Designing tube amplifiers. Encyclopedia of radio electronics and electrical engineering

Encyclopedia of radio electronics and electrical engineering / Tube Power Amplifiers

 Comments on the article

The article discusses the features of building single-ended tube amplifiers for high-quality sound reproduction. The author recommends the most suitable tubes and designs of output transformers for such amplifiers.

Probably, every radio amateur and every audiophile heard opinions about the superiority of tube amplifying equipment over transistor ones, but not everyone could be convinced of this for himself. There are several reasons for this: tube amplifiers are not often found in our time, and most importantly, in order to hear a noticeable superiority, you need to use high-quality phonograms that have not been spoiled by repeated processing and re-recording. If you're listening to someone like Eminem or Celine Dion, you're unlikely to notice the benefits of tube gear. Moreover, listening to some recordings, you can come to completely opposite conclusions. But if someone has ever felt the advantage of single-ended tube amplifiers, he will forever "get sick" with tubes.

It is said that tube amplifiers do not reproduce rock music well. However, more recently, in some discos, a power amplifier with four 6P45S lamps at the output of each channel, operating in class B, was successfully used. This amplifier had a maximum power of 200 ... 300 W and was let down only by its poor reliability.

Opponents of tube amplifiers rightly criticize them for "loose", "vague" bass, but the reason for this phenomenon has already been considered in the literature, for example, in [1]: the increased output impedance of a tube amplifier, which does not sufficiently dampen the low-frequency section of the speaker system to suppress the main resonance of the emitter. Therefore, the best, although not an easy, solution to the problem is to calculate and adjust the speakers, matching it with a specific amplifier, and even adjust the amplifier to this speaker. As a result, you can listen to the same Pink Floyd, enjoying the beauty of the guitar solo, and be surprised at the clarity of localization and the depth of sound of the bass register instruments. And how heartfelt the old recordings of the 40-60s, made using simple tube equipment, will sound!

The reasons for the advantages of tube amplifiers operating in class A have been repeatedly analyzed in the literature [2, 3]. It is possible to formulate the "first law of Hi-End'a": the audio signal should undergo as few transformations as possible, amplified by as few cascades as possible. And this is the best fit for lamps - an amplifier with a sensitivity of 0,1 ... maybe not for all DACs). This eliminates the analog filters collected on the op-amp.

In addition to the high power gain and high linearity of the lamps, two more fundamental advantages should be noted: the constancy of the interelectrode capacitances, as well as the independence of the characteristics from temperature and, consequently, from the level of the amplified signal.

Having once realized the advantage of linear amplification (in class A), the arguments of the supporters of push-pull cascades in the UMZCH become completely incomprehensible. The compensation of the second harmonic declared by them is not always an advantage, since it has been repeatedly proven that the second harmonic, if it does not exceed 2 ... 3% of the main signal, does not spoil the sound, rather the opposite. And the need for a phase inverter for a push-pull cascade generally causes a number of problems. You can read about all this in more detail in the above-mentioned articles and in [4].

This article is devoted to single-cycle lamp UMZCH, their circuits, used lamps and transformers.

There are two main varieties of single-cycle tube UMZCH: in one of them, the output stage is built on a triode without a common OOS, in the second - on a pentode or beam tetrode with a depth of up to 16 dB covering the last two stages of the OOS. As examples in Fig. 1 and 2 show amplifier circuits, which are discussed in more detail below. By the way, we note that in output triodes, such as classic 2AZ and 300V, internal feedback, which is customary to be silent about in modern literature, has approximately the same depth - 12 ... 16 dB. Sometimes you can read in articles that only triode cascades are capable of providing the highest class sounding amplifiers, but this is not entirely true. So, Audio Note produces several models of amplifiers with tetrodes and a common OOS, for example, OTO Line SE, Soro Line SE. The latter, by the way, has been used as a reference by audio experts from St. Petersburg for several years.

(click to enlarge)

(click to enlarge)

The output stage on tetrodes with a constant voltage on the second grid is somewhat more economical and has the advantage that several tetrodes can be connected in parallel to increase power, even with some difference in their characteristics.

Let us pay attention to one particular, but often discussed question about shunting cathode auto-bias resistors with blocking capacitors. It is usually argued that shunting should always be done, despite the fact that any oxide capacitor in the audio signal path is additional distortion. Let's look at the objective reasons for this or that decision.

