Direct conversion transceiver circuit on digital microcircuits. Basic parameters of the receiving path. Basic parameters of the transmission path

Direct conversion transceivers (DCT) are distinguished by their simplicity of design with sufficient good parameters and have long attracted the attention of radio amateurs. To a large extent, this was facilitated by articles and books by the famous designer and popularizer of direct conversion technology V.T. Polyakov RA3AAE, especially, which has become a reference book and textbook for entire generations of radio amateurs.

Previously, Radio magazine had already published several successful designs of single-band TPPs with phase suppression of the mirror sideband, built using traditional, now classic, circuitry based on LC low-frequency phase shifters (LFPS). The main disadvantages of such solutions include single-band, low (by today's standards) suppression of the mirror sideband, labor-intensive winding of multi-turn coils and adjustment of the low-frequency waveform, susceptibility to magnetic interference, which presented certain difficulties when repeating the design by radio amateurs, especially beginners. I would especially like to note the TPP on 160m, in which, at the cost of certain compromises, the author managed to remove labor-intensive elements and create an easily repeatable design, which greatly contributed to the introduction of hundreds of beginning radio amateurs to amateur radio communications on HF.

Thanks to the widespread availability of new high-speed digital microcircuits and high-quality low-noise op-amps, it became possible to implement new approach in the construction of single-sided TPPs, using digital switches as a mixer and using well-developed circuitry of functional units on the op-amp in the rest of the circuit.

The version of the main board of the TPP offered to your attention is a logical continuation and implementation of this approach in the construction of single-sided TPP, described in detail in. The author set himself the task of making a design on a modern element base, easily repeatable at home and not requiring any complex adjustment and tuning work or a fleet of measuring instruments - an ordinary digital multimeter, preferably with a capacitance measurement function, is sufficient. Successful repetition requires only accuracy and patience. If serviceable parts of the required rating are used and there are no errors in installation, the main board of the TPP starts up immediately, providing very high parameters, at least not worse than those stated.

Basic parameters of the receiving path

• Operating frequency ranges, MHz - 1.8, 3.5, 7 and 14
• Bandwidth of the receiving path (level - 6 dB), Hz - 400-2500
• Sensitivity of the receiving path from the mixer input (bandwidth 2.1 kHz, S/N ratio - 10 dB), µV, no worse - 0.3*
• Maximum total gain - 250 thousand
• Self-noise voltage at the ULF output at maximum Kus and a resistance of 50 ohms connected at the input of the TPP, no more than, mV - 25
• Allowable range of input signals in the passband, dB, not less than - 100
• Dynamic range for cross modulation (DD2) at 30% AM and 50 kHz detuning, not less, dB
  • On the 160m range – 116*
  • On the range 80m – 110*
  • On the 40m range – 106*
  • On the 20m range – 106*
• Selectivity in the adjacent channel (with a detuning from the carrier frequency of -5.5 kHz + 3.0 kHz), no less, dB – 80
• Suppression of specular sideband, not less, dB
  • On the 160m range – 54*
  • On the 80m range – 52*
  • On the 40m range – 46*
  • On the 20m range – 48*
• Squareness coefficient of end-to-end frequency response
  • (at levels -6, -40dB) — 1.4
  • (at levels -6, -60dB) — 3.2
  • (at levels -6, -80dB) — 4
• AGC adjustment range when the output voltage changes by 12 dB, no less, dB - 72 (4000 times)
• RRU range, no less, dB - 84 (16,000 times)
• output power LF path at a load of 8 ohms, at less than 0.5 W
• Current consumed from an external stabilized power source 13.8V, no more, A - 0.3

Basic parameters of the transmission path

• Output voltage (at a load of 50 Ohms) in CW mode, not less than Veff - 0.7
• Suppression of signal carrier frequency, dB - no worse than 50*

* the indicated figure is limited by the capabilities of the equipment used for measurements and, in reality, may be higher.

1. To obtain a large dynamic range of the receiving path and efficient operation of the AGC, the cascade distribution of the gain coefficients of unregulated cascades has been optimized and the permissible levels of input signals in the passband have been expanded.
2. To obtain high selectivity, the principle of sequential selection is used, when in addition to the main active bandpass filter, in fact, in each amplifier stage the passband is limited at the level of 300-3000 Hz by the corresponding choice of values ​​of interstage isolating capacitors and in the OOS circuits.
3. To suppress the mirror sideband, a method is used, described in detail in and based on the use of a multi-link LF phase shifter in a 4-phase signal system, allowing relatively by simple means, despite the increased number of elements, obtain good suppression and high temperature and time stability of parameters. To obtain a 4-phase signal system, a digital phase shifter is used, which greatly simplifies the creation of multi-band designs.
4. Due to the fact that differential signal amplification is used in all critical (due to large structural dimensions and low signal levels) components (mixer-detector, preliminary ULF, low-frequency phase shifter - polyfuser), the design has good noise immunity, including interference from electrical networks.
5. To reduce the total number of transceiver parts and, accordingly, the size of the main board structural scheme The TPP was chosen in such a way that the most complex and bulky units (eight-link LF FV and the main FSS) are used both for receiving and transmitting signals.
6. Electronic switching of all operating modes of the transceiver is used.
7. The design is single-board, which eliminates the possibility of errors during the installation of parts and assemblies, and also ensures, in the author’s opinion, an optimal layout and good overall and mutual shielding of the main functional units. Application of single-sided printed circuit board printed conductors(the second side serves as a common wire - screen) allows you to make a high-quality board at home using the so-called “laser ironing” technology.

A possible functional diagram of the TPP is shown in Fig. 1. It consists of five structurally complete units. Node A1 consists of a four-band, switchable relay, low-pass filter, and broadband amplifier power, for which any known design that has been repeatedly described in amateur radio literature can be used, for example. Node A3 contains a two-link attenuator (the first link has an attenuation of -10 dB, the second -20 dB, which allows, with appropriate switching, to obtain four attenuation values ​​of 0, -10 dB, -20 dB, -30 dB and thereby optimally match the dynamic range of the receiving path TPP with real levels of input antenna signals), useful when working on a full-size antenna, and quad-band bandpass filter, for which you can use any of the known designs of 50-ohm three-circuit PDFs, also repeatedly described in amateur radio literature. Node A4 is a local oscillator based on a single, non-switchable generator at frequencies of 56-64 MHz, tunable mechanically using a control unit or with electronic frequency tuning using a multi-turn resistor, and a controlled frequency divider with a variable division ratio of 1,2,4,8. The necessary stability with the help of a DAC and digital frequency reading is provided by node A2, made on the basis of a ready-made digital scale “Makeevskaya”, which can be purchased in many regions of Ukraine and Russia and is not described here as an option for self-made we can recommend the well-proven development of A. Denisov.

