replaced with similar pinouts, taking into account the features of each. Purely theoretically, the circuit works with any MS built on the principle of a powerful op-amp. Attention! On the MC body there is a minus power supply!
Lamp R.O. 1 6N23P-EV is changed to 6N23P or an imported analogue ECC88. It is installed in a ceramic or any other socket designed for mounting on a printed circuit board or on a UMZCH chassis and is connected to the board with copper conductors.
In addition, taking into account modern trends in design, separate amplifier blocks have been developed for L.M. 3886 , which are installed on the radiator inside the UMZCH housing, and the lamp is installed in a special socket located on the housing cover. In this version, the entire llama harness ( R 1 , R 2 , 2x R 3 , C 3 , C 4 ) is carried out by mounted mounting directly on the socket terminals. And then it is connected to the power amplification units using a shielded signal cable. Don't forget to ground the lamp shield.
The printed circuit board of one PA channel is given on Figure 6:
Since it takes about 5 s to warm up the lamp, all these 5 s the amplifier input “hangs in the air”. At this time, all imaginable interference and a very noticeable rumble are present at the output. This can be avoided in two ways - by using a mute circuit or a relay to delay the turn-on. In both cases the control signal will be bipolar transistor with an RC divider in the base. If the delay is not enough, simply increase the value R 1 .
A diagram of such a delay is given in Figure 7:
In addition, at the time of modeling I had relays lying around TR 81 companies TTI . She was divorced for them printed circuit board. Its drawing can also be used as a guide for wiring for any relay you like with a normally open contact group. The board layout is given on Fig.8.
Details:
VR 1 – to the supply voltage of the relay winding. You can take it a little higher (about 2 V - drop across the transistor). In my case 12 V, i.e. stabilizer 7812..7815 .
C2 – on the voltage of the PA supply arm.
C1 – higher than stabilization voltage VR 1
This protection is connected to the positive side of the PA power supply (powerful transformer). The negative power terminal and the mute circuits of both amplifier channels (or all, if there are more channels) connected together are connected to the relay.
So finally SOUND
Fans of “tube sound” will really like this amplifier. What immediately catches your eye is the excellent vocals, the stage design and its incredible depth for transistor amplifiers. Unlike the typical sound of the LM3886, the HF is not washed out in this inclusion. They sound very subtle and precise. Silver and crystal do not smudge, as in a non-inverting inclusion. It is also impossible not to note the presence of a dense, collected and powerful, but extremely well-developed bass, which has always been so difficult to achieve from LM. Jazz and Blues sound so soulful that when listening, I often found myself getting goosebumps running down my spine.
The sound of this amplifier cannot be called absolutely accurate with a multi-frequency signal, but this sound is much more pleasant to the ear than various “super-linear” designs with distortion coefficients of thousandths of a percent.
To summarize: This amplifier is intended for music, not for measurement systems. Its objective properties are doubtful, but its sound and dynamic range so fascinating that when you hear the word “vector nonlinear distortion meter” you want to spit.
Moscow 2006 ( Lincor_ nobox@ inbox. ru)
The idea of building a high-quality tube amplifier for headphones has been in my head for a long time. The idea is not bad, but one thing stopped me. From the technical side, assembling this product was not difficult. Many schemes in this area have been reviewed. As it turned out, there were no more than a dozen similar schemes on the Internet; after reviewing and studying each one in detail, I came to a disappointing conclusion: at best, two out of ten schemes, it seemed to me, were more or less similar to the truth.
The rest were composed illiterately or, in principle, could not provide decent sound due to the tubes used. A lot of time was spent repeating the found circuits in order to check for quality and performance. Ultimately, I chose a circuit based on a 6N6P lamp, which had proven itself well according to reviews from radio amateurs who repeated this device. You can view the diagram at the link: Diagram
I already had one 6N6P lamp and, naively thinking of buying another one, I got down to business, but as it turned out, it was not possible to buy such a lamp - they simply weren’t there. Then, having reviewed the found circuits again, it was decided to use 6N3P; on the recommendation of one of the authors of the found circuits, he was written that very good results are obtained with this lamp. That's what I decided on.
But before assembling this system, I came across several more schemes with hybrid solution, lamp + transistor. I looked and repeated two of the ones I found. I'll tell it like it is: the result was not impressive. The author of one circuit suggested using 6n23p in the driver and two IRFs in the final stage. Moreover, power the entire circuit with a voltage of 35 Volts, referring to the fact that this lamp is capable of operating at ultra-low voltages. The lamp works, of course, but how... in other words, in the passport of this lamp there is a completely different value for the permissible minimum anode voltage. She is much taller.
There is probably no need to explain that there simply cannot be normal emission with such a power supply, and as a result the lamp is constantly in a half-locked state and no tricks will help to open it as expected, which was confirmed by own experience. I think it’s clear why transistors were used to match the voltage with a low-impedance load. A kind of getting rid of output transformers. Of course I thought about this moment.
The lamp has a high-impedance output, and the headphones available in stores have a maximum of 52 ohms. Accordingly, I abandoned this scheme. Having assembled another hybrid one using CT transistors with a push-pull output, I was also not delighted. Here the lamp was powered as it should, but the output stage worked in mode B. Well, it would have been all right if there had been germanium transistors. Anyone who has heard their sound will understand, of course. I could take part of that scheme and this one and combine it.
