|
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 Transistor UMZCH with increased dynamic thermal stability. Encyclopedia of radio electronics and electrical engineering
Encyclopedia of radio electronics and electrical engineering / Transistor power amplifiers The article describes the UMZCH, which uses technical solutions that improve the dynamic thermal stability of the output stage on bipolar transistors. In such a cascade, switching distortions are eliminated due to the elimination of current cutoff in high-power transistors. In the second part of the article, recommendations are given for expanding the amplifier bandwidth from below, which has a beneficial effect on the quality of sound reproduction. A similar UMZCH was presented by E. Aleshin at the Russian Hi-End 1998 exhibition, where it adequately competed with tube amplifiers.
The main source of heat release in the UMZCH is the output stage, and in the development of transistor power amplifiers, great attention has always been paid to its thermal stabilization. In the 80-90s, in high-quality UMZCH (for example, [1 - 3]), the output stage circuit, simplified in Fig. 1, was most widely used. 2. Its advantages include satisfactory thermal stability (when transistors VT4, VT5, VTXNUMX are placed on a common heat sink), a high cutoff frequency of the transfer coefficient, and low output resistance. However, the cutoff of the current of the passive arm, as well as the dynamic instability of the quiescent current of the output transistors due to fluctuations in the temperature of the junctions of the transistors when the signal level changes, contribute to an increase in switching distortions. These features worsen the subjective assessment and reliability of sound reproduction. About dynamic mode stabilization A few years ago, the Khabarovsk inventor E. Aleshin proposed a method for stabilizing the operating mode (quiescent current) of transistor cascades [4,5, 6], which made it possible to reduce dynamic temperature instability by an order of magnitude, eliminate current cutoff in the UMZCH push-pull output cascade and make the current redistribution in it more accurate ( as in a "parallel" amplifier [XNUMX]).
On fig. 2 shows a simplified circuit of an amplifier with current feedback [2] (A1 is a push-pull follower), where, unlike the prototype, the operating point of the output stage is stabilized using the node proposed by E. Aleshin. The quiescent current stabilizer is made on the elements VT3, VT4 and VD1, VD2. When a through current flows through powerful transistors VT5, VT6 and non-linear elements connected in series with them - diodes VD1, VD2 - a voltage drop forms on the latter, which, when the opening threshold of transistors VT3, VT4 is reached, causes their base and collector current to appear, reducing the input current of the transistors VT5, VT6. As a result, the through current through the transistors of the output stage is limited and, accordingly, the current through the diodes VD1, VD2 - current sensors. Static (long-term) thermal stability is achieved, as in the scheme in Fig. 1, providing thermal contact of transistors VT3, VT4 with diodes VD1, VD2. Dynamic stabilization is much better due to less heat generation on diodes than on powerful transistors, and the effect is achievable if the crystals of these diodes and transistors are comparable in volume. In the presence of a signal, a smooth redistribution of current through the load and between the diodes VD1 and VD2 is obtained due to the logarithmic CVC of the diodes. Moreover, the current through them never decreases to zero, except for the current cutoff of the output transistors. The current through the passive shoulder can be significantly increased by including a resistor between the bases of transistors VT3, VT4 (i.e., parallel to VD1, VD2). At the same time, neither the temperature of powerful transistors nor the voltage drop across the resistors (if any) in the base and emitter circuits of these transistors affect the quiescent current and its distribution between the arms in the presence of a signal. It may seem difficult to choose diodes and transistors connected in parallel with them by an emitter junction in order to provide a stabilization condition: Σ UBe = Σ UVd. In fact, it is enough just to find suitable types of devices, selection of copies is not required. In addition, there is a simple way to