Simple high quality umzch. Ultra-linear high-end UMF with transistors (80W) High-quality UMF
A transistor UMZCH with a differential stage (DC) at the input is traditionally built according to three cascade schemes: DC input voltage amplifier; voltage amplifier; output two-stroke current amplifier. In this case, it is the output stage that makes the greatest contribution to the distortion spectrum. These are, first of all, "step" type distortions, switching distortions aggravated by the presence of resistances in the emitter (source) circuits, as well as thermal distortions, which until recently were not given due attention. All these distortions, being phase-shifted in negative feedback circuits, contribute to the formation of a wide range of harmonics (up to 11th). This is what determines the characteristic transistor sound in a number of unsuccessful developments.
For all cascades, a huge set of circuit solutions has been accumulated to date, from simple asymmetric cascades to complex fully symmetrical ones. However, the search for solutions continues. The art of circuitry lies in the fact that simple solutions achieve a good result. One of such successful solutions is published in . The authors note that the mode of operation of the most common output stages with a common collector is set by the voltage at the emitter junctions, which strongly depends on both the collector current and temperature. If in low-power emitter followers it is possible to stabilize the emitter base voltage by stabilizing the collector current, then in high-power class AB output stages this is almost impossible to do.
Thermal stabilization circuits with a temperature-sensitive element (most often a transistor), even when the latter is installed on the case of one of the output transistors, are inertial and can track only the average change in the temperature of the crystal, but not instantaneous, which leads to additional modulation of the output signal. In some cases, thermal stabilization circuits are a source of soft excitation or sub-excitation, which also gives the sound a certain coloration. For a fundamental solution to this problem, the authors proposed to implement the output stage according to the OE scheme (the idea is not new, see for example). As a result, unlike the traditional three-stage construction (each stage with its own cutoff frequency and its own spectrum of harmonics), we got only a two-stage amplifier. Its simplified scheme is shown in Fig.1.
The first stage is made according to the traditional DC scheme with a load in the form of a current mirror. Symmetric pickup of the signal from the DC using a current mirror (counter dynamic load) makes it possible to obtain twice as much amplification while simultaneously reducing noise. The output impedance of the cascade with such a signal pickup is quite high, which determines its operation in the current generator mode. In this case, the current in the load circuit (the base of the transistor VT8 and the emitter of the transistor VT7) depends little on the input resistance and is determined mainly by the internal resistance of the current source. The emitter currents of transistors VT8, VT9 are basic for transistors VT10, VT11. The current generator I2 and the level shift circuit on transistors VT5 VT7 set and stabilize the initial current of transistors VT8 VT11, regardless of their temperature.
Let us consider in more detail the operation of the current control circuit of the output transistors. Transitions base emitter transistors VT5 VT8 form two parallel circuits between the output of the current source I2 and the base of the transistor VT10. This is nothing more than a complex scale current reflector. The principle of operation of the simplest current reflector is based on the fact that a specific value of the collector (emitter) current corresponds to a well-defined voltage drop at its base-emitter junction and vice versa, i.e. if this voltage is applied to the base-emitter junction of another transistor with the same parameters, then its collector current will be equal to the collector current of the first transistor. The right circuit (VT7, VT8) consists of base-emitter junctions with different collector (emitter) currents. For the "current reflector" principle to work, the left circuit must be mirrored with respect to the right, i.e. contain identical elements. In order for the collector current of transistor VT6 (aka current generator current I2) to match the collector current of transistor VT8, the voltage drop at the base-emitter junction of transistor VT5, in turn, must be equal to the voltage drop at the base-emitter junction of transistor VT7.
To do this, in a real circuit (Fig. 2), the VT5 transistor is replaced by a composite transistor according to the Shiklai scheme. Based on the foregoing, the following conditions are met:
- static current transfer coefficients of transistors VT7, VT8, VT11 (VT12) must be equal;
- static current transfer coefficients of transistors VT9 and VT10 should also be equal to each other, and even better if all 6 transistors (VT7 VT12) have the same characteristics, which is difficult to do with a limited number of transistors available;
- as transistors VT8, VT9, it is necessary to select transistors with a minimum base-emitter voltage (taking into account the spread of parameters), since these transistors operate at a reduced emitter-collector voltage;
- products of static current transfer coefficients of transistors VT11, VT13 and VT12, VT14 should also be close.
Thus, if we want to set the collector current of transistors VT13, VT14 to 100 mA and have output transistors with h21e=25, then the current generator current on transistor VT6 should be: Ik(VT6)/h21e=100/25=4 mA, which and determines the resistance of the resistor R11 about 150 ohms (0.6 V / 0.004 A = 150 ohms).
Since the output stage is controlled by the output current of the DC, the total emitter bias current is chosen to be large enough, about 6 mA (determined by resistor R6), it also determines the maximum possible output current of the DC. From here you can calculate the maximum output current of the amplifier. For example, if the current gain product of the output transistors is 1000, then the maximum output current of the amplifier will be close to 6 A. For the declared maximum output current of 15 A, the current gain of the output stage must be at least 2500, which is quite realistic. Moreover, in order to increase the load capacity of the DC, the total emitter bias current can be increased to 10 mA by reducing the resistance of the resistor R6 to 62 ohms.
The following amplifier specifications:
- Output power in the band up to 40 kHz at a load of 8 ohms - 40 watts.
- Pulse power at a load of 2 ohms - 200 watts.
- The amplitude value of the undistorted output current is 15 A.
- Harmonic coefficient at a frequency of 1 kHz (1 W and 30 W, Fig. 3) - 0.01%
- Output voltage slew rate - 6 V/μs
- Damping coefficient, not less than - 250
The plot of the harmonic coefficient at an output power of 1 W (curve a) and at an output power of 30 W (curve b) at a load of 8 ohms is shown in Fig. 3. The comments on the circuit state that the amplifier has high stability, there are no "switching distortions", as well as higher-order harmonics.
Before assembling a prototype amplifier, the circuit was mocked up virtually and studied using the Multisim 2001 program. Since the output transistors indicated in the circuit were not found in the program database, they were replaced by the closest analogues domestic transistors KT818, KT819. Studies of the circuit (Fig. 4) gave results somewhat different from those given in. The load capacity of the amplifier turned out to be lower than the declared one, and the harmonic coefficient was more than an order of magnitude worse. The phase safety factor of only 25° was also insufficient. The slope of the frequency response in the region of 0 dB is close to 12 dB / oct., which also indicates the lack of stability of the amplifier.
For the purpose of experimental verification, an amplifier layout was assembled and installed in the guitar amp of the rock group "Aphasia". To increase the stability of the amplifier, the correction capacitance is increased to 2.2 nF. Field tests of the amplifier in comparison with other amplifiers confirmed its merits and the amplifier was highly appreciated by musicians.
Technical parameters of the amplifier
- Bandwidth at 3dB-15Hz-190kHz
- Harmonic distortion at 1 kHz (25 W, 8 ohms) - 0.366%
- Unity gain frequency - 3.5 MHz
- Phase Margin - 25°
Strictly speaking, the above arguments about the current control of the output stage are valid for an amplifier with an open feedback loop. With a closed CNF, in accordance with its depth, not only the output impedance of the amplifier as a whole, but also of all its cascades decreases, i.e. they essentially begin to work as voltage generators. 
Therefore, in order to obtain the declared specifications the amplifier was modified to the form of Fig. 5, and the result of its study is shown in Fig. 6. As can be seen from the figure, only two transistors have been added to the circuit, which form a push-pull hybrid class A repeater. The introduction of a buffer stage with a high load capacity made it possible to more effectively use the voltage amplifying properties of the DC and significantly increase the load capacity of the amplifier as a whole. An increase in gain with a broken OOS had a positive effect on a decrease in the harmonic distortion coefficient.
Increasing the correction capacitance from 1 nF to 2.2 nF, although it narrowed the bandwidth from above to 100 kHz, but increased the phase margin by 30 ° and ensured a frequency response slope in the unity gain region of 6 dB / oct., which guarantees good stability of the amplifier.
A meander-type signal with a frequency of 1 kHz (calibration signal from an oscilloscope) was applied to the amplifier input as a test signal. The output signal of the amplifier had neither a rollover nor spikes at the signal fronts, i.e. fully consistent with the input.
Specifications of the modified amplifier
- 3 dB bandwidth - 8 Hz - 100 kHz
- Unity gain frequency - 2.5 MHz Phase margin - 55°
- Gain - 30 dB
- Harmonic coefficient at a frequency of 1 kHz (25 W, 8 Ohm) - 0.007%
- Harmonic coefficient at a frequency of 1 kHz (50 W, 4 Ohm) - 0.017%
- Harmonic coefficient at Ku=20 dB - 0.01%
For the purpose of full-scale testing of the modified amplifier, two samples were made in the dimensions of the Lorta 50U 202S amplifier board (aka Amfiton 001) and installed in the indicated amplifier. At the same time, the volume control was finalized in accordance with.
As a result of refinement, the owner of the amplifier completely abandoned the tone control, and full-scale tests showed its clear advantage over the previous amplifier. The sound of the instruments has become more pure and natural, apparent sound sources (SIS) have become more clearly formed, they have become, as it were, more "tangible". Noticeably increased and undistorted output power amplifier. The thermal stability of the amplifier exceeded all expectations. After a two-hour test of the amplifier at an output power close to the maximum, the side heat sinks turned out to be practically cold, while with the previous amplifiers, even in the absence of a signal, the amplifier, being left on, warmed up quite strongly.
Construction and details
The board (with elements in the light) of the amplifier intended for installation in the Lorta amplifier is shown in Fig. 7. The board provides places for installing a diode bridge and resistor R43 from the old circuit, as well as places for installing current equalizing base and emitter resistors for paired output transistors. At the bottom of the board, places are reserved for installing elements of an active current source (AIT) in the form of a current reflector, consisting of a current-setting resistor with a resistance of 75 kOhm from the PA output, two KT3102B transistors and two 200 Ohm resistors to actively turn off the lower arm of the amplifier (in the prototype was not installed). Capacitors C4, C6 type K73 17. The capacitance of capacitor C2 can be painlessly increased to 1 nF, while the cutoff frequency of the input low-pass filter will be 160 kHz.
Transistors VT13, VT14 are equipped with small aluminum flags 2 mm thick. Transistors VT8 and VT12 for better thermal stabilization of the amplifier are installed on both sides of the common flag, and the transistor VT8 through a mica gasket or an elastic heat-conducting insulator of the "Nomakon Gs" type TU RB 14576608.003 96. As for the parameters of the transistors, they are discussed in detail above. As transistors VT1, VT5, you can use transistors KT503E, and instead of transistors VT2, VT3 transistors like KT3107 with any letter index. It is desirable that the static current gains of the transistors be equal in pairs with a spread of no more than 5%, and the gains of transistors VT2, VT4 be slightly greater than or equal to the gains of transistors VT1, VT5.
As transistors VT3, VT6, you can use transistors of the types KT815G, KT6117A, KT503E, KT605. Transistors VT8, VT12 can be replaced by transistors of the KT626V type. In this case, the transistor VT12 is attached to the box, the atransistor VT8 to the transistor VT12. Under the head of the screw on the side of the transistor VT8, place a texto-lithic washer. As a transistor VT10 from domestic field-effect transistors, a transistor of the type KP302A, 2P302A, KP307B (V), 2P307B (V) is best suited. It is advisable to select transistors with an initial drain current of 7-12 mA and a cut-off voltage in the range of (0.8-1.2) V. Resistor R15 of type SP3 38b. Transistors VT15, VT16 can be replaced, respectively, KT837 and KT805, as well as KT864 and KT865 with higher frequency characteristics. The board was developed for the installation of paired output transistors (KT805, KT837). For this purpose, the board provides places for installing both basic (2.2-4.3 Ohm) and emitter (0.2-0.4 Ohm) current equalizing resistors. In the case of installing single output transistors, instead of current equalizing resistors, solder the jumpers or immediately solder the wires of the output transistors to the appropriate places on the board. The "native" output transistors were left in the experimental sample, only they had to be swapped.
In the amplifier, it is desirable to increase the power capacitances (in the original amplifier, 2.2200 uF.50 V in each arm). At a minimum, it is desirable to add another 2200 uF to each arm, and even better, replace it with a 10000 uF capacitor. 50 V. At 50 V, foreign capacitors are relatively cheap.
