Country radio amateur transceiver. Homemade transceiver Latest developments of transceivers and their circuits

Somehow there was a desire to make an SDR transceiver. And the search for information and diagrams on SDR transceivers began. As it turned out, there are practically no completed transceivers, with the exception of various versions of the SDR-1000. But for many, this transceiver is both expensive and complicated. Also published various options main boards, synthesizers, etc. ,those. separate functional units. Tasa YU1LM, which also made the complete “AVALA” transceiver, did a lot in the field of development and popularization of simple SDR technology, and we can recommend its designs for beginners in this field and those who want to try what SDR is at minimal cost.

In the end, I decided to make my own, as simple as possible and at the same time high-quality SDR transceiver. YU1LM materials and other publications were used during development. It was decided to make the mixer on 74HC4051 - Sergei’s direct conversion receiver was once made US5MSQ ,with a mixer on this chip. And the use of 74HC4051 in a transceiver allows you to make a very simple mixer - common for both the receiving and transmitting paths. The quality of work of this mixer is quite satisfactory. The entire history of the development of the transceiver can be read in detail at forum SKR (Krasnodar site). And if you intend to make this or any other simple SDR transceiver, then I highly recommend reading the forum - my entire path from the idea to make a transceiver to a completed and working design and a lot of other useful information that simply cannot be included in this article are described in detail.

The transceiver is built using a direct conversion scheme from an operating frequency to an audio frequency for signal processing by a computer sound card... Therefore, much that has been written about the direct conversion technique also applies to SDR. In particular, the need to suppress the non-working sideband (in the SDR mirror channel) using the phase method.

It was decided to make a simple single-board single-band transceiver, with a quartz oscillator at the main frequency and QRP power, i.e. a completely finished device. I chose the 14 MHz band as the most interesting for me. If desired, it will not be difficult to make a transceiver for any other low frequency range. The transceiver was not tested at frequencies above 14 MHz, and more low frequencies should work well. The resulting transceiver has the following parameters:

• Operating frequency range 14.140 - 14.230 MHz. (When using a quartz resonator at a frequency of 14.185 MHz and sound card with a sampling rate of 96 kHz)
• The sensitivity is about 1 µV and highly depends on the quality of the sound card.
• Dynamic range intermodulation is more than 90 dB - there was nothing more precise to measure.
• Carrier suppression for transmission is more than 40 dB (I got 45 - 60 dB) and depends on the specific instance of the 74HC4051, as well as on the quality of the tuning.
• Suppression of the mirror channel is more than 60 dB with the correction program.
• Output power is about 5 W.

It is clear that for an SDR transceiver it is necessary control program, and my choice fell on the M0KGK program because of the program’s ability to correct amplitude and phase throughout the entire operating range of the sound card and memorize calibration points. This is very important. This property of the program allows you to suppress the mirror channel very well. Due to the lack of ability to store calibrations at several sound card frequencies in the program, I refused to use it - this program works great with SDR transceivers with built-in frequency synthesizers, where the frequency tuning is done by the synthesizer, and not by the sound card frequency.

The circuit diagram is simple and I will not describe the operating principle. You can read this from Tasa YU1LM, though at English language. No errors were found in the printed circuit board. For ease of soldering, I signed the values ​​of the elements in the figure. printed circuit board, and not serial numbers of elements.

The transceiver practically does not need configuration, and if installed correctly, it starts working immediately. With the correct settings of the M0KGK program, of course. This data can also be read on the forum.

It is clear that many will have difficulties purchasing a quartz resonator. Therefore, in case of its absence or because of the desire to have the entire 20 m range, you can simply use an external VFO or synthesizer at the operating frequency, the signal from which must be fed to the 1st pin of the 74HC04 through a 10 nF coupling capacitor. Do not install capacitors C63 and C64.

Working with this transceiver is very pleasant and convenient. All control computer mouse. The entire spectrum in the 96 kHz band is visible, and by simply indicating or “dragging” the program filter, we instantly tune to the station of interest. Very quickly and clearly. After working on this transceiver, working on a regular one is already missing something - visual information about the situation on the band.

Sergey 4Z5KY

Schematic diagram of a simple transceiver on 80m

Contour winding data. Coil L2 has an inductance of 3.6 μH - that's 28 turns on an 8 mm frame, with a sub-frame core. The throttle is standard.

How to set up a transceiver

The next step is setting up the transceiver for transmission. Instead of an antenna, we hang an equivalent - a 50 Ohm 1 W resistor, connect an RF voltmeter in parallel to it, at the same time turn on the transceiver for transmission (by pressing the key), begin to rotate the core of the L2 coil according to the readings of the RF voltmeter and achieve resonance. That's basically all, I want to add that the author himself wrote that you should not install a powerful output transistor; with an increase in power, all sorts of whistles and excitations appear. This transistor plays two roles - as a mixer when receiving and as a power amplifier when transmitting, so kt603 it will be a steal here. And finally, a photo of the structure itself:

Since the operating frequencies are only a few megahertz, any RF transistors of the appropriate structure can be used. The design of this transceiver was repeated and configured by Comrade. Radiovid.

