|
Arabic
Bengali
Chinese
English
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Malay
Polish
Portuguese
Spanish
Turkish
Ukrainian
Vietnamese|
ENCYCLOPEDIA OF RADIO ELECTRONICS AND ELECTRICAL ENGINEERING Four channel cassette recorder. Encyclopedia of radio electronics and electrical engineering
Encyclopedia of radio electronics and electrical engineering / Audio equipment In a modern portable studio, to perform primary sound recording, there must be a recorder, the functions of which can be performed by a multi-channel analog cassette recorder. The authors made an attempt to create a simple four-channel apparatus. Its feature is the adaptability of the recording path to the signal spectrum, as a result, the overload capacity of the path in the high audio frequency region was significantly increased. Subsequent signal processing using computer programs for noise reduction makes it possible to achieve a signal-to-noise ratio of 75...80 dB without phonogram companding. The high stability of the movement of the magnetic tape is provided by a speed stabilizer with a quartz oscillator. The design of the nodes described in the article is intended for the manufacture of a recorder based on the Mayak MP-249S LPM. Such a device, together with a portable mixing console, will be quite suitable for recording "live" concerts of musical ensembles and choirs that exist in many cities, and will become a useful addition to the equipment of amateur music studios. Digital methods of sound reproduction have firmly entered our everyday life. The same cannot be said for digital recorders - R-DAT tape recorders and CD recorders. These devices are still less accessible to a wide range of recording enthusiasts. A major disadvantage of these devices is the impossibility of high-quality recording of more than two channels. The 32-channel recording option available on some DAT recorders uses only 12kHz sampling rate and 45500-bit non-uniform quantization, which is not compatible with the Hi-Fi standard (DIN 8). At the same time, most mixing consoles have a four-channel output, and when recording, for example, "live" music, multi-channel recording provides additional opportunities for significantly improving the final stereo sound due to separate signal processing in the channels. There are digital multi-track recording systems, from the eight-channel AKAI DR-2430 ($24) to the 2424-channel Tascam MX-6290 ($XNUMX), but these are obviously not available to many. At the same time, the possibilities of analog multichannel sound recording are far from being exhausted. This is evidenced by the ongoing production of analog studio reel-to-reel tape recorders: A-820 from STUDER (Switzerland) and MTR-15 from ATARI (Japan). These are multi-speed tape recorders, complex and expensive, but they also have high technical characteristics: a frequency band of 40 ... 28000 Hz with a signal-to-noise ratio of 75 ... 78 dB. Also available is the Fostex X-34 port studio ($550), which provides four-channel recording on a compact cassette. The main disadvantages of analog sound recording are the insufficient signal-to-noise ratio: 50...56 dB (unweighted, on IEC-1 tape), as well as the falloff of the magnetic tape and large non-linear distortions when recording at frequencies above 6...8 kHz. An increase in the signal-to-noise ratio by 10 ... 15 dB is provided by various compander noise reduction systems: Dolby A, B, C, dbx, Hicom, Super D, etc. An alternative to companding today is the use of modern computer noise reduction algorithms available in Sound Forge, Cool Edit and other sound editors. These algorithms use the FFT and implement noise reduction not in two or four frequency bands, but in several hundred - thousand (set by the user) with a preliminary definition of noise reduction thresholds in each of the frequency bands. Such processing of a phonogram makes it possible to improve the signal-to-noise ratio by 15...20 dB and the signal-to-regular noise ratio by 40...50 dB. Attempts to improve analog biased treble recording have been made in a variety of ways. This includes limiting the depth of RF correction when recording high-frequency signals of a high level (ADRS devices from Akai and DYNEQ from Tandberg), and the use of dynamic bias. Of particular interest is the article by O. Zaitsev [1], which proposes a combination of the methods mentioned above for a reel-to-reel tape recorder operating at a tape speed of 9,53 cm/s. The proposed article presents the main components of a four-channel cassette recorder - a recorder for recording "live" music at a speed of 4,76 cm / s. An increase in the output of a magnetic tape, a decrease in the nonlinearity of the recording path at high frequencies is achieved by adapting the depth of high-frequency correction in the recording amplifier (US) and the high-frequency bias current. In order to save space, the article shows schematic diagrams of only one recording and playback channel (the rest are identical) and printed circuit boards for two channels, which is associated with the use of the K157UD2 chip. A four-channel version of the UV and US recorder will require a double set of printed circuit boards. The erasure and bias generator (GSP) ensures the operation of four recording channels. To reduce the bias current (when using IEC-1 magnetic tapes), the supply voltage is usually reduced. This leads to a deterioration in erasure and a change in the frequency of the GSP, which entails a disruption in the operation of the trap filters for oscillations with a bias frequency. We have developed a GSP based on a quartz resonator (clock) with a frequency multiplier by three (frcn = 98,3 kHz), operating at a constant supply voltage. The high-frequency bias modulator is made on the basis of a parallel oscillatory circuit with a variable quality factor. Oscillations of the quartz oscillator after the corresponding frequency division are also used in the digital PLL unit for stabilizing the speed of rotation of the motor shaft of the LPM, which is used as a collector DC motor with a tachogenerator (from the VCR "Electronics VM-12"). A functional diagram of the main components of a cassette tape recorder in a two-channel (stereo) version is shown in fig. 1.