Establishment
Before connecting the output transistors, it is necessary to temporarily solder any diodes in place of the base-emitter junctions of the output transistors. medium power(for example, KD105, KD106), apply power to the board and, without connecting the load, make sure that the amplifier is working middle point. Apply a signal to the input of the amplifier and check with an oscilloscope that it is amplified without distortion and excitation at idle. This indicates the correct installation and serviceability of all elements of the amplifier. Only after that you can solder the output transistors and proceed to setting their quiescent current.
To set the quiescent current, it is necessary to set the slider of the resistor R15 to the lower position according to the diagram, remove the fuse in one of the arms of the amplifier and turn on the ammeter instead. The current consumption is set under the trimmer resistor R15 in the range of 110-130 mA (taking into account the DC current of about 6 mA and the buffer follower current of about 3-5 mA). Then the sensitivity of the amplifiers is checked and, if necessary, the OS resistors are adjusted.
After that, you can proceed to various studies, if, of course, the equipment of the radio amateur laboratory allows. For this purpose, you can use the direct input of the amplifier by removing the jumper plug from it on the back of the amplifier.
Literature
- Digest UMZCH//Radiohobby. 2000. No. 1. S.8 10.
- Petrov A. Superlinear EP with high load capacity//Radioamator. 2002. No. 4. C.16.3.
- Dorofeev M. Mode B in AF power amplifiers//Radio. 1991. No. 3. S.53 56.
- Petrov A. Refinement of the volume control of the amplifier "Lorta 50U 202S"//Radioamator. 2000. No. 3. p.10
given in recent times preference for tube output power amplifiers audio frequency for sound reproduction of high fidelity, it is difficult to understand, based on their objective comparison with transistor UMZCH. Indeed, in all measured characteristics, the modern UMZCH on transistors is significantly superior to the lamp one. In our opinion, the usually measured non-linear distortions (NI) do not exhaust those distortions that determine the quality of sound reproduction.
In the most advanced designs of transistorized UMZCH, the NI level is brought almost to the hearing threshold and even lower, so it is doubtful that they can be perceived by ear, especially when masked by a useful signal.
The point, apparently, is that NI is usually measured in the steady state, when the transient process after applying the measuring signal to the input of the amplifier under test is already completed both at the input and at the output of the amplifier, and in a closed loop of a common negative feedback (OOS ) a stationary oscillatory process is established, which responds with greater or lesser accuracy to the input signal.
Obviously, the nonlinearity of the amplifier manifests itself much more strongly during the transient process (the duration of which can be significant due to the signal delay in the CNF circuit), especially at its initial stage, when the CNF action is the least effective (due to the mentioned delay).
In contrast to dynamic distortions, which lead to an overload of the input stage throughout the entire duration of an input signal that is unfavorable in terms of parameters, the transient NIs under consideration are present even when there are no dynamic ones, but only until the transient process is completed.
And if we take into account that real sound programs are very far from stationarity and actually cause an almost continuous transient process in the UMZCH, then when playing such programs, the NI can be much higher than those measured by conventional methods in the same instance of the amplifier.
Due to the short duration of the transient process compared to the time of laboratory measurements, they still "escape" from experimental study (this requires the development of special methods) and at the same time are easily perceived by ear throughout the entire soundtrack.
From this point of view, the advantage of tube amplifiers becomes clear: although they have a higher measured NI level (this applies only to stationary mode), in real conditions, lamps, as much more linear devices, provide lower NI than transistors (although, of course, greater than those the same tubes in stationary mode), which determines the best sound of tube amplifiers.
However, such shortcomings of tube amplifiers as inconvenience in operation, bulkiness and large mass, significant power consumption at relatively low efficiency and output power are obvious.
In this regard, it would be tempting to create a transistor amplifier with a real NI level no worse than that of a tube amplifier. The latter means that the NI level of such an amplifier, measured by conventional methods, must be reduced by one or two orders of magnitude (!) in comparison with the best samples (more can be), so that the NI in the nonstationary mode has an acceptable value.
However, the linearization methods currently used transistor amplifiers, apparently, have already exhausted themselves and will not allow reaching the required NI coefficient (0 = 0.0001 ... 0.00001%).
Therefore, the task was to study the possibility of obtaining such a record low level own NI transistor UMZCH, without stopping at the complexity of circuit solutions, and then decide whether such an approach is justified, whether it brings a gain in sound quality compared to existing circuits.
The design presented in this work is addressed primarily to the most demanding connoisseurs of high-quality sound reproduction. It is developed on the basis of the principle set out in, which is an improvement on the well-known distortion reduction method described in.
Rice. 1-3. Block diagrams of amplifiers.
Figure 1 shows a block diagram of a two-stage amplifier with the transfer function of the first stage K1 and the second K2, the transfer function b of the general feedback circuit covering the entire amplifier, and the transfer function g of the local positive feedback circuit (MPOS) covering the first stage. The resulting transfer function of such a device is described by the expression К=К1К2/(1-тК1+рК1К2). (1)
If we set the gain in the MPOS loop mK1 = 1, then it turns out that, unlike an amplifier with one OOS, in which K \u003d K1K2 / (1 + |ZK1K2) and only approximately K \u003d 1 / p (at |ZK1K2 "1), the transfer function of this amplifier will be exactly 1/p.
In this case, the depth of the FOS should be greater than the depth of the MES, i.e. |ЗК1К2>уК1, which is a necessary (but not sufficient) condition for stability. Thus, when yK1=1, all distortions that occur in the second stage and are caused by the variability of its transfer function are suppressed (since K=1/|3 and does not depend on K2).
However, absolutely complete suppression of distortion is possible only with an ideal first stage. In reality, both nonlinear and frequency distortions are inherent in it, leading to a deviation of the transfer function K1 from the optimal value. In addition, it changes due to fluctuations in supply voltages, temperature drift, and changes in part parameters over time.
The problem is also to ensure the joint stability of such a complex system with the combined action of the FOS and the SSP (the second condition for stability), since the introduction of the SSP reduces the stability margin of the original system.
On the other hand, it is desirable (to obtain the greatest linearity) that the depth of both the FOS and the FOS be constant in the operating frequency range, i.e. so that the first pole of the frequency response of the system with open feedback was at a frequency f>20...30 kHz, and the cutoff frequency in the POS loop was also no less.
Meanwhile, it is not at all easy to fulfill the latter requirements and at the same time ensure a reliable margin of stability, and deviation from them significantly reduces the effectiveness of the method. Apparently, therefore, the author does not know examples of using the described principle of distortion suppression for the purposes of high-quality sound reproduction.
The fundamental disadvantage of the device shown in Fig. 1 is, as the analysis shows, that the MFOS loop is connected in series to the OOS circuit. The operation of the device can be significantly improved by connecting the MPOS loop to the OOS loop in parallel, i.e. by connecting the input of the second stage not to the output of the first stage (point 2 of Fig. 1), but to its input (point 1).
The block diagram of the device proposed in is shown in Fig.2. The most important advantage of such a device is the smaller phase shift introduced into the OOS loop by the elements of the MPOS circuit (from the input of the device to the input of the second stage).
This is clear from a comparison of Fig. 2 with Fig. 1, since it is obvious that the phase of the signal at point 2 lags behind the phase at point 1 (Fig. 1) by the phase shift introduced by the first stage (and this shift can be quite significant at frequencies 0.2 ... 1 MHz and higher, in the area of which the stability of the device must be ensured).
This advantage is decisive for the use of this method of distortion compensation in high-quality UMZCH, since the minimum phase shifts introduced during its use make it possible to obtain a sufficient margin of stability and thereby ensure reliable operation of the amplifier with MPOS.
The advantage of the device shown in Fig. 2 is also the possibility of a more independent (although this independence is relative, since the loops still interact with each other) and optimal choice parameters of the MPOS and OOS loops in accordance with their functional purpose, which is significantly different.
This greater independence is evident from the expression for the transfer function of the improved system K = K2/(1 -7KI + |3K2), (2) which, unlike (1), does not contain mixed products of the transfer functions of elements related to different loops.
Such a separation is impossible in the device shown in Fig. 1, where the first stage is a common part of the MOS and FOS loops, as a result of which its parameters determine both the FOS properties and the FOS properties at the same time. The requirements for these parameters are largely contradictory, which also makes it difficult to solve the problem of maximum distortion suppression.
The advantages of parallel connection of the MFOS loop to the NFB loop make it possible to practically implement a device with not even one, but with two MFOS, mutually reinforcing each other's action and thereby improving distortion compensation. A block diagram of such a device is shown in Fig. 3, where K1, K2, KZ are the transfer functions of the three stages of the main amplifier channel; c - transfer function of the OOS circuit; а1у1 and а2у2 are the transfer functions of the first and second loops of the MPOS, respectively, and the equalities а1у1=1 and а2у2=1 are set with the greatest possible accuracy. From its transfer function К = К1К2К3/[(1-а1у1)(1-а2у2)+рК1К2К3] (3) it follows that since 1-а1у1<<1, то степень подавления искажений, зависящая от выражения (1-а1у1)(1-а2у2), значительно больше, чем в устройстве с одной петлей МПОС, в котором эта степень определяется одним членом 1 -а1у1<<(1-а1у1)(1-а2у2).
However, the most remarkable thing is that with one MOS, the minimum achievable level of NI cannot be made less than the distortions introduced by the elements of the MOS loop itself, and in a device with two (or more) MOS loops, as calculation shows, the own NI of each MOS loop is suppressed by the action of the other, those. it is possible to reduce the NI below the level determined by the most linear unit of the device, which should be the MOS circuit.
This is a significant advantage of this distortion compensation method over others, which make it possible to reduce distortion only to the limit determined by the intrinsic nonlinearity of the compensation circuit.
Note that all of the above fully applies to those distortions that are due to the variability of the transfer functions (except for nonlinear ones, for example, amplitude-frequency ones). Such distortions are compensated in any parts of the device, except for the feedback circuit b.
It can be shown that these distortions are compensated if they occur in the parts of the device located between the MOS loop and the device output, including the output itself, and those that occur between the device input and the MOS loop are not compensated. Therefore, the noise level of the device shown in Fig. 3 is determined mainly by the noise properties of the front end.
Power Amplifier Specifications
- Rated input voltage 0.3V;
- Rated output power at a load of 8 ohms (4 ohms) - 40 (80) W;
- The frequency range with blockages at the edges is not more than 0.5 dB - 15-100000 Hz;
- Input resistance - 50 kOhm;
- Output impedance - 0 Ohm;
- (with MPOS circuits) Intermodulation distortion factor, no more than 0.005%;
- Noise level (weighted) -105 dB (with MOS loops).
Schematic diagram of UMZCH
Schematic diagram of the UMZCH, corresponding to Fig. 3, is shown in Fig. 4. To obtain the lowest possible level of NI, the main channel of the amplifier (without MOS) is conceived as a fairly linear UMZN.
Rice. 4. Schematic diagram of a transistor LF power amplifier for 80W Hi-End class.
To do this, all amplifier stages are made push-pull on complementary pairs of transistors, which made it possible to make both arms symmetrical with respect to the common wire and obtain a more linear amplitude characteristic.
All transistors operate in mode A, with the exception of the output stage with a floating input bias (super-A), which is set by the circuit on the elements VT15-VT18, R38-R41, VD15, VD16. This provides a non-switching off mode of operation of the terminal transistors with their low quiescent current.
The input stage is made according to the cascade scheme (VT1, VT3, VT2, VT4). The operation mode of its transistors is chosen so that they do not enter the cutoff or current limiting mode when signals with an amplitude several times higher than the rated input voltage are applied at the input, even when the OOS is turned off.
This compares favorably with the traditional differential cascade. The R19, R18, C7 chain with a cutoff frequency of 90 kHz limits the amplification of the highest frequency components of the pulse signals, preventing overload and subsequent amplifier stages.
Thanks to these measures, as well as high speed due to the rejection of the use of transistors with a common emitter in cascades and advance correction (capacitors C5, C6), there are no dynamic distortions in the amplifier, which is especially important for the stable operation of the system with POS.
The OOS voltage from the output of the amplifier is fed to the connection point of the resistors R11 and R12, which together with R10 and R13 determine the operating current VT1 and VT2. At the same time, R10 and R13 as part of the dividers R14 / R10C3 and R15 / R13C4 set the transfer function of the OOS circuit.
The constant component of the output voltage is supplied to the emitters of the input transistors through R10R11 and R12R13, and not only through R14 and R15, therefore, the depth of the CNF for direct voltage is much greater than for alternating voltage, and the direct voltage component is rigidly stabilized at the output of the UMZCH.