The main processing of the signal in the reception and transmission modes - its conversion, suppression of the mirror sideband and filtering - is performed by node A5 - the main board of the TPP.

In the receiving mode, the signal from the PDF output is sent to the mixer-detector U3, which uses half of a high-speed dual four-channel switch FST3253 with an average switching time of 3-4nS. The second half of this switch is used as a mixer-modulator U2 when operating for transmission.

The use of a four-channel switch FST3253 as a mixer made it possible to simplify the circuit, since part of the functions of the phase shifter is performed by the internal control logic of the switch, the address inputs of which receive control signals from counter 4 (node ​​U4). Switching of the working sideband occurs when a USB/ULB signal is supplied from the control circuit by changing the order of incoming control pulses from the counter to the switch. In this case, the local oscillator frequency should be four times higher than the operating frequency. As a result, a four-phase system of signals is formed at the output of the mixer, which, after preliminary filtering by single-stage low-pass filters Z3...Z6 and pre-amplification by differential amplifiers A3 and A4, through closed contacts electronic switch SA3.2...SA3.5 are supplied to the low-frequency phase shifter U6. Differential amplifiers A5, A6 are connected to the output of the latter, compensating for the attenuation of signals in the phase shifter. Next, the useful sideband signals, which received a zero phase shift, are added on the adder A10, and the mirror sideband, which received a 180° phase shift, are subtracted and suppressed. The main active bandpass filter is connected to the output of the adder through the closed contacts of the electronic switch SA3.6, which is a series-connected normalizing amplifier A8, FSS Z7, consisting of a third-order low-pass filter and a sixth-order low-pass filter and a buffer amplifier with a differential output A7.

The filtered useful signal through the closed contacts of the electronic switch SA3.1 is supplied to the ULF, consisting of a voltage-controlled amplifier A6 and the final ULF A5, to the output of which is connected the loudspeaker BA1, the AGC detector U5 and gain and volume controls. The TPP enters transmission mode either when the pedal is pressed or when the key is pressed.

In the first case, a +TX signal is generated in the control circuit U7, which switches the contacts of the electronic switch SA3 to the opposite position, turns off the mixer-detector U3 and activates the mixer-modulator U2. The microphone path is turned on. To increase the energy efficiency of the transmitter by 8-9 dB (6-8 times in power), compression of the dynamic range of the speech signal is used using a sequential phase limiter, consisting of an amplifier-limiter A12, a single-link phase shifter U9 and a cleanup limiter U8. Next, the generated signal, through the closed contacts of the electronic switch SA4 and SA3.6, enters the main active bandpass filter, which is a series-connected normalizing amplifier A8, FSS Z7, consisting of a third-order low-pass filter and a sixth-order low-pass filter and a buffer amplifier with a differential output A7. The useful signal, filtered from harmonic residues, from the direct and inverse outputs of the FSS through the closed contacts of the electronic switch SA3.2 ... SA3.2 is supplied to the inputs of the low-frequency phase shifter U6 combined in pairs, which is necessary for the correct phasing of the modulating quadrature signals resulting from the output of the latter. These signals pass through differential amplifiers A5, A6, which compensate for the attenuation of the signals in the phase shifter, and are fed to a quadrature mixer-modulator U2, at the output of which the signals of the useful sideband, which received a zero phase shift, and the mirror sideband, which received a phase shift of 180°, are subtracted and are suppressed.

In the second case, when you press the key, in the control circuit U7, in addition to “+TX”, two more signals are generated - “+MIC off”, which turns off the microphone path and connects the telegraph signal generator G2 by switching the contacts of the electronic switch SA4, and the signal “+KEY ", directly controlling the keying of this generator. The tone telegraph signal through the normally closed contacts of the electronic switch SA4 and SA3.6 enters the main active bandpass filter and passes the same path as the microphone one.

Schematic diagram node A5 - the main tract of the Chamber of Commerce and Industry is shown in Fig. 2. As you can see, some components are already known to us and are described in detail in, some features of their operation and requirements for parts are also given there. Therefore, we will not describe them in detail here.

In the initial position, with contacts X13, X15 not connected to the common wire, the path operates in receive mode. A low level of the +TX signal goes to pin 1 of DD2 and allows the operation of the mixer-detector, while through the inverter DD1.1 74AC86, a high level goes to pin 15 of DD2, prohibiting the operation of the mixer-modulator. When switching to transmit mode, the +TX signal high level(approximately +8.0...8.5 V) is supplied through a divider on resistors R2R3, matching voltage levels, to pin 1 of DD2 and prohibits the operation of the mixer-detector, while through the inverter DD1.1 a low level is supplied to pin 15 of DD2, allowing operation of the mixer-modulator.

So, in the receiving mode, the signal from the PDF output through the C4R7 circuit is supplied to the four-phase (quadrature) mixer DD2, made on the lower half of the four-channel switch FST3253 (it is possible to use SVT3253 and other analogues produced by different manufacturers with a slightly modified name). To increase performance, the switch is powered by an increased voltage of +6 V from the VR1 stabilizer. Resistor R7 improves balancing and equalizes resistance public keys(typical approximately 4 ohms with a technological spread of ±10%). A bias voltage from the divider R1R11 equal to +3V is supplied to the input of the switch through resistor R10, which ensures operation of the mixer in the most linear part of the characteristic. Control signals (heterodyne) to the switch come from a synchronous counter-divider by 4, made on D-flip-flops of the DD3 74AC74 microcircuit. They have a meander shape with a 90-degree phase shift. They are finally formed by the internal control circuit of the switch itself so that the four keys are opened one by one. For clarity, in Fig. 2, opposite the corresponding pins of the DD1 microcircuit, the phases of the output signal are indicated. Elements DD1.2, DD1.3, included in the feedback circuits of the synchronous counter, control the order of arrival of control pulses to the switch and are intended for selecting the working sideband. In the initial position, this is the top one, and when contact X3 is closed to the common wire, the bottom one is allocated.

Load capacitors (C21C28, C22C29, etc.) are connected to the output of each of the four channels of the quadrature detector, limiting the detector bandwidth to approximately 3000 Hz.