The BG1 universal head unit is connected by the SA1 switch to a two-channel playback amplifier or to a recording amplifier. The playback amplifiers provide for electronic switching of time constants of 120 and 70 μs (for a tape based on Fe2 03 or Cr02) and blocking of the output in all modes of operation of the CVL, except for playback. The operation modes of the blocks are controlled by logic voltage levels of 0 and +5 V applied to the corresponding keys. In order to simplify the diagram, the control device and the power supply are not shown on it. Their structure depends on the type of CVL used and the requirements for a tape recorder. A low-pass filter with a cutoff frequency of 20...22 kHz is installed at the input of the recording channel. From the output of the ultrasound signal is fed to the amplitude detectors AD1, AD2 and through the filter plug LfSf, tuned to the frequency of high-frequency bias (HFF), to the recording head. The VChP voltage modulator is connected to the universal head through the tuning capacitor Sp. The output voltage of AD1 controls the Mod 1 VChP modulator: with an increase in the level and frequency of high-frequency components in the recorded signal (7 ... 20 kHz), the VChP voltage at the modulator output decreases. The voltage from the AD2 output is supplied to the high-frequency correction depth adaptation unit (link LkCkRkVT1), which reduces the high-frequency correction depth as the recording level and signal frequency increase. The GSP is designed as a generator with external excitation and consists of a frequency multiplier by three and a power amplifier, the load of which is the erasing head BS1. The input of the multiplier receives fluctuations in the meander shape with a frequency of 32,768 kHz from a quartz oscillator located in the digital PLL of the LPM engine. The erasing head enters the oscillating circuit at the PA output, from which the erasing voltage is supplied to the modulators Mod 1 and Mod 2 of the recording channels (in the four-channel version and to the modulators of channels 3, 4). The speed stabilizer unit for the driving motor, made on the basis of a digital PLL, includes a quartz self-oscillator at a frequency of 32768 Hz, two frequency dividers (FC), a frequency-phase PFD detector that proportionally integrates the PIF filter, a DC amplifier of the UPC collector motor with a TG tachogenerator and a limiting amplifier UO. Stabilization of the engine speed is carried out due to feedback on signals from the TG. The sinusoidal voltage from the motor TG output in the limiting amplifier is converted into a sequence of rectangular pulses, which, after frequency division in DC2, are compared in frequency and phase in the PFD with the crystal oscillator pulses that have passed through DC1. The error signal from the output of the proportionally integrating circuit is amplified in the UPT and fed to the electric motor, as a result, the shaft speed changes until the frequency and phase of the pulse sequences at the PFD inputs match. Such a construction of the block makes it possible to obtain high stability of the average speed of the belt (no worse than ±0,05%) and to ensure the minimum coefficient of fluctuations in the speed of rotation of the capstan, which depends only on the accuracy of manufacturing rotating parts. Schematic diagram of the playback amplifier (UV) is shown in fig. 2. Here the scheme of one HC channel is described; other channels are arranged similarly. In playback mode, the signal from the universal head BG1.1 through the contacts of connector X2 and relay K1 is fed to the base of a low-noise amplifier made on a transistor VT4. Common to both channels are the relay control unit K1, K2, made on transistors VT1 - VT3, a parametric voltage regulator of -2,2 V on VD3, HL1, R12, C4 and voltage stabilizers ± 9,5 V of the op-amp supply, made respectively on the elements VT5, VD5, R24 and VT8, VD4, R28.
To reduce low-frequency noise, a direct connection of the head with the base of the low-noise amplifier transistor was used. The stabilization of the emitter current VT4 is performed using a resistor R10 connected to a stabilizer - 2,2 V. The high-frequency correction in the SW is achieved due to resonance in a parallel oscillatory circuit formed by the inductance of the head BG1.1 and capacitor C1. The circuit is tuned to the upper limiting frequency of the tape recorder 18 ... 20 kHz, and the resistor R7 provides the desired quality factor. Capacitor C3 reduces the level of high-frequency noise entering the stage input. The resistor R13 regulates the amplification of the cascade, changing the depth of the OOS through the elements C6, R11, R13 to set the nominal level of the output voltage of the SW. Elements C2, R8 additionally filter the power in the VT4 collector circuit. From the resistor R9, the amplified signal through the capacitor C5, the resistor R14 is fed to the non-inverting input of the op-amp DA1.1. The C7L1 series oscillatory circuit is tuned to the bias frequency and is a notch filter. This circuit is necessary for the simultaneous operation of the HC and the recording channel in the overwrite mode in tape recorders with two CVLs. When using one LSM, the contour elements are not installed. Op-amp DA1 is covered by OOS for direct current through resistor R18. For alternating current, the frequency-dependent OOS, which forms the necessary frequency response correction, operates through the R20R21 divider and the R19C11R17R16C8 circuit. The VT7 transistor switch connects the R23C14 circuit, providing for the Fe203 tape a change in the time constant of the corrective circuit from 70 to 120 μs. Capacitor C9 prevents excitation of the amplifier at ultrasonic frequencies. The signal from pin 13 of the OU through resistors R22, R25 (with a private key on VT6) goes to the output. Transistor VT6 is open in all modes of operation of the LPM, except for the playback mode, and blocks the passage of switching noise and other noise to the output of the tape recorder. A schematic diagram of one recording channel is shown in fig. 3.