The use of electrolytic capacitors C3, C4 does not lead, as follows from the measurements, to a significant increase in distortion, since they are polarized by a constant voltage of about 4 V (the variable component is much less), so their mode of operation is almost linear.
The second stage on transistors VT5-VT8, connected according to the OK-OB scheme, is a buffer between two MPOS circuits. Diodes VD3-VD6 set the bias voltage at the bases of emitter followers VT9, VT10, and diodes VD7, VD8 protect it from too much increase in case of malfunctions in the amplifier or blown one of the fuses.
The voltage amplifier (VT11, VT13 VT12, VT14) is also made according to the cascode scheme. The supply voltage of the first stages is about 21 V and is set by the stabilizer (VT23, VT24, VD17, VD18). The output transistors operate with a low quiescent current, so they do not require thermal stabilization.
The frequency correction elements R19R18C7, R27C10, R22C8, R23C9 form the frequency response of the amplifier, ensuring its stability under the action of OOC. At the same time, R19 and R27 serve as the load of the input and buffer stages, respectively, as well as the load of the MPOS loops, determining their gain.
Field-effect transistors are used in the MPOS circuits to minimize their own distortion of the circuits. Each MPOS circuit is an amplifying stage with a transfer coefficient of about one, which can be changed by trimming resistors R58 and R67.
By directly connecting the output of the cascade to its input, 100% PIC is performed. The R57C15 and R66C16 chains correct the frequency response of the cascades, improving the compensation accuracy at frequencies in the audio range. The MPOS circuits are connected to the main channel at the nodal points A, B and to the common wire.
The operating points of the transistors of the first stages and circuits of the MPOS are rigidly stabilized by high-resistance resistors in their emitter (source) circuits. This achieves the constancy of the characteristics of the cascades connected to points A and B.
In addition, transistors VT3VT4 and VT27VT28, VT7VT8 and VT31VT32 are a dynamic load for each other, and emitter followers VT5VT6, VT9VT10 and field-effect transistors VT25VT26 and VT29VT30 have a high input impedance, so the load resistance for the MPOS loops is determined by resistors R19, R27 (at audio frequencies ).
Thanks to this, it was possible to achieve high stability of the gain in the MCO loops, which does not depend on temperature and does not change over time.
Setting up the amplifier
Then, with trimmers R7, R20 and R31, set zero voltage at the output of the amplifier and at the nodal points A and B, respectively. Check the total voltage drop across the pairs of diodes VD3VD4, VD5VD6, VD11VD12, VD13VD14, which should be about 2 V. After that, check the quiescent current of the output transistors
VT21, VT22, which should be within 20 ... 30 mA. Its value must be set by selecting resistors R38, R39, in which there are no step-type distortions.
A dummy load with a resistance of 4.8 ohms is connected to the output of the amplifier and the operation of the floating bias circuit of the final stage is checked.
To do this, connect the oscilloscope to the bases VT19 and VT20 and a sinusoidal signal with a frequency of 100 Hz is fed to the input of the amplifier. The oscillogram should be in the form of a pulsating voltage (like a "rectified" sinusoid) with an amplitude of about 5 V at the nominal output voltage and a load resistance of 4 ohms. As the load resistance increases or the input signal decreases, this amplitude should decrease.
Check the passage through the amplifier of rectangular pulses. There should be no surges on the output voltage waveforms, otherwise the capacitance of capacitors C5 and C6 will increase. On this, the setup of the main channel can be considered complete.
Note that already the basic amplifier (without MCO circuits) has the following rather high characteristics (see the beginning of the article).
The MPOS circuits are tuned by connecting them to the circuit and setting the R58, R67 sliders to the position of maximum resistance, i.e. minimum loop gain of the MPOS circuits.
The voltage between the drain and source of the field-effect transistors should be no more than 10 V (the maximum allowable for the KP103 transistor), but not too small, otherwise the desired value is achieved by selecting resistors R51, R52, R60, R61. It is desirable that complementary transistors be matched in pairs with close values of the initial drain current and cutoff voltage.
The input of the amplifier is shorted, an acoustic system (AC) or a measuring device is connected to the output, and the signal from the source (signal generator or source of a musical program rich in low and high frequency components) with a high-impedance output is fed to the nodal point B, simulating a distortion signal.
The common wire of the source is connected to the common wire of the amplifier. By adjusting R58, maximum attenuation of the signal at the output of the amplifier is achieved. The selection of R57C15 improves the suppression of the high-frequency components of the signal spectrum.
Having configured the first circuit of the MCO, it is disconnected from point A, and the source-simulator of distortions - from point B. The output of the simulator is connected in parallel with the resistor R35 and the second circuit of the MCO is configured similarly to the first one. After that, the first circuit of the MPOS is reconnected and additional signal suppression is observed.
At the final stage, a direct check of the suppression of NI in the amplifier is carried out. It is sufficient to measure only the coefficient of intermodulation distortion of the OI, since at its sufficiently small values the coefficient of harmonic distortion is obviously acceptable.
In accordance with the technique, two sinusoidal signals with a frequency of 25-30 kHz and a frequency difference of 1 kHz are fed to the input of the amplifier at the same amplitude, not exceeding half the nominal, and the sound level reproduced by the speaker is estimated.
When the MPOS circuits are disconnected, a very quiet sound can be heard (corresponding to 0I = 0.005%), which completely disappears when they are connected.
For a visual demonstration of NI suppression, you can temporarily increase the non-linearity of the base amplifier by connecting a series-connected diode in the conductive direction (for example, D9) and a 47 kΩ resistor in parallel with resistor R9.
In this case, the RI of the base amplifier increases to approximately 0.5%, the combination frequency becomes clearly distinguishable, and one can more confidently judge its suppression when connecting the MOS circuits.
It follows from such measurements that each of the MOS circuits suppresses distortion by at least 30 dB, and both of them together by almost 60 dB, so that it is impossible to measure the NOR of the entire amplifier by conventional methods due to their extremely small value, but one can only estimate taking into account the OI of the base amplifier, reduced by three orders of magnitude, which gives a fantastic value of 0I = 0.00001%)!
It should be noted one more positive side of the use of MOS in the amplifier. Since when the general OOS is terminated, the gain tends to increase due to the action of the POS, then with signal delays in the OOS circuit, the MOS circuits actually become forcing corrective devices that speed up the processes in the system and reduce the phase shift between the input and output signals. This improves the quality of the transient process, which also helps to reduce distortion.
The subjective impression of the work of this amplifier is difficult to convey in words, you need to hear the purity and transparency of its sound. In this respect, it is not only not inferior to tube amplifiers, but also noticeably surpasses them, without introducing practically anything “from itself” into the sound picture.
The experience of its operation for 5 years has shown the reliability of the design, and periodic checks have shown good tuning stability and maintaining the accuracy of distortion compensation within the specified limits without additional adjustments.
Parts and circuit board
The printed circuit board is designed with common requirements in mind. MPOS blocks on VT25-VT32 transistors are made on two separate small boards and in the form of modules and are fixed perpendicular to the main amplifier board near the nodal points A and B.
Rice. 5-6. Printed circuit boards for a high-quality LF power amplifier circuit.
The amplifier uses MLT-type resistors, trimmers of the SPZ-29M type, capacitors K50-16 (SZ, C4, C11-C14), K73-I7 (C1, C2), KD1, KT1 - the rest. The heat sinks of transistors VT21, VT22 are located near the elements of the floating bias circuit of the final stage to compensate for the temperature instability of the quiescent current of the output transistors.
Printed circuit boards are made of foil textolite. The size of the main channel board (Fig. 5) is 150 x 105 mm, MPOS modules (Fig. 6) is 105 x 30 mm.
After desoldering all the parts, the MPOS modules are installed on the main board along the directions indicated by the arrows in Fig.1. The corresponding printed conductors of the boards are connected according to the circuit diagram using wire jumpers. Common wire busbars can be connected using guy wires that hold the boards in a mutually perpendicular position.
Disabling and connecting the MPOS circuits during configuration is carried out by jumpers between the nodal points A, B and the corresponding points of the MPOS modules.
For a stereo amplifier, the boards of the main channel and MPOS modules have twice the width - not 105, but 210 mm, and two identical patterns are applied to them.
The layout of the amplifier should be given special attention. The wires connecting the amplifier to the power supply should be as short and as large as possible.
This is especially true of the wire connecting the common wire bus of the printed circuit board to the "zero" of the power supply - the connection point of the filter capacitors.
If for some reason the last requirement is not feasible, then it is better not to connect the “ground” terminals of capacitors C13, C14 to a common wire on the board, but, having shorted each other, connect to the “zero” of the power supply with a separate wire. The wires from the acoustic systems are also connected to the same place, as shown in Fig. 7.
Rice. 7. Zero wiring and speaker connection in the amplifier.
The quality of the stereo amplifier layout is easy to check by loading one of its channels with a 4-ohm load equivalent and applying a meander with a frequency of 2000 Hz to the input of this channel, and control is carried out through the speakers of the second channel, the input of which is shorted. With the correct layout of the signal with the frequency of the meander in the speakers should not be.
Literature:
- Matyushkin V.P. - Linear amplifier.
- Design of transistor audio frequency amplifiers - N.L. Bezladnov, B.Ya.Gertsenshtein, V.I. Kozhanov and others - M .: Communication, 1976.
- Kostin V. - Psychoacoustic sound quality criteria and choice of UMZCH parameters. Radio 1987-12.
- Khlypalo E.I. - Calculation and design of non-linear corrective devices in automatic systems, 1982.
Answers Matyushkin V.P. to the questions of those who want to repeat the design of the amplifier
- What is the slew rate of the output voltage? Answer: The slew rate of the output voltage is at least 20 V / µs with the OOS turned on.
What is the gain value? Answer: The value of Ku is determined by the value of the transmission coefficient of the OOS circuit (inverse to it) and at audio frequencies - mainly by the ratio R14 / R10 (R15 / R13). Its measured value is about 86.
- What is the maximum voltage allowed at the input of the amplifier without degrading its performance?
Answer: When limiting the signal peaks in the output stage, the distortions are not compensated, since the "correcting" voltage of the MPOS links can no longer change the output. At such times, the amplifier parameters correspond to the amplifier without MFOS in clipping mode, and distortion is significant. Therefore, ivv should not be more than the nominal value.
- Is it possible to avoid the use of emitter followers, i.e. shorten the signal path?
Answer: It is impossible to do without emitter followers. They are necessary to match the high Rout of the buffer stage and the MCO link with the relatively low Rin of the voltage amplifier. In addition, EPs are needed to amplify the current signal, because. only they, together with VT11, VT12, determine the buildup current of the final stage (VT13, VT14 do not amplify the current, because they are included according to the OB circuit).
- Is it possible to lower the signal-to-noise ratio through the use of field-effect transistors in the UMZCH. If so, which ones and in what cascades?
Answer: In the first stages of the amplification channel, it is necessary to use complementary pairs of field-effect transistors with a cutoff amplification frequency of at least 200 MHz. It is quite possible to use low-frequency transistors in the MPOS links, but they are not suitable for the main channel.
In principle, the entire UMZCH can be performed on field-effect transistors, but this will be a different design.
- Is it possible to increase the output power of the UMZCH, i.e. number of output transistors?
The simplest option is to use more modern and powerful KT8101, KT8102 instead of VT21, VT22 and increase the supply voltage to ± 46 V. Then, as VT13, VT14, you need to use KT502E, KT503E. The resistance of resistors R46, R47 must be increased to 1.5 kOhm, and R36, R37 - up to 5.1 kOhm.
It is desirable to increase the capacitance of the capacitors in the power supply. It may also be necessary to change the values of the corrective elements C5, C6, C8, C9, R18 to ensure stability. As a result, the power rating rises to at least 150W into a 4 ohm load at a nominal input voltage of ~0.4V.
- What should be the UMZCH power supply: stabilized or not?
Answer: The power supply is an unstabilized bipolar rectifier with 10,000 uF filter capacitors. The use of switching power supplies is undesirable, since they create significant RF interference on the UMZCH circuit.
- What should be the area of heat sinks of transistors VT19-VT22?
Answer: The surface area of the heatsinks of the output transistors must be at least 400 cm2. In a more powerful version of the UMZCH (see above), it should be increased to 600 cm2. In this case, small heat sinks made of 1.5 mm thick aluminum sheet 2x3 cm2 in size and transistors VT19, VT20 should be provided.
- What diodes can replace KD520A?