Normal lamp power plus mode A on IRF. But the scheme would be quite complicated. On top of that, I burned a couple of transistors instantly, and their price is 175 rubles apiece.
I still pursued the goal of collecting a high-quality teaching accessible for repetition. And if on lamps, then on lamps without any transistors. Having spent another week on these experiments, annoyed by the lack of any worthwhile results, I took the socket and the lamp and threw the rest off the balcony under the construction site so as not to be upset anymore. And he began to assemble on 6N3P.
I assembled it in a day. I listened and was very pleased with the result. Sounds absolutely amazing! But, as was said in all the articles, to put it in simple words This solution with a dynamic load does not pull the low end at all. Only at low volume. At maximum there is complete blockage and wheezing. It’s understandable why this happened. The difference in the resistance of the lamp and the load, what nonsense! But stupidity will remain stupidity if you don't touch it. So I decided, I’ll spend another week, but I’ll achieve decent sound.
The first thing that came to my mind was to use a transformer, as it should be in these circuits. Now there is another nonsense. How to make a compact amplifier so that transformers are not visible? After thinking about it, I tried to rewind the TVZ after recalculating the secondary. What can I say... it sounds great, there's plenty of low end, but it's cumbersome. This option immediately disappeared. I took the trances from the old Soviet receiver Alpinist 404.
Exhausted, I wound the primary with 0.08 wire, but missed the point that the secondary should have been laid as the first layer and the last one too. When I realized my mistake, it was already too late, and there were simply no nerves to unwind. Therefore, two secondary coils were wound with 0.25 wire and paralleled. The result was not bad, even good. But as it turned out, they disappeared high frequencies, because it was wound incorrectly. I no longer had enough patience and, having given up everything, I was in thought for a couple of days.
Amplifier circuit
The decision came unexpectedly. If it doesn’t work out with transformers, then you need to make sure that the lamp has at least half the output resistance. In the end, this is the diagram we got. It doesn't need any description. Both lamps operate in parallel.
Now about choosing lamps
Having experimented with what lamps were available, I used 6N1P and 6N23P. It was this combination that gave the best result. Before the final result there were ratios 6n1p+6n2p, 6n3p+6n2p, 6n1p+6n6p, 6n23p+6n2p... and a few more. Each option had its own obvious shortcomings. Insufficient gain, distortion at low volumes, whistling, metallic sound, etc. Subsequently, two versions of the 6N1P+6N23P amplifier were assembled, a four-tube and a two-tube. In the latter, the result is much worse, since the lamps work in the usual way and the low-frequency rolloff still remains, albeit significantly less than 6n3p or 6n6p... The four-lamp version makes me happy to this day. Good bottoms, well-drawn tops. I'm posting photos of both options.
Setting up schemes
A few words about setting up both schemes. Important condition: the voltages at the 6N23P cathode must be the same and the output voltage relative to minus must not exceed 125 Volts. Otherwise, a cracking sound appears, as if there is poor contact on the cathodes, 3.3-8 Volts are acceptable. It all depends on the lamps. The older, the higher on the cathode. These values were selected experimentally.
A little about the lamps used. It is advisable to install not used ones or at least two halves that worked equally well. If the lamp has a difference in operating time, then an alternating current background will be heard in the absence of a signal. I would like to warn you right away: do not immediately connect the headphones the first time you turn them on after assembly. Measure the output voltage: it should be no more than 0.3-0.5 Volts. If this value is higher, then the capacitor has a leak and requires replacement. Typically this is an electrolyte.
Non-polar capacitors play an important role: they enrich and emphasize high frequencies. Therefore, be as scrupulous as possible in choosing the latter. You should not install square ones in a plastic case. And MBMs are absolutely no good. Most a good choice- these are our domestic brown mica ones, I don’t remember the brand. But it is not possible to find one for 1 microfarad, and it is impractical to fence several pieces in parallel. The best option is K73-17. Electrolytes are better imported. Best of all, the Rubicon company says so on them. I don’t offer other more branded ones, since the price per piece is astronomical.
A few words about headphones
Don't even try Chinese plugs. Don’t spare three four thousand, go to the store and choose the most sensitive and high-resistance ones. And best of all, with a diffuser as large as possible. You won't get any difference or sound from low-quality headphones. Ideally, the most wonderful option is studio professional high-impedance headphones from 300 Ohms and above. The price for such a product is measured in tens of thousands, which is simply unthinkable. Therefore, he acquires the highest quality available. It also sounds very good.
Amplifier power
I don't touch on food. Any options are possible, but do not try to use electronic transformer as a booster. And for incandescence it couldn’t be better suited, after a little modification. Throw off the extra turns.
Amplifier design
Well, as the last thing to the finished device: design. I didn’t start making anything and making it from scrap materials, as many comrades use cases from CDs and old amplifiers, from all kinds of devices. I wanted the item to look vintage. After spending another week searching the shops of the city, I bought 8 brass candlesticks and two boxes, one of which was made of gold metal. The candlesticks fit exactly the size of the panel.
I glued it with supermoment. We disassemble the candlestick and drill out the existing holes as large as possible. We take a telescopic antenna, select the diameter of the elbow, cut it to the required length and assemble the candlestick. We solder it at the top and onto the nut at the bottom, in the version with a wooden box. The remaining parts of the candlesticks were assembled into three separate ones as a design to complement the structure.
Loading...
