TRX multiband Antek SQ9GL

I present here an interesting design of radio-based TRX TRX SQ9GL Antek ( AVT2310 kit ). This is my version of a multi-band known from several publications Antek's TRX SP5AHT . TRX is a small portable QRP power of approximately 6W (80m band), ultimately 10W, SSB working in five bands: 80m, 40m, 20m, 10m and 6m SSB (LSB, USB, CW) and having a digital input on emissions (eg BPSK31). .

At the beginning I started with the construction of a complete kit assembled AVT 2310 version 1.2 (corrected some errors in the PCB). Kit has been assembled with the default components and tuned for 80m. For this I decided to make a module synthesis of DDS2 by Jarek SP3SWJ (synthesis of two generators and two VFO VFO as BFO for SSB), which is next.

The whole cabinet decided to put in the PC power supply, which Antek kit fits very well.


Speaking of design from scratch, I started with the least loved by designers that is the case of electronics casing. As I said before radio was to be packed in the casing of the PC power supply.
After removing the top cover shows the front and rear wall. On the front (side with holes inlet adapter) had to be glued to the front plate. By contrast, the back wall (wall with a hole for the power supply fan) had to be bolted back plate, with attached heat sink for the power amplifier, antenna socket and power switch and the DB-9 connector for connecting the PC interface.

I made the back wall with a piece of aluminum sheet with a thickness of 0.5 mm. Dimensions matched so that the wall exactly fit over the entire surface of the back of the adapter.
The wall that cut the holes for the power switch, the interface connector (DB-9), and two UC-type antenna socket. This also cut a hole for the transistor HF amplifier.
In addition to the rear of the heatsink is screwed to the rear of the power transistor. I use the heatsink from the processor PC. In case of strong heating of the fan can be screwed. The heat sink is screwed by means of threaded screws M3 earlier holes.
Thus prepared the back plate is screwed to the adapter in place of the original holes.

Time for a front plate. It is also done with a piece of aluminum sheet with a thickness of 0.5 mm. I cut holes in it to the display of synthesis, speaker, a keyboard. Drilled hole in the middle of the pulse tuning. Additionally, in a later stage was cut a hole for a microphone jack and a hole drilled in the LF amplifier potentiometer.

As the panel prepared a pre-glued to the housing supply in the front. I used epoxy glue to two-component metal available in most shops with products screw. Adhesive after curing has led to additional stiffening of the faceplate.

Before bonding the front plate smeared with glue the entire surface of the wall. Front panel has been pressed with clamps and leave to dry.

Synthesis DDS

In the meantime, proceeded to build a synthesis of DDS2 by SP3SWJ based on DL4JAL . The exact description of the implementation of the synthesis is found at Jarek at:
This is a DDS synthesizer system with optional dual VFO (two chips DDS - DDS1 and DDS2), which can work as a VFO and BFO, and therefore ideal for the construction of our Antek. The use of BFO synthesis is not necessary, but to work in the USB, it is necessary detuning BFO. Adding a second generator of the synthesis of perfectly fulfilled this condition. DDS synthesizer is built on the basis of the ready synthesis AD9850/51 and AVR processor type AT89S52 . I selected AD9851 chips due to possible frequency multiplier x6 unlock and use easy to come by quartz generator 25-30MHz. Moreover, in the future could try to run an additional 144 MHz band, and therefore need a frequency above 100MHz. This generator has a lot harder to get.
The construction started from the performance synthesizer tiles using the "iron". To this end, I cut two-sided laminate for a bit more dimension, cleaned it with fine sandpaper (density 1000), iron warms up to about, 150C (2 dots), printed on a laser printer drawing on paper with colored magazine and put on it on the plate. After 3 minutes of annealing put the plate in hot water with dishwashing liquid. After 15 minutes he left the paper with ease. PCB etched in ferric chloride. At the end of the drill holes drilled and milled 1mm thicker drill holes from the elements on the screen. Finally, I dragged the plate rosin dissolved in methylated spirit.
I proceeded to mount components. At the beginning of the brazed components, SMD and through hole components. Processor and memory put on stands.