As I already noted in the above-mentioned article, the dynamic range of mixers made on the basis of modern high-speed switches (74NS405x, FST3253) is limited not by the mixer, but by the preliminary ULF from above due to direct detection of AM interference in it, and from below by its noise. DD2 can be improved by another 10...20 dB by installing additional low-pass filters after the mixer. This idea is implemented in the Chamber of Commerce and Industry by installing single-stage low-pass filters (R30C34, R31C35, etc.) with a cutoff frequency of approximately 6 kHz. In this circuit design, the use of preliminary ULF resistive filters at the input did not lead to any noticeable deterioration in sensitivity (at least I was not able to detect this instrumentally), but had the most positive effect on improving the overall or, if you like, real selectivity.

On the one hand, this provides good suppression of out-of-band interference, on the other hand, it introduces a noticeable additional phase shift into the useful signal, so the corresponding resistors and capacitors in all four channels must be thermally stable and selected for capacitance with an accuracy of no worse than 0.2% (here and further implies the accuracy of the selection of elements of four channels among themselves; the absolute value can have a scatter of up to 5%).

Op-amps DA3, DA4 NE5532, connected according to a differential measuring amplifier circuit, improve the symmetry of signals and suppress common-mode interference (AM detection products, interference with the mains frequency, etc.) proportionally Kus = 19 times. Such preliminary amplification is optimal, in the author’s opinion, in order to ensure high sensitivity and compensate for losses in the low-frequency phase shifter in the receive mode, without compromising the permissible range of input signals in the passband. Resistors in the feedback circuits R45, R46, R49-R52 must be selected with an accuracy of no worse than 0.5%.

Since the low-frequency filter is used for both reception and transmission, electronic switches DD4,DD5 HCF4066 are used to switch its inputs (can be replaced with similar ones from the CD4000 series or domestic 1561KT3). The outputs of the differential preamplifier through the electronic keys of the DD4 switch, open in the receiving mode (in this case, the +TX control signal is low and the electronic keys DD5 are closed) are connected to a four-phase eight-bar low-frequency RC phase shifter on elements R69-R126 and C57-C109. When switching to the transmission mode, a high level (approximately +8...8.5 V) of the +TX signal opens the electronic switches of the DD5 switch, connecting the LF PV inputs to the antiphase outputs of the FSS (pins 7 DA5.1 and DA2.2). In this case, transistor VT1, inverting the +TX control signal to a low level (approximately +0...0.5 V), closes the electronic keys of switch DD4, thereby turning off preamplifiers from LF PV and, accordingly, from the transmission path.

This LF PV, despite the increased number of elements, is simple in design. Thanks to the mutual compensation of phase and amplitude imbalances of individual chains, it is possible to use elements with a tolerance of ±5% (of course, the accuracy of the selection of quadruple elements should be no worse than 0.5%) while maintaining high accuracy of the phase shift. To facilitate the selection of elements, the option of low-frequency PV on identical capacitors was chosen. This option, compared to the one used, has slightly greater attenuation, which is easily compensated by increasing the gain of the preliminary stage. The value of the capacitance itself may be different - the optimal values ​​are in the range of 10-33 nF - with a larger capacitance, the pre-ULF can be overloaded, and with a smaller capacitance, the LF PV circuits become high-impedance and the danger of interference and interference increases. Options for possible resistor values ​​depending on the selected LF PV capacity are given in Table 1.

Table 1.

From the output of the low-frequency waveform, the signals are supplied to op-amps DA7, DA8, also connected according to the differential measuring amplifier circuit, which further improves the symmetry of the signals and suppresses common-mode interference (AM detection products, interference with the mains frequency, etc.) in proportion to Kus = 7 times. Such amplification is sufficient, in the author’s opinion, to compensate for losses in low-frequency waveform in the transmission mode. Resistors in the feedback circuits R130-R135 also need to be selected with an accuracy of no worse than 0.5%. Since in transmission mode the outputs of this differential stage are connected to a low-resistance load - a modulator (it is turned off during reception), the outputs of the op-amp DA7, DA8 are powered by pairs of complementary transistors VT8VT9, VT10VT11, etc. (any serviceable ones will do, for example KT315, 361 or KS547, 557). It would be more optimal to use high-quality medium-power op-amps, but they are not available in our area and, as experience has shown, the solution used works efficiently and reliably.

Next, the four-phase signal is fed to the inputs of a classic adder at the DA9.1 op-amp, where, thanks to the resulting phase shifts, the signals of the lower sideband are added and amplified, and those of the upper sideband are subtracted and suppressed. The signal from the output of the adder through a passive bandpass filter R160C127R161C128 is supplied to the first key (pins 1-2) of the electronic switch DD6 HCF4066 (can be replaced with similar ones from the CD4000 series or domestic 1561KT3), which is controlled by the second key (pins 8-9), turned on by the control inverter signal +TX. In the receiving mode, the +TX signal has a low level, so the first switch is open and the useful signal freely enters the input of the normalizing amplifier DA6.2. The main task of this cascade is to ensure optimal signal levels in both the receiving and transmitting paths of the TPP. In reception mode, its Kus = R122 / (R161 + R160) = 1.3 is selected small, which is necessary to ensure the maximum range of permissible signal levels in the passband. Capacitor C105 limits the bandwidth of this stage to approximately 3 kHz. When switching to the transmission mode, a high level (approximately +8...8.5V) of the +TX signal closes the first key and opens the third electronic key (pins 3-4) of the DD6 switch, thereby disconnecting the adder output from the normalizing amplifier and connecting parallel-connected microphone and telegraph outputs. If the microphone path is active (this is determined by the control signals MICoff and +KEY, but more on this below when describing the corresponding nodes), the gain of the normalizing amplifier Kus = R122/R140, and for the telegraph path Kus = R122/R129. This allows you to set the optimal levels of the modulating signal separately for the microphone and telegraph paths using trimming resistors R129, R140 during setup.

Further, in the receiving mode, the signal is received by an active main signal frequency filter (FSF), made on three 3rd order links connected in series - one high-pass filter with a cut-off frequency of 350 Hz on the DA5.2 op-amp and two low-pass filters with a cut-off frequency of 2900 Hz on Op amp DA6.1 and DA5.1.

To improve isolation and reduce interference in the power supply circuit, the stages of differential amplifiers DA3, DA4, DA7, DA8 and the rest of the small-signal part of the path (adder, FSS, MSO, etc.) are powered by separate integrated stabilizers VR2, VR3. Supply voltage dividers R72R73, R86R119, R96R153 create a bias voltage for the op-amp of the corresponding nodes with a unipolar supply.