Discuss the article SIMPLE TRANSCEIVER

Homemade transceiver

UR0VS

The transceiver was made taking into account its development in a month and a half. Moreover, on weekdays from 20:00 to 24:00, and on weekends he received attention until lunch. Therefore, its construction can be recommended for not very experienced radio amateurs. The scheme does not stand out for its originality. Due to my busy schedule, I did not reinvent the “bicycle” (I really wanted to go on the air again), but put together the contents of my “boxes” with radio components, and well-proven previously developed units. For the same reasons, any service like VOX, detuning, etc., was not developed. True, I had the case and I limited myself to only drilling holes in the right places for attaching the boards.

The circuit and PCBs were designed using the OrCad 9.0 design system. The quartz filter was calculated using an excellent, in my opinion, program from UA1OJ. I didn’t even have to finish it after the calculation.

Performance characteristics

Power - 7-10 watts (depending on the range). With a 100-watt tube amplifier, nearby TVs do not “jump”.

Sensitivity is sufficient :) even without UHF (node ​​A5).

The blockage is normal (what kind of blockage is there:), there are almost no radio amateurs left).

In short, for everyday phone work in rural areas, this is what you need. And most importantly, it is more modern than UW3DI.

Transceiver circuit

Block A1 is the main board. It consists of mid-level diode mixers (D1 - D4, D6, D8 - D10), an IF amplifier (Q3, Q1, Q4), switching its direction using a relay (K1 - K2), a low-frequency amplifier (U1), and an AGC circuit ( Q7 - Q8). Emitter followers are assembled on transistors Q2 - Q5 to match local oscillators with mixers. The reference local oscillator is assembled in transistors VT1, Q6. Microphone amplifier Q9 - Q10. Final ULF Q11 - Q13.

The printed circuit board of block A1 was laid out in two versions. The difference between the options lies in the quartz used. I have quartz in B1 cases with a frequency of 9050 kHz, but it is possible to install small quartz, for example from PAL/SECAM decoders, at a frequency of 8865 kHz.

Block A2 - GPA. Something similar is used in the Druzhba transceiver. It's just a little easier here. Assembled in a tinned copper box from some old radio station. Only the frequency divider is assembled on the printed circuit board. Everything else is on ceramic stands. Broken MLT resistors can be used as racks (this idea came to my friend UR0VF), you just need to clean the “black” layer. The circuit is ceramic with waxed copper from the same r-st. I do not provide a full description of this node for the reason described below.

Block A3 - bandpass filters. There is no point in commenting on this node for a very simple reason. As a rule, for radio amateurs, the contents of the “boxes” are different for everyone, and if you try to use all the details that the author has, then any design turns into a “project of a lifetime.” Feel free to take this unit from any design for which you have the equipment (this also applies to the GPA). If these are PFs from the “drozdiver”, then the device will have even better characteristics. And in this case, node A5 can be completely abandoned. I’ll just say that I used the same PFs as in the Ural-84 transceiver.

Block A4 is a “power booster”. All transformers are wound on K10 x 5 rings with PEV 0.3 - 0.5 twisted wire and have 12 turns. Transformer T3 is wound with 3 wires. The selection of parts in this unit is not so large. You can vary it with other transistors in the final stage. KT921 work very well; they are precisely designed to work in linear amplifiers. There was experience using KT606A medium power transistors in this cascade (due to careless inclusion). The power in this case was the same in all ranges, but really not very much. About 4.5 watts! For those who are “afraid of transistors,” we can recommend a well-proven lamp circuit. More on this below.

Block A5 – switchable UHF. There seems to be nothing to comment

There is one more block. This is a digital scale (OUT2 is provided for it in the GPA). Here, too, I didn’t invent anything, I “created” a very simple scale on a PIC controller and ALS318, designed by RA3RBE. True, I had to finish it a little. There was very strong interference in the HF ranges. It only disappeared when I installed an emitter follower at its input. I draw your attention to the word emitter, source does not give anything!

The power supply is very simple. This is KR142EN8B, standing on the wall of the case, and a constant voltage of about 17-18 volts to this microcircuit is used to power the final stage of the PA. Another requirement is that the power supply transformer must provide a current of about 2.5A.

All resistors are MLT type 0.125 - 0.25. Ceramic capacitors, types KM - 5, KM - 6. Coils L1 and L4 in block A1 are wound on frames from SMRK blocks of old televisions. They have a 6mm diameter with 4mm carbonyl cores. For a frequency of 9 MHz, L1 - 20 turns. Wire PELSHO 0.25. The communication coil has 5 turns of the same wire. C16 in this case is 240 pf. L4 – the same wire is wound until it is full. Transformers T1, T2 and T4, T5 are wound on rings with a permeability of 600 - 100 Nm with an outer diameter of 7 - 10 mm in three wires with a twist of 4 - 5 twists per centimeter, the same wire as the circuit. T3, T6 - the same wire, also twisted, only in two wires. The beginning and end of the windings can be seen in the figure from the installation side.