Answer: They can be replaced by other silicon diodes, for example, series KD503, D219, D220. Since they determine the operating points of the corresponding transistors, you need to check the collector current VT11, VT12, VT13, VT14 in silent mode, the value of which should be about 5 mA and no more.
If it is much smaller, you can increase the number of series-connected diodes compared to the circuit, if the current is greater, reduce the resistance of resistors R28, R29 (to reduce 1k VT11, VT12) and increase the resistance of resistors R32, R35 (to reduce 1k VT13, VT14).
- Is it possible to replace trimming resistors R7, R20, R31, R53, R67 with wire type SP-5?
- What should be the resistance of the signal source to tune the amplifier?
Answer: The output impedance of the signal source connected to the nodal point must be at least tens of kiloohms, but if Rout is too large, the recorded signal decreases. I tuned the amplifier by connecting the signal source through a 16-20 kΩ resistor.
When tuning the second circuit, Rout should be reduced to ~2 kΩ, and the output voltage of the source should be increased to several volts, since in this case the recorded signal is significantly less than when tuning the first circuit.
- What is the permissible level of the constant component at the output of the amplifier at points A and B?
Answer: At the output of the UMZCH, the level of the constant component should be as close to zero as possible. Permissible can be considered 20-50 mV. At points A and B, the level of the constant component can be zero only if the pairs of transistors VT5, VT6 and VT9, VT10 are ideally complementary.
Since in fact the spread of the input characteristics reaches tenths of a volt, then the mentioned level should also differ from zero by the amount of this spread, if the priority (as in this case) is to maintain the same collector currents in each of the pairs of transistors. The presence of a constant component at these points is not of fundamental importance.
- Is it possible to adjust the collector currents of transistors VT11, VT12 with resistors R33, R34 (tuning with resistors R28, R29 is not possible)?
Answer: It is possible, but not desirable, since the gain channel gain strongly depends on the resistances of the resistors R33, R34, and changing them can lead to self-excitation, which will require changing the values of other correction elements.
Proceed as directed in RA2/99 (page 12). I note that with R28 = R29 = 0 1k transistors VT11, VT12 will also be equal to zero, therefore, it is always possible to reduce the collector current by reducing the resistances of resistors R28 and R29. It is important to change the resistances equally and simultaneously. If this fails, then either the transistors are faulty, or the potential at point B is too high and must be adjusted using R31.
- What is the reason that the second circuit of the MPOS (VT29-VT32) cannot be configured? The tests were carried out in both channels of the amplifier, all elements of the MPOS are in good order, the voltages on the transistors correspond to those recommended in the article.
Answer: It is more difficult to set up the MPOS B-loop, although the principle of setting is the same. First, it is difficult to get a significant signal level at the output of the amplifier. Secondly, when the simulator is connected to the voltage amplifier and the final stage, self-excitation easily sets in, and even with a slight excitation, R67 practically does not work. Therefore, when setting up, you need to control the absence of generations.
The B-loop can be tuned to minimize non-linear distortion during the experiment described at the end of the article. The ratings of the circuit elements are chosen so that even without tuning the accuracy of setting a1, y1 is about 10%, and the task is reduced to achieving the maximum possible effect.
- Is it necessary to select transistors by gain?
Answer: Bipolar transistors (in the main amplification channel) do not need to be selected. Field-effect transistors (in MPOS circuits) should preferably be selected according to the values of the initial drain current and cut-off voltage.
Answer: At first, one UMZCH was assembled. After fine-tuning the circuit, it was repeated as the second channel of the stereo amplifier. It was efficient and had characteristics close to the first without selection of elements (not counting field-effect transistors). This indicates a good repeatability of the design.
A radio amateur from Zhytomyr R. Dubchenko assembled an amplifier, listens to it with S-90 acoustics and is pleased with the sound. He reported that he had succeeded in almost all experiments with the MOS circuits (tuning and suppression of distortions) described in the article.
Answer: Judging by the symptoms, the problems are not in the amplifier itself, but from its incorrect docking with the signal source (IC), power supply unit (PSU) and load. The input impedance of the amplifier is relatively large, so its input is sensitive to interference.
Under no circumstances should the "ground" output of the load be transferred to the common bus of the printed circuit board. The collector wire of each output transistor must be twisted into one bundle with the emitter, the base wire should be left free. If the length of the wires is more than 10 cm, they should be shortened.
The noise disappears after the first circuit of the MPOS is connected to point A. Before that, it is really noticeable. However, until the amplifier is adjusted, the MPOS circuits should not be connected. First you need to achieve stable operation of the amplifier for a load dummy and only then connect the speakers.
- What transistors of the KP103 and KP303 series can be used, what is the permissible spread of their parameters and what is the nominal voltage between drain and source?
Answer: You can use transistors KP103E, Zh, I; KP303A, B, Zh with a spread of parameters of 20-30%. isi.nom ~9 V. We also provide the author's answers to questions on the article by V.P. Matyushkin "Physiological regulation of timbre" (see below)
- What functional dependence should have a variable resistor R15 (Fig. 4, a)?
Answer: It is better to use variable resistors R14, R15 with a linear regulation characteristic.
- What circuits of the preamplifier, volume controls and stereo balance did the author use?
Answer: You can use any scheme of these devices.
- Are the curves in the graph of Fig. 4b in the high-frequency region a continuation of the curves in the low-frequency region (curves 0, 1, 2)?
Answer: The high-frequency parts of the frequency response in Fig. 4b are shown at various positions of the R15 slider to illustrate their characteristic shape. Their form at f>>1 kHz practically does not depend on the position of switch SA1. In other words, the bass and treble tone controls are independent of each other, as in conventional tone controls.
Figure 1 The LANZAR power amplifier circuit is completely based on bipolar transistors.
INCREASE
Figure 2 LANZAR power amplifier circuit using field-effect transistors in the penultimate stage.
INCREASE
Figure 3 Schematic of the LANZAR power amplifier from the MS-8 simulator. INCREASE
|
LIST OF ELEMENTS INSTALLED IN THE LANZAR AMPLIFIER |
|
|
FOR BIPOLAR VERSION |
FOR THE OPTION WITH FIELD WORKERS |
| C3,C2 = 2 x 22µ0 C4 = 1 x 470p C6,C7 = 2 x 470µ0 x 25V C5,C8 = 2 x 0µ33 C11,C9 = 2 x 47µ0 C12,C13,C18 = 3 x 47p C15,C17,C1,C10 = 4 x 1µ0 C21 = 1 x 0µ15 C19,C20 = 2 x 470µ0 x 100V C14,C16 = 2 x 220µ0 x 100V R1 = 1 x 27k VD1,VD2 = 2 x 15V VT2,VT4 = 2 x 2N5401 |
C3,C2 = 2 x 22µ0 C4 = 1 x 470p C6,C7 = 2 x 470µ0 x 25V C5,C8 = 2 x 0µ33 C11,C10 = 2 x 47µ0 C12,C13,C18 = 3 x 47p C15,C17,C1,C9 = 4 x 1µ0 C21 = 1 x 0µ15 C19,C20 = 2 x 470µ0 x 100V C14,C16 = 2 x 220µ0 x 100V R1 = 1 x 27k VD1,VD2 = 2 x 15V VT8 = 1 x IRF640 |
For example, let's take the supply voltage equal to ±60 V. If the installation is done correctly and there are no faulty parts, then we will get a voltage map shown in Figure 7. The currents flowing through the elements of the power amplifier are shown in Figure 8. The dissipated power of each element is shown in Figure 9 (on transistors VT5, VT6, about 990 mW is dissipated, therefore, the TO-126 package requires a heat sink).
Figure 7. LANZAR power amplifier voltage map ENLARGE
Figure 8. Power Amplifier Current Map ENLARGE
Figure 9. Amplifier power dissipation map
A few words about the details and installation:
First of all, you should pay attention to the correct installation of parts, since the circuit is symmetrical, errors are quite common. Figure 10 shows the layout of the parts. The adjustment of the quiescent current (the current flowing through the terminal transistors with the input closed to a common wire and compensating the current-voltage characteristic of the transistors) is performed by resistor X1. When you first turn on the resistor slider must be in the upper position according to the diagram, i.e. have maximum resistance.
Quiescent current should be 30...60 mA. There is no point in putting it higher - neither the instruments nor the tangible changes occur by ear. To set the quiescent current, the voltage is measured at any of the emitter resistors of the final stage and is set in accordance with the table:
|
VOLTAGE AT THE OUTPUTS OF THE EMITTER RESISTOR, V |
QUIET CURRENT TOO LOW, STEP DISTORTION POSSIBLE, NORMAL QUIET CURRENT, HIGH QUIET CURRENT - EXCESSIVE HEATING, IF THIS IS NOT AN ATTEMPT TO CREATE A CLASS "A", THEN THIS IS EMERGENCY CURRENT. | |||||
|
QUIET CURRENT OF ONE PAIR OF FINAL TRANSISTORS, mA |
||||||
|
|
|
|
||||
|
|
|
|
||||
Figure 10 Location of parts on the power amplifier board. Shown are the places where the most common installation errors occur.
The question was raised about the advisability of using ceramic resistors in the emitter circuits of terminal transistors. You can also use MLT-2, two pieces connected in parallel with a nominal value of 0.47 ... 0.68 Ohm. However, the distortions introduced by ceramic resistors are too small, but the fact that they are interrupted - when overloaded, they break off, i.e. their resistance becomes infinite, which quite often leads to the rescue of terminal transistors in critical situations.
The area of the radiator depends on the cooling conditions, Figure 11 shows one of the options, it is necessary to fasten the power transistors to the heat sink through insulating gaskets
. It is better to use mica, since it has a rather small thermal resistance. One of the options for mounting transistors is shown in Figure 12.
Figure 11 One of the radiator options for a power of 300 W, subject to good ventilation
Figure 12 One of the options for mounting power amplifier transistors to a heatsink.
Insulating pads must be used.
Before mounting power transistors, as well as in case of suspicion of their breakdown, power transistors are checked by a tester. The limit on the tester is set to test the diodes (Fig. 13).
Figure 13 Checking the terminal transistors of the amplifier before installation and in case of suspicion of breakdown of transistors after critical situations.
Is it worth it to select transistors for coffee. amplification? There are quite a lot of disputes on this topic and the idea of selecting elements has been going on since the deep seventies, when the quality of the element base left much to be desired. Today, the manufacturer guarantees the spread of parameters between transistors of one batch of no more than 2%, which in itself speaks of the good quality of the elements. In addition, given that the terminal transistors 2SA1943 - 2SC5200 are firmly established in sound engineering, the manufacturer began to produce paired transistors, i.e. both direct and reverse conduction transistors already have the same parameters, i.e. the difference is not more than 2% (Fig. 14). Unfortunately, such pairs are not always found on sale, however, several times we happened to buy "twins". However, even having a parsing of coffee. gain between transistors of direct and reverse conduction, it is only necessary to ensure that transistors of the same structure are of the same batch, since they are connected in parallel and the spread in h21 can cause an overload of one of the transistors (for which this parameter is higher) and, as a result, overheating and exit from building. Well, the spread between transistors for positive and negative half-waves is fully compensated by negative feedback.
Figure 14 Transistors of different structures, but of the same batch.
The same applies to differential stage transistors - if they are of the same batch, i.e. purchased at the same time in the same place, the chance that the difference in parameters will be more than 5% is VERY small. Personally, we prefer FAIRCHALD transistors 2N5551 - 2N5401, however, STs sound quite decent.
However, this amplifier is also assembled on the domestic element base. This is quite real, but let's make an adjustment for the fact that the parameters of the purchased KT817 and found on the shelves in my workshop, bought back in the 90s, will differ quite a lot. Therefore, here it is better to use the h21 meter available in almost all digital testers. True, this lotion in the tester shows the truth only for low power transistors. It will not be entirely correct to select the transistors of the final stage with its help, since h21 also depends on the current flowing. That is why separate test benches are already being made to reject power transistors. from adjustable collector currents of the tested transistor (Fig. 15). The calibration of a permanent device for rejecting transistors is carried out in such a way that the microammeter deviates half the scale at a collector current of 1 A, and completely at a current of 2 A. When assembling an amplifier only for yourself, you don’t have to make a stand; two multimeters with a current measurement limit of at least 5 A are enough.