The filtered signal from the FSS output is supplied through the R53C48 separating circuit (single-level high-pass filter with a cutoff frequency of approximately 300 Hz) to the input of the adjustable amplifier stage at the DA2.1 op-amp. Its gain is determined by the ratio of the total resistance of the resistor R29 connected in parallel in the OOS circuit and the channel resistance of the field-effect transistor VT3 KP307G (any transistors from the KP302, KP303, KP307 series, having a cut-off voltage of no more than 3.5 V at the highest possible initial drain current, are suitable here) to resistance of resistor R53. When the bias voltage at the VT3 gate changes from 0 to +4.5 V, Kus changes from 40 to 0.002, i.e. from +32 to – 54 dB, which provides effective automatic (AGC) and manual (RRU) control of the overall gain of the receiver . Figure 3 shows a graph of the dependence of the voltage at the output of the ULF on the voltage at the input of the DFT of the author's copy of the TPP, illustrating the operation of the AGC. Circuit R27R34С33 supplies half the signal voltage to the gate of transistor VT3, which improves the linearity of the adjustment characteristic, resulting in even input signal 2 Veff (the maximum possible signal at the output of the main bandpass filter), the level of nonlinear distortion does not exceed 0.1%.

In parallel with the drain and source terminals of transistor VT3, an electronic switch VT2 on a KP307G transistor is connected (possible replacements are the same as for VT3). When switching to the transmit mode, the high-level +TX signal (approximately +8.0...8.5 V) enters through a divider on resistors R28R37, which reduces the voltage level at the VT2 gate to +4.3...4.5 V, which leads to its full opening. The low channel resistance (approximately 50-80 Ohms) of the open transistor VT2 strongly shunts the resistor R29 of the OOS circuit, which leads to a decrease in the ULF Kus by about 16-20 thousand. A small residual ULF transmission coefficient (Kus = 0.1-0.15 times) practically does not interfere when working with a microphone and allows you to get a quiet but clear signal of self-control when working with a telegraph. The D6R38C38 circuit ensures fast (fractions of msec) opening of the VT2 key when switching to transmission and its slow (approximately 50 msec, determined by the time constant R38C38) closing when switching to receiving, which eliminates the appearance of loud clicks in phones when switching operating modes.

The signal from the output of op-amp DA2.1 is supplied through a single-link low-pass filter R23C16 to the input of the final low-pass filter DA1 LM386N with Kus = 80 and further, from output DA1 to the output of the board to the volume control and through the chain R16R17С14 to the AGC detector, made on diodes VD1-VD5 KD522 (you can use any silicon KD510, KD521, 1N4148, etc.) and has two control circuits - an inertial one with capacitor C26 and a fast-acting one with capacitor C19, which allows improving the operation of the AGC in conditions of pulsed noise. The common connection point of the AGC detector elements is connected to the divider R19R20R36.0R2, which creates the initial bias voltage of the field-effect transistor. Using a trimming resistor R19, it is set optimal for a particular instance of the transistor and, if necessary, the overall gain of the receiver is adjusted. Resistor 0R2 (it is outside node A5) quickly regulates the overall gain when listening to the air. In fact, this adjustment is equivalent to changing the RF or IF gain in superheterodynes.

A microphone amplifier with a serial phase limiter (SLP) is made using a DA10 NE5532 op amp, designed for the use of an electret microphone. +9 V power is supplied through the chain R165, C133, R166. Resistor R165 determines the current (in this case, approximately 0.75 mA, which is suitable for many types of computer headsets and can be adjusted if necessary), and, accordingly, the microphone operating mode. Capacitors C74, C129 are used to protect against RF interference. The signal from the microphone is fed to the input of the amplifier-limiter (pin 3 of DA10.1) through a passive high-pass filter C134, R163, R156 with a cutoff frequency of approximately 5.5 kHz, which provides a rise in the high-frequency components of the spectrum by about 6 dB/octave, which significantly improves quality and intelligibility generated signal. The use of such a passive correction circuit leads to attenuation of the microphone signal (by approximately 14 dB at a frequency of 1 kHz), but taking into account the fact that electret microphones They produce a high-level signal at the output (on average -5-15 mV and up to 50-70 mV amplitude in loud “A” mode), which allows you to significantly simplify the circuit without loss of signal quality. The gain of the amplifier-limiter DA10.1 is determined by the ratio of resistors R152, R162 and in this case is equal to approximately 1000, which, taking into account the attenuation by the correction circuit by 5 times (by about 14 dB at a frequency of 1 kHz, for which we are calculating), gives an overall gain = 200 . The limiting threshold of diodes D19,20 (you can use any silicon KD522, KD521,1N4148, etc.) is approximately 600 mV, therefore the beginning of the limitation for the microphone signal is approximately 3 mV. If, during tests with a particular microphone, it seems to you that this gain is excessive, this can be easily corrected by proportionally increasing resistor R162. After testing this MOU, I came to the conclusion that such amplification is optimal, because will allow you to work with many types of microphones without additional adjustment. If desired, you can introduce operational adjustment of the clipping level in the range of 0-30 dB, for which you need to put a 1-2.2 kOhm variable resistor in series with R162, preferably with a logarithmic characteristic, which can be displayed on the front panel.