The input signal through the capacitor C1 is fed to the base of the emitter follower on the transistor VT1 and then to the active low-pass filter with the approximation of the Zolotarev-Kauer frequency response [2], assembled on the elements R4, R5, R7, C4 - C6 and VT2. The cutoff frequency is chosen equal to 20 kHz, the slope of the frequency response in the suppression band is about 30 dB per octave. The divider R1R2 provides a voltage based on VT1, at which the bias voltage at the non-inverting input of the op-amp DA1.1 is close to zero. The LPF suppresses the ultrasonic components of the input signal that create audible beats with the oscillations of the GPS. Such components exist in the signal at the outputs of a stereo tuner (in the form of oscillations of the 31,25 or 38 kHz subcarrier frequency and their harmonics), as well as a CD player (as pulses of the sampling frequency of 44,1 kHz and its harmonics). The recording amplifier is assembled on the K157UD2 op-amp, the feedback circuit of which includes elements of low-frequency correction R10, R13, C10, C7, R8 and adaptive high-frequency correction C8, L1, R9, VT3. The depth of the RF correction is determined by the total resistance of the resistor R9 and the output resistance of the transistor stage at VT3. At low input signal levels, the transistor VT3 is close to saturation due to the base current flowing through the resistors R12, R27 and R25. The quality factor of the C8L1 circuit is maximum, the depth of the RF correction reaches 14 dB. The output of the recording amplifier (terminal 13 DA1) is connected through a low-pass filter R16C12, an isolation capacitor C17, a bias voltage filter plug C20L2, a resistor R31 that stabilizes the recording current, to connector X4, from which the signal is fed to connector X1 (see Fig. 2) and then through X2 to the universal head BG1. In addition, a divider of the signal R17R21C13 supplied to the recording level indicator, as well as the input of the detector on the elements C15, VD1, R23, VT7, R26, C19, which controls the high-frequency bias modulator, and the input circuit C11, R14 of the inverter on the transistor VT4, are connected to the output of the ultrasound. Resistor R26 provides the initial bias of the diode VD1 and the emitter-base junction VT7, increasing the linearity of the initial section of the detection characteristic. In the absence of RF components in the input signal of the detector, the voltage at the top terminal of the detector capacitor C19 according to the circuit is +1 V. The detector that controls the RF correction depth during recording through the VT3 transistor is made according to a full-wave circuit in the form of two emitter detectors C14R19VT5 and C16R22VT6 connected in parallel at the output, the inputs of which are supplied with antiphase voltages. The load of the detector is the elements R25 and C18. Resistor R24 limits the peak discharge current of capacitor C18. Resistor R27 creates the initial bias of the emitter-base junctions of transistors VT5, VT6. Parallel connection of these detectors doubles the frequency of the envelope ripple and reduces the distortion of the regulated signal due to the absence of even harmonics. As the level and frequency of the signal increase, the voltage across the capacitor C18 of the detector changes from +0,9 V to -2 V, causing the transistor VT3 to close and the depth of the RF correction to decrease. The bias voltage modulator is made on the basis of a parallel oscillatory circuit C22L3R32 with a quality factor controlled by changing the average loss resistance of the circuit by the transistor VT8 of the modulator. It is known that at the resonant frequency the resistance of the reactive elements of the circuit is Q times (Q is the quality factor of the circuit) greater than the series loss resistance. The role of the loss resistance is performed by the parallel connected elements R32, VD2 and the collector-emitter resistance of the transistor VT8. Since the current flowing in the inductive branch of the circuit is the same for inductance and equivalent loss resistance, the voltage drops across these elements are proportional to their resistances. So, with the quality factor of the circuit QE = 10 and the voltage amplitude on the circuit, for example, 50 V, the voltage amplitude across the loss resistance will be only 5 V, and a low-power low-voltage transistor can be used to change the quality factor of the circuit. To prevent opening with negative half-waves of voltage across the resistor R32 of the base-collector junction of the transistor VT8, a diode VD2 is used. Thus, the change in the quality factor of the oscillatory circuit is carried out by changing the output resistance of the transistor modulator VT8 with positive half-cycles of the voltage on its collector. It is known that the equivalent resonant resistance of a parallel circuit (at f = fo) is calculated by the formula Rer = QeVL3/C22 and will also change when Qe changes. Considering that the voltage from the GSP is supplied to the described circuit through the capacitor C23, we obtain a voltage divider in which the role of the lower arm is played by a parallel oscillating circuit L3C22 with elements R32, VD2, VT8 with a variable quality factor. Thus, the bias voltage is modulated. At low levels of the RF components of the signal at the output of the recording amplifier, a voltage of +1 V at the emitter VT7 of the detector saturates the transistor VT28 through resistor R8. In this case, the loss resistance of the circuit is minimal, and the bias voltage on the L3C22 circuit is maximum. Through the capacitor C21, it enters the universal head circuit. As the level of the RF components and (or) their frequency increases, the voltage at the upper output of the capacitor C19 according to the circuit decreases, the output resistance of the transistor VT8 increases (with positive half-waves of the voltage on the collector). In this case, the average loss resistance of the circuit over the period increases, and its quality factor and equivalent resonant resistance decrease. As a result, the bias voltage on the L3C22 circuit decreases. Elements R28, R29, R30 ensure the linearity of the modulation characteristic of the modulator on VT8 when the voltage on the circuit drops to 1/3 of the maximum. The advantages of the proposed modulator are high linearity of control, additional filtering of the bias voltage, simplicity, the possibility of modulating the bias voltage with an amplitude of up to 100 V when using low-voltage low-power transistors (lk max<100 MA, Uke max<20...30 V), for example, KT315B. The disadvantages include the presence of inductance L3 and the need to tune the L3C22 circuit to the GSP frequency. A schematic diagram of the erasure and bias generator is shown in fig. four.
