|
Arabic
Bengali
Chinese
English
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Malay
Polish
Portuguese
Spanish
Turkish
Ukrainian
Vietnamese|
ENCYCLOPEDIA OF RADIO ELECTRONICS AND ELECTRICAL ENGINEERING The truth and tales of high-quality sound reproduction. Encyclopedia of radio electronics and electrical engineering
Encyclopedia of radio electronics and electrical engineering / Audio equipment Over the past five or six years, the market for household sound reproducing equipment in the CIS countries has clearly been saturated with various foreign equipment, and today not every audiophile, especially a radio amateur, will take the "word" of advertising about its merits, even if we are talking about the High End class. Unfortunately, many people, having invested a considerable amount in the purchase of, for example, UMZCH of a particular company, find that its main advantage is just a beautiful design, but by no means the sound quality. A well-known role in the disorientation of buyers is made by the colorful magazines for audiophiles that have appeared in recent years. In them, almost all materials talk about the features of the equipment, based on data provided by advertisers, as a rule, in "pink" color. Well, advertising, as you know, is the engine of trade, but a true recording enthusiast with a critical mindset is always able to figure out what is "good" and what is "bad". Muscovite Nikolai Klimenko, one of the readers of "Radio", took with great doubt the reasoning and unfounded criticism of the experts of the magazine "AUDIO MAGAZIN" (hereinafter "AM") about UMZCH high fidelity (hereinafter - UMZCH VV), described in [1]. In particular, he asked to comment on some judgments (under the heading "Mail" - "AM", 1996, No. 4, p. 3,4). After reviewing the notes in "AM", I can note that experts V. Zuev and S. Kunilovsky, in my opinion, understand circuitry, to put it mildly, not very well. So, for example, V. Zuev, evaluating the UMZCH VV circuitry, tried to prove that (I quote) "the microcircuit at the amplifier input ... will surely steal the virtual depth of the stereo panorama, so necessary to create the effect of presence" (meaning the high-speed K574UD1 op-amp with input cascade on field-effect transistors). It is appropriate to ask why exactly this op-amp "steals depth", and a dozen op-amps through which the sound signal passes to the UMZCH in a tape recorder, CD player or any other signal source (even in "tube" CD players, the DAC is made, as it should be known and an expert, on a solid-state IC, inside of which there are several op-amps), will they behave "decently" and will not "steal" anything? Further, the AM expert tries to convince us of "it is practically impossible to obtain good sound in amateur conditions", since "for good sound reproduction, expensive "hi-fi" conductors, switches, complex methods of their connection (oxygen-free soldering, special solders) are required )". It justifies the "ridiculous" price of Audio Note's ($120400) 17W and Kegon's ($247000) 45W amplifiers, as well as, apparently, several hundred dollar non-crystal interconnect cables. It is known from the course of physics that any metal-to-metal contact (in the presence of at least the thinnest oxide film) can be considered as a non-linear element of an electrical circuit. And this non-linearity can degrade the sound of high fidelity systems. But, for example, it’s hard for me to believe that V. Zuev heard the real operation of the UM3CH VV and, moreover, is somehow familiar with its circuit, since it was precisely the elimination of the nonlinearity of the connecting wires, connector contacts and relays that was given special attention during the development of this amplifier. In particular, a special cascade was introduced into the amplifier, which compensates not only for non-linearity, but also for the active and reactive components of the distributed resistance of the connecting wires, and the common OOS circuit is designed in such a way that it compensates for the non-linearity of the "cold" contacts of the UMZCH output switching relay and connectors. In other words, those negative factors that V. Zuev mentions and that can worsen the sound are eliminated in the UMZCH VV in the most effective way - circuitry. I cannot agree with the statement that "amateurism in sound engineering cannot compete with branded equipment ... in terms of sound quality." If we are talking about the design and execution of the case - yes, it is difficult for an amateur to compete with the industry. But if we talk about sound quality, then today even a radio amateur with average training is quite capable of collecting UMZCH in the price category of $300-500, while spending only $40...50. But for this you need to be a radio amateur and not follow the advice of V. Zuev "it is better to buy a ready-made device." Somewhat pretentious, I think, is the opinion of the "AM" expert that "Mr. Sukhov, very late, drew attention to the schematic exoticism of some foreign companies that do not differ in the sound quality of their products (meaning Kenwood and Akai. - Note by the author ) and... late by about 10 years". But why then "AM" discusses the design of seven years ago as the most popular and still unsurpassed in terms of parameters? For the world of electronic technology, this is a long time. Concluding the presentation of my opinion about the notes in "AM", I want to note that in themselves such journals are, of course, useful. But many of the statements of individual authors of articles may seem indisputable only to those readers who, excuse me, are not able to distinguish a transistor from a resistor. For people who understand the circuitry of audio equipment, some articles in "AM" make a miserable impression. I am convinced that it is possible to teach someone when you yourself thoroughly, in the smallest detail, know what you are writing about. In his letter to "Radio" N. Klimenko was also interested in the "philosophy" that I adhered to when developing UM3CH VV, and in conducting expert auditions. So, this amplifier was developed as the final link of the stand for the subjective examination of the sound of CD players on the instructions of one of the testing laboratories. The task was to make a design on the domestic element base and provide an output power of 100 W at a load of 8 ohms (JBL studio monitors) with a level of distortion and noise 10 ... 20 dB lower than that of CD players. Having repeated on domestic elements up to a dozen UMZCH options from leading Western firms, I was convinced that on complementary transistors of the KT818, KT819 series with a low cutoff frequency, it would not be possible to obtain an acceptable (according to TK - no more than 0,001%) level of non-linear distortion at the highest frequency of the audio range. The phase shift created by these transistors already at audio frequencies (i.e., one or two orders of magnitude lower than those of foreign ones) forced the introduction of a deeper frequency-phase correction to ensure stability, which, in turn, limited the depth of the OOS at higher frequencies and worsened the linearity. The problem was solved by completely abandoning the inclusion of transistors according to the OE scheme. A lead correction was introduced, compensating for the pole formed by the transistors of the output stage on the frequency response of the amplifier with an open OOS. As a result, the customer's requirements for linearity were met with a large margin over the entire audio range, and the amplifier was put into operation. But then it turned out (I participated as a "listener" in most subjective tests) that the CD being played sounds differently through monitors (studio speakers) connected to the UMZCH with different cables! Then, having carefully studied the phenomenon, we realized that those thousandths of a percent of the distortion that the UMZCH gave were nothing compared to the distortions created by connecting cables with connectors. Replacing the connectors with gold-plated ones, and ordinary connecting wires with special ones with a "non-crystalline" structure ($250 for a twisted pair 4 m long), only partially solved the problem - the distortions decreased several times, but did not disappear. Then, after a series of experiments with Kenwood studio amplifiers with the "Sigma Drive" system, I tried to introduce into the UMZCH cascades to compensate for the full impedance of the wires and the non-linearity of "cold" contacts. The result exceeded all expectations - the distortions disappeared, and regardless of the quality (and price!) of the connecting wires and connectors. This is how the construction described in Radio No. 6, 7 of 1989 was born. By the way, I strongly recommend that all lovers of high-quality sound install the mentioned compensation scheme in their UMZCH. This is easy to do: you only need three precision (or precisely matched) resistors and one op-amp. Its type does not matter much, it can be K140UD6, IK157UD2. On fig. 1 shows the functional diagrams of typical UMZCH: fig. 1, a - with an input stage on discrete elements, fig. 1, b - with the input stage on the op-amp, the rest of the stages are "hidden" in block A2. The input of the compensation circuit is connected to a common output directly at the loudspeaker terminal, and the output through the resistor Rdop, the resistance of which must be exactly equal to the resistance of the resistor R2 in the UMZCH common OOS circuit, is connected to the inverting input of the input stage. Resistors in the compensator should be used precision (with an error of no more than 1%).