The printed circuit board is made of double-sided PCB and the top layer is used as a “ground” wire, thus providing excellent shielding. W1,W2 are pieces of thin coaxial cable.

In the GPA, all tuning capacitors with an air dielectric with a capacity of 1 - 10 pf. As a variable dual KPI, you can use capacitors from old receivers with a capacity of 5 - 495 pf, only in this case, capacitors of about 25 - 33 pf must be connected in series with them. All frequency-setting capacitors must have negative TKE - M47, M75. The schematic arrangement of parts in the GPA housing is shown in the figure.

Assembly - setup

It’s not for nothing that I combined these two concepts. Since, for example, the main board is a multifunctional unit (this applies to transceivers of any design), the concept, as many write, “with working parts..., etc.”, will not “work” here. I advise you to do it this way. Start with the final bass amplifier. Apply power, if necessary, adjust by selecting the current of the output transistors within 15 - 20 mA. Next, you can assemble a microphone amplifier. Connect the microphone and apply power to it and to the ULF. Listen to yourself. Then you can begin assembling the crystal oscillator. Check the generation using at least a voltmeter. If the radio amateur does not have an HF generator, then the voltage from the CG can be used to pre-set the L1 circuit, the IF amplifier. Next are mixers, AGC and buffer stages for mixers. A quartz filter can be built at any stage. There are dozens of configuration methods. How the author set it up is described at the beginning of this “writing”. Two more words about capacitor C14. On the printed circuit board it stands apart. When adjusting the balance of the mixer, due to the difference in diode capacitances, it may have to look for a connection point to another diode.

Sufficient information about setting up the remaining nodes can be gleaned from a variety of other sources. It will be necessary to set the quiescent current in the PA to about 150-200mA. Depends on the pair of transistors used. For KT606, the current should be 50-60mA.

In the original version, the transceiver operates only on five bands, this is due to the lack of an antenna system for operation on all bands. However, those wishing to enter all ranges should not encounter any difficulties.

Today we will talk about the Radio-76 transceiver, or more precisely about its modernization; with the permission of the author of the diagram, I will not call it that, since there is little left of the Radio-76 transceiver.

The fact is that I had a long period of creative crisis, so to speak, and I did not engage in radio sports, due to moving from the countryside to the city, and I did not have the opportunity to install an antenna on at least one band, I postponed my favorite thing for a long 7 years. But thoughts about my favorite hobby did not leave me, and I decided to assemble a transceiver for myself, but another problem arose about choosing a circuit, and then the choice fell on the transceiver “Reversing path on bipolar transistors based on the R-76”, the author of which is Sergei Eduardovich US5MSQ http://us5msq.com.ua

P.S. In secret))) On the forum, Sergei Eduardovich actively answers all questions that arise during the assembly process, for which we must pay tribute, since not all the authors of their “brainchild” are so active in answering especially stupid questions. Verified personally

Below I will post the text of all the questions and answers from the author of the diagram that other radio amateurs who assembled this transceiver had. On my own behalf, I will say that if you assemble carefully, you should not have any questions, since I can get everything working right away, not counting my mistakes in installation.

Below are clippings from posts from the forum where radio amateurs discussed this transceiver. Since there is no complete description of this scheme, I will do it this way.

Characteristics:

• The overall level of self-noise is about 35-45mV
• The total value from the mixer inlet is approximately 340-350 thousand.
• The noise level referred to the input is approximately 0.12 μV, and the sensitivity from the mixer input at c/noise = 10 dB is about 0.4 μV

The AGC starts to operate at a level of about 4-5 µV (S5-6), while actually holding the signal to at least 15 mV (+50 dB).

And so let's get down to the scheme itself.

At the end of the article there will be an archive with all the diagrams for downloading in full size.

Fig. 1 Diagram of the main board with voltage map

I’ll add on my own behalf that if you follow all the voltages indicated in the diagram, the adjustment issues will disappear on their own.

Fig. 2 Diagram of bandpass filters with an attenuator and a swing amplifier on VT1.

Fig.3 GPA diagram.

Rice. 4 Low-pass filter and SWR meter circuit.

Clippings of messages from the forum

US5MSQ: As for the winding data of transformers, it is possible to use any that you have ferrite rings with a diameter of 7-12 mm and a permeability of 600-3000, it is important to ensure that the inductance for the first mixer is at least 50 μH (about 60-80) and for the detector/modulator at least 170 (). You can calculate the specific number of turns for your ring using standard formulas; it is convenient to use the tablet developed by Yu. Morozov.

It is important to ensure that the windings in the transformer itself are identical. I did this - I measured three identical conductors with a ruler (16cm for Tr1 and Tr2 and 24cm for Tr3 and Tr4), stripped and tinned the ends, soldering one side in the form of a needle (this side will be used for winding in the future), clamped it in a vice and twisted it by hand to the level of approximately 3 twists per cm. We wind the winding evenly by laying the turns until completely filled - on rings 2000NN 7x4x2 (for Tr3 and Tr4, 2 are glued together) we get about 15-16 turns. Before winding, do not forget to smooth out the sharp edges of the rings with sandpaper or a file.