To perform a rejection, you should take any transistor from the rejected batch and set the collector current to 0.4 ... 0.6 A for the penultimate stage transistors and 1 ... 1.3 A for the terminal stage transistors with a variable resistor. Well, then everything is simple - transistors are connected to the terminals and, according to the readings of the ammeter included in the collector, transistors with the same readings are selected, not forgetting to look at the readings of the ammeter in the base circuit - they should also be similar. A spread of 5% is quite acceptable; for dial indicators on the scale, you can make marks of the "green corridor" during calibration. It should be noted that such currents do not cause bad heating of the transistor crystal, and given that it is without a heat sink, the duration of measurements should not be stretched in time - the SB1 button should not be held down for more than 1 ... 1.5 seconds. Such a rejection, first of all, will allow you to select transistors with a really similar gain coefficient, and checking powerful transistors with a digital multimeter is only a check to calm your conscience - in the microcurrent mode, powerful transistors have a gain coefficient of more than 500, and even a small spread when checking with a multimeter in real current modes can turn out to be huge . In other words, when checking the gain coff of a powerful transistor, the multimeter reading is nothing more than an abstract value that has nothing to do with the transistor gain coff through the collector-emitter junction, at least 0.5 A flows.
Figure 15 Rejection of powerful transistors by gain coefficient.
Feed-through capacitors C1-C3, C9-C11 are not quite typical inclusion, in comparison with the factory analogues of amplifiers. This is due to the fact that with this inclusion, it is not a polar capacitor of a rather large capacity that is obtained, but the use of a 1 μF film capacitor compensates for the not entirely correct operation of electrolytes at high frequencies. In other words, this implementation allowed for a more pleasant sounding amplifier, compared to a single electrolyte or a single film capacitor.
In older versions of Lanzar, instead of diodes VD3, VD4, 10 ohm resistors were used. The change in the element base allowed us to slightly improve the performance at signal peaks. For a more detailed consideration of this issue, let's turn to Figure 3.
In the circuit, not an ideal power source is modeled, but closer to the real one, which has its own resistance (R30, R31). When playing a sinusoidal signal, the voltage on the power rails will look like shown in Figure 16. In this case, the capacitance of the power filter capacitors is 4700 uF, which is somewhat small. For normal operation of the amplifier, the capacitance of the power supply capacitors must be at least 10,000 microfarads per channel, it is possible and more, but a significant difference is no longer noticeable. But back to Figure 16. The blue line shows the voltage directly on the collectors of the transistors of the final stage, and the red line shows the supply voltage of the voltage amplifier if resistors are used instead of VD3, VD4. As can be seen from the figure, the supply voltage of the final stage has dipped from 60 V and is located between 58.3 V in the pause and 55.7 V at the peak of the sinusoidal signal. Due to the fact that the capacitor C14 not only becomes infected through the decoupling diode, but also discharges at the peaks of the signal, the power supply voltage of the amplifier takes the form of a red line in Figure 16 and fluctuates from 56 V to 57.5 V, i.e. it has a range of about 1.5 AT.
Figure 16 voltage waveform when using decoupling resistors.
Figure 17 The shape of the supply voltages on the terminal transistors and voltage amplifier
Replacing the resistors with diodes VD3 and VD4, we get the voltages shown in Figure 17. As can be seen from the figure, the amplitude of the ripples on the collectors of the terminal transistors has not changed much, but the supply voltage of the voltage amplifier has taken on a completely different look. First of all, the amplitude decreased from 1.5 V to 1 V; by about 0.5 V, while when using a resistor, the voltage at the peak of the signal sags by 1.2 V. In other words, by simply replacing the resistors with diodes, it was possible to reduce the supply ripple in the voltage amplifier by more than 2 times.
However, these are theoretical calculations. In practice, this replacement allows you to get "free" 4-5 watts, since the amplifier comes at a higher output voltage and reduces distortion at signal peaks.
After assembling the amplifier and adjusting the quiescent current, you should make sure that there is no constant voltage at the output of the power amplifier. If it is higher than 0.1 V, then this already definitely requires adjustment of the operating modes of the amplifier. In this case, the easiest way is to select a "supporting" resistor R1. For clarity, we give several options for this rating and show the changes in the constant voltage at the output of the amplifier in Figure 18.
Figure 18 Variation of the DC voltage at the output of the amplifier depending on the noman R1
Despite the fact that on the simulator the optimal constant voltage was obtained only at R1 equal to 8.2 kOhm, in real amplifiers this value is 15 kOhm ... 27 kOhm, depending on which manufacturer the VT1-VT4 differential stage transistors are used.
Perhaps it is worth saying a few words about the differences between power amplifiers completely on bipolar transistors and using field workers in the penultimate cascade. First of all, when using field-effect transistors, the output stage of the voltage amplifier is VERY heavily unloaded, since the gates of field-effect transistors have practically no active resistance - only the gate capacitance is a load. In this version, the amplifier circuitry begins to step on the heels of class A amplifiers, since the current flowing through the output stage of the voltage amplifier almost does not change over the entire range of output powers. An increase in the quiescent current of the penultimate stage operating on a floating load R18 and the base of emitter followers of powerful transistors also varies within small limits, which ultimately led to a rather noticeable decrease in THD. However, there is also a fly in the ointment in this barrel of honey - the efficiency of the amplifier has decreased and the output power of the amplifier has decreased, due to the need to apply a voltage of more than 4 V to the gates of the field workers to open them (for a bipolar transistor, this parameter is 0.6 ... 0.7 V ). Figure 19 shows the peak of the sinusoidal signal of the amplifier, made on bipolar transistors (blue line) and field devices (red line) at the maximum amplitude of the output signal.
Figure 19 Changing the amplitude of the output signal when using different element base in the amplifier.
In other words, a decrease in THD by replacing field-effect transistors leads to a “shortage” of about 30 W, and a decrease in the THD level by about 2 times, so it’s up to everyone to decide exactly what to set.
It should also be remembered that the THD level also depends on the amplifier's own gain. In this amplifier gain coefficient depends on the values of resistors R25 and R13
(at the used ratings, the gain coefficient is almost 27 dB). Calculate gain factor in dB can be given by the formula Ku = 20 lg R25 / (R13 +1), where R13 and R25 - resistance in Ohms, 20 - multiplier, lg - decimal logarithm. If it is necessary to calculate the gain coefficient in times, then the formula takes the form Ku = R25 / (R13 + 1) . This calculation may be necessary when manufacturing a preamplifier and calculating the amplitude of the output signal in volts in order to exclude the operation of the power amplifier in the hard clipping mode.
Decreasing your own coffee. gain up to 21 dB (R13 = 910 ohms) leads to a decrease in the THD level by about 1.7 times with the same output signal amplitude (increased input voltage amplitude).
Well, now a few words about the most popular mistakes when assembling an amplifier yourself.
One of the most common mistakes is installation of 15 V zener diodes with incorrect polarity, i.e. these elements do not work in voltage stabilization mode, but like ordinary diodes. As a rule, such an error causes a constant voltage to appear at the output, and the polarity can be both positive and negative (more often negative). The voltage value is based between 15 and 30 V. In this case, no element is heated. Figure 20 shows the voltage map with incorrect installation of zener diodes, which was issued by the simulator. The erroneous items are highlighted in green.
Figure 20 Power amplifier voltage map with incorrectly soldered zener diodes.
The next popular mistake is mounting transistors upside down, i.e. when they confuse the collector and emitter in places. In this case, there is also constant tension, the absence of any signs of life. True, the reverse switching on of the differential cascade transistors can lead to their failure, but then how lucky. The voltage map for "inverted" inclusion is shown in Figure 21.
Figure 21 Voltage map with "inverted" switching on of differential stage transistors.
Often transistors 2N5551 and 2N5401 are confused, and they can also confuse the emitter with the collector. Figure 22 shows the voltage map of the amplifier with the "correct" mounting of the transistors interchanged, and in Figure 23, the transistors are not only swapped, but also turned upside down.
Figure 22 Transistors of the differential stage are interchanged.
Figure 23 Transistors of the differential stage are swapped, besides, the collector and emitter are swapped.
If the transistors are mixed up in places, and the emitter-collector is soldered correctly, then a small constant voltage is observed at the output of the amplifier, the quiescent current of the window transistors is regulated, but the sound is either completely absent or at the level “it seems to be playing”. Before mounting the transistors soldered in this way on a board, they should be checked for operability. If the transistors are interchanged, and even the emitter-collector is interchanged, then the situation is already quite critical, since in this variant for differential stage transistors the polarity of the applied voltage is correct, but the operating modes are violated. In this embodiment, there is a strong heating of the terminal transistors (the current flowing through them is 2-4 A), a small constant voltage at the output and a barely audible sound.
It is rather problematic to confuse the pinout of the transistors of the last stage of the voltage amplifier when using transistors in the TO-220 package, but transistors in the TO-126 package are quite often soldered upside down, swapping the collector and emitter. In this embodiment, a highly distorted output signal is observed, poor regulation of the quiescent current, and no heating of the transistors of the last stage of the voltage amplifier. A more detailed voltage map for this power amplifier mounting option is shown in Figure 24.
Figure 24 The transistors of the last stage of the voltage amplifier are soldered upside down.
Sometimes the transistors of the last stage of the voltage amplifier are confused. In this case, there is a small constant voltage at the output of the amplifier, the sound, if any, is very weak and with huge distortions, the quiescent current is only regulated upwards. An amplifier voltage map with such an error is shown in Figure 25.
Figure 25 Erroneous mounting of the transistors of the last stage of the voltage amplifier.
The penultimate cascade and terminal transistors in the amplifier are confused too rarely, so this option will not be considered.
Sometimes the amplifier fails, the most common reasons for this are overheating of the final transistors or overload. Insufficient heat sink area or poor thermal contact of the transistor flanges can lead to heating of the final transistor crystal to the temperature of mechanical destruction. Therefore, before the power amplifier is fully put into operation, it is necessary to make sure that the screws or self-tapping screws that fasten the terminals to the radiator are fully tightened, the insulating gaskets between the flanges of the transistors and the heat sink are well lubricated with thermal paste (we recommend the good old KPT-8), as well as the size of the gaskets over the size of the transistor by at least 3 mm on each side. If the heat sink area is insufficient, and there is simply no other, then you can use 12 V fans, which are used in computer technology. If the assembled amplifier is planned to operate only at above-average capacities (cafes, bars, etc.), then the cooler can be switched on for continuous operation, since it will still not be heard. If the amplifier is assembled for home use and will be operated at low power, then the operation of the cooler will already be audible, and there is no need for cooling - the radiator almost does not heat up. For such modes of operation, it is better to use controlled coolers. Several options for controlling the cooler are possible. The proposed options for controlling the coolers are based on the temperature control of the radiator and are switched on only when the radiator reaches a certain, controlled temperature. You can solve the problem of failure of window transistors either by installing additional overload protection, or by carefully mounting the wires going to the speaker system (for example, use oxygen-free wires to connect speakers to an amplifier, which, in addition to reduced active resistance, have increased insulation strength, resistant to shock and temperature ).
For example, consider several options for the failure of terminal transistors. Figure 26 shows a voltage map in the event that the reverse terminal transistors (2SC5200) go open, i.e. transitions are burnt out and have the maximum possible resistance. In this case, the amplifier maintains operating modes, the output remains close to zero, but the sound quality definitely wants better, since only one half-wave of the sinusoid is reproduced - negative (Fig. 27). The same will happen if the direct terminal transistors (2SA1943) break, only a positive half-wave will be reproduced.
Figure 26 Reverse terminal transistors burnt out to a break.
Figure 27 The signal at the output of the amplifier in the case when the 2SC5200 transistors burned out completely
Figure 27 is a voltage map in a situation where the terminals are out of order and have the lowest possible resistance, i.e. shorted. This variant of a malfunction drives the amplifier into VERY harsh conditions and further burning of the amplifier is limited only by the power source, since the current consumed at this moment can exceed 40 A. The surviving parts instantly gain temperature, in the arm where the transistors are still working, the voltage is slightly more than in where the short circuit to the power bus actually occurred. However, it is this situation that belongs to the easiest diagnostics - before turning on the amplifier, it will be enough to check the resistance of the transitions between each other with a multimeter, without even unsoldering them from the amplifier. The limit of the measurement set on the multimeter is DIOD TEST or BEEP. As a rule, burnt transistors show resistance between junctions in the range from 3 to 10 ohms.
Figure 27 Power amplifier voltage map in case of burnout of terminal transistors (2SC5200) to a short circuit
The amplifier will behave in exactly the same way in the event of a breakdown of the penultimate stage - when the outputs are cut off, only one half-wave of the sinusoid will be reproduced, with a short circuit of the transitions - huge consumption and heating.