adjust the operating point, which is shown later in the description of the proposed UMZCH. About thermal distortion Here it is appropriate to talk a little about thermal distortions and methods for their elimination when designing transistor amplifiers. Thermal distortions are changes made to the signal when it passes through an electrical circuit or amplifying stage, due to the thermal effect of the signal itself (current) on the temperature-sensitive parameters of the amplifier elements. An example of thermal distortion in passive circuits is signal compression in dynamic heads due to heating of the voice coils (especially in high-power, high-temperature heads). In semiconductor devices, an increase in the crystal temperature under the action of a flowing signal current causes a change in such basic parameters as, for example, the forward voltage of diodes (-2,2 mV/K), the base-emitter voltage of bipolar transistors (-2,1 mV/K), static current transfer coefficient of bipolar transistors (+0,5%/K), etc. Thermal processes have an inertial nature, due to the real heat capacity of the crystal and the device case. Therefore, electrothermal processes in transistors not only lead to changes in the instantaneous values of parameters, but also create a "memory" effect in electrical circuits and amplifying stages. Thermal memory in amplifying cascades manifests itself as time-varying parameters after exposure to a powerful signal: displacement of the operating point of the cascades, change in the transfer coefficient (non-stationary multiplicative error); shift of the constant component of the signal (non-stationary additive error). The latter is similar to the manifestation of the absorption of the dielectric of a capacitor in the signal path circuit. These processes create linear and non-linear signal distortions that degrade the quality of the reproduced sound [7]. It should be especially noted that conventional thermal stabilization is not able to significantly improve the dynamic thermal stability of cascades due to the much larger time constant of thermal processes in the device compared to the time constant of thermal processes inside a semiconductor device. This is partly true even for monolithic microcircuits. Obviously, to eliminate the problems associated with the thermal memory of semiconductor devices, it is necessary to use circuit solutions that reduce temperature fluctuations in device crystals or their effect on amplifier parameters. Such solutions can be: - isothermal mode of operation of a semiconductor device [8];
Description of the UMZCH scheme The power amplifier is made according to the circuit diagram (Fig. 3), corresponding to the shown block diagram.
Main Specifications Rated input voltage, V...............1
An R1C2 low-pass filter is installed at the input to reduce RF interference at the input. The same circuit includes an input voltage limiter on the elements R3, R4, C1, C3, VD1 -VD4 to protect against overload of the input stages of the amplifier. The input signal from the volume control (RG) through the low-pass filter goes to the "parallel" follower VT1, VT2, VT4, VT5 (called in [10] a pseudo-push-pull emitter follower). Resistors R5, R6 are used to balance the input current, i.e., to eliminate the constant component of the current through the RG, which occurs due to the difference in the static current transfer coefficients of the input bipolar transistors and creates a bias voltage at the input. Capacitor C6 prevents self-excitation of the input stage at radio frequencies. The static mode of operation of the repeater is stabilized by the supply voltage by parametric stabilizers R7VD5, R12VD6 and is set by resistors R8-R11, R16, R17T8K, so that at rest the difference in thermal power between the transistors of the repeater stages is small. The dynamic thermal regime determined by the elements R13, R14, R24, R25 in combination with the static regime is chosen so as to minimize power fluctuations on the repeater transistors in the presence of a signal and the difference in the instantaneous powers of transistors VT1 and VT4 (VT2 and VT5), thus obtaining , the minimum instantaneous temperature difference between their crystals. This is done so that the thermal voltage fluctuations of the IBE transistors of the first and second stages are subtracted and the signal voltage at the output of the repeater, and therefore at the output