Well, one more important point regarding the calculation and manufacture of communication coils. They are wound, as a rule, over the middle of the contour, over the edge of the contour closer to the grounded end, or, if the frame is sectional, in the section adjacent to the grounded end. In these cases, to more accurately reflect the coupling coefficient (mutual induction), we introduce a correction factor - for the 1st case of the order of 1-1.05, the second - 1.1-1.2 and the third -1.3-1.4. Thus, if we wind a communication coil with a number of turns 1/10 of the contour one, in reality it will approximately correspond to coefficients of 1/10, 1/11 and 1/13.

US5MSQ: coils for PDF can be made on almost any frames you have, and the results (main parameters of PDF) will be almost the same with fairly small losses, of course we are talking about correctly designed ones, and the majority of those published are.

The reason is that the relative width of modern bands (160, 80, 40m) reaches 9-10%, which means that the loaded quality factor of the circuits will be about 8-10, and even the most “left-handed” coils have a design quality factor of at least 40- 50, so losses even in three-circuit PDFs usually do not exceed 3 dB.

Our choice of three-loop DFTs is determined solely by the desire to get the SLR suppression as high as possible, for example, on the 80 m band at an IF of 500 kHz it is about 38-40 dB (80-100 times), a little of course, but two-loop ones are generally useless here (no more than 24-26 dB or only something like 15-20 times).

US5MSQ: DFT setting. If there is no GCH, then the DFT can be adjusted by the GSS (HF generator) and even simply to the maximum of air noise. If you are not sure that the antenna (or GSS) is matched, i.e. has an output impedance of 50-75 ohms, then you can turn on a standard -20dB attenuator at the input, which will ensure a consistent mode at the PDF input for any signal source. We set the receiver to the middle of the range, connect a speaker (phones) and some kind of output indicator (oscilloscope, AC voltmeter, etc.) to the ULF output. Volume control to maximum. During the setup process, in order to avoid the influence of the AGC, by adjusting the output of the GSS or standard RRU (when working with an antenna), we maintain the output voltage of the order of 0.3-0.4V. To obtain the correct (optimal) frequency response in this DFT, all circuits must be tuned to resonance in the middle of the range. There are many methods for tuning without GKCh described (including on this thread). One of the simplest consists of two steps:

Temporarily bypass the middle circuit coil with a 150-220 ohm resistor and adjust the first and third circuits to the maximum signal in the middle of the range, remove the shunt
- to tune the middle circuit to resonance, we shunt the coils of the primary and third circuits with the same resistors, and remove the shunts.

That's all!

US5MSQ: The S-meter drank a lot of blood, in the original version it was not even a display meter - due to the high steepness of the AGC control, the needle stood almost motionless when the signal changed by 70 dB. The R-76M2 took the path of slightly reducing the control steepness, but this did not improve the situation much. I refused to reduce the steepness, because... Now I like the work of the AGC - I don’t have to worry and don’t jerk to the volume control, even if my neighbor with a “kilowatt” switches on next to me.

Several options for expanders were tested, the best results (both in linearity and simplicity of the circuit and adjustment) were shown by the last circuit (on T5) - now we set only the S9 level (50 μV) to the middle of the scale, while the scale is sufficiently linear up to levels of +40 dB. In principle, +50, +60dB are reflected a little, but this is of no practical value.

The readings of this simple S-meter do not correlate in any way with the RRU settings, which allows for a comparative reading of levels (the most frequently requested function) at any gain settings, although the accuracy will be low + - kilometer. Of course, a sufficiently accurate reading of absolute levels, as well as a comparative reading, will be possible only at the gain at which the calibration was carried out, in this case at Kusmax.

US5MSQ: To obtain good selectivity of the circuits, especially the first one, and stable operation of the amplifier, the inductance of the coil cannot be any, much less excessively (several times) greater than the optimal one (in our case, 100 μH).

US5MSQ: We are considering the latest version of the main board. The circuit uses electronic switching of RX/TX modes, for which transistors T11, T13 are connected to a common emitter resistor R39. In receive mode, the supply voltage is not supplied to the microphone amplifier, so T11 is closed by a small (about 0.28V) blocking voltage drop across R39 caused by the flow of collector current T13, the value of which is selected for the following reasons.

The input resistance of this stage, connected according to the circuit with OB, is equal to Rin[ohm]=0.026/I[mA]. To ensure matching with the mixer/detector, the required 50 ohms are obtained at a current of 0.5 mA. By the way, this also results in low pre-LF noise, which is also important. In this case, the voltage at the collector will be about 4.7+-0.5V, and at the emitter T14 it will be about 0.7V less, respectively 4+-0.5V. If necessary, you can more accurately select the collector current T13 using resistor R47

When switching to TX mode, the microphone amplifier is supplied with +9V TX SSB voltage. The current of the emitter follower T11 of the order of 9 (+-1) mA, flowing through the common R39, creates a voltage drop of 5 (+-0.5) V on it, completely blocking T13, thereby turning off the ULF. Naturally, in this case, the voltages on the collector T13 and emitter T14 will be close to the supply voltage.