It is highly desirable to shunt a resistor in the output stage on a triode in order not to increase the output resistance of the stage and maintain its maximum sensitivity.

In the output stage on a tetrode with a constant voltage on the second grid, shunting the cathode resistor is necessary, but the reason here is completely different. The FOS created by this resistor linearizes only the cathode current. The anode current is the cathode current minus the current of the second grid, which has a relatively non-linear dependence on the same cathode current. As a result of the introduction of such an OOS, we get a cascade with somewhat smaller, but more unpleasant to the ear distortions, while losing approximately twice in sensitivity.

In the pre-terminal (driver) stage, followed by the triode output stage, it is not necessary to shunt the resistor, but it is desirable. Here, the criterion becomes the condition for combining the output resistance of this stage with the input capacitance of the next input capacitance of the triode stage

Svx \u1d Csk + CCA (K + XNUMX),

where CC is the grid-cathode capacitance; SSA - mesh-anode capacity; K is the voltage transfer coefficient of the cascade.

For example, if the driver stage is assembled on a 6N2P triode with an unshunted cathode resistor and has an output resistance of 50 kOhm, then with an input capacitance of the output stage of 200 pF, the upper cutoff frequency

f=1/(2πRC) = 16kHz!

In the pre-terminal stage, followed by the output stage on the tetrode, the cathode resistor cannot be shunted, since it is often fed a feedback signal from the output of the amplifier.

In the input stage, if it needs to have a gain less than μ/2 or introduce frequency correction, for example, uneven speaker characteristics in the low-frequency region, the cathode resistor should not be shunted; this will increase the stability of the gain or correction parameters.

Now let's talk about the choice of tubes for the amplifier. The author carried out studies of various lamps on the spectrum of harmonics of the output signal in the mode of small and large signals to the limiting mode. Along with this, the effect of the distortion spectrum on the quality of sound reproduction was assessed by auditory examination (listening). Particular attention was paid to the correlation of subjective and metrological assessments. The results of such comparative studies basically confirmed the information known from modern literature. Let's pay attention to the most suitable specific tubes for various stages of amplifiers.

Among the lamps for the output stage on a tetrode, the classic beam tetrode 6P6S turned out to be the leader in "musicality". This coincides with the statements of the article [5]. The second place should be taken by 6PZS (close analogues - 6L6 6P7S, G-807), one and a half times more powerful beam tetrode with a very similar spectrum, but with a slightly higher level of high harmonics.

Output beam tetrodes - 6P14P, EL34 (6P27S - an analogue, but a museum rarity), 6550 (KT88) - come with some lag. The 6P1P finger lamp is an analogue of the 6P6S octal lamp, but it is better to use an octal lamp, and it is easier to find it. They say that the 6F6S pentode is linear and "musical", but it is rare, and its output power is too small (3,2 W).

There is an opinion that horizontal scanning television lamps are unsuitable for UMZCH (we are talking about 6P45S, 6P44S and the like). This is not so: they can be used, but not in a typical mode, but with a voltage halved on the second grid. For example, a 6P44S lamp in such an atypical mode is very similar in sound to a 6P14P in a typical mode, but one and a half times more powerful.

The leader in the group of lamps for the output stage on a triode and, in general, the absolute leader, quite unexpectedly, turned out to be the 6P44S beam tetrode in a triode connection. In terms of the delicacy of handling sound, this lamp surpassed even the 6C4C triode, which should be put in second place. The composition of harmonics of the anode current 6P44S, measured at the maximum signal immediately before the limitation, is given in the table.

Recommended lamp operating mode: UAK = 250 V, IA ≤ 90 mA, RH = 2450 Ohm, UCK = -34...-37 V, RK = 400 Ohm. The output power of the cascade with this lamp is 5 W (measured after the transformer with losses up to 8%); this is one and a half times more output power with a 6С4С triode. By the way, some articles give overestimated output power values ​​​​for a 6C4C lamp: 5, 10 and even 20 W This is not so: in class A mode at a nominal power dissipated by the anode, 15 W (250 V and 60 mA) output power with a 6C4C triode is 3,7 W excluding losses in the transformer. The same power value is indicated in [6, p. 132]. The amplitude of the control signal for 6P44S is 36 V versus 43 V for 6S4S

Next, we should name, of course, the famous 300V triode. In terms of "musicality" this lamp (produced by the Svetlana association) is slightly inferior to the 6C4C triode, but many audiophiles prefer it because it allows you to get an output power of at least 8 W from a single lamp.