The design of the input circuits of the MSO allows, if necessary, to easily perform a fairly large and flexible correction of the frequency response and vary the pre-emphasis, which may be required when optimizing the quality of the generated sound depending on the characteristics of a particular microphone and the timbre of the operator’s voice. For example, with a low, dull voice, you can select R162 = 6.8 Ohm and C132 = 22 µF, which will provide an additional increase in sound frequencies from approximately 1000 Hz. And if at the same time you install a capacitor C129 = 47 nF, which, together with R163 = 1 kOhm, forms a low-pass filter with a cutoff frequency of approximately 3 kHz. The resulting frequency response of the input circuit will receive a noticeably pronounced resonant shape with a peak at frequencies of approximately 2.5-2.7 kHz, which will have a positive effect on signal intelligibility.
The signal, limited almost to a rectangular one, is fed to a single-link phase shifter made on the DA10.2 op-amp. The natural frequency of the phase-shifting circuit R145, C115 is selected to be approximately 400Hz - as the experiment has shown, this provides slightly better results than the usually recommended 500-600Hz. At the same time, the phase method effectively suppresses the harmonics of limited signals in the frequency range from 500 to 1000 Hz, and above 1000 Hz it suppresses the harmonics of the main FSS no less effectively. For proper operation of the phase shifter, resistors R142, R144 must have the same values ​​(preferably no worse than +-1%), the value itself is not critical and can be in the range of 3.3-100 kOhm. When a limited low-frequency signal passes through the phase shifter, the harmonics receive a phase shift of about 70-100 degrees. relative to the fundamental frequency. In this case, the shape of the rectangular signal is greatly distorted and the harmonics, which previously formed steep fronts, now form surges near the peaks of the sinusoidal voltage of the fundamental frequency. These emissions are cut off by the second limiter, made on diodes D17, D18.. Here I would like to draw the attention of my colleagues to a very important point, on which I myself stumbled during the first tests - the efficiency or, if you like, the quality of work of such an MOU, consisting of two (sometimes more) successive limiters, very much depends on the degree (rigidity) of the limitation of the first and the conjugation of the limitation levels of the first and second limiter. Moreover, the more we limit the signal, the more pronounced the effect of phase suppression of harmonics is. This is well confirmed by the results of the experiments shown in Fig. 4 – when limited to 30-40 dB, the level of nonlinear distortion at frequencies of 500-900 Hz is practically the same and does not exceed 8.5%. The best results are obtained if the level of the second limiter is equal to 0.5-0.7 of the level of the first, so I used KD514 diodes in the second. It is quite acceptable to replace it with KD522, 1N4148 - measurements showed that nonlinear distortions increased slightly - to about 11-12%, but the signal sounds quite decent.

Electronic switches on the VT16 KP307G transistor (possible replacements are the same as for VT2, VT3), shunting the OOS circuit of the DA10.2 op-amp and the fourth element (pins 10-11) of the DD6 switch, connecting the MOU output to the common wire, serve to disconnect the microphone path in receiving or telegraph operating modes, for which a high-level control signal is used (voltage approximately +8.0...8.5 V) +MICoff. This two-stage, or two-key, control ensures reliable muting of the microphone and completely eliminates the appearance of interference from it in the receiving and telegraph modes.

The telegraph signal generator is made on a DA9.2 op-amp according to a circuit with a Wien bridge R98R107C87C95 in the positive feedback circuit. The generation frequency is determined by the formula f = 0.159/R98C87, in this case it is approximately equal to 1000 Hz and can be changed if necessary. At the specified frequency value, the main FSS effectively suppresses harmonics, resulting in a crystal clear tone signal at the output of the TPP. Rigid stabilization of the amplitude of the generated oscillations is carried out using back-to-back diodes D14, D15 (any silicon KD522, KD521, 1N4148, etc. can be used) at a level of approximately 0.25 Veff. Next, the generator signal, through a single-link low-pass filter, which reduces the level of harmonics, is sent to the electronic switch VT7 KP307G (possible replacements are the same as for VT2, VT3), which directly manipulates the telegraph signal when a high-level control signal enters the gate circuit (approximately +8. 0…8.8V) +KEY. This signal arrives through a divider on resistors R114R121, which reduces the voltage level to +4.3...4.5V at the VT7 gate. The D16R120R128C110 circuit is designed to form a trapezoidal control signal from the +KEY square signal in the gate circuit with a rise time of approximately 15 mSec and a fall time of approximately 20 mSec. Such values ​​are optimal, in the author’s opinion, for average transmission speeds of 90-120 characters per minute. If you like to work at higher speeds, it is advisable to choose the C110 capacitance equal to 47 nF. In this case, the duration of the rise and fall of the generated telegraph message will be approximately 7 and 10 mSec, which corresponds to the traditionally recommended values ​​in the domestic literature. Thanks to the quadratic current-voltage characteristic of the field-effect transistor, the shape of the envelope of the generated pulses becomes close to the optimal, bell-shaped, which provides a narrow spectrum of telegraph transmission radiation, of course, provided that the PA cascades have a sufficiently linear amplitude characteristic. In the inactive mode (control signals + MICoff or + TX low level), the operation of the master oscillator is blocked by the current flowing through the D8D9R61 D15 chain. The small differential resistance of diode D15, open by the flowing current, bypasses resistor R106 of the OOS circuit, which eliminates the possibility of generation. A constant voltage from the output of the generator (pin 1 of DA9.2) of approximately +5 V is supplied to the source of VT7, and at its gate the +KEY signal level is low, so it is closed. This two-stage control ensures reliable shutdown of the telegraph generator and completely eliminates the appearance of interference from it in the receiving and microphone modes.

The transceiver is switched to the microphone or telegraph transmission mode by a special control circuit made on four two-input Schmidt triggers of the DD7 HCF4093 microcircuit (K1561TL1 can be used), which generates the necessary control signals. IN original condition, receive mode - until the key or pedal is pressed, at pins 3.10 DD7 (signals +KEY. +TX) there is a low voltage (approximately +0.3...0.8V), and at pin 11 DD7 (signal +MICoff) is high voltage (approximately +8.0…8.8V).

When you press the pedal or in some other way close the X15 pin of the main board to the common wire at pin 10,12 DD7, a high level of the +TX control signal is simultaneously formed, switching the transceiver to the transmit mode, and a low level of the +MICoff control signal, allowing the operation of the microphone path and blocking telegraph generator. If the key is pressed while the pedal is pressed (pin X13 of the main board is shorted to a common wire), the high level of the +TX control signal, switching the transceiver to the transmit mode, will remain, and at pin 11 DD7 (signal +MICoff) a high voltage level will appear, allowing operation telegraph generator and blocking microphone path. At the same time, a high level of the +KEY control signal is generated at pin 3 of DD7, forming a telegraph message.

If you operate the key without pressing the pedal, it becomes possible to listen to the broadcast in pauses between telegraph messages (the so-called “full half-duplex” mode - QSK). When you press the key for the first time, a high-level voltage at pin 3 of DD7, forming a high level of the +KEY control signal, quickly (fractions of a msec) charges capacitor C46 through resistor R48. A high voltage level on this capacitor leads to the appearance of a low level voltage at pin 4 of DD7, which initiates the formation of a high level control signal +TX and +MICoff by logic elements DD7.3, DD7.4. The holding time of the transceiver in transmit mode after releasing the key is approximately 0.1 seconds and is determined by the time constant of the R44C46 circuit. If the switching circuits external devices(for example, a tube Mind with relay switching) cannot withstand such a “rate of fire”, the holding time can be increased by proportionally increasing the value of resistor R44, for example, if you select 1Mohm, the holding time will be approximately 1 second.