Rectangular oscillations with a duty cycle of 2 and a frequency of 32,768 kHz are fed through the C1R1 circuit from the quartz oscillator of the digital PLL unit of the leading motor to the input of the C2L1 oscillatory circuit. To multiply the frequency, the third voltage harmonic of the "meander" shape is used, to the frequency of which the circuit is tuned. Elements R2, VD1, C3 provide the necessary mode of operation of the subsequent cascades of the GSP and their temperature stabilization. The emitter follower on the transistor VT1 matches the high resonant resistance of the L1C2 multiplier circuit with the input impedance of the power amplifier. The inclusion of the GSP is carried out by applying a voltage of +5 V to the connection point of the elements R2, R3, C4. The GSP power amplifier consists of an emitter follower on a VT2 transistor and a resonant amplifier on VT3, made according to a common emitter circuit with incomplete inclusion of the C6C7L2BS1 oscillatory circuit in the collector circuit. Resistor R4 is used to set the critical operating mode of the generator at a cutoff angle of the collector current close to 90 degrees. The role of the inductance of the oscillatory circuit is performed by the inductor L2 and the erase head BS1, the inductance of which is about 360 μH. Capacitor C7 is used to fine-tune the oscillator circuit to a frequency of 98,3 kHz. Resistor R7 serves to measure the emitter current (practically equal to the collector current) and, being an element of the OOS circuit, slightly increases the input resistance of the final stage, additionally stabilizes its mode. Elements C8, L3, C9 form an oscillation filter with the frequency of the GSP along the power circuit. Switch SA1 with resistor R8 changes the voltage (and current) of erasure and bias for various types of tapes - with normal ("Fe203") and high ("Cr02") bias levels. Incomplete inclusion of the oscillatory circuit (turn-on factor p \u0,22d 6) achieves a voltage swing on capacitor C85 of at least 8 V at a supply voltage on capacitor C12 of 1 V (for a tape with a normal bias level, switch SA110 is open) and about 2 V with closed contacts. If necessary, this voltage can be increased by reducing the inductance of the inductor L6. The voltage from the capacitors C7, C1 of the circuit is supplied to the bias voltage modulators that are part of the recording channels (see Fig. 3 and XNUMX). Schematic diagram of the digital PLL block the lead motor of the LPM is shown in fig. 5. It is made in accordance with the functional diagram (see Fig. 1). On transistors VT1, VT2 and a quartz "clock" resonator ZQ1 (FKB = 32768 Hz), a reference frequency generator is made, the oscillations of which from the resistor R7 are fed to the GSP unit and to the input of the frequency divider DCH1 {input CN1 DD1). It is made on digital microcircuits DD1, DD2 and the "AND" element on diodes VD1-VD4, which set the division ratio, as well as elements R14, R15, C9.
For the frequency division factor N1 indicated on the diode switching circuit, the frequency division factor N202 is 1. When the contents of the counter on DD202 reaches the value 2 = 8 + 64 + 128 + 12, logical "14s" will appear on pins 5, 6, 1, 1 of the DD1 microcircuit, the diodes VD4-VD14 will close and the reset pulse through the integrating circuit R9C1 will reset the counters DD2.1, DD1 to the initial state at the input R. By installing additional diodes at the outputs DD2, DD1, any value of the coefficient N2 from 511 to XNUMX can be dialed with a binary code. Pulses with a comparison frequency of 32768/202 = 162,2 Hz from pin 11 DD2 are fed to the input From the first trigger of the DD3 chip, on which the frequency-phase detector is assembled. The second input ChfD - input From the lower trigger circuit of the same DD3, which receives pulses from the second frequency divider ДЧ2, made on the other half of the counter DD2 (output - pin 5 DD2). The frequency division factor is selected N2 = 8. The input DF2 (pin 1 DD2) receives pulses from the output of the limiting amplifier, assembled on transistors VT3, VT4. A sinusoidal voltage from the tachogenerator of the DPLT electric motor acts at the input of the CR, the frequency of which is related to the engine speed by the ratio ftg = 38fdv. When the PLL is in the capture mode, the frequencies of the pulse sequences at the PFD inputs are equal, i.e. fqv/N1 = ftrg/N2 = 38fmot/N2 = 162 Hz. The reset inputs R DD3 receive pulses from the direct trigger outputs through the "AND" element on the diodes VD5 and VD6. The inverted output of the upper trigger according to the circuit (pin 2) is connected through a resistor divider R20R21 to the input of the key on VT8, and the direct output of the lower trigger (pin 13) through the divider R22R23 is connected to the key input on VT9. The output voltage of the PFD from the connection point of the current-limiting resistors R24, R25 is fed to a proportionally integrating filter R26C14R29C15, from the output of which the smoothed voltage through two emitter followers (VT10, VT5) is fed to a power amplifier based on transistors VT6, VT7. The load VT6 is a collector DC motor of the DPLT type with a tachogenerator, used in the VCR "Electronics VM-12". Transistor VT7 with resistor R19 dampens the motor and reduces the time of transients, chokes L1, L2 together with capacitors C12. C13 reduce collector switching noise. The described construction of the PLL block allows you to change the motor shaft speed exactly twice by simply switching the DD2 outputs. So, when connecting pin 11 DD3 to pin 4 DD2, the speed (and the speed of the tape) is halved, and when using pin 6 DD2, the speed of the LPM engine doubles. Let us present a method for calculating the division factor N1 using the example of a CVL of a Mayak M-249S-1 cassette tape recorder. Initial data: capstan shaft diameter dT = 3 mm, flywheel diameter dM = 91,2 mm, engine pulley diameter dsh = 13,5 mm, belt speed \/l = 47,625 mm/s. For the case of the absence of slippage of the belt, a calculation formula has been obtained that relates the above parameters:
We round the obtained value to the nearest integer N1 = 202, while the engine speed will be more than the nominal by (202,084/202 -1) 100% = 0,041%, which is quite acceptable. The oscillation frequencies at various points of the PLL block are as follows: fkv = 32768 Hz, ftg = N2fkv / N1 = 1297,7 Hz, fav = fqv/N1 = 162,2 Hz, fmotor = ftrg /38 = 34,151 Hz, p = f 60 = 2049 rpm. For n \u2049d 5,6 rpm, the voltage supplying the DPLT motor at idle is Udv \u5,8d XNUMX ... XNUMX V. The calculation of the coefficient N1 can be performed for other parameters of the CVL, and the found value of N1 is then typed in binary code using diodes at the outputs of the counters DD1 and DD2 (see Fig. 5, coefficient designations in DD1 and DD2). Construction and details. The blocks of the cassette recorder are made on printed circuit boards made of one-sided foil-coated fiberglass with a thickness of 1,5 mm. On fig. 6 shows the recording channel board,
in fig. 7 - GSP board (click to enlarge),
in fig. 8 - playback channel board,
in fig. 9 - board of the digital PLL of the LPM engine (click to enlarge).