Some more recommendations on the use of the 6P44S lamp. To get a triode amplification mode, it is necessary to connect the second grid of the lamp to the anode through a 100 Ohm resistor, otherwise self-excitation will appear at the RF. To increase the output power, you can use two or more 6P44S lamps connected in parallel. But in this case, it is absolutely necessary to select them according to the parameter μ with a difference in the operating point of no more than 1 ... 2%. Slope matching (S) is optional. Each lamp must have its own "anti-parasitic" resistors in the control and second grid circuits (with a resistance of 1 kOhm and 100 Ohm, respectively), as well as a separate auto-bias resistor shunted with a 470 microfarad capacitor at 63 V.

By the way, the opinion that triodes should not be connected in parallel is quite justified. However, if it is possible to accurately select lamps for μ, triodes can be connected in parallel, and there is a lot of evidence for this. For example, the 6S4S (2AZ) lamp, beloved by many, contains two parallel-connected triodes inside the cylinder, and some expensive Audio Note models have an output stage on two parallel-connected triodes.

Unfortunately, it was not possible to find a suitable mode for the 6P45S lamp in triode switching. Easily delivering 10 W to the load (more than the famous 300V triode), this lamp has a poor harmonic spectrum - the third harmonic spoils the sound, starting with a power of 2,5 W. And the reliability of this lamp is not great. Lamps 6P44S, on the contrary, proved to be quite reliable: some samples have been working for 15 years. Moreover, in the process of setting up, their anodes sometimes became red-hot, and this did not affect their further work in the least.

Triodes designed for voltage stabilizers (such as 6S19P, 6S3ZS, 6N13S) should not be used in single-ended amplifiers due to noticeable non-linearity. Of course, there are still powerful triodes: 211, 845 and the domestic GM-70, but this is a completely different safety technique - the anode voltage reaches 1000 V or more, and it is extremely difficult to make an output transformer for such lamps at home.

There are still many excellent output triodes that have not been covered by research due to their exorbitant prices: these are 300V manufactured by Western Electric, a single-anode version of 2AZ (there is one), a pre-war German AD1 similar to it, a domestic triode of the same time UB-180, modern W30B and so on.

Driver stage lamps must provide a large signal amplitude with a minimum output impedance. Article [4] lists four types of double triodes: 6N1P, 6N2P, 6N8S and 6N9S. Indeed, these triodes have the longest linear section of the characteristic, but in terms of output resistance they are not the best tubes. In many cases, the 6N23P double triode turns out to be the most optimal. With the correct mode (UA= 120 V, IA= 14 mA, UCK= -2,25 V, RA= 12 kOhm, RK- 160 Ohm), it develops a signal amplitude of 57 V quite linearly, having an output resistance of only 2...2,5 .200 kΩ and thus providing a bandwidth of about 80 kHz. But if you need to get a signal amplitude of 300 V, for example, to build up a 6V triode, it is better, of course, to use the 8H6C triode in the following mode: IA \u6d 1 mA, UCK \u50d -6 V, RK \u12d XNUMX kOhm, RA \uXNUMXd XNUMX kOhm. There is another very interesting lamp XNUMXFXNUMXP. Both the triode and the pentode in this lamp have remarkable characteristics - you can experiment.

The most important node of the tube amplifier is the output transformer. For some reason, some secrets of its proper manufacture are not mentioned in the literature. The fact that the transformer of a high-quality amplifier must be multi-sectional is probably no secret to anyone. And for some reason, they don’t write anywhere about the fact that spacers should be placed between the sections of the primary and secondary windings, as well as between the layers of the primary winding to reduce the capacitance. Moreover, the thickness of these spacers should vary in direct proportion to the variable component of the stress between the layers to be separated. The best available insulating material for gaskets is PTFE-4. In extreme cases, and also as an additional material, dry Whatman paper is suitable, but not capacitor paper, as is sometimes found in some descriptions. The thickness of the spacers and the number of winding sections can be calculated, but due to its complexity, only some specific designs will be given in this article.