On transistors VT4, VT5, VT6 a key amplifier-former of control signals +13.8RX and +13.8TX is made for switching external components (PDF, PA, low-pass filter, attenuator, etc.). The power of transistors VT5,VT6 determines the permissible load. With the specified KT814 (replacement with KT816 with V>50 is possible), a load of up to 0.5A is permissible. If the load current does not exceed 0.25A, then you can successfully install KT208, KT209, KT502 with any letter index.

Requirements for parts, possible replacements and their selection, if necessary, are set out in the text along the description of the relevant components of both the main transceiver path considered here, and in the text of the description of the receiver, which we strongly recommend that you read.

Most of the TPP parts are located on a printed circuit board (Fig. 5) made of double-sided foil fiberglass. The top side serves as a common wire and shield. The holes around the leads of parts not connected to the common wire should be countersunk with a drill with a diameter of 2.5-3.5 mm. The terminals of parts connected to the common wire are marked with a cross. The common wire of the ULF power part (pin 4 DA1) is connected to the upper side of the common wire at only one point - contacts X10, X22, which are soldered on both sides. The common wire from the power supply is also supplied here. Due to the high density of parts, installation is recommended to be done in the following sequence: first, all jumper wires made of thin mounting insulated wire are installed on the board; then passive and active elements are mounted, having leads soldered to a common wire, and only then the remaining components.

Before applying power to the board, carefully check the installation again. If everything is done without errors and using serviceable parts, the main board starts up immediately. After applying the supply voltage, the current consumption in the receive mode (without the VFO signal, the key and pedal in the open position) should be close to 100 mA, and a quiet and uniform noise should be heard from the speaker. It is useful to check the operating modes of the DC cascades - at the output of all op-amps there should be a voltage close to +4.5 V, at the terminals of logical elements and switches there should be control voltage levels corresponding to the description of the operating logic of these nodes.

The first stage in the setup is setting the AGC threshold of the receiving path. To do this, the slider of resistor 0R1 Volume is set to the top position according to the diagram, and the sliders of resistor 0R2 Gain and trimming resistor R19 (see Fig. 2) are set to the left position according to the diagram. Connect a 50 ohm resistor to the receiver input. Connect the VFO. A speaker or telephones are connected to the output (pins X9, X10) of the receiver; if desired, you can connect an oscilloscope or avometer in AC voltage measurement mode. By moving the slider of trimming resistor R19, find the position at which the noise begins to decrease, and from this position move the slider slightly in the opposite direction. This will be the optimal setting of the AGC threshold.

Setting up the transmission path can be done in two stages. First, by connecting an oscilloscope or multimeter in AC voltage measurement mode to the negative terminal of one of the electrolytes (C117, C120, C126 or C131), we close the key contacts and switch the TPP to the telegraph signal transmission mode. Using the trimming resistor R129, we set the level of the modulating signal to approximately 1.7 Veff (amplitude 2.3 V). In this case, the self-control signal should be clearly audible in the speaker. Connect the microphone and press the pedal. In the loud “A” mode, by rotating the tuning resistor R140, we set the level of the modulating signal to about 1.1 Veff (amplitude about 2.2 V). Pre-setting transmission path is completed.

In Fig. Figure 6 shows a diagram of the distribution of transmission coefficients, a diagram of the cascade levels of the signals of the receiving and transmitting paths, which will help to better understand the principle of operation of the TPP and, if necessary, configure it more carefully.

Literature

1. Polyakov V. Direct conversion receiver at 28 MHz. - Radio, 1973, No. 7, p. 20.
2. Polyakov V. SSB direct conversion receiver. - Radio, 1974, No. 10, p. 20.
3. Polyakov V.T. Single sideband modulator-demodulator. - Radio engineering, vol. 29, 1974, No. 10.
4. Polyakov V. Direct conversion receiver mixer. - Radio, 1976, No. 12, p. 18.
5. Polyakov V. Direct conversion receiver. - Radio, 1977, No. 11, p. 24.
6. Polyakov V. Phase limiters of speech signals. - Radio, 1980, No. 3, p. 22
7. Polyakov V., Stepanov B. Mixer of a heterodyne receiver. - Radio, 1983, No. 4, pp. 19-20
8. Polyakov V. Direct conversion receivers. ― M.: DOSAAF, 1981
9. Polyakov V. Direct conversion transceivers. ― M.: DOSAAF, 1984
10. Polyakov V. Radio amateurs about direct conversion technology. ― M.: Patriot, 1990.
11. Drunk Yu. Direct conversion transceiver. - Radio, 1979, No. 7, p. 14
12. Luts E. Direct conversion transceiver at 28 MHz. - Radio, 1988, No. 1, p. 16
13. Polyakov V. Direct conversion transceiver on 160m. - Radio, 1982, No. 10, pp. 49-50, No. 11, pp. 50-53
14. . - Radio, 2005 No. 10, pp. 61-64, No. 11, pp. 68-71.
15. Abramov V. (UX5PS), Telezhnikov S. (RV3YF). Shortwave transceiver “Druzhba-M”. - http://www.cqham.ru/druzba-m.htm.
16. Denisov A. Digital scale-frequency meter with LCD indicator and automatic frequency adjustment. - http://ra3rbe.qrz.ru/scalafc.htm.
17. Titze U., Schenk K. Semiconductor circuitry. ― M.: Mir, 1982.
18. Horowitz P., Hill W. The Art of Circuit Design, vol. 1. - M.: Mir, 1983.

What changed in the transceiver after its publication in the “RADIO” Magazine No. 9.11 2006.

There are few changes. If possible, instead of pairs of capacitors (ceramics C21+ film C28), it is better to put imported MCTs, MCRs of 0.1 μF in each channel, naturally selected with an accuracy of no worse than 0.2% (as the experiment showed, the accuracy of this four directly determines the quality of suppression of the side, because if you remove them (reduce to 3.3-4.7 nF), the suppression in the low frequency ranges increases to 60-63 dB!!!, but unfortunately they are needed, otherwise the resistance to AM interference decreases), which made it possible to slightly improve the suppression mirror side at 7 MHz and 14 MHz. The AGC circuits have also been slightly optimized (this is already reflected in the TPP diagram (Fig. 2) version 11.0), now there are no pops with sharp and loud signals, it works softly and imperceptibly, and at the same time it suppresses impulse noise well, almost completely. changes to the printed circuit board are minimal, if the board (For the signet drawing posted on pages 23 and 78 of the forum on modern Chamber of Commerce and Industry) is already ready - close with jumper R167 and transfer the connection to the upper leg of capacitor C19, adjusting the tracks with a cutter. I did it simpler - it was a pity to cut the tracks - I soldered the specified conductor on the side of the printed conductors. If the board has not yet been prepared, then during manufacturing it is better to use the already corrected drawing (this is already reflected in the printed circuit board drawing Fig. 5 version 8.0). In this version I also slightly changed the ground routing around LM386. Therefore, the “ground” pin C16 must be soldered on both sides.