of the amplifier, is minimally subject to thermal distortions, interpreted as "signal voltage memory" (non-stationary additive error) . The voltage from the output of the amplifier through the divider R26R16 and R27R17 is fed to the output of the "parallel" follower - the emitters VT4, VT5, changing the current through them, i.e., an error current is formed, proportional to the deviation of the output voltage of the amplifier, divided by the UMZCH gain, from the input. The antiphase error current through the current follower VT3 (VT6) is supplied to the current amplifier VT13 (VT14). Its output is loaded on resistors R39, R40 and the input impedance of the output follower VT15, VT16, on which voltage is allocated (i.e., this is an impedance conversion stage) and fed to the load (AC) through the output follower. Resistor R41 determines the quiescent current of the error current amplifier (VT13, VT14) and is chosen so as to exclude the closing of the passive arm of this stage due to the current flowing through R39, R40. The latter shift the first pole up in frequency in the general NF loop. Frequency correction in the OOS loop is carried out by capacitors SYU, C11, connected between the impedance conversion stage and the output of the "parallel" follower. Their inclusion improves the transient response of the amplifier when it is loaded on a low-impedance load, i.e. on speakers [2]. Phase advance correction is performed by the R28C7 and R29C8 circuits. The trimmer resistor R15 serves to eliminate the offset at the UMZCH DC output. The emitter current of the output stage flows through current sensors - diodes VD11-VD14. The voltage from the diodes, containing information about the instantaneous value of the through current of the output stage, is fed through the divider R42R36R37R43 to the differential amplifier VT11, VT12 and is converted into current. From the collectors VT11, VT12, the current through the current mirror VT7, VT9 (VT8, VT10) is fed to the input of the error current amplifier, reducing its input current. Since the change in this current is in-phase in both arms (unlike the error current from the "parallel" follower), it leads to a change in the through current of the error amplifier, and hence the output stage, but does not change the output voltage. Thus, the quiescent current of the output stage is stabilized. The R38C13 circuit prevents parametric excitation of the stabilization unit, and also, together with R42, R43, performs frequency correction in the OOS loop. Connecting the stabilization unit is somewhat different from the diagram in Fig. 2, but this is not important, and in amplifiers of various structures it can be implemented in different ways. In this case, however, it must be taken into account that the dynamic temperature fluctuations of the stabilization feedback transistors (VT3, VT4 in Fig. 2 and VT11, VT12 in Fig. 3) also affect the thermal stability of the operating point of the output stage, but shift it in the opposite direction compared to diodes - current sensors. Diodes VD7-VD10 are protective, they prevent opening of the OOS of stabilization of the quiescent current during transients (for example, when the power is turned on or strong impulse noise), while turning into a POS with an uncontrolled increase in the through current in the output stage. DiodeYu9 (VD10) also creates an additional voltage drop across the current mirror transistor VT7 (VT8), bringing it to a more linear section of the characteristic. Construction and details The amplifier is assembled by the author on a universal breadboard. Powerful transistors of the output stage are installed on a common heat sink with a thermal resistance of not more than 2 K / W through insulating heat-conducting spacers. Powerful diodes, together with transistors VT11, VT12, are placed on a separate heat sink connected to a common wire, with a thermal resistance of not more than 15 K / W. It is better to install transistors on the reverse side of the plate heat sink, opposite the diodes with the highest forward voltage (if they are of different types, as in Fig. 3), i.e. in this case, VT11 is opposite VD12, and VT12 is opposite VD13. Transistors VT13, VT14 are installed on small heat sinks with a thermal resistance of 20...30 K/W. They can also be placed on a heat sink with output stage diodes, but this will worsen the static thermal stability of the quiescent