For an amplifier with an output power of 10 ... 15 W, it is best to use a magnetic circuit and a frame from an OSM-0,25 kVA (SHL32x50) transformer. The transformer must be disassembled, the edges of the frame, on which the first layer of the winding lies, should be rounded, with a radius of 1,5 mm, and additional holes for the leads should be drilled in its cheeks. It is necessary to wind very carefully, each section should contain an integer number of layers filled from cheek to cheek.

Below is information about the transformer for the output stage on two parallel-connected 6P44S tetrodes in a triode connection. Its primary winding consists of four sections of 325 turns connected in series, for a total of 1300 turns of wire with a diameter of 0,355 mm. Each section consists of two layers with a 0,2 mm thick PTFE gasket between them. The secondary winding for a load with a resistance of 4 ohms consists of five sections of 77 turns connected in parallel. Each section contains one layer of wire with a diameter of 0,77 mm. Two more sections are wound over the second and fourth sections of this winding without spacers, each with 32 turns in two wires with a diameter of 0,56 mm (the placement of the windings is shown in Fig. 3).

These sections must be wound with a gap between the turns so that a uniform filling of the layer from cheek to cheek is obtained. All four wires of 32 turns are connected in parallel, and the resulting winding is connected in series with the winding of 77 turns. Thus, a winding of 109 turns is obtained for a load of 8 ohms. Between the four sections of the primary winding and the five sections of the secondary are eight spacers, the thickness of which varies approximately in an arithmetic progression from 1,3 mm (first spacer) to 0,2 mm (last spacer) as the alternating voltage component between winding sections I and II decreases .

When assembling the transformer, it is necessary to put rigid insulating gaskets 0,18 ... 0,19 mm thick into the gaps of the magnetic circuit.

The output stage with such a transformer has a reproducible frequency band of 4 Hz ... 200 kHz with a small signal, and 20 Hz ... 200 kHz at maximum power.

Let's talk now about the design features of the power transformer. Since the current drawn by the amplifier in class A mode remains virtually unchanged, the power transformer delivers a lot of power all the time. The methods given in the books for calculating a transformer operating on a rectifier with a filter are either too complicated or too simplified. Below are fairly accurate and simple formulas for calculating a transformer operating on a rectifier with a filter starting with a large capacitor.

Let's start with the simplest formulas. The no-load voltage of the secondary winding of the transformer is U2 = 220(n2/n1) [V] - this is understandable, although it is better to rely on the real average or maximum voltage in the network.

Let's denote the resistance

R=RB+RT.

where RB is the rectifier resistance (see below) and RT is the transformer resistance reduced to the secondary winding:

Rt= R2+R1 (n2/n1)2, where and R2 are winding resistances:

R1= 0,017 (Ii[m]/Si[mm2]).

The next step is to calculate the voltage drop VU. It is calculated from a system of two equations:

ΔU = √2·U2(1-cos φ);

ΔU = 1,5I R(90°/φ), where I is the direct current drawn by the amplifier.

The easiest way to solve this system of equations is by fitting (iterations), taking for the first approximation the cutoff angle φ within 20...30°.

The amplitude of the no-load voltage of the secondary winding of the transformer, which all filter and interstage capacitors must withstand, is determined from the equality and the rated voltage after heating the lamps on the first filter capacitor

U = √2 U2-ΔU - UB, what is UB, see below.

And the last formula is for the thermal power released in the transformer:

P = 0,8 I ΔU(RT/R).

When simplifying the formulas, some approximations were used, but they contribute to the error, as a rule, a smaller contribution than the discrepancy between the sine of the real form of the voltage in the network. In particular, the current-voltage characteristic of the rectifier was considered linear:

U(t) = UB+RB I(t).

For a rectifier bridge with silicon diodes, RB=0, UB=1,5 V can be considered, and for a 5TsZS kenotron, for example, RB=160 Ohm, UB=11 V.

The above method did not take into account the winding (windings) of lamp incandescence. It can be calculated independently from the calculation of the step-up winding, considering the voltage loss in it as the product of current and its resistance, and taking into account that the loss of effective AC voltage in the primary winding is usually about 2%.

The next important question is how to make a powerful transformer that does not create an acoustic background? In the article [7], some reasons for the "hum" of transformers were considered and a completely correct conclusion was made that it is necessary to increase the number of turns per volt by 15 ... 20% in comparison with the calculated value. This measure reduces the buzz only of the magnetic circuit, and even then not always. The acoustic background created by a loaded winding, on the contrary, grows with an increase in the number of turns. The method of dealing with winding buzz is surprisingly simple - this is sectioning, the same as in the output transformer. Sometimes it is enough to place the primary winding between the halves of the secondary, and the acoustic background is reduced to an acceptable level.