The principle of operation of such a compensator is to measure the voltage drop on one of the connecting wires, doubling it and "adding" it to the normal signal at the UMZCH output, which is equivalent to eliminating the wires between the amplifier and loudspeakers. Such a circuit design does not require any adjustment when replacing connecting cables or speakers. Try it and you will see that the effect will exceed all your expectations (of course, if your amplifier, signal source and especially speakers are of high enough quality). Answering the question about the subjective comparison of the sound of UMZCH VV, I want to note that I recognize only "anonymous" tests conducted according to the system of the so-called A-B-X examination, during which the compared devices A and B are invisible to experts and switch randomly (say ,"A" then "B" and subsequent "X" switches are not declared). So, in the course of the A-B-X examination of the comparison, the UMZCH VV was better or no worse than the Kenwood KA-500, Quad 405, Yamaha A-1 cost category $ 400 - 1000 available at the disposal of the testing laboratory and much better than "Brig", "Odyssey" -010" or lamp "Surf". By the way, it was the A-B-X examination that made it possible to see firsthand how many High End connoisseurs lost the ability to distinguish the components of the Hi-Fi and High End classes as soon as the object of their boundless, but "blind" love disappeared behind a black partition. Of course, I don’t have a perfect ear for music, but, in my opinion, much of what is now “spinning” around the word “High End” looks like a religious dispute (“I believe it or not”), and the excitement is artificially pumped up with the sole purpose is to stimulate sales. In this regard, I recall the case with the release of a "special variant" of the popular tape recorder "Nakamichi 1000 ZXL" by Nakamichi, in which all the details, up to the power supply radiators, were gilded! Whether this added quality to the sound - readers will guess for themselves, but the price has increased by about three times compared to the standard model. Speaking of modern high-quality sound reproduction, I can't help but share some observations that also do not correspond to "pink" shades. Tube amplifiers. They, indeed, for the most part sound nicer than transistor ones. But "more pleasant" does not mean more accurate. The output transformer is a device with much greater (due to the hysteresis loop and finite saturation induction of the magnetic circuit) non-linearity, frequency and phase distortion than a transistor in linear mode. "Pure lampoviks", who understand the problem, created transformerless UMZCH on 6C3ZS, but this is an exception to the rule. It is precisely because of the large phase distortions that the tube UMZCH is difficult to cover with deep OOS, which ultimately manifests itself in a relatively large output resistance (units of ohms, for transistors - usually hundredths of an ohm), as well as a relatively smooth limitation during overload (in Fig. 2 curves 1 and 2 represent the typical amplitude characteristics of tube and transistor amplifiers, respectively).
Try to artificially increase the output impedance of any "medium" transistor UMZCH to 2 ... 4 ohms (for this, it is enough to turn on a 10-20-watt resistor with such a resistance in series with the speaker system) and do not exceed a quarter of its rated power so that short-term signal peaks do not cut off. You will be convinced that the sound in 95% of cases will acquire "tube softness". The reason lies in the fact that many (but not all!) Loudspeakers provide a minimum of intermodulation distortion (in terms of sound pressure) not with an output impedance of the UMZCH close to zero, but with its value of at least 3 ... 5 Ohm *. However, such resistance violates the linearity of the frequency response and phase response of passive crossover filters of acoustic systems, which are usually designed based on the zero value of the output impedance of the UMZCH. But these are not problems of amplifiers, but of acoustic systems! It is acousticians who should take care when developing systems not only about the linearity of the frequency response and phase response in terms of sound pressure on a sinusoidal signal, but also about minimizing acoustic intermodulation distortion at Rout = 0 or, worse, normalize Rout, say, by 3 Ohm and calculate crossover filters for is the source resistance. Another common misconception among audiophiles is that compact discs (CDs) provide more dynamic range than analog compact cassettes (CCs). In this case, the formula for calculating quantization noise is given as the main argument: Nkv=6N+1,8 [dB], where N is the quantization bit depth by level. For CD, N=16 is accepted, therefore, the theoretical level of quantization noise NKBKd =6X16+1,8=97,8 dB. With someone's light hand, this value is taken as the dynamic range of the CD. Given that the best QCs have a signal-to-noise ratio (without noise reduction systems) of the order of 55 dB, it is concluded that the CD gain is more than 40 dB. But we must not forget that the principles of analog QC and digital QD are fundamentally different, so it is incorrect to use QC measurement methods to assess the dynamic range of QD. In QC, the dynamic range from below is indeed determined by the noise level, but this does not mean that the same is the case with CD! Looking at fig. 3, which shows the typical dependences of the coefficient of nonlinear distortion Kni KK and KD as a function of the signal level, you can easily see that in analog recording, with a decrease in the level, KNI monotonously decreases, while in digital recording it increases, tending to 40% (since relative size of the quantization step).