current. In this case, the thermal resistance of the total heat sink should be no more than 10 K/W. Fixed resistors - metal-film, tuning - multi-turn. Resistors R8-R11, R16-R18, R23, R26, R27, R32, R35 - with a tolerance of ±1%; they can be selected from ordinary ones with a tolerance of ± 5% or precision closest to the indicated ratings from the E96 series. The remaining fixed resistors have a tolerance of ±5%. Oxide capacitors C14, C15 - low-impedance (low ESR) used in switching power supplies; non-polar with the specified rated voltage - film. Capacitors C2, C10, C11 are preferably used with a polystyrene or polypropylene dielectric, the rest are ceramic for a voltage of 25 or 50 V with an X7R dielectric (or NPO, COG groups for C6 C8). Zener diodes VD5, VD6 are precision, they have a tolerance of ± 1%, you can also use others with a tolerance of ± 2% (for example, BZX55B) or select from a range of ± 5% (BZX55C). Diodes VD7-VD10 - ultrafast (ultrafast) for an average current of 1 A, with a forward voltage of 0,6 ... 0,7 V at a current of 0,1 A. The output stage diodes can be any powerful Schottky diodes or ultrafast for an average current of not less than 10 A. Any combination of types and number of diodes in the arm is acceptable; it is only important that the total voltage drop for a given quiescent current flowing through them be within 0,7 ... 0,9 V. For example, the VD12 (VD13) diode can be replaced by two MBR1045 or MBR1035 connected in series. It is preferable to use diodes for currents up to 20 A or more, as having a larger crystal volume, and therefore capable of providing better dynamic thermal stability. Transistors BC550C, BC560C in the "parallel" repeater can be replaced by BC550B, BC560B or BC549, BC559 with letter indices C or B, and in other positions also by BC547, BC557 or BC546, BC556 with letter indices C or B. Transistors VT11, VT12 - low-power high-frequency ones with low junction capacitance, permissible direct collector current of at least 0,1 A and collector-emitter voltage of at least 60 V. 2SA1540, 2SC3955 or BC546, BC556 with any letter index are also suitable, in the latter case, the stability margin of the stabilization unit will decrease somewhat. Transistors VT13, VT14 - high-frequency medium power, with a permissible direct collector current of at least 1 A and a collector-emitter voltage of at least 60 V; it is preferable to use instances with a large value of h2ia-Output transistors can be 2SA1302, 2SC3281, preferably group O (with a large value of parameter h213). It is desirable to select complementary pairs of transistors of all stages according to a close value of h213. Transistors of the "parallel" follower are best used from the same batch, the same applies to current mirror transistors. When selecting radioelements, one can be guided by the recommendations set forth in [3] (No. 1, pp. 18-20). UMZCH nutrition may be unstabilized. Installation of a common wire and power supply is carried out according to well-known rules. We only note that the elements C1-C5, R2, VD3-VD6 and the screen of the cable connecting the amplifier input to the volume control are assigned to the input local "ground".
Setting up and measuring parameters Before turning on for the first time, the fusible links in the power circuits are replaced with resistors with a resistance of 22 ... 33 Ohms and a power of 5 W, and the trimming resistor sliders are set to the middle position (for the resistor R37 - to the position of maximum resistance). The load is off, the input is closed. Slowly increasing the supply voltage, control the current consumption in both power circuits; it should not exceed 0,15 A. Bringing the voltage on the capacitors C14, C15 to +/-18 V, check the voltages indicated in the diagram: the diodes VD3, VD4 should be 1,5 ... 1,7 V each; on zener diodes VD5, VD6 - 7,4 ... 