Another possible reason for the buzz of the power transformer is the saturation of the magnetic circuit with the constant voltage component, which, although small, is often present in the network. This reason manifests itself, as a rule, only in toroidal transformers with a continuous magnetic circuit, and the saturation effect increases with an increase in the number of turns and with a decrease in the resistance of the primary winding. There is only one method of dealing with this phenomenon - installing a filter in series with the primary winding of the transformer that delays the direct current component. The filter circuit for a network transformer for power up to 300 W, borrowed from the American amplifier LAMM M1.1 developed by V. Shushurin [8], is shown in fig. 4. If the transformer is more powerful, then the capacitance of oxide capacitors must be proportionally increased, and the resistance of the resistor must be reduced.

On fig. 1 and 2 show two practical circuits of tube single-ended amplifiers: a power of 10 W on tetrodes in a triode connection and 12 W on tetrodes. The output transformer for the first of them is described above, and the transformer for tetrodes is assembled on the same magnetic core, but has slightly different windings. Its primary winding - 1512 turns of wire with a diameter of 0,35 mm - consists of five sections: 168, 336, 504, 336 and 168 turns. Between them there are four sections of the secondary winding for a load with a resistance of 4 ohms - 77 turns of wire with a diameter of 0,77 mm, connected in parallel. Over the second and third sections of this winding, without spacers, two sections of 32 turns of wire with a diameter of 0,72 mm are wound, connected in parallel. This winding is connected in series with a winding of 77 turns; this is how the secondary winding for a load of 8 ohms is obtained. The gaskets between the primary and secondary windings and between the layers of the primary, as well as the gaskets in the gaps of the magnetic circuit, are the same as in the transformer for the triode amplifier.

The output impedance of an amplifier with triodes at the output for an 8 ohm load is 2,4 ohms, and with tetrodes it is 1,6 ohms. At the output for a load of 4 ohms - exactly two times less.

Finally, a note about the choice of capacitors for signal circuits. For use in high-quality amplifiers, capacitors with a polypropylene dielectric (K78-6, K78-2) and with a paper dielectric (K40U-9, MBM) for a voltage of at least 400 V are most suitable. A low-capacity capacitor (C6 in Fig. 2) - mica KSO-1. Oxide capacitors should be selected from products of well-known foreign companies (TC series, SK Jamicon and similar); it is permissible to use domestic K50-35. In power filter circuits, capacitors K50-20, K50-32 can be used.

Literature

1. Vrublevsky A., Chumanov D. Acoustic systems for tube amplifiers. - Audio Shop, 2001, No. 1, p. 183-186.
2. The road to one-stroke paradise (interview with P. Quartrup). - Audio Shop, 1996, No. 3, p. 68-70.
3. Vrublevsky A., Chumanov D. Single-ended tube amplifier from available parts. - Audio Shop, 2000, No. 4, p. 185-188.
4. Gendin G. Features of the design of modern tube ultrasonic frequencies. - Radio, 2003, No. 1, p. 12, 13.
5. Troshkin N. Triode from improvised materials. Details. - Class A, 1998, March, p. 20-27.
6. Barkan VF, Zhdanov VK Design of radio engineering devices. - M.: Oborongiz, 1963.
7. Polyakov V. Reduction of the transformer stray field. - Radio, 1983, No. 7, p. 28, 29.
8. Kunilovsky S. Power amplifier LAMM Audio Laboratory M1.1 - Audio Shop, 2000, No. 2, p. 61-65.

Author: A.Ivanov, Ivanovo

See other articles Section Tube Power Amplifiers.

Read and write useful comments on this article.