In case of overheating, when it is considered that the radiator for the transistors of the last stage of the voltage amplifier is not needed (transistors VT5, VT6), they can also fail, and both go to an open or a short circuit. If the VT5 junctions burn out and the transition resistance is infinitely high, a situation arises when there is nothing to maintain zero at the amplifier output, and the ajar 2SA1943 terminal transistors will pull the voltage at the amplifier output to minus the supply voltage. If the load is connected, then the value of the DC voltage will depend on the set quiescent current - the higher it is, the greater the negative voltage value at the amplifier output. If the load is not connected, then the output will have a voltage very close in magnitude to the negative power bus (Fig. 28).
Figure 28 The voltage amplifier transistor VT5 "broke".
If the transistor in the last stage of the voltage amplifier VT5 is out of order and its transitions are closed, then with the load connected, the output will have a rather large constant voltage and a direct current flowing through the load, of the order of 2-4 A. If the load is turned off, then the output voltage amplifier will be almost equal to the positive power rail (Fig. 29).
Figure 29 The voltage amplifier transistor VT5 "closed".
Finally, it remains only to offer a few waveforms at the most coordinate points of the amplifier:
The voltage at the bases of the differential stage transistors at an input voltage of 2.2 V. The blue line is the VT1-VT2 bases, the red line is the VT3-VT4 bases. As can be seen from the figure, both the amplitude and phase of the signal practically coincide.
Voltage at the connection point of resistors R8 and R11 (blue line) and at the connection point of resistors R9 and R12 (red line). Input voltage 2.2 V.
The voltage on the collectors VT1 (red line), VT2 (green), as well as on the upper output of R7 (blue) and the lower output of R10 (purple). The voltage dip is caused by work on the load and a slight decrease in the supply voltage.
The voltage on the collectors VT5 (blue) and VT6 (red. The input voltage is reduced to 0.2 V, so that it can be seen more clearly, there is a difference of about 2.5 V in direct voltage
It remains only to explain at the expense of the power supply. First of all, the power of the mains transformer for a 300 W power amplifier should be at least 220-250 W and this will be enough to play even very hard compositions. You can learn more about the power of the power supply of power amplifiers. In other words, if you have a transformer from a tube color TV, then this is the IDEAL TRANSFORMER for one amplifier channel that allows you to easily play musical compositions with power up to 300-320 watts.
The capacitance of the power supply filter capacitors must be at least 10,000 microfarads per arm, optimally 15,000 microfarads. When using capacitances higher than the specified value, you simply increase the cost of construction without any noticeable improvement in sound quality. It should not be forgotten that when using such large capacities and a supply voltage above 50 V per arm, the instantaneous currents are already critically huge, so it is strongly recommended to use soft start systems.
First of all, before assembling any amplifier, it is strongly recommended to download descriptions of manufacturers' factories (datasheets) to ALL semiconductor elements. This will make it possible to get acquainted with the element base closer and, if any element is not on sale, find a replacement for it. In addition, you will have the correct pinout of transistors at hand, which will significantly increase the chances of a correct installation. Particularly lazy people are invited to VERY carefully familiarize themselves with at least the location of the terminals of the transistors used in the amplifier:
Finally, it remains to add that not everyone needs a power of 200-300 W, so the printed circuit board was redesigned for one pair of terminal transistors. This file was made by one of the visitors of the forum site "SOLDERING IRON" in the program SPRINT-LAYOUT-5 (DOWNLOAD THE BOARD). Details about this program are located.
12V battery in increased bipolar - you can proceed to the power amplifier itself. There are several channel amplifiers in the design.
TDA2005
- 20-25 watts are bridged. They are assembled on two separate boards for easy installation. Each of the amplifiers is activated when plus 12 volts is applied to the remote control output, this closes the relay and the amplifier is powered. Input capacitors can be selected to taste. Microcircuits are screwed onto a common heat sink through insulating gaskets.
Zener diodes for 15 volts with a power of 1-1.5 watts, make sure they are installed correctly, when connected back they will work like a diode, there is a danger of burning the differential stage. Differential cascade - made on low-power complementary pairs, which can be replaced with others that are as similar as possible in terms of parameters. It is in this stage that the sound is formed, which is subsequently amplified and fed to the terminal (output stage). If you plan to make an amplifier for 100-150 watts, then you can exclude the second pair of the output stage, since the power of the amplifier directly depends on the supply voltage. With one pair of outputs, it is not advised to increase the supply voltage above +/-45 volts. If you are planning to assemble a subwoofer amplifier, then this circuit is what you need! A variable resistor adjusts the quiescent current of the amplifier, the further service life of the circuit depends on it.
Before soldering the tuning resistor R15, it must be “unscrewed” so that its impedance is soldered into the gap of the track. You need to take a multi-turn resistor, they can very accurately adjust the quiescent current, it is also very convenient for further tuning. But of course, if it is not already there, then you can get by with an ordinary trimmer, but it is advisable to remove it from the common board with wires, since after installing all the components, tuning will be almost impossible.
The quiescent current is adjusted after "heating the circuit", in other words, turn it on for 15-20 minutes, let it play, but don't get carried away! The quiescent current is an important factor, without proper tuning the amplifier will not last long, the correct operation of the output stage and the level of constant at the output of the amplifier depend on it. The quiescent current can be found by measuring the voltage drop across a pair of emitter resistors (set the multimeter to the limit of 200mV, the probes to the VT10 and VT11 emitters). Calculation according to the formula: Ipok \u003d Uv / (R26 + R26). Next, smoothly rotate the trimmer and look at the readings of the multimeter. You need to set 70-100mA - this is equivalent to a multimeter reading (30-44) mV. We check the level of constant voltage at the output. And now everything is ready - you can enjoy the sound of an amplifier assembled by yourself!
A small addition. Having assembled the UMZCH, you need to think about heat sinks. The main heat sink was taken from a domestic amplifier RADIO U-101 STEREO- it almost does not heat up during operation. Low-power diff-cascade transistors heat up, but overheating is not terrible, so they do not need cooling. The output transistors are screwed to the main heat sink through insulating gaskets, it is also desirable to use thermal paste, which I did not.
All other transistors can be installed on small separate heat sinks, or you can use a common one (for each stage), but in this case you need to screw the transistors through spacers. IMPORTANT ! All transistors must be screwed to the radiators through insulating gaskets, there should not be any short circuits on the bus, therefore, before turning on, carefully check with a multimeter whether the transistor leads are closed to the heat sink. We can consider the assembly of the device completed, but for today I say goodbye to you - AKA KASYAN.
Discuss the article AMPLIFIER WITH YOUR HANDS - BLOCK UMZCH
Viktor Zhukovsky, Krasnoarmeysk, Donetsk region
UMZCH BB-2010 is a new development from the well-known line of amplifiers UMZCH BB (high fidelity) [1; 2; five]. A number of technical solutions used were influenced by the work of Ageev SI. .
The amplifier provides Kr of the order of 0.001% at a frequency of 20 kHz with Рout = 150 W at a load of 8 ohms, a small signal bandwidth at a level of -3 dB - 0 Hz ... 800 kHz, an output voltage slew rate of -100 V / μs, signal-to-noise ratio and signal/background -120 dB.
Due to the use of an op amp operating in a light mode, as well as the use of only stages with OK and OB in the voltage amplifier, covered by deep local OOS, UMZCH BB is highly linear even before the general OOS is covered. In the very first high-fidelity amplifier back in 1985, solutions were used that until then were used only in measuring technology: a separate service node supports direct current modes, to reduce the level of interface distortions, the transient resistance of the AC switching relay contact group is covered by a common negative feedback, and a special node effectively compensates for the influence of the resistance of the AC cables on these distortions. The tradition has been preserved in UMZCH BB-2010, however, the general environmental protection also covers the resistance of the output low-pass filter.
In the vast majority of designs of other UMZCH, both professional and amateur, many of these solutions are still missing. At the same time, the high technical characteristics and audiophile advantages of the UMZCH BB are achieved by simple circuit solutions and a minimum of active elements. In fact, this is a relatively simple amplifier: one channel can be assembled slowly in a couple of days, and the setting consists only in setting the required quiescent current of the output transistors. Especially for beginner radio amateurs, a method has been developed for node-by-node, cascade-based performance testing and adjustment, using which you can guarantee to localize the places of possible errors and prevent their possible consequences even before the UMZCH is fully assembled. For all possible questions about this or similar amplifiers, there are detailed explanations, both on paper and on the Internet.
At the input of the amplifier, an R1C1 high-pass filter with a cutoff frequency of 1.6 Hz is provided, Fig. 1. But the efficiency of the mode stabilization device allows the amplifier to work with an input signal containing up to 400 mV of DC voltage. Therefore, C1 is excluded, which realizes the age-old audiophile dream of a path without capacitors © and significantly improves the sound of the amplifier.
The capacitance of the capacitor C2 of the input low-pass filter R2C2 is chosen so that the cutoff frequency of the input low-pass filter, taking into account the output resistance of the preamplifier 500 Ohm -1 kOhm, is in the range from 120 to 200 kHz. The frequency correction circuit R3R5C3 is placed at the input of the op-amp DA1, which limits the band of processed harmonics and interference coming through the CUS circuit from the output side of the UMZCH to a band of 215 kHz at a level of -3 dB and increases the stability of the amplifier. This circuit reduces the difference signal above the cutoff frequency of the circuit and thus eliminates unnecessary overload of the voltage amplifier with high-frequency interference, noise and harmonics, eliminating the possibility of dynamic intermodulation distortion (TIM; DIM).
Next, the signal is fed to the input of a low-noise operational amplifier with field-effect transistors at the input DA1. Many "claims" against the UMZCH BB are made by opponents regarding the use of an op-amp at the input, which allegedly degrades the sound quality and "steals the virtual depth" of the sound. In this regard, it is necessary to pay attention to some quite obvious features of the operation of the OS in the UMZCH VV.
Operational amplifiers of preamplifiers, post-DAC op-amps are forced to develop several volts of output voltage. Since the gain of the op amps is low, ranging from 500 to 2,000 times at 20 kHz, this indicates that they operate with a relatively large difference signal voltage - from several hundred microvolts at low frequencies to several millivolts at 20 kHz, and a high probability of introducing intermodulation distortion from the op amp input stage. The output voltage of these op amps is equal to the output voltage of the last voltage amplification stage, usually made according to the OE scheme. An output voltage of several volts indicates the operation of this cascade with rather large input and output voltages, and as a result, it introduces distortions into the amplified signal. The op-amp is loaded with the resistance of the OOS circuit and the load connected in parallel, sometimes amounting to several kilo-ohms, which requires up to several milliamps from the output follower of the output current amplifier. Therefore, changes in the current of the output follower of the IC, the output stages of which consume a current of no more than 2 mA, are quite significant, which also indicates that they introduce distortions into the amplified signal. We see that the input stage, the voltage amplification stage and the output stage of the op-amp can introduce distortions.
But the high-fidelity amplifier circuitry, due to the high gain and input resistance of the transistor part of the voltage amplifier, provides very gentle operating conditions for the op-amp DA1. Judge for yourself. Even in the UMZCH, which has developed a rated output voltage of 50 V, the input differential stage of the op amp operates with differential voltage signals from 12 μV at frequencies of 500 Hz to 500 μV at a frequency of 20 kHz. The ratio of the high input overload capacity of the differential stage, made on field-effect transistors, and the meager voltage of the difference signal provides a high linearity of signal amplification. The output voltage of the op-amp does not exceed 300 mV. which indicates a low input voltage of the voltage amplification stage with a common emitter from the operational amplifier - up to 60 μV - and the linear mode of its operation. The output stage of the op-amp gives to the load about 100 kOhm from the side of the VT2 base an alternating current of not more than 3 μA. Consequently, the output stage of the op-amp also operates in an extremely lightweight mode, almost at idle. On a real musical signal, voltages and currents are, most of the time, an order of magnitude less than the given values.
From a comparison of the voltages of the difference and output signals, as well as the load current, it can be seen that, in general, the operational amplifier in the UMZCH BB operates hundreds of times easier, and, therefore, in a linear mode than the op-amp mode of preamplifiers and post-DAC op-amps of CD players that serve as sources signal for UMZCH with any depth of environmental protection, as well as without it at all. Consequently, the same op amp will introduce much less distortion as part of the UMZCH BB than in a single inclusion.
Occasionally there is an opinion that the distortions introduced by the cascade are ambiguously dependent on the voltage of the input signal. This is mistake. The dependence of the manifestation of the nonlinearity of the cascade on the voltage of the input signal may obey one law or another, but it is always unambiguous: an increase in this voltage never leads to a decrease in the introduced distortions, but only to an increase.