7,6 V each. The output voltage must be within ± 0,3 V, and the currents consumed from power sources must be the same. By increasing the supply voltage to +/-25 V (at C14, C15), the indicated voltages and current consumption are again checked. By controlling the output voltage with an oscilloscope, they are convinced that the amplifier is not self-excited. Then set the minimum constant voltage at the output trimmer resistor R15. Next, set the quiescent current of the output stage with a tuning resistor R37, if necessary, select R36. By controlling the output voltage with a millivoltmeter, the input is opened and the trimming resistor R6 sets the same voltage at the output as before opening. Then, closing the input again, minimize the bias voltage at the output with resistor R15 as accurately as possible. Having opened the input, they again check the voltage at the output and, if necessary, bring it to zero with resistor R6. On test signals - a sinusoid and a meander with a frequency of 1 kHz - they check the absence of self-excitation at various amplitudes, up to limitation. Three types of self-excitation are possible (for example, due to the use of other types of transistors). The first, as a rule, is associated with an excessive phase shift in the common OOS loop, which is eliminated by an increase in the capacitance of capacitors C10 and C11; in this case, it is necessary to take into account the corresponding decrease in the frequency of the first pole in the CNF loop and the maximum rate of voltage rise at the output. The second is due to a phase shift in the OOS loop of the quiescent current stabilization unit; it is reduced by reducing the resistance of the resistor R38. The third type is parametric excitation in the quiescent current stabilization unit, which is clearly visible at the output in the absence of a signal (in this case, a current of up to several amperes flows through the output stage if there are no current-limiting resistors in the power circuits). It is eliminated by increasing the resistance R38. As you can see, the requirements for this resistor are contradictory, therefore (if necessary) to determine the optimal resistance, you need to find its upper and lower limits, at which self-excitation does not yet occur, and calculate the optimal value as the arithmetic mean. You can use a tuning resistor for this procedure if you solder it directly to the board, without wires, so that parasitic connections and inductances do not distort the result. The ratio of the found upper and lower bounds must be greater than 3 in order to provide a sufficient margin of stability. Otherwise, it will be necessary to replace transistors VT11, VT12 with other types. Another way is to increase the capacitance of the capacitor C13, but this is undesirable, since it reduces the speed of the quiescent current stabilization unit. Now you can install fusible links and connect a load equivalent - a 4 ohm 50 W resistor. Again check the absence of self-excitation on the test signals. Lastly, if it is possible to use a spectrum analyzer, the trimming resistor R30 minimizes the level of the second harmonic when a test signal with a frequency of 1 kHz and a load power of 40 W is applied to the input. If at the same time a voltage offset appears at the output (in the absence of a signal), then you need to minimize it again using R15. In extreme cases, harmonic tuning can be omitted by excluding resistors R30, R31 and setting R26 of the same rating as R27 After tuning, the amplifier has the following parameters. With an input voltage of 1 V, the output power at a load with an impedance of 4 ohms (with a phase shift of up to 60 degrees) is 50 watts. Output voltage slew rate - not less than 100 V/µs. The level of harmonic distortion in the frequency band 10 Hz ... 22 kHz with an output power of 40 W at a load of 4 ohms - no more than 0,02%, with an output power of 20 W at a load of 8 ohms - no more than 0,016%. The level of intermodulation distortion (frequencies 19 and 20 kHz in an amplitude ratio of 1:1) at a peak output power of 40 W at a load of 4 ohms is 0,01%, at a peak output power of 20 W at a load of 8 ohms - 0,008%. The noise level, weighted according to the IEC-A characteristic, with a signal source resistance of 0,13 and 26 kOhm is slightly different - -101, -89, -85 dB, respectively. Suppression of supply voltage ripples (more than +/-17 V) at a frequency of 100 Hz - at least 70 dB. The first pole in the common OOS loop with a load resistance of 4 ohms is at a frequency of 20 kHz. The margin of stability of the overall OOS modulo with a load resistance of at least 2 ohms is more than 12 dB. Figures 4 and 5 show the total harmonic distortion (THD) and even (EVEN) and odd (ODD) harmonic distortion versus output power at 1 kHz with a load impedance of 4 and 8 ohms, respectively, in fig. 6 and 7 - the same, on the frequency at an output power of 40 W at a load of 4 ohms and 20 W at a load of 8 ohms.