<< Back

Latest news of science and technology, new electronics:

Machine for thinning flowers in gardens 02.05.2024

In modern agriculture, technological progress is developing aimed at increasing the efficiency of plant care processes. The innovative Florix flower thinning machine was presented in Italy, designed to optimize the harvesting stage. This tool is equipped with mobile arms, allowing it to be easily adapted to the needs of the garden. The operator can adjust the speed of the thin wires by controlling them from the tractor cab using a joystick. This approach significantly increases the efficiency of the flower thinning process, providing the possibility of individual adjustment to the specific conditions of the garden, as well as the variety and type of fruit grown in it. After testing the Florix machine for two years on various types of fruit, the results were very encouraging. Farmers such as Filiberto Montanari, who has used a Florix machine for several years, have reported a significant reduction in the time and labor required to thin flowers. ... >>

Advanced Infrared Microscope 02.05.2024

Microscopes play an important role in scientific research, allowing scientists to delve into structures and processes invisible to the eye. However, various microscopy methods have their limitations, and among them was the limitation of resolution when using the infrared range. But the latest achievements of Japanese researchers from the University of Tokyo open up new prospects for studying the microworld. Scientists from the University of Tokyo have unveiled a new microscope that will revolutionize the capabilities of infrared microscopy. This advanced instrument allows you to see the internal structures of living bacteria with amazing clarity on the nanometer scale. Typically, mid-infrared microscopes are limited by low resolution, but the latest development from Japanese researchers overcomes these limitations. According to scientists, the developed microscope allows creating images with a resolution of up to 120 nanometers, which is 30 times higher than the resolution of traditional microscopes. ... >>

Air trap for insects 01.05.2024

Agriculture is one of the key sectors of the economy, and pest control is an integral part of this process. A team of scientists from the Indian Council of Agricultural Research-Central Potato Research Institute (ICAR-CPRI), Shimla, has come up with an innovative solution to this problem - a wind-powered insect air trap. This device addresses the shortcomings of traditional pest control methods by providing real-time insect population data. The trap is powered entirely by wind energy, making it an environmentally friendly solution that requires no power. Its unique design allows monitoring of both harmful and beneficial insects, providing a complete overview of the population in any agricultural area. “By assessing target pests at the right time, we can take necessary measures to control both pests and diseases,” says Kapil ... >>

Random news from the Archive Monkeys are capable of long reflections 30.04.2023 A study by neuroscientists at the University of Pittsburgh showed that monkeys, like humans, are capable of long-term thinking and careful choice of decisions. A few decades ago, Nobel Prize winner Daniel Kahneman revolutionized the field of behavioral economics with his proposition theory. In his book Thinking Fast and Slow, he suggested that people use two different systems of thinking: one is almost instantaneous, which occurs automatically, and the other is conscious-logical, requiring great time and mental effort. Kahneman called the first type of thinking "fast" and the second - "slow". Slow, expensive thinking allows us to compose music, build scientific hypotheses, or balance the balance sheet. Neuroscientists challenged two rhesus monkeys with a combinatorial optimization problem. This is the "knapsack problem". It is formulated like this. There is a knapsack (or box) and a number of items that need to be placed in the knapsack. Items of different sizes (or values) and depending on the order in which we put things in a knapsack, more or less things can fit in it. In general, the problem is not solved, that is, there is no optimal stacking algorithm. For a small number of items, there are such algorithms, but they are rather complicated. The scientists rewarded the monkeys based on the value of the solution they found: the more items they managed to pack, the more valuable the solution. The monkeys gradually stopped rushing and began to think. The performance and speed of thinking of the monkeys depended on the complexity of the problem, and the solutions found were close not only to the solutions of humans, but also to the optimal choices of optimization computer algorithms. The conclusion made by scientists: monkeys can think.

Other interesting news:

▪ Corals push Japan's borders

▪ Thermoid converts heat into electricity

▪ Ocean under Beijing

▪ Epson 10000 lumens installation projectors

▪ TV on the wrist

News feed of science and technology, new electronics

Interesting materials of the Free Technical Library:

▪ section of the site Palindromes. Article selection

▪ Mother Teresa article. Famous aphorisms

▪ article Who Invented the Telephone? Detailed answer

▪ article Work on machines for binding packs of books such as MOP, AMPAG, ZIBIK, CYCLOP, etc.. Standard instructions for labor protection

▪ UHF Converter article. Encyclopedia of radio electronics and electrical engineering

▪ article COM port radio extender. Encyclopedia of radio electronics and electrical engineering

Leave your comment on this article:

All languages ​​of this page

Arabic	Hebrew	Polish
Bengali	Hindi	Portuguese
Chinese	Italian	Spanish
English	Japanese	Turkish
French	Korean	Ukrainian
German	Malay	Vietnamese

Home page | Library | Articles | Website map | Site Reviews

www.diagram.com.ua
2000-2024