But let's return to the microphone amplifier. If necessary (large deviation), the required mode T11 is selected by resistor R46. The voltage on the collector T12 will be about 6.2 (+-0.6) V.

Resistor R40 performs a dual function - it increases the output resistance of the emitter follower to the 50-60 ohms required for normal matching of the modulator and attenuates (divides) the output signal of the MCU (the maximum amplitude at the limiter output is about 0.25-0.28V) to a level of 0.15- 0.18V, eliminating overload of the modulator at any levels from the microphone and positions of the R45 engine.

US5MSQ: You must follow certain rules before turning on for the first time!

You must carefully check the installation for errors!

We set all controls (RRU, VOLUME, TX Level) to maximum, SA1 to the SSB position. Having applied the supply voltage, it is advisable to control the total current consumption - it should not exceed 30mA. Next, we check the cascade modes by DC- on the emitters T3, T4, T7, T8 there should be about +1...1.2V, on the emitter T13 - about +0.26V (if necessary, we achieve the required by selecting R47).

We check the operation of the support - at the right terminal of R50 there should be an alternating voltage of 0.7 Veff (+-0.03 V) with a frequency of 500 kHz. If there is no generation, we shunt the quartz with a capacity of about 10-47 nF and with core L4 we set the generation frequency to about 500 kHz and remove the shunt - the frequency should be set to exactly 500 kHz (+-50 Hz). if there is a strong difference in the voltage required, we achieve it by selecting R58 and, possibly, C59. If generation does not appear even when the quartz is shunted, it is necessary to cross the terminals of the communication winding L4 and then according to the above method.

A sign of normal operation of the detector is a noticeable decrease in noise at the ULF output when the left (according to the circuit) terminal of resistor R50 is closed.

Setting up the IF tract can be done traditionally using the GSS (if it exists), but you can also do it with your own standard means. To do this, first set up the CW generator - switch SA1 to the CW position, close the PEDAL and KEY contacts. By adjusting R11 we set the emitters T3, T4, T7, T8 to about +1...1.2V, i.e. For now, during setup, we set the IF gain in TX mode to maximum. By selecting C34 (roughly) and trimmer C39 (precisely), we achieve a generation frequency of about 500.8-501 kHz (more precisely, we select the tonality to suit our taste (hearing), while the self-control signal should be audible in the dynamics). The signal level at the T10 emitter should be 0.7 Veff + -0.1 V - if necessary, select R33. We connect the oscilloscope through a high-resistance divider or a 10-15pF capacitor to the coupling coil L1 and by sequentially adjusting the cores of the coils L2 (we control this resonance by increasing the volume of self-control), L1 and then trimmers C22, C18, we achieve maximum oscilloscope readings. With these adjustments, the resonance should be clear and not at the limit of the adjusting elements - if this is not the case, it will be necessary to more precisely select the capacitances C35, C5, C25 and C16, respectively.

This completes the initial setup, you can open the PEDAL and KEY contacts and enjoy the reception

US5MSQ: Let's look at setting up the transmission path; it is quite simple thanks to the applied circuit solutions.

We connect a configured PDF to the output (this is important, because without PDF, the mixer output signal is a hellish mixture of the remains of the VFO, the main and mirror components), loaded at 50 Ohms. The decisive requirement is to obtain the maximum level of the useful signal and eliminate overload (provide linear mode) of the modulator and mixer. With a GPA (reference) voltage of about 0.6-0.7, sufficient linearity is maintained at a signal level of no more than 200 mV, optimally about 120-150 mV. To protect the modulator from overload at any level from the microphone, a diode limiter D6, D7 is used, limiting the amplitude at the T11 emitter to a level of about 0.25V, and taking into account R40, no more than 150mV is supplied to the modulator. Using the R45 trimmer we set the required level of limitation (or lack thereof) for a particular microphone.

When setting up, it is enough to move the R45 engine up in the diagram, i.e. to maximum gain and apply a modulating signal of about 20-50 mV and a frequency of 1-2 kHz to the input (not critical). By adjusting the IF and EMF circuits we achieve the maximum. We set the optimal level of amplification of the transmission path with trimmer R11, achieving a voltage of about 50-60 mV at the load - this ensures optimal performance mixer We switch to CW and select C40 to achieve about 70-80mV at the PDF output. That's all the setup.

US5MSQ: Regarding the operating modes of RRU/AGC. The depth of adjustment depends on how much we can reduce the collector current of the amplifier transistors (at least to 10-20 μA), while preventing them from completely blocking. Those. lower level of control voltage supplied to the bases of transistors to obtain maximum efficiency The RRU/AGC must be fixed at the optimal value for a particular type of transistor; diodes D1 (RRU) and D2 (AGC) are responsible for this. For diodes of the 1N4148 type, with the ratings 0R1 and R2 indicated in the diagram, this is usually ensured. If necessary, the modes can be adjusted - for example, if the transistors are completely blocked in the RRU mode, then the voltage drop across D1 is not enough - it can be slightly increased by increasing the current through the diode (for example, by connecting an additional resistor in parallel), if not enough, then by replacing it with a better diode .