Nonlinearity measurements were carried out with a signal source resistance of 13 kΩ, so the measurement results also take into account the input non-linearity (in fact, it is much less than the total). The signal source resistance of 13 and 26 kΩ corresponds to the middle position of the volume control slider with a nominal resistance of 50 and 100 kΩ, respectively. When the supply voltage is turned on and off, the transient process in the UMZCH is insignificant, so the speakers can be connected without a turn-on delay unit. In the author's design with an unstabilized power supply, the amplitude of this process when turned on does not exceed ±40 mV for about 20 ms, and when turned off, it does not exceed ±60 mV for up to several seconds. The suppression of the supply voltage ripple can be increased by replacing the parametric stabilizers with low-noise integrated ones [3] on LM317, LM337 and setting the stabilization voltage to 7,5 ± 0,1 V. The quiescent current of the output stage is chosen somewhat overestimated in order to obtain a stable low non-linearity and the absence of switching distortions, as well as in order to reduce the so-called format distortions (FI). The essence of FI lies in the non-monotonic nonlinearity of the transfer characteristic, i.e., in different sections of the characteristic, it is described by different functions or the function has different parameters. As a result, the signal, shifted along the transfer characteristic by oscillations of the low-frequency component, changes its spectrum of harmonics and intermodulation; when the signal amplitude changes, the harmonic envelope does not correspond to the signal envelope, which can be distinguished by hearing as changes in the fine structure of the sound. Comparative measurements of the dynamic thermal stability of the quiescent current of the output stage, carried out in the described UMZCH and an amplifier with a stage according to the scheme of Fig. 1, ceteris paribus (modes and components) showed its improvement by three to four times. The best result, as noted above, can be obtained by using more high-current diodes. Dynamic thermal stability was determined by comparing the instantaneous value of the quiescent current before and after a short (up to 1 s) pulsed impact on the output stage by the load current. About lowering the bandwidth limit The power amplifier can be used without an isolation capacitor at the input, thus obtaining a bandwidth limit from zero hertz (another idea by E. Alyoshin in relation to the entire audio path). In this case, to improve the stability of zero at the output, it is advisable to use servo control - DC feedback. A possible scheme of such a device in an amplifier is shown in Fig. 8; this is a variant of the implementation of a non-linear direct current feedback [11, 12] with a linear section near the zero of the transfer characteristic. The first stage on the op-amp DA1.1 amplifies the voltage from the output of the UMZCH and symmetrically limits it, and for small signal amplitudes the stage is almost linear. The second one - on the op-amp DA1.2 - is an integrator, from the output of which the current through the resistors R5, R6 is fed to the summation points of the currents of the general OOS of the power amplifier. Transistors VT1, VT2 form a stabilized supply voltage for the op-amp (+/-6,8 V). If integrated stabilizers are installed in the UMZCH (see above), these transistors can be eliminated by supplying power to the op-amp from the stabilizers through resistors (10 ohms, 0,125 W).
Op-amps can be any with field-effect transistors at the input, supply voltage from +/-6,5 V, providing an output current of at least 3 mA for DA1.1 and 30 mA for DA1.2. Transistors - any medium power, with h21E more than 60. If they are in the TO-220 package, then a heat sink is not needed, and if smaller, then a heat sink is needed for each, capable of effectively dissipating 0,6 W. Schottky diodes - any low-power with a minimum forward voltage (less than 0,4 V at 2 mA), having a junction capacitance of less than 100 pF at a reverse voltage of 1 V. Capacitor C1 - film (polyethylene terephthalate), the rest - ceramic with an X7R dielectric and a rated voltage of 25 B (or 50). The tuning resistor can be any small-sized one, but it is more reliable to use a multi-turn one. Setting up a non-linear OOS node via a FET, connected to an established UMZCH, comes down to setting zero at the amplifier output when a tone signal is applied to its input - a sinusoid with a frequency of 1 kHz - with an amplitude several volts less than the output limit voltage. More precisely, you need to set the same voltage as in the absence of a signal (a few millivolts). A load (equivalent) must be connected. The output voltage is measured with a DC millivoltmeter connected to the output through a low-pass filter (R = 10 kOhm, C = 1 μF). The test signal should not contain more than 1% even harmonics. The tuning process can be accelerated by temporarily reducing the capacitance of the capacitor