It is known that the level of distortion products attributable to a given frequency decreases in proportion to the depth of negative feedback for this frequency. The idle speed gain, up to the coverage of the feedback amplifier, at low frequencies cannot be measured due to the smallness of the input signal. According to calculations, the idle amplification developed up to the NOS coverage makes it possible to achieve an OOS depth of 104 dB at frequencies up to 500 Hz. Measurements for frequencies starting from 10 kHz show that the depth of feedback at a frequency of 10 kHz reaches 80 dB, at a frequency of 20 kHz - 72 dB, at a frequency of 50 kHz - 62 dB and 40 dB - at a frequency of 200 kHz. Figure 2 shows the amplitude-frequency characteristics of UMZCH BB-2010 and, for comparison, UMZCH similar in complexity to Leonid Zuev.
High gain before coverage of the OOS is the main feature of the circuit design of VV amplifiers. Since the goal of all circuitry tricks is to achieve high linearity and high gain for maintaining deep feedback in the widest possible frequency band, this means that circuitry methods for improving amplifier parameters are exhausted by such structures. Further reduction of distortion can only be ensured by constructive measures aimed at reducing the pickup of harmonics of the output stage on the input circuits, especially on the inverting input circuit, the gain from which is maximum.
Another feature of the UMZCH BB circuitry is the current control of the output stage of the voltage amplifier. The input op-amp controls the voltage-to-current conversion stage, performed with OK and OB, and the received current is subtracted from the quiescent current of the stage, performed according to the OB circuit.
The use of a linearizing resistor R17 with a resistance of 1 kOhm in the differential stage VT1, VT2 on transistors of different structures with serial power increases the linearity of the conversion of the output voltage of the op-amp DA1 to the collector current VT2 by creating a local OOS with a depth of 40 dB. This can be seen from a comparison of the sum of the intrinsic resistances of the emitters VT1, VT2 - approximately 5 ohms each - with the resistance R17, or the sum of the thermal voltages VT1, VT2 - about 50 mV - with a voltage drop across the resistance R17, which is 5.2 - 5.6 V .
Amplifiers built according to the considered circuitry have a sharp, 40 dB per decade of frequency, gain decay above a frequency of 13 ... 16 kHz. The error signal, which is a distortion product, at frequencies above 20 kHz is two to three orders of magnitude smaller than the useful audio signal. This makes it possible to convert the linearity of the differential stage VT1, VT2, which is excessive at these frequencies, into an increase in the gain of the transistor part of the UN. Due to slight changes in the current of the differential stage VT1, VT2, when weak signals are amplified, its linearity does not deteriorate significantly with a decrease in the depth of the local OOS, but the operation of the op-amp DA1, on the operating mode of which the linearity of the entire amplifier depends on the operating mode of which at these frequencies, the gain margin will facilitate, since all voltages, The distortions that determine the distortions introduced by the operational amplifier, starting from the difference signal to the output signal, decrease in proportion to the gain in gain at a given frequency.
The phase advance correction circuits R18C13 and R19C16 were optimized in the simulator in order to reduce the difference voltage of the op-amp to frequencies of several megahertz. It was possible to increase the gain of UMZCH BB-2010 compared to UMZCH BB-2008 at frequencies of the order of several hundred kilohertz. Gain gain was 4 dB at 200 kHz, 6 dB at 300 kHz, 8.6 dB at 500 kHz, 10.5 dB at 800 kHz, 11 dB at 1 MHz, and 10 to 12 dB at frequencies above 2 MHz. This can be seen from the simulation results, Fig. 3, where the lower curve refers to the frequency response of the UMZCH BB-2008 lead correction circuit, and the upper one to UMZCH BB-2010.
VD7 protects the emitter junction VT1 from reverse voltage arising from the flow of recharging currents C13, C16 in the voltage limiting mode of the UMZCH output signal and the resulting limit voltages with a high rate of change at the output of the op-amp DA1.
The output stage of the voltage amplifier is made on a transistor VT3, connected according to a common base circuit, which excludes the penetration of a signal from the output circuits of the stage into the input circuits and increases its stability. The cascade with OB, loaded on the current generator on the transistor VT5 and the input impedance of the output stage, develops a high stable gain - up to 13,000 ... 15,000 times. The choice of the resistance of the resistor R24 half the resistance of the resistor R26 guarantees the equality of the quiescent currents VT1, VT2 and VT3, VT5. R24, R26 provide local OOS that reduce the effect of the Earley effect - the change in p21e depending on the collector voltage and increase the initial linearity of the amplifier by 40 dB and 46 dB, respectively. The supply of the UN with a separate voltage, modulo 15 V higher than the voltage of the output stages, makes it possible to eliminate the effect of quasi-saturation of transistors VT3, VT5, which manifests itself in a decrease in n21e when the collector-base voltage drops below 7 V.
The three-stage output follower is assembled on bipolar transistors and does not require any special comments. Don't try to fight entropy © by saving on the quiescent current of the output transistors. It should not be less than 250 mA; in the author's version - 320 mA.
Prior to the operation of the relay for switching on AC K1, the amplifier is covered by OOS1, implemented by turning on the divider R6R4. The accuracy of maintaining the resistance R6 and the consistency of these resistances in different channels is not essential, but to maintain the stability of the amplifier it is important that the resistance R6 is not much lower than the sum of the resistances R8 and R70. By actuating relay K1, the OOS1 is turned off and the OOS2 circuit, formed by R8R70C44 and R4, comes into operation, and covers the contact group K1.1, where R70C44 excludes the output low-pass filter R71L1 R72C47 from the OOC circuit at frequencies above 33 kHz. The frequency-dependent OOS R7C10 generates a decline in the frequency response of the UMZCH to the output low-pass filter at a frequency of 800 kHz at a level of -3 dB and provides a margin in depth of the OOS above this frequency. The frequency response decay at the AC terminals above the frequency of 280 kHz at a level of -3 dB is provided by the combined action of the R7C10 and the output low-pass filter R71L1 -R72C47.
The resonant properties of loudspeakers lead to the emission of damped sound vibrations by the diffuser, overtones after impulse action and the generation of its own voltage when the turns of the loudspeaker coil cross the magnetic field lines in the gap of the magnetic system. The damping coefficient shows how large the amplitude of the diffuser oscillations is and how quickly they decay when the AC is loaded as a generator on the impedance from the UMZCH. This coefficient is equal to the ratio of the AC resistance to the sum of the output resistance of the UMZCH, the transient resistance of the contact group of the AC switching relay, the resistance of the inductor coil of the output LPF usually wound with a wire of insufficient diameter, the transient resistance of the AC cable clamps and the resistance of the AC cables themselves.
In addition, the impedance of loudspeakers is non-linear. The flow of distorted currents through the wires of AC cables creates a voltage drop with a high degree of non-linear distortion, which is also subtracted from the undistorted output voltage of the amplifier. Therefore, the signal at the AC terminals is much more distorted than at the UMZCH output. These are the so-called interface distortions.
To reduce these distortions, compensation of all components of the total output impedance of the amplifier was applied. The own output resistance of the UMZCH, together with the contact resistance of the relay contacts and the resistance of the wire of the inductor of the output low-pass filter, is reduced by the action of a deep general OOS taken from the right output of L1. In addition, by connecting the right output of R70 to the “hot” AC terminal, you can easily compensate for the transient resistance of the AC cable clamp and the resistance of one of the AC wires, without fear of generating UMZCH due to phase shifts in the wires covered by the OOS.
The AC wire resistance compensation unit is made in the form of an inverting amplifier with Ky = -2 on the DA2, R10, C4, R11 and R9 op-amps. The input voltage for this amplifier is the voltage drop on the "cold" ("earth") wire of the speaker. Since its resistance is equal to the resistance of the "hot" wire of the AC cable, to compensate for the resistance of both wires, it is enough to double the voltage on the "cold" wire, invert it and through the resistor R9 with a resistance equal to the sum of the resistances R8 and R70 of the OOS circuit, apply to the inverting input of the op-amp DA1 . Then the output voltage of the UMZCH will increase by the sum of the voltage drops on the AC wires, which is equivalent to eliminating the influence of their resistance on the damping coefficient and the level of interface distortion at the AC terminals. Compensation for the drop in the resistance of the AC wires of the non-linear component of the back-EMF of loudspeakers is especially needed at the lower frequencies of the audio range. The signal voltage at the tweeter is limited by a resistor and capacitor connected in series with it. Their complex resistance is much greater than the resistance of the wires of the AC cable, so the compensation of this resistance at the RF is meaningless. Based on this, the integrating circuit R11C4 limits the operating frequency band of the compensator to 22 kHz.
Of particular note: the resistance of the "hot" wire of the AC cable can be compensated by covering it with a common OOS by connecting the right terminal of R70 with a special wire to the "hot" AC terminal. In this case, only the resistance of the "cold" AC wire will need to be compensated, and the gain of the wire resistance compensator must be reduced to the value Ku \u003d -1 by choosing the resistance of the resistor R10 equal to the resistance of the resistor R11.
The current protection unit prevents damage to the output transistors during short circuits in the load. Resistors R53 - R56 and R57 - R60 serve as a current sensor, which is quite enough. The amplifier output current flowing through these resistors creates a voltage drop that is applied to the divider R41R42. A voltage with a value greater than the threshold opens the transistor VT10, and its collector current opens the VT8 trigger cell VT8VT9. This cell goes into a steady state with open transistors and shunts the HL1VD8 circuit, reducing the current through the zener diode to zero and locking VT3. Discharging C21 with a small base current VT3 can take a few milliseconds. After the trigger cell is activated, the voltage on the lower plate of C23, charged by the voltage on the HL1 LED to 1.6 V, rises from the level of -7.2 V from the positive power rail of the UN to the level of -1.2 V 1, the voltage on the upper plate of this capacitor also rises to 5 V. C21 is quickly discharged through the resistor R30 to C23, the transistor VT3 is locked. Meanwhile, VT6 opens and through R33, R36 opens VT7. VT7 shunts the zener diode VD9, discharges capacitor C22 through R31 and turns off transistor VT5. Not receiving a bias voltage, the output stage transistors are also locked.
Restoring the initial state of the trigger and turning on the UMZCH is done by pressing the button SA1 "Reset protection". C27 is charged by the VT9 collector current and shunts the VT8 base circuit, locking the trigger cell. If by this time the emergency has been eliminated and VT10 is locked, the cell goes into a state with stably closed transistors. VT6, VT7 are closed, a reference voltage is applied to the bases VT3, VT5 and the amplifier enters the operating mode. If the short circuit in the UMZCH load continues, the protection is activated again, even if the capacitor C27 is connected to SA1. The protection works so effectively that during the adjustment of the correction, the amplifier was de-energized several times for small soldering ... by touching the non-inverting input. The resulting self-excitation led to an increase in the current of the output transistors, and the protection turned off the amplifier. Although this crude method should not be offered as a rule, but due to current protection, it did not harm the output transistors.
The work of the compensator for the resistance of AC cables.
The efficiency of the UMZCH BB-2008 compensator was tested by the old audiophile method, by ear, by switching the compensator input between the compensating wire and the common wire of the amplifier. The improvement in sound was clearly noticeable, and the future owner was eager to get an amplifier, so no measurements of the effect of the compensator were carried out. The advantages of the cable-cutter scheme were so obvious that the compensator + integrator configuration was adopted as the standard assembly for installation in all developed amplifiers.
It's amazing how much unnecessary debate about the usefulness / uselessness of cable resistance compensation has flared up on the Internet. As usual, those to whom the extremely simple cable-cleaning scheme seemed complicated and incomprehensible, the costs for it - exorbitant, and the installation - time-consuming ©, especially insisted on listening to a non-linear signal. There were even suggestions that, since so much money is being spent on the amplifier itself, it’s a sin to save on the sacred, but you need to go the best, glamorous way that all civilized mankind goes and ... buy normal, human © super-expensive cables made of precious metals. To my great surprise, the statements of highly respected experts about the uselessness of the compensation unit at home, including those specialists who successfully use this unit in their amplifiers, added fuel to the fire. It is very unfortunate that many fellow radio amateurs were distrustful of reports about improving the sound quality at low and medium frequencies with the inclusion of a compensator, avoided this simple way to improve the operation of the UMZCH with all their might, than robbed themselves.