If the RRU operates normally, then in the AGC mode, if necessary, the adjustment depth is adjusted by selecting R2.

As for the VFO, I didn’t make it, or rather assembled it, but due to the size of my case, I abandoned it and assembled a frequency synthesizer.

A little video about the operation of the transceiver when it was still at the setup stage.

Download the archive with documentation of printed circuit boards in LAY format

Development of UV7QAE.
Synthesizer for HF (160m, 80m, 40m, 20m, 15m, 10m) transceiver with down conversion.

STM32F100C8T6B controller in LQFP48 package. Synthesis on Si5351a. Color screen 1.8" (ST7735), black and white NOKIA 5510 (economy version).
We decided not to install the encoder on the board; this will allow us to use an encoder of any size and place it anywhere in the structure.
You can abandon the encoder altogether since you can control the frequency with the INC and DEC buttons.

The circuit is designed to connect an optical encoder, so if anyone repeats it with a mechanical encoder, install an RC filter at the encoder inputs.

Printed circuit board 85mm x 45mm in Sprint-Layout 6 format for buttons measuring 6x6mm synthesizer_si5351_buttons_6x6M.lay

To enlarge the diagram, click with the left mouse button. Or just download

Output CLK0 - VFO frequency.
CLK1 output - SSB BFO frequency.
CLK2 output - CW BFO + CW TONE frequency.
You can set frequency reverse during transmission in the "SYSTEM MENU" option "TX REVERSE".
Option "TX REVERSE" = ON,

Buttons.
Up, Dn - Up, down ranges, menu.
Mode - Change LSB, USB, CW in operating mode, in the menu for quick frequency entry.
Menu - enter/exit the menu.
Selecting button functions in the "SYSTEM MENU" option "BUTTON MODE".
VFO, Step - Switching VFO A/B, Frequency tuning step. Changes values ​​in the menu.
Or.
Inc(+), Dec(-) - frequency tuning in operating mode. Changes values ​​in the menu.

Enter "USER MENU" by briefly pressing the Menu button.

Entering the "SYSTEM MENU" by pressing and holding the Menu button for more than 1 second.

USER MENU.

SYSTEM MENU.

To control the decoder/multiplexer, pins BAND 160, BAND 80, BAND 40, BAND 20 are used (see diagram).

Control outputs.
Pin BAND 160 = DATA1/A
Pin BAND 80 = DATA2/B
Pin BAND 40 = DATA4/C
Pin BAND 20 = DATA8/D

Binary code for decoder.

Firmware

Source: https://ut5qbc.blogspot.com

I present to your attention a power amplifier for a HF transceiver using IRF510 field-effect transistors.

With an input power of about 1 watt, the output is easily 100-150 watts.

I immediately apologize for the quality of the diagram.

The amplifier is two-stage. Both stages are made on popular and cheap key mosfets, which distinguishes this design from many others. The first stage is single-ended. Input matching with a 50 Ohm signal source was not the best, but in a simple way- using a 51 Ohm resistor R4 at the input. The load of the cascade is the primary winding of the interstage matching transformer. The cascade is covered by a negative circuit feedback to equalize the frequency response. L1, which is part of this circuit, reduces the feedback in the higher frequencies and thereby increases the gain. The same goal is pursued by installing C1 in parallel with the resistor at the source of the transistor. The second cascade is push-pull. In order to minimize harmonics, separate displacement of the cascade arms is applied. Each shoulder is also covered by an OOS chain. The load of the cascade is transformer Tr3, and matching and transition to an asymmetric load is provided by Tr2. The bias of each stage and, accordingly, the quiescent current are set separately using trimming resistors. Voltage is supplied to these resistors through the PTT switch on transistor T6. Switching to TX occurs when the PTT point is shorted to ground. The bias voltage is stabilized at 5V by an integrated stabilizer. In general, a very simple scheme with good performance characteristics.

Now about the details. All amplifier transistors are IRF510. Others can be used, but with them you can expect an increase in the gain rolloff in the frequency range above 20 MHz, since the input and pass-through capacitances of the IRF-510 transistors are the lowest of the entire line of key mosfets. If you can find MS-1307 transistors, you can count on a significant improvement in the performance of the amplifier in the higher frequencies. But they are expensive... The inductance of chokes Dr1 and Dr2 is not critical - they are wound on rings of 1000NN ferrite with 0.8 wire in one layer until filled. All capacitors are SMD. Capacitors C5, C6 and especially C14, C15 must have sufficient reactive power. If necessary, you can use several capacitors connected in parallel. To ensure high-quality operation of the amplifier, special attention must be paid to the manufacture of transformers. Tr3 is wound on a 600NN ferrite ring with an outer diameter of 22 mm and contains 2 windings of 7 turns each. It is wound into two wires that are slightly twisted. Wire - PEL-2 0.9.