If the analog recording in the distortion spectrum is dominated by the third and fifth harmonics, which are not very sharp, then the situation is much worse in the digital recording - many combinational components do not form the harmonic series familiar to the ear, and their effect becomes noticeable already at levels of about 1%. It is easy to see that at signal levels of the order of -50 dB and below, the distortion of CD signals crosses the threshold of acceptable 1%. From below, its dynamic range is limited not by quantization noise, but by non-linear distortions. And out of the theoretical 97,8 dB, only 50 remains. But that is not all! When the QC is overloaded, the non-linear distortions are proportional to the square of the recording level (when the level is doubled, the harmonic coefficient increases only four times), so their short-term appearance at the signal peaks is imperceptible by ear. For CD, when the nominal input level of the analog-to-digital converter (ADC) is exceeded by only 2 ... 3 dB, nonlinear distortions increase thousands of times, therefore, in real digital recording equipment, a level of 12 ... 15 dB is taken as nominal (i.e. .to the crest factor of a real music signal) is less than the input limit for the ADC. As a result, out of the original 97,8 dB, only 35 ... 37 dB of real remains, which is 20 dB less than that of QC. That is why, despite the subjective absence of a "thorn", many phonograms played back from a CD lead to rapid fatigue and have a noticeably worse "stereo panorama depth" than the same phonogram played back from an analog vinyl record or a high-quality CD. By the way, modern records made using Direct Metal Mastering technology provide a dynamic range of 60...65 dB and are highly valued by audiophiles. It is impossible not to mention two more "attacks" on KK - by the developers of the DCC digital compact cassette and the MD mini-disk. Since the advent of DCC (1989) and MD (1993), Philips, the developer of DCC, has been trying to convince audiophiles that it is DCC that will completely replace CC in 1-2 years. Sony, the developer of MD, also made a similar statement, but in relation to MD. But ... time passed, and KK is still the main household carrier of audio programs with the ability to record. Moreover, if at first the DCC format was supported by the world giant Matsushita and a number of other well-known companies, today DCC only produces Philips, and even then only a few models (against the background of dozens of KK models). Sony, also dismayed by the subjective assessment of sound quality by the German magazine "Audio", which placed the MD in last place with 45 points out of 100, behind the CD player (1 points) and cassette recorder ( 2 points) and took 85rd-85th places of the vinyl record player (3 points) and DCC tape recorder (4 points), began feverishly improving the digital audio data compression system, as a result of which four (!) Versions of the compression algorithm were born in 80 years ATRAC 80 - ATRAC 4, and the previous ones are not compatible with all subsequent ones (i.e. "old" MD players are not able to play "new" records)... It's time to remember that in DCC and MD, as in CD, 16-bit level quantization is used, but to reduce the flow of data written to the media, digital compression is used according to the PASC (Precision Adaptive Subband Coding) and ATRAC (Adaptive TRansform) algorithms, respectively. Acoustic Coding), which reduce the digital data stream from 2 Mbps to 384 kbps and 300 kbps, i.e. both DCC and MD reproduce sound fundamentally less accurately than CD. Forecasting is a thankless task, but in fairness, let's remember the fate of another (theoretically superior in quality to CD) R-DAT format, which at the time of its appearance in 1987 was also predicted to be the heir to KK. Indicative in this sense is the fairly accurate forecast of the author of these lines, published in [2]. While almost all foreign and domestic press wrote that by 1991 R-DAT would completely replace KK, this was perhaps the only publication in which R-DAT was given a modest place, perhaps in semi-professional recording studios. In conclusion, I take this opportunity to express my deep gratitude to all correspondents and admirers, whose moral, informational and material support made it possible to develop many of my designs. Note * "Should the UMZCH have a low output impedance?" in "Radio", 1997, No. 4, p. 14-16. Literature
Author: N.Sukhov, Kiev, Ukraine
Machine for thinning flowers in gardens
02.05.2024 Advanced Infrared Microscope
02.05.2024 Air trap for insects
01.05.2024
▪ France to get rid of coal-fired power plants ▪ Mouse Virtual Reality Headset ▪ 5" plastic-based flexible AMOLED display ▪ NXP Opens Design Center in Singapore to Develop W-USB Technology
▪ section of the site Personal transport: land, water, air. Article selection ▪ article Where and when did the fighting stop every day to feed the ducks? Detailed answer ▪ article Work on calenders. Standard instruction on labor protection
Home page | Library | Articles | Website map | Site Reviews
www.diagram.com.ua |
See other articles Section 
Latest news of science and technology, new electronics:
All languages of this page