Little research has been done to document the truth. A number of frequencies were supplied from the GZ-118 generator to the UMZCH BB-2010 in the region of the AC resonant frequency, the voltage was controlled by an S1-117 oscilloscope, and Kr at the AC terminals was measured by INI C6-8, Fig. 4. 
Resistor R1 is installed to avoid interference at the input of the compensator when switching it between the control and common wires. The experiment used common and publicly available AC cables with a length of 3 m and a core cross section of 6 square meters. mm, as well as the GIGA FS Il speaker system with a frequency range of 25 -22.000 Hz, a nominal impedance of 8 ohms and a rated power of 90 W from Acoustic Kingdom.
Unfortunately, the circuitry of the harmonic signal amplifiers from the C6-8 composition provides for the use of high-capacity oxide capacitors in the environmental protection circuits. This causes the low-frequency noise of these capacitors to affect the resolution of the device at low frequencies, as a result of which its resolution at low frequencies deteriorates. When measuring Kr of a signal with a frequency of 25 Hz from GZ-118 directly from C6-8, the instrument readings dance around a value of 0.02%. It is not possible to get around this limitation using the GZ-118 generator notch filter in the case of measuring the compensator efficiency, because a number of discrete values of the tuning frequencies of the 2T filter are limited at low frequencies by the values of 20.60, 120, 200 Hz and do not allow measuring Kr at the frequencies of interest to us. Therefore, reluctantly, the level of 0.02% was taken as zero, the reference.
At a frequency of 20 Hz with a voltage at the AC terminals of 3 Vpp, which corresponds to an output power of 0.56 W into an 8 ohm load, Kr was 0.02% with the compensator on and 0.06% after it was turned off. At a voltage of 10 V amps, which corresponds to an output power of 6.25 W, the Kr value is 0.02% and 0.08%, respectively, at a voltage of 20 V amps and a power of 25 W - 0.016% and 0.11%, and at a voltage of 30 In the amplitude and power of 56 W - 0.02% and 0.13%.
Knowing the relaxed attitude of manufacturers of imported equipment to the values of inscriptions regarding power, and also remembering the miraculous, after the adoption of Western standards, the transformation of the 35AC-1 speaker system with a subwoofer power of 30 W into S-90, long-term power of more than 56 W was not supplied to AC.
At a frequency of 25 Hz at a power of 25 W, Kr was 0.02% and 0.12% with the compensation unit on / off, and at a power of 56 W - 0.02% and 0.15%.
At the same time, the necessity and effectiveness of covering the output LPF of the general OOS was checked. At a frequency of 25 Hz at a power of 56 W and connected in series to one of the wires of the AC cable of the output RL-RC low-pass filter, similar to that installed in the superlinear UMZCH, Kr with the compensator turned off reaches 0.18%. At a frequency of 30 Hz at a power of 56 W Kr 0.02% and 0.06% with the compensation unit on / off. At a frequency of 35 Hz at a power of 56 W, Kr is 0.02% and 0.04% with the compensation unit on / off. At frequencies of 40 and 90 Hz at a power of 56 W, Kr is 0.02% and 0.04% with the compensation unit on / off, and at a frequency of 60 Hz - 0.02% and 0.06%.
The conclusions are obvious. There is a presence of non-linear distortion of the signal at the AC terminals. The deterioration of the linearity of the signal at the AC terminals is clearly recorded with its inclusion through an uncompensated, uncovered OOS resistance of a low-pass filter containing 70 cm of a relatively thin wire. The dependence of the level of distortion on the power supplied to the AC suggests that it depends on the ratio of the signal power and the nominal power of the AC woofers. Distortions are most pronounced at frequencies near the resonant one. The back EMF generated by the speakers in response to the impact of an audio signal is shunted by the sum of the output resistance of the UMZCH and the resistance of the wires of the AC cable, so the level of distortion at the AC terminals directly depends on the resistance of these wires and the output impedance of the amplifier.
The cone of a poorly damped woofer itself emits overtones, and in addition, this loudspeaker generates a wide tail of harmonics and intermodulation distortion products that a midrange loudspeaker reproduces. This explains the deterioration of the sound at medium frequencies.
Despite the assumption of a zero Kr level of 0.02% due to the imperfection of the IRI, the effect of the cable resistance compensator on signal distortion at the AC terminals is clearly and unambiguously noted. It can be stated that the conclusions made after listening to the operation of the compensation unit on a musical signal and the results of instrumental measurements are in full agreement.
The improvement that is clearly audible when the cable cleaner is turned on can be explained by the fact that with the disappearance of distortion on the AC terminals, the midrange loudspeaker stops reproducing all this dirt. Apparently, therefore, by reducing or eliminating the reproduction of distortions by a mid-frequency loudspeaker, a two-cable AC connection circuit, the so-called. "biwiring", when the LF and MF-HF links are connected by different cables, has an advantage in sound compared to a single-cable circuit. However, since in a two-cable circuit the distorted signal at the terminals of the LF section of the AC does not disappear anywhere, this circuit loses to the option with a compensator in terms of the damping coefficient of the free vibrations of the cone of the low-frequency loudspeaker.
You can't deceive physics, and for a decent sound it is not enough to get brilliant performance at the output of the amplifier with an active load, but it is also necessary not to lose linearity after the signal is delivered to the speaker terminals. As part of a good amplifier, a compensator made according to one scheme or another is absolutely necessary.
Integrator.
The effectiveness and possibility of reducing the error of the DA3 integrator was also tested. In UMZCH BB with op-amp TL071, the output DC voltage is in the range of 6 ... 9 mV, and it was not possible to reduce this voltage by including an additional resistor in the non-inverting input circuit.
The effect of low-frequency noise, characteristic of a DC-input op-amp, due to the coverage of deep feedback through the frequency-dependent circuit R16R13C5C6, manifests itself in the form of an instability of the output voltage of a few millivolts, or -60 dB relative to the output voltage at rated output power, at frequencies below 1 Hz , not reproducible speakers.
On the Internet, it was mentioned about the low resistance of the protective diodes VD1 ... VD4, which allegedly introduces an error into the operation of the integrator due to the formation of a divider (R16 + R13) / R VD2 | VD4 . . To check the reverse resistance of protective diodes, a circuit was assembled in Fig. 6. Here, the op-amp DA1, connected according to the inverting amplifier circuit, is covered by the OOS through R2, its output voltage is proportional to the current in the circuit of the tested diode VD2 and the protective resistor R2 with a coefficient of 1 mV / nA, and the resistance of the circuit R2VD2 - with a coefficient of 1 mV / 15 GΩ. To eliminate the influence of the additive errors of the op-amp - bias voltage and input current on the results of measuring the diode leakage current, it is necessary to calculate only the difference between the intrinsic voltage at the output of the op-amp, measured without the diode under test, and the voltage at the output of the op-amp after its installation. In practice, a difference in the output voltages of the op-amp of several millivolts gives the value of the reverse resistance of the diode of the order of ten to fifteen gigaohms at a reverse voltage of 15 V. It is obvious that the leakage current will not increase with a decrease in the voltage across the diode to a level of several millivolts, which is characteristic of the difference voltage of the op-amp of the integrator and compensator .
But the photoelectric effect inherent in diodes placed in a glass case really leads to a significant change in the output voltage of the UMZCH. When illuminated with an incandescent lamp of 60 W from a distance of 20 cm, the constant voltage at the output of the UMZCH increased to 20 ... 3O mV. Although it is unlikely that a similar level of illumination can be observed inside the amplifier case, a drop of paint applied to these diodes eliminated the dependence of the UMZCH modes on illumination. According to the simulation results, no drop in the frequency response of the UMZCH is observed even at a frequency of 1 millihertz. But the time constant R16R13C5C6 should not be reduced. The phases of the alternating voltages at the outputs of the integrator and the compensator are opposite, and with a decrease in the capacitance of the capacitors or the resistance of the resistors of the integrator, an increase in its output voltage can worsen the compensation of the resistance of the AC cables.
Amplifier sound comparison. The sound of the assembled amplifier was compared with the sound of several foreign industrial amplifiers. The source was a Cambridge Audio CD player, the Radiotekhnika UP-001 pre-amplifier was used to build up and adjust the sound level of the terminal UMZCH, the Sugden A21a and NAD C352 used regular adjustment controls.
The first to check was the legendary, outrageous and damn expensive English UMZCH "Sugden A21a", operating in class A with an output power of 25 watts. Remarkably, in the accompanying documentation for VCL, the British considered it good not to indicate the level of non-linear distortion. Say, it's not about distortions, but about spirituality. "Sugden A21a>" lost to UMZCH BB-2010 with comparable power both in terms of level and clarity, confidence, nobility of sound at low frequencies. This is not surprising, given the peculiarities of its circuitry: just a two-stage quasi-symmetric output follower on transistors of the same structure, assembled according to the circuitry of the 70s of the last century with a relatively high output resistance and an electrolytic capacitor switched on at the output that further increases the total output resistance - this is the last the solution itself degrades the sound of any amplifiers at low and medium frequencies. At medium and high frequencies, UMZCH BB showed higher detail, transparency and excellent stage elaboration, when singers and instruments could be clearly localized in sound. By the way, speaking of the correlation of objective measurement data and subjective impressions of the sound: in one of the magazine articles of Sugden's competitors, its Kr was determined at the level of 0.03% at a frequency of 10 kHz.
The next was also the English amplifier NAD С352. The general impression was the same: the pronounced "bucket" sound of the Englishman at low frequencies did not leave him any chances, while the work of the UMZCH BB was recognized as impeccable. Unlike NADa, whose sound was associated with thick bushes, wool, cotton wool, the sound of BB-2010 at medium and high frequencies made it possible to clearly distinguish the voices of performers in the general choir and instruments in the orchestra. In the work of NAD C352, the effect of better audibility of a more vociferous performer, a louder instrument, was clearly expressed. As the owner of the amplifier himself put it, in the sound of the UMZCH BB, the vocalists did not “shout-nod” each other, and the violin did not fight in the power of sound with a guitar or trumpet, but all the instruments peacefully and harmoniously “made friends” in the overall sound image of the melody. At high frequencies, the UMZCH BB-2010, according to figurative audiophiles, sounds like “as if drawing a sound with a thin, thin brush.” These effects can be attributed to the difference in intermodulation distortion of the amplifiers.
The sound of the UMZCH Rotel RB 981 was similar to the sound of the NAD C352, with the exception of better performance at low frequencies, yet the UMZCH BB-2010 remained out of competition in the clarity of AC control at low frequencies, as well as transparency, delicacy of sound at medium and high frequencies.
The most interesting in terms of understanding the mindset of audiophiles was the general opinion that, despite the superiority over these three UMZCH, they bring “warmth” to the sound, which makes it more pleasant, and UMZCH BB works smoothly, “it is neutral to the sound.”
The Japanese Dual CV1460 lost in sound immediately after being turned on in the most obvious way for everyone, and they did not waste time listening to it in detail. His Kr was in the range of 0.04 ... 0.07% at low power.
The main impressions from the comparison of amplifiers in general terms were completely identical: UMZCH BB was ahead of them in sound unconditionally and unambiguously. Therefore, further tests were considered unnecessary. As a result, friendship won, everyone got what they wanted: for a warm, intimate sound - Sugden, NAD and Rotel, and to hear what was recorded on the disc by the director - UMZCH BB-2010.
Personally, I like high-fidelity UMZCH with a light, clean, impeccable, noble sound, it effortlessly reproduces passages of any complexity. As my friend, an audiophile with great experience, put it, he works out the sounds of drum kits at low frequencies without options, like a press, at medium frequencies he sounds as if he does not exist, and at high frequencies he seems to paint the sound with a thin brush. For me, the non-irritating sound of UMZCH BB is associated with the ease of operation of cascades.
Literature
1. Sukhov I. UMZCH high fidelity. "Radio", 1989, No. 6, pp. 55-57; No. 7, pp. 57-61.
2. Ridiko L. UMZCH BB on a modern element base with a microcontroller control system. "Radiohobby", 2001, No. 5, pp. 52-57; No. 6, pp. 50-54; 2002, No. 2, pp. 53-56.
3. Ageev S. Superlinear UMZCH with deep environmental protection "Radio", 1999, No. 10 ... 12; "Radio", 2000, No. 1; 2; 4…6; 9…11.
4. Zuev. L. UMZCH with parallel environmental protection. "Radio", 2005, No. 2, p. 14.
5. Zhukovsky V. Why do we need the speed of UMZCH (or "UMZCH BB-2008"). "Radiohobby", 2008, No. 1, pp. 55-59; No. 2, pp. 49-55.