Tr1 and Tr2 are made according to the classic design of a single-turn SHPT (aka “binoculars”). Tr1 is made on 10 rings (2 pillars of 5 each) made of 1000NN ferrite with a diameter of 12 mm. The windings are made of thick MGTF wire. The first contains 5 turns, the second - 2 turns. Good results are obtained by making windings from several wires of smaller cross-section connected in parallel. Tr2 is made using ferrite tubes taken from the monitor signal cords. Copper tubes are tightly inserted inside their holes, which form one turn - primary winding. A secondary winding is wound inside, which contains 4 turns and is made of MGTF wire. (7 wires in parallel). This circuit does not have elements to protect the output stage from high SWR, except for the built-in structural diodes, which effectively protect the transistors from “instantaneous” overvoltages at the drains. Protection against SWR is handled by a separate unit, built on the basis of an SWR meter and reducing the supply voltage when the SWR increases above a certain limit. This diagram is the topic of a separate article. Resistors R1-R4,R7-R9,R17,R10,R11 - type MLT-1.R6 - MLT-2. R13,R12 - MLT-0.5. The rest are SMD 0.25 W.

A little about constructive:

Good day! In this article I will add in parts a video review of the assembly of a transceiver from the 60s. Vladimir Semyashkin did a great job of designing and detailed video report, transceiver assembly from the 60s.

What impressed me most was the build quality and the placement of all the components in the case.

Part No. 1

Part No. 2

Part No. 3

Part No. 4

Part No. 5

Part No. 6

Part No. 7

Part No. 8

Part No. 9

Part No. 10

All because it was my first transceiver that worked the first time it was turned on, but then due to circumstances I had to move to the city and there was no longer an opportunity to deploy the antenna to 160m. Well, somehow the 160 meter band became empty; everyone began to rise higher in frequency. I have already published this diagram on my website. And here we will talk about improvements.

Disadvantages noticed when repeating the transceiver:

1. The use of a rather expensive field-effect transistor in the output stage.
2. Lack of AGC system
3. Poor carrier suppression (you have to select microcircuits)
4. Long delay when switching from transmitting to receiving
5. Lack of Smeter.
6. Use of SB cups in bandpass filter circuits
7. No tone generator.

Output stage

When repeating the transceiver, first of all, an output stage was used, using widely available transistors, which made it possible to obtain an output power of about 15 watts. With an input power of about 30 watts. The use of the KT 805A transistor ensures high reliability of the cascade, since the collector-emitter voltage of this transistor is about 160 volts, which allows it to withstand a load break during operation, and a not too high cut-off amplification frequency has a beneficial effect on the stability of the output stage to self-excitation. When using the KT805AM transistor, the power will have to be reduced somewhat.

The output stage transistor is fixed to the rear aluminum panel of the case through a mica gasket, the preliminary stage transistor is fixed directly to the chassis, since the collector is grounded. During testing and operation, the transceiver operated without matching device on various pieces of wire of arbitrary length, without any load at all, on a 220V 100 watt incandescent lamp and no failure of transistors was observed.

The output stage diagram is shown in Fig. 1

The inductor (nominal value not indicated in the diagram) is wound with 0.5-0.7 mm pel wire (on a ferrite ring or on a piece of ferrite, the number of turns of 20-25 is not critical). The use of transistors of different conductivities made it possible to simplify the circuit.

Tone generator, AGC amplifier, S-meter and antenna current indicator.

The next inconvenience is the lack of a tone generator during tuning and the lack of AGC when receiving stations. I provide a diagram of this block (Fig. 2)

As a tone generator and amplifier, Aru uses a circuit taken from the UW3DI-II transceiver (it is easily repeated and works well. The installation of this unit and the power amplifier was carried out on the patches and depended on the location on the chassis since the devices were all small and the chassis design was very different. The device shows the signal strength in receiving mode and the current in the antenna in transmitting mode (when connecting a matching device, we achieve the maximum)

The input of the AGC amplifier is connected to the output of the ULF microcircuit, and so that manual adjustment of the ULF does not affect the readings of the S meter, the regulator is installed after the low-frequency amplifier in front of the telephones.

In Fig. 3 I show a modified diagram of the main board.

Drawings of modified printed circuit boards are shown in Fig. 4

Output 14 of the main board is connected through the pedal contacts (receive-transmit toggle switch) and is grounded during transmission.

Poor suppression of the carrier signal during transmission.