Due to the high mounting density and one-sided arrangement of printed conductors, some of the connections (mainly power circuits) are made with wire jumpers soldered from the side of the printed conductors. The blocks used constant resistors MLT-0,125, tuning resistors - SPZ-1 (playback channel), SP5-16 (GSP). The deviation from the ratings of most of the elements indicated in the diagram should not exceed ± 10%. For resistors R17, R19, R20, R21, R23 in the playback channels, as well as R4, R5, R7 in the recording channels, the deviation is allowed no more than ±5%. The resistors on the printed circuit board of the recording path are installed perpendicularly, and the leadless resistors R24 (R24') are placed on the side of the printed conductors. Capacitors of filters and correction circuits C11, C14 (in playback channels) and C4, C6, C8 (in recording channels) - K73-17 series with a deviation of no more than ± 5%. Capacitors C6 (K31 -10), C7 in the GSP and C20-C22 in the recording channels must have an operating voltage of at least 100 V. Oxide capacitors - K50-16 or K50-35, capacitor C14 in the PLL - K53-4, the rest - from the KTM, KM series. The inductance coils L2 in the recording channels, as well as L1 in the GSP, each contain 80 turns of PELSHO 0,12 wire and are placed in armored ferrite magnetic cores OB-14, the cups of which are glued with a gap formed by two layers of tracing paper. Coils L1 in the playback channels have 185 turns, and L1 in the recording channels - 130 turns of the same wire and are placed in the same magnetic circuits. The L3 coils in the recording channels are placed in the OB-19 magnetic circuit and contain 80 turns of PELSHO 0,22 wire each. The cups of the magnetic circuit are glued with a similar gap. Before gluing the coils, it is desirable to measure their inductance (at frequencies corresponding to the working ones) and, if necessary, adjust the number of turns. As L2, L3 (GSP) chokes DPM-0,1 are used, as L1 (in the PLL) - a choke type DM-0,6. The L2 filter coils (PLL unit) are wound on a K16x10x4,5 ferrite ring of the 2000NM brand with a PELSHO 0,22 wire folded in half and contain 2x80 turns. The value of this inductance is not critical. Filter elements C12, L2, C13 (PLL) are placed near the motor on a small printed circuit board. Transistors KT3102E (VT4 in the recording channels) can be replaced with KT3102D, preferably in metal cases. Other transistors can be used with other letter indices. Instead of diodes of the KD522 series, diodes KD521A are applicable, and instead of microcircuits of the K561 series - KR1561. ZD24.12002 was used as a universal head in a two-channel (stereo) version, a four-track block 7N10S (BB45), an erasing head of the ZS12.4210 type from the Mayak cassette recorder was used in the four-channel version. Due to the absence of erasing heads for the entire width (3,81 mm) of the tape, four-channel recording should be done on a pre-demagnetized (for example, by a choke) tape of a compact cassette. Relays RES-1 are used as switches K2, K49. The manufacture and adjustment of the tape recorder units can, of course, be prepared by radio amateurs who have measuring instruments: a low-frequency oscillation generator (frequency 20 Hz ... 200 kHz), an electronic oscilloscope with a frequency range of 0 ... 1 MHz, a millivoltmeter (with limits of 1 mV ... 1 V) and an electronic frequency meter (frequency range 20 Hz ... 200 kHz). Establishment start with the digital PLL block of the lead engine LPM. A C12L2C13 filter and a motor collector circuit are connected to the assembled block. The winding of the tachogenerator is connected with one terminal to the common wire, the other - to the left terminal of the capacitor C13 according to the scheme. Resistor R27 is temporarily unsoldered, and resistor R26 is replaced with a variable one with a maximum resistance of 300 ... 500 kOhm. The unit is supplied with a supply voltage of +15 V. Using an oscilloscope, they are convinced of the presence of oscillations of a quartz oscillator (on the VT2 collector). In their absence, reduce the resistance of the resistor R2 until stable oscillations are obtained. If there are no oscillations at a resistance close to zero, then the quartz resonator is replaced. The frequency meter checks the oscillation frequency, which should be within 32768 ± 20 Hz. Using an oscilloscope and a frequency meter, the presence of rectangular pulses and their frequency are checked at the output of the first frequency divider (pin 3 DD3). The pulse amplitude is about 10 V, the frequency is 162,2 ± 0,1 Hz. By reducing the resistance of the variable resistor included instead of R26, the voltage on the engine is increased to 5,6 ... 5,8 V. It is desirable that the engine be installed in the LPM and a belt is put on its pulley. The initial setting is carried out at idle speed of the LPM (the cassette is not inserted, the pressure roller does not touch the capstan). An oscilloscope checks at the output of the tachogenerator for the presence of sinusoidal oscillations with a swing of about 0,5 V and rectangular pulses with an amplitude of 9 ... 10 V on the VT4 collector. By adjusting the variable resistor, a pulse repetition rate of 1298 Hz is achieved, while at the output of the second frequency divider (pin 5 DD2), the pulse frequency should be equal to 162,2 Hz. Then turn off the power to the unit, unsolder the variable resistor, measure its resistance with a digital device and solder the constant resistor of the closest value in place of R26. Install the previously removed resistor R27 and turn on the power. The electric motor must have a shaft speed of 2049 rpm, while the pulse frequency at terminals 3 and 11 of DD3 must be equal to 162,2 Hz, which does not change when the LPM flywheel is braked with a finger. With an increase in load, the voltage on the motor and the current consumption should only increase from 60 ... 70 mA (at idle) to 300 ... 