C1 to 0,1 uF. According to available information, in particular from [13], such a node can improve the sound quality of recordings made on equipment with a lower bandwidth limit significantly higher than 0,02 Hz. Apparently, this happens due to the "cutting" of relatively slow parasitic signal shifts in the recording that occur in differentiating circuits (for example, an interstage capacitor) when a pulse signal passes through them, which is sound (musical) information in the electronic path [12] - see below. To do this, the integration constant in the cascade on the DA1.2 should be small enough, but not so small as to noticeably reduce the low-frequency content in the reproduced sound at low volume. For the scheme in fig. 8, this corresponds to a capacitance C1 of the order of 0,1 µF. Repeaters of this node should experiment by changing the constant of integration at different volume levels. The idea of "0 Hz", or more precisely "almost 0 Hz", as the frequency boundary of the sound path band from the microphone to the speakers, implies the rejection of the commonly used circuits that differentiate low-frequency and infra-low-frequency signals - interstage capacitors and integrators in the OOS circuit, which are from practical considerations have relatively small values of the time constant. As a result of the use of such filters, linear distortions are introduced into a non-stationary signal (sound, music), which have a negative impact on the subjective perception of the reproduced sound. On fig. Figure 9 shows how a symmetrical non-stationary signal changes when passing through six first-order differentiating circuits (thick line) having a cutoff frequency an order of magnitude lower than the frequency of the first period of signal oscillation. The exponential section of the transient process is shown by a dashed line. Distortions arise due to the leading phase shift created by the filter in the LF region, which leads to "blurring" of the sound attack [14]. That is, the envelope of sound vibrations is distorted, to which the sensitivity of hearing increases with decreasing frequency, since time factors prevail in the analysis of the signal in the auditory system in the LF region. The phase shift between the harmonic components of the sound can also change the perception of the timbre [15]. In this case, the signal amplitude increases, which increases its dynamic range by several decibels and, accordingly, reduces the dynamic range of the path by the same value, which is the greater, the higher the cutoff frequency of the HPF in relation to the signal frequency. In the limit, the amplitude increase is +6 dB on a square wave (in reality it is always less) Another consequence of the advanced phase shift affects the quality of sound reproduction indirectly. It lies in the fact that the phase shift and the change in the amplitude of the LF and LF components lead to fluctuations in the signal center line relative to zero. The dotted line in fig. 9 shows the "sliding" of the middle line, which was not in the original signal.
To understand the connection of this "slip" with the deterioration of the sound, it is necessary to take into account that the transfer characteristic of amplifying stages, especially the power amplifier, is not only non-linear, but, as a rule, has a non-monotonic non-linearity (i.e., FI takes place). This means that the signal, being "sliding" along the transfer characteristic, has a changing spectrum of harmonics and intermodulation, i.e., the non-linearity with respect to the signal becomes non-stationary. The latter circumstance, according to the observations of the author of the idea E. Alyoshin, significantly degrades the sound quality, preventing hearing from adapting to the nonlinearity of the path Another negative consequence of the "slip" of the signal is manifested during electroacoustic conversion. When such a "sliding" signal is reproduced by a sound emitting head, a shift in the sound spectrum occurs due to the Doppler effect. When playing a real sound signal, this causes additional frequency modulation (detonation) of the sound, which, as is known, also worsens the subjective quality of sound reproduction. References:1. Sukhov N. UMZCH high fidelity. - Radio, 1989, No. 6, p. 55-57; No. 7, p. 57-61.
Publication: radioradar.net
Machine for thinning flowers in gardens
02.05.2024 Advanced Infrared Microscope
02.05.2024 Air trap for insects
01.05.2024
▪ Smart dressing for the treatment of chronic wounds ▪ Underwater volcanoes can warm entire continents ▪ Mercedes electric cars with wireless charging ▪ Measured the electronic work of the brain
▪ section of the site Intercoms. Article selection ▪ Visakhadatta article. Famous aphorisms ▪ article How did the first lamps appear? Detailed answer ▪ article Boat for every day and vacation. Travel Tips
Comments on the article: chrism There are references to sources [x] in the text, but there are no sources themselves. Diagram 2mir Thanks, fixed.
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