,

The transceiver is easy to manufacture and does not contain any scarce parts. The transceiver is based on the "SSB direct conversion receiver". Works by telegraph CW And SSB on range 80 m. Transmitter output power - 1.5 W. Receiver sensitivity at signal-to-noise ratio 10 dB - 1 µV. Suppression of carrier and non-working sideband - no less 30 dB.

The received signal from the input circuit C2L1C4 through the capacitor C3 and the bandpass filter L8C32C30L9C33 is supplied to the input of the RF amplifier, which is made using transistors V11-V13. Gain by high frequency regulated by variable resistor R28. Through coils L11, L12, the signal is supplied to a balanced mixer assembled using diodes V14-V17. The local oscillator voltage is supplied to the mixer from a two-stage local oscillator on transistors V5, V6. Required phase shift of 90° local oscillator voltage in the operating frequency range 3.5-3.65 MHz provides the L4C67R7 wideband phase shifter.

Rice. 1a. Schematic diagram of the transceiver

Rice. 1b. Schematic diagram of the transceiver (continued)

As a result of frequency mixing, an audio frequency signal is released, arriving at the shoulders of the low-frequency phase shifter, formed by circuits L13C39, L14C41 and resistors R34-R37. Transistor V19 is connected to one of the arms, the collector junction of which in receive mode is opened by the voltage supplied through resistor R42. The low-frequency phase shifter provides a phase shift of 90° within the audio frequency range.

At the output of the low-frequency phase shifter, sound signal lower side stripe. Through the low-pass filter C46L15C47L16C48, the audio signal is fed to a low-pass amplifier made using transistors V20-V23. To attenuate frequencies above 3 kHz The first two stages of the low-frequency amplifier are covered by frequency-dependent negative feedback. To adjust the bandwidth, a positive Feedback(via elements C55, R43, C56). The passband is changed with variable resistor R43. The gain of the low-frequency signal is controlled by resistor R54.

In transmission mode, the signal from the microphone is sent to an amplifier made of transistors V7-V9. The amplified low-frequency signal is fed to a low-frequency phase shifter. In order to highlight the lower sideband during transmission, an additional phase shift of 180° is required, which is provided by transistor V19 (during transmission it is switched to amplification mode). After mixing the signals in balanced mixers at a common load, a lower sideband signal is formed. The necessary suppression of the carrier frequency is set by resistors R32, R33.

The selected single-sideband signal through the coupling coil L10 is supplied to the RF amplifier on transistors V1-V4 and through the P-circuit C2L1C4 is fed to the antenna. When working with telegraph, a tone generator on transistor V10 is used. The same generator, turned on by button S2, serves to adjust the final stage according to the brightness of the light bulb H1, which is an indicator of the collector current of the output transistor V1.

Details and design

The transceiver is assembled on a printed circuit board measuring 180x50 mm using single-sided mounting. The winding data of the coils is given in table. The RF amplifier of the receiver and transmitter, as well as the GPA, are separated by shielding partitions.

Coils L1-L9 are wound with PEV 0.1 wire on frames with a diameter of 7 mm, a trimmer made of M400NN ferrite with a length of 12 and a diameter of 2.8 mm. Coils L10-L12 are wound with PEV 0.1 wire on an SB-9a core. Coils L13-L16 are wound with PEV 0.07 wire on a Sh6x8 permalloy magnetic core. Chokes L17-L19 can be of any inductance 50-100 µH. Coils L12-L14 are wound into two wires. Then connect the end of one half of the winding to the beginning of the other.

Diodes for a balanced mixer should have similar reverse current values. Transistors V2-V6, V11-V13 can be any high-frequency, and V7-V10, V21-V23 can be any low-frequency. Transistor V20 must be low noise. Diodes V14-V17 - any from the D311 series. Somewhat worse results are obtained by using D18 diodes. The TM-2M microtelephone capsule was used as a microphone. Lamp H1 - any low voltage, for example, 6.3 V 0.28 A.

Setting up the transceiver

Setting up the transceiver must begin with setting up the low-frequency phase shifter. To do this you will need an oscilloscope and a sound generator. Phase shifter arms ( Fig.2 ) connect to the inputs " X" And " Y"oscilloscope. An audio frequency signal is supplied to the input of the phase shifter. Resistors R34 and R35 achieve the presence of a circle on the oscilloscope screen when the generator frequency changes in the interval 300-3000 Hz. Further adjustment of the phase shifter is carried out when it is connected to the transceiver.

Rice. 2

To configure high-frequency circuits, you will need an RF generator and an SSB signal receiver. The adjustment begins with the receiving part, having previously set the frequency of the smooth range generator within the range. Circuits L8C32 and L9C33 are configured to average frequency range.

Trimmer resistors R32 and R33 are set to the middle position. resistors R36, R37 and R7 achieve maximum suppression of the upper sideband. Resistor R39 does not significantly affect operation in receive mode. It is necessary to make sure that there is no excitation in the low-frequency amplifier at different positions of resistor R43. If it is, then select capacitors C55, C56.

In transmission mode, the operation of the bass amplifier and sound generator is first checked. The L2C65 circuit must be tuned to the mid-range frequency. By adjusting resistors R32 and R33, maximum suppression of the carrier frequency is achieved, and by adjusting resistor R39, maximum suppression of the upper sideband in transmit mode.

When the transmitter is excited, check the thoroughness of the shielding and the presence of decoupling capacitors on the “negative” buses. The transceiver was tested at a collective radio station UK3ACR. Contacts were made with Soviet radio amateurs in regions 1-6 and foreign correspondents.

"Radio" No. 10/1978

The transceiver has separate high-frequency and low-frequency paths for reception and transmission; common to both modes are a mixer-modulator and a smooth range generator.