350 mA while maintaining the specified speed. The final setting of the block is made when playing back the recording of the measuring tape (part "E"). The signal frequency at the playback channel output should be within 3150±20 Hz (±0,6%). If the obtained frequency value does not correspond to the nominal one, it is necessary to calculate a new division factor N,, set it using diodes VD1 - VD5 and re-measure the signal frequency from the measuring tape. GPS setting produced in the following order. Open switch SA1. The base of the transistor VT2 is connected to a common wire through a 0,01 μF capacitor and the maximum resistance of the variable resistor R4 is set. A measuring generator is connected to the input of the block, on which the effective value of the voltage is set to 1 V and the frequency is 98,304 kHz (controlled by a frequency meter). Connect the Y input of the oscilloscope to the emitter of the transistor VT1. The recording mode is turned on by applying power and control voltage and using the L1 coil trimmer, tune the L1C2 circuit to resonance (according to the maximum signal swing). If it is impossible to adjust the circuit with a trimmer, you can change the capacitance of capacitor C2. At the end of the tuning, they are convinced of its correctness by tuning the frequency of the generator. The amplitude of the signal at the emitter VT1 should decrease both with increasing and decreasing frequency. The coil trimmer L1 is fixed with hot glue. Next, the output of the 0,01 μF capacitor is disconnected from the common wire and connected to the output of the measuring generator, on which the signal swing is set to not more than 0,5 V. The erase head is connected to the unit and capacitor C7 is soldered from the unit. An oscilloscope using a divider 1:10 (input capacitance - no more than 15 pF) is connected to the output of the GSP. +15 V power supply and +5 V control voltage are turned on. By changing the generator frequency, determine the frequency f( of the resonance of the C6L2BS1 circuit (by the maximum voltage, the swing of which should be 30 ... 60 V). The value of f1 must be greater than the nominal f0 = 98,304 kHz. By changing the frequency of the generator, make sure that the C6C7L2BS1 circuit is accurately tuned to a frequency of 98,3 ± 0,5 kHz. After turning off the power, connect the GSP input to the output of the PLL crystal oscillator (resistor R7). The PLL unit and the +15 V GSP supply voltage are turned on. The oscilloscope is connected to the GSP output. By reducing the resistance of the resistor R4, the amplitude of the signal at the output of the GSP is not less than 80 V. The shape of the collector current pulses VT3 (at the resistor R7) is close to cosine: the current amplitude is not more than 0,15 A, and the cutoff angle is 70 ... 80 degrees. The voltage swing on the erase head must be at least 70 V when the supply voltage on the capacitor C8 is about +12 V. The erase voltage shape may differ from sinusoidal. Setting the playback path (described in a two-channel version) consists in setting the angle of inclination of the working gap of the universal head, the nominal level of the output signal, checking the phasing of the channels and setting the RF correction. A universal head is connected to the X2 connector of the playback channel board, a millivoltmeter and an oscilloscope are connected to the X1,2 connector (pins 5). +1 V voltage is applied to resistors R27 and R15. The supply voltage +15 V and -5 V are turned on. A cassette with a measuring magnetic tape (part "H") is installed in the tape recorder's LPM and the working stroke is turned on. The position of the GU with the help of an adjusting screw is set to the maximum return at frequencies of 14 ... 0 kHz. The nominal output level of 550 dB (1 mVrms) was determined by the authors by playing back a 45 kHz auxiliary recording of a SONYTC-K4 tape recorder. This tape recorder was factory tuned using a SONY P-81-L-333 test tape (0 Hz, 3 dB) [550]. The rated voltage of 333 mV at a frequency of 400 (13) Hz, when adjusted by the measuring tape, is set by the resistor R2, first in the first channel (output 1 HZ), then in the second (output XNUMX HZ). The phasing of the channels is checked on a 1 kHz signal (part "U") by connecting pins 1, 2 of the XZ connector. With the correct phasing of the channels, the output voltage will not change or decrease slightly (by no more than 1 ... 2 dB), if it is incorrect, it will be close to zero. In the latter case, you need to swap the conclusions of one of the heads (BG1.1 or BG1.2). The RF correction is adjusted individually in each of the channels by selecting the capacitor C1 according to the minimum frequency response unevenness in the region of 5 ... 14 kHz when playing frequency packets (the “Ch” part) of the measuring cassette. At a frequency of 10 kHz, the frequency response should not exceed 3 dB. In conclusion, the channel is blocked by applying a voltage of +5 V to the anode of the VD6 diode and switching the time constant of 70/120 μs by temporarily turning off the voltage of +5 V from the resistor R27. RџSЂRё establishing a recording path first, they check the cutoff frequency of the low-pass filter, set the frequency of the high-frequency correction circuits to 18 kHz, adjust the L2C20 notch filters (see Fig. 3) to the bias frequency, and tune the L3C22 circuits of the VChP modulator. Then, the optimal bias current and the limits of its adaptation are set, as well as the nominal recording level and recording current. The rms value of the input voltage of the recording channels, equal to 110 mV, was chosen as the maximum input level. This level corresponds to 0 dB of the recording channel characteristics given below. For adjustment, a measuring generator is connected to the inputs of the recording channels and its output voltage is set to 110 mV. Turn on the power and check the cutoff frequency of the input low-pass filter (on pins 2 and 6 of the DA1 chip) at a level of -3 dB, it should be 20 ... 22 kHz. Attenuation in the LPF at a frequency of 44,1 kHz must be at least 36 dB. The constant component of the voltage at the output DA1 (terminals 13, 9) should not exceed ±0,5 V, otherwise the resistor R2 should be selected. Then the generator voltage is reduced by 20 dB (up to 11 mV) and the frequency of the maximum frequency response rise is determined (terminals 13, 9 DA1), which should be 17 ... 18 kHz. If the frequency does not correspond to this value, the capacitance of the capacitor C8 is selected. By switching the generator frequency to 1 and 18 kHz while maintaining the input level of 11 mV, the correction depth is determined, which should be within 14 ± 1 dB. On fig. 10 shows the family of frequency response of the recording channel, measured at various levels of the input signal (from 0 to -24 dB). Due to the action of the auto-regulation circuit, the depth of high-frequency correction with increasing input signal level decreases to 2 dB, which prevents tape overload at high frequencies. It is not necessary to measure all these characteristics due to the high complexity of the point-by-point measurement process. We measured these characteristics in automatic mode using a PC, which will be described in more detail below. It is enough to measure the rms voltages at pins 13 and 9 at frequencies of 1 and 10 kHz. They should be 1,2 and 1,6 V, respectively, with an input voltage of 110 mV.