When repeating the transceiver, poor suppression of the carrier signal was observed. The reason for poor suppression lies in the high sensitivity of mixer microcircuits, which leads to interference and direct input of the local oscillator signal, both through the mounting capacitances and through the contact capacitances of the local oscillator switching relay. To eliminate it, it is necessary to introduce additional resistors that shunt the windings of the main board mixer transformers; the resistor ratings should be the same for both mixers from 100 to 200 ohms, which completely eliminated this drawback, while paying attention to the sameness of the ferrite rings. It is advisable to take these rings from the same source (you can use cups from the IF circuits of a transistor receiver, but they should be from the same receiver, grind off the bottoms on an emery stone, leaving only the “skirts”). Transformers are wound with two PEL wires twisted together (3-5 twists per 1cm) before winding, the ring is insulated with fluoroplastic or cellophane tape. Also, these resistors are a load for both local oscillators and allow you to reduce the voltage at the mixer input to an acceptable value. The 500 kHz voltage on the balanced modulator should have a level of 50-100 mV (selected by resistor R7), the GPA voltage 100-150 mV (selected by changing the value of capacitor C54 of the GPA board, usually downward). During manufacturing, it is advisable to install sockets for K174PS1 microcircuits, since very often when purchasing you come across defective microcircuits and you may have to pick them up.

If the balanced modulator does not balance at all during transmission, replace the chip. Also, for smoother balancing, you can make a balancing resistor out of 3 resistors; as a rule, making these changes is quite sufficient.

Long delay when switching from transmitting to receiving.

It is caused by the slow discharge of the electrolytic capacitor C39 of the ULF microcircuit, which, during transmission, is charged through resistor R17 and a diode to a voltage of + 12V, which locks the ULF microcircuit. This can be eliminated by installing an additional resistor from the 2nd leg of the microcircuit to ground (10*k), which will allow the capacitor to discharge more quickly and switch to reception.

The preamplifier of the output stage is often driven.

The reason is the KT603 transistor and the inductor in the collector circuit. To eliminate this, replace this transistor with a KT 3102 and the choke with a 100-150 ohm resistor.

Quite a high level of variable background when receiving stations.

This can be eliminated by installing additional electrolytic capacitors and an additional resistor in the microphone power circuit.

Using scarce 12V relays on the main board in the presence of +33V voltage

More affordable relays with a supply voltage of 24-27V are used; they are powered from a 33V power source; through an additional 30-500 ohm resistor, they are selected so that the voltage on the relay windings in transmission mode is equal to the rated voltage of the relay.

Use of SB cups in bandpass filter circuits.

In the manufacture of several transceivers, circuits on sectioned frames from the MV or DV circuits of transistor receivers were used. The circuits were installed on the main board and did not need to be shielded. The circuit winding is evenly distributed among the sections of the frame; instead of a tap, an additional communication winding is used (wound in a section with a grounded terminal), which makes it possible to more accurately select the connection between the receiving path and the antenna. Coils L2 and L3, 50 turns each; communication coils L1* and L4, 8-10 turns each, PEL wire 0.25

If you want to build your first transceiver! then this diagram is for you, my first transceiver was.

The basis of this transceiver was the SA612 chip. The components used in the transceiver were taken from other devices, so there is nothing new or original here.

Click to enlarge

For reception and transmission, the “Radio-76” “TORS-160” principle is used, which has reduced the number of microcircuits. Naturally, you shouldn’t expect anything beyond the parameters, but “it” works, which is quite enough for a start.

The telegraph part was taken from the "UT2FW" transceiver, the ULF from YES-97, the idea of ​​AGC for IF from RW4HDK, and other components were taken from different schemes as simple and easy to repeat. The AGC circuit itself can be taken from these transceivers.

OEP-13 in the open state has a resistance of about 100 ohms and has practically no effect on sensitivity (variable resistors are used as attenuators). You can get by with just one LM386 for ULF, but when working on a speaker, “it won’t be enough.” The quartz filter is a standard 6-resonator filter at 9 megahertz. In principle, if the transceiver is needed only for SSB, the telegraph local oscillator can be used as a reference.

Lay PCB File

The development of the topic in transceiver equipment is the diagram of the main block of the transceiver for the amateur radio range of 160 m. The diagram is shown in the figure below (click on the picture to enlarge).

The device is a full-fledged transceiver using single-sideband modulation. For its practical use, it is enough to connect an external ULF and a PA - output signal power amplifier.

The unit's local oscillator operates in the frequency range 2300-2500 kHz. The output of the device generates a single-sideband signal in the range 1800-2000 kHz (160 m). To switch from reception to transmission, a voltage of 12 V is applied to relays K1 and K2.

Bandpass filter coils are placed in SB-9 armored cores. Coils L2, L3, L6 and L7 each contain 30 turns of PEV 0.2 with a tap from the 10th turn (except for L3, it has a tap from the 15th turn). The L4 local oscillator coil is wound on a plastic frame with a diameter of 8 mm with an adjusted SCR core (from the UPCH circuit of a black-and-white tube TV). It contains 40 turns of PEV 0.2. Coils L1 and L5 are chokes on the SB-9, they have 100 turns of PEV 0.09 each.

Pin assignments of the SA612A chip:

1,2 - input of the amplifier;
3 - general;
4 - mixer output;
5 - output of the local oscillator circuit;
6, 7 - input of the AM UHF path;
8 - demodulator output;
9 - ULF input;
10 - ULF blocking;
11 - general;
12 - ULF output;
13 - food;
14 - demodulator input;
15 - output of the amplifier;
16 - AGC blocking (amplifier output).