The smooth range generator (VFO) is made on two field-effect transistors VT5 and VT6 with source coupling. It operates at a frequency equal to half the frequency of the received or transmitted signal. When operating for reception and transmission, the output circuits of the GPA are not switched and the load on the GPA does not change. As a result, when switching from receiving to transmitting or vice versa, the VFO frequency does not deviate. Adjustment within the range is carried out using a variable capacitor with an air dielectric SJ, which is part of the GPA circuit.

The transceiver is designed to transmit and receive SSB and CW in the 28-29.7 MHz range. The device is built according to a direct conversion scheme with a common mixer-modulator for reception and transmission.

Specifications:

• sensitivity in receiving mode with a signal/noise ratio of 10 dB, no worse than........1 µV;
• dynamic range of the receiving path, measured using the two-signal method, about......80 dB;
• bandwidth of the receiving path at a level of -3 dB.........2700 Hz;
• spectrum width of single-sideband radiation during transmission........2700 Hz;
• the carrier frequency and non-operating sideband are suppressed no worse than ........ 40 dB;
• transmitter output power in telegraph mode at a load of 75 Ohms......7 W;
• The local oscillator frequency drift after 30 minutes of warming up after switching on is no more than.....200 Hz/h.

In SSB transmission mode, the signal from the microphone is amplified by operational amplifier A2 and fed to a phase shifter using elements L10, Lll, C13, C14, R6, R7, which provides a phase shift of 90° in the frequency range 300-30-00 Hz.

In the L4C5 circuit, which serves as the general load of the mixers on diodes VD1-VD8, the upper sideband signal is allocated in the range of 28-29.7 MHz. The L6R5C9 high-frequency wideband phase shifter provides a 90° phase shift in this range.

The selected single-sideband signal is fed through capacitor C6 to a three-stage power amplifier using transistors VT7-VT9. The stage of pre-amplification and decoupling of the output circuit of the mixer-modulator is made using a VT9 transistor. The high input impedance combined with the low capacitance of C6 ensures minimal impact of the power amplifier on the C5L4 circuit. The VT9 collector circuit includes a circuit configured to the middle of the range. The intermediate stage on the VT8 field-effect transistor operates in class B mode, and the output stage operates in class C mode.

The U-shaped low-pass filter on the C25L13C26 cleans the output signal from high-frequency harmonics and ensures that the output impedance of the output stage is matched to the characteristic impedance of the antenna. Ammeter PA1 is used to measure the drain current of the output transistor and indicates the correct settings of the P-circuit.

The telegraph mode is ensured by replacing amplifier A2 with a sinusoidal signal generator with a frequency of 600 Hz (Fig. 21). Switching CW-SSB is done using switch S1. The telegraph switch controls the bias of VT11 of the generator preamplifier and, therefore, the supply of a low-frequency signal to the modulator.

In receive mode, 42 V power is not supplied to the transmitter stages, and the power amplifier and microphone amplifier are turned off. At this time, a voltage of 12 V is supplied to the stages of the receiving path.

The signal from the antenna enters the input circuit L2C3 through the coupling coil L1; it matches the loop impedance to the antenna impedance. The transistor VT1 is used for AMP. The gain of the cascade is determined by the bias voltage at its second gate (divider across resistors R1 and R2). The load of the cascade is circuit L4C5, the connection of the RF cascade with this circuit is carried out through the communication coil L3. From the coupling coil L5, the signal is supplied to a diode demodulator using diodes VD1-VD8.

Coils L8, L9 and a phase shifter on L10 and L11 highlight signal 34 in the frequency band 300-3000 Hz, which is fed through capacitor C15 to the input of operational amplifier A1. The gain of this microcircuit determines the basic sensitivity of the transceiver in receive mode. Next comes amplifier 34 on transistors VT2-VT4, from the output of which signal 34 is sent to small-sized speaker B1. The reception volume is adjusted using a variable resistor R15. In order to eliminate loud clicks when switching “reception-transmission” modes, power is supplied to the UMZCH on transistors VT2-VT4 both during reception and transmission.

Most of the transceiver parts are installed on three printed circuit boards, sketches of which are shown in Fig. 22-24, On the first board there are parts of the input RF receiving path (on transistor VT1), parts of the mixer-modulator with phase-shifting circuits, as well as parts of the local oscillator. The second board contains low-frequency stages on microcircuits A1 and A2 and transistors VT2-VT4. The third board houses the power amplifier of the transmitting path.

The board with the mixer-modulator, RF amplifier and GPA is shielded. The “reception-transmission” modes are switched by a pedal, which turns on and off the 42 V voltage and controls two electromagnetic relays, one of which switches the antenna, and the second supplies 12 V voltage to the receiving path. The relay windings are powered by a voltage of 42 V, and in the de-energized state the relay contacts switch on the receiving mode.

To power the transceiver, a basic stationary power supply is used, from which a constant stabilized voltage of 12 V with a current of up to 200 mA and a constant unstabilized voltage of 42 V with a current of up to 1 A are supplied.

Winding data of transceiver coils Table 4

The transceiver uses fixed MLT resistors for the power indicated in the diagrams. The adjusted resistor is SPZ-4a. Loop capacitors are necessarily ceramic, tuning capacitors are KPK-M. Electrolytic capacitors - type K50-35 or similar imported ones. Variable capacitors of the local oscillator and output circuit are with an air dielectric.

To wind the contour coils of the URCH, mixer and transmitter, ceramic frames with a diameter of 9 mm with tuning cores SCR-1 are used (plastic frames from the UPCH paths of old tube TVs are also possible, but their thermal stability is much worse than that of ceramic ones). Low-frequency mixer-modulator coils L8 and L9 are wound on ring cores K16x8x6 made of 100NN or higher frequency ferrite (100HF, 50HF). Coils L10 and L11 are wound on OB-ZO frames made of 2000NM1 ferrite. The coils of erasure and magnetization generators of semiconductor reel-to-reel tape recorders were wound on such cores. The winding data of the transceiver coils are given in table. 4.

KPZZG transistors can be replaced with KPZOZ with any letter index or with KP302. The KP350A transistor can be replaced with KP350B, KP350V or KP306. Transistor KP325 - on KT3102. Powerful field-effect transistors KP901 and KP902 can be with any letter indices. Any silicon and germanium (respectively) transistors of the appropriate structure are suitable for UMZCH. Diodes KD503 can be replaced with KD514, and diode D9 with D18.

Literature: A.P. Family man. 500 schemes for radio amateurs (Radio stations and transceivers) St. Petersburg: Science and Technology, 2006. - 272 pp.: ill.