Check the frequency response of the VChP adaptation detector, made on the elements C15, VD1, R23, VT7, R26, C19. A voltage of 110 mV with a frequency of 400 Hz is applied to the input of the recording channel. Measure the constant voltage at the emitter VT7, which should correspond to 1 V. Increase the frequency of the input signal to 7,9 kHz, the voltage at the emitter VT7 should become close to zero. With a further increase in frequency (up to 16 ... 20 kHz), the voltage drops to -1,2 ... -1,6 V. If the measurement results do not match the given data, the value of the capacitor C15 should be selected within 390-910 pF. Next, the GSP outputs to the modulators are temporarily connected to contacts 1, 2 of the X4 connector of the recording board. Solder capacitors C21, C21'. Turn on the power of the recording board and GPU. The filter plugs L2C20 are adjusted to the minimum voltage of the VChP on the capacitor C12 (range 1 ... 2 V). Turning off the power of the GPU and the recording board, switch the outputs of the GPU to the right (according to the scheme) plates of capacitors C23, C23. Set the capacitors C21, C2G with a nominal value of 75 pF and the voltage at the output of the measuring generator is equal to zero. After turning on the power of the units, connect an oscilloscope to the capacitor C22 through a 1:10 divider and tune the L3C22 circuit to a frequency of 98,3 kHz at the maximum voltage, using the trimmer L3. If it is impossible to tune into resonance, capacitor C22 should be selected. With fine tuning, the voltage swing across the capacitor C22 is 80 ... 100 V. Next, set the frequency of 16 kHz on the measuring generator and smoothly increase its output voltage from 0 to 110 mV. The voltage swing across the capacitor C22 should decrease to 30 ... 40 V. An important operation is to set the optimal bias current for small signals. The generator voltage is set to 11 mV and oscillations with frequencies of 1 and 10 kHz are alternately recorded in one of the channels for different capacitances of the capacitor C21 (22 ... 110 pF). The recording is played back and the option is noted in which the voltages with frequencies of 1 and 10 kHz are the same. The value of C21 corresponding to this option is the optimal one. The procedure is repeated for the other channel. The final operation is to adjust the sensitivity of the recording level meter and set the nominal recording current. A signal is recorded with a frequency of 1 kHz and an RMS value at the input of 110 mV for various values of the resistor R31. At the same time, the upper terminal of the resistor R21 is connected to the input of the recording meter (preferably peak). Selecting the resistance R21, achieve a meter reading of 0 dB. During playback, a recording option is noted that provides a voltage of 550 mV at the output of the playback channel. The value of the resistor R31, corresponding to this option, is optimal. The end-to-end frequency response of the tape recorder is measured in the range of 20 ... 20000 Hz for various recording levels: 0, -6, -12, -18 dB. To measure the final end-to-end frequency response of the tape recorder, we used the following technique: the generation of test signals, their registration and processing were performed on a PC. The test signal was formed using the Cool Edit Pro 1.2 program. The test signal consisted of three parts: the first two parts were tone signals with a duration of 1,5 each with a frequency of 1 kHz and levels of 5 and -30 dB, respectively. The third part is a signal with a duration of 20 s with an exponentially changing frequency in the range of 20000...30 Hz. To generate a signal with an exponentially changing frequency, the Generate Tones command was used with the following settings: Duration 20 seconds, Initial Settings 20000 Hz, Final Settings XNUMX Hz, Log Sweep, Flavor Sine. Two tonal pulses with different levels are used to calibrate the final characteristics visualization program. To take into account the uneven frequency response of the used sound cards, the test signal was corrected using a 30-band graphic equalizer in the Cool Edit Pro program. The test signal was output from the PC through a Creative SB 128 sound card. The test signal recorded on magnetic tape was input into the PC during playback using a YAMAHA YS-724 sound card. The uneven frequency response of input-output devices (without a tape recorder), measured in the frequency range of 20...20000 Hz, did not exceed ±0,5 dB (after correcting the frequency response of sound cards in the test signal). Further, the recorded file was processed in order to determine the signal envelope and record the measurement results in the usual coordinates along both axes. For this purpose, a program for visualizing the results of measuring the frequency response was written in the Delphi language. A simplified block diagram of the program operation algorithm is shown in fig. eleven.
The test signal envelope was calculated using the moving average method. To do this, the following actions were performed on the test signal: the modulus was calculated, then the points of the resulting frequency response were calculated by averaging the data over a given time interval. The averaging time of the envelope changes quickly within 0,1...2 s. Typical values of time intervals were 0,1...0,4 s. The program has a simple graphical interface, which provides for the possibility of arbitrary scaling of the displayed frequency response along both coordinate axes, saving the calculation results both in a graphical format and as an array. This program also works with test signals in the form of segments of narrow-band (1/3 and 1/6 octave) noise, connected without phase breaks and covering the range of 20...20000 Hz. These signals were used to measure the frequency response of acoustic systems and microphones by sound pressure. On fig. 12-15 shows the amplitude-frequency characteristics of the recording-playback channel for the following cases: - standard recording method (with fixed high-frequency correction and bias current) - fig. 12;
- recording method with adaptive high-frequency correction (fixed bias current) - fig. 13;
- recording method with bias adaptation (fixed depth of high-frequency correction) - fig. 14;
- recording with adaptation of high-frequency correction and bias - fig. fifteen
Turning off the adaptation of high-frequency correction is carried out by connecting the VT3 collector to a common wire, turning off the adaptation of the VChP - by soldering one of the terminals of the capacitor C15 from the board. Measurements of the parameters of the recording-playback path were carried out using a BASF Fe 1 type IEC-1 tape. The upper cut-off frequency of the end-to-end frequency response with the standard recording method with a signal level of 0 dB is only 8 kHz (3 dB decay), the return at a frequency of 15 kHz drops below -24 dB. In the frequency range 15,6. ..18 kHz there is an overtone caused by beats of the 5th harmonic of the recorded signal and the bias frequency, with a level of -16,5 dB (15%). The waviness of the frequency response in the frequency range of 20...160 Hz (the so-called "snake") is explained by the commensurability of the recording wavelength with the dimensions of the working surface of the used head [4]. Since the shape of the frequency response below 3 kHz is practically independent of the recording level, the graphs in Fig. 13-15 are given in the range of 2,5 ... 20 kHz. Comparison of recording methods can be made according to various criteria, we have chosen the level of return of the magnetic tape in the through channel at frequencies of 10 and 15 kHz. In table. 1 shows the levels in dB for the four methods studied.
At 10 kHz, HF-only adaptation is preferable to high-frequency correction adaptation, but at 15 kHz, these adaptation methods (individually) give the same result (return -16,5 dB). The joint use of the adaptation of the correction of the HF and HF at a frequency of 15 kHz allows you to get a return of -6 dB, which is 10,5 dB higher (!) than when using these methods separately. The non-linearity of the tape recorder was estimated using the third-order difference tone method [4]. The measuring signal was formed using the Cool Edit Pro 1.2 program as a sum of two harmonic oscillations: one with amplitude A and frequency f1, the other with amplitude A/2 and frequency f2, with f2 = 2f1 - 500. The product of the nonlinearity of the magnetic recording path (including, in addition to the electronic part, also a universal head and a magnetic tape) in the form of a difference combination tone with a frequency of 500 Hz was measured by a spectrum analyzer at the output of the left playback channel. To do this, the signal was entered into a computer and analyzed by the Audio Tester 1.4 program (spectrum analyzer mode). The tolerance curve was measured by varying the frequencies of the test signal and maintaining a constant difference tone level. The latter was chosen to be 2,5% (-32 dB) of the nominal output level (550 mV). Naturally, as the frequencies f1, f2 of the test signal increase, the amplitudes of its components (A and A/2) decrease. The measurement results are given in table. 2 for the component frequencies and the peak-to-peak test signal at the output of the recording amplifier (in volts and dB relative to the nominal peak-to-peak of 3,4 V).
It is noted in [4] that for "good" recording-playback channels, the slope of the curve does not exceed 15 dB at a tape speed of 19 cm/s at the highest frequency of the range. Using the adaptation of the bias and the depth of the high-frequency correction during recording, this decay was only 3,2 dB at a tape speed of 4,76 cm / s (!). It should be noted that this article describes a tape recorder with a deeper regulation of the bias current (up to 10 dB) than in the known systems of dynamic bias (4...6 dB) and dynamic regulation (2,6 dB) [1]. A subjective assessment of the sound quality of phonograms recorded on this tape recorder from CDs showed a high overload capacity of the path. The maximum recording levels measured by the peak indicator (τint = 1 ms, τrep = 350 ms) reached +6 dB without noticeable distortion. For recording, soundtracks with sharp beats, cymbals and a powerful bass line were used. The recorded phonogram has undistorted "bass", does not lose brightness and richness, differing from the original one only by the appearance of a small tape noise (unweighted signal-to-noise ratio 52...54 dB) in pauses. To suppress the noise of four-channel phonograms made on a cassette recorder, the Cool Edit Pro program was used after entering them into a computer. Noise suppression in each channel is performed in two stages: at the first stage, the "noise profile" is determined as statistical information necessary to optimize the noise suppressor; on the second - there is actually a suppression of noise components in the processed phonogram. Typical settings for high quality squelch performance are: Snapshots in profile: 300; FFT Size: 4096; Precision factor: 7 Smoothing amount: 1.25 Transition width: 3. Typical signal to noise improvement is 15...20 dB. For regular interference, the improvement can reach 40...50 dB. Literature
Authors: A.Filatov, K.Filatov, Taganrog, Rostov region.
Machine for thinning flowers in gardens
02.05.2024 Advanced Infrared Microscope
02.05.2024 Air trap for insects
01.05.2024
▪ Perfectionism leads to insomnia ▪ 2,5" 3TB Toshiba portable drives ▪ Dangerous hydroelectric power station
▪ section of the site Funny puzzles. Article selection ▪ article You can't serve two masters. Popular expression ▪ article Brown bear. Travel Tips ▪ article Electronic notebook. Encyclopedia of radio electronics and electrical engineering ▪ Ratchet article. Focus Secret
Home page | Library | Articles | Website map | Site Reviews
www.diagram.com.ua |
See other articles Section 
Latest news of science and technology, new electronics:
All languages of this page