After turning on the synthesizer removed the circuit boards and other small underdeveloped. The synthesizer at the beginning I used 100MHz crystal oscillator. When I try to 144MHz band and 30MHz can be used to enable replication frequency x6. For the system to work properly, the generator was necessary to edit the original program for the AVR and enter the appropriate multiplier the last line of code HEX. I have used this popular program ICPROG. Detailed description on the webpage Jarek SP3SWJ.

Finally, I turned the motherboard plate synthesizer DDS2 - the second VFO as the BFO. I plugged the other VFO, and measured the frequency of my simple frequency meter.

For the first run I prepared a simple synthesizer keyboard in the form of four DIP switches. In addition, I connected the pulser. I used the sheer pulser available including in AVT. The total cost of performing the synthesis of closed within less than 100 PLN, which, combined with AVT Antek ready kit offers the opportunity to make multiband QRP SSB transceiver on HF bands in the range of decent 6m performance at a price below 300PLN. Ambitious plan, but it seems reasonable!


Then proceeded to do a small keyboard. Like a simple thing, but not the end. For the normal run of synthesis enough 4 basic keys and the pulser. But they want to make full use of its ability to perform a complete keyboard consisting of 15 +1 keys. You can make the keyboard by projects on the webpage Jarek SP3SWJ. My radio was to fit in a small package, so the large keyboard, there was no room, taking also into account aesthetics. So I decided to design a plate of small keyboard that would fit on the right side panel, cut a rectangle in size 30x35mm . The figure below his draft bilateral printed circuit board.
Print version:

WARNING! The pulsed diode with an anode represent the beginnings of a drawing!

Double-sided PCB I made a similar method as PCBs synthesizer, but in this case, first printed out, punched TOP page. Then etched using this site on the surface of the plate, the plate placed on the surface of ferric chloride, so that the flows freely, being careful not to splash the faces. Then, after cleaning, etching plates back on the iron carefully cut to put a figure on the other hand already BOTTOM. For nothing is moved first, I laid the tile and I put the drawing and then turned on the iron plate at the same time pushing the figure.

PCB should be drilled two drill bits - drill guides section 0.5-0.8 mm, 1 mm holes for the buttons. Additionally, in the corners of the holes for screws M3. Places with rings to solder on both sides with wire. I used a 1N4148 diode in SMD version. DIP switches on the keyboard used the dimensions 6.5 x 6, available such as AVT in the network.

After the printed circuit board keyboard, but before the assembly, I made a mount for these plates on the front in the form sticky M3 screws. To do this, the screw is screwed to the plate and a synthesizer keyboard. Drilled in the plate synthesizer 3 holes for the screws in the keyboard fourth.

Then I put both PCBs in the front panel carefully folding horizontally so that the tiles were fairly evenly matched to the plate - LCD display is properly positioned over the hole. I put the keyboard to insert the four buttons at the four corners, to facilitate adjustment. Then smeared with epoxy glue and screw heads once again I put the plate carefully aligning screw into place. So I left PCB located the glue dries.
When I put the adhesive on the PCB screws. Everything fits as planned. Any slight shift can be corrected slotted holes and shift them to the studs.
Finally cutted the keyboard. Before soldering the components thoroughly checked the presence of faults, which I removed. I repeated the action after welding the wire guides. I plugged the keyboard and all sail!
After matching the PCB gently cleaned with abrasive paper (1000) and the panel prepared for painting with paper glued inside, to protect against excessive splashing paint. Then I painted the housing black matte acrylic paint. The effect of medium, but in the beginning enough. After the paint dries I started attaching the speaker.

First, I cut a piece of material (for this I used a piece of black stockings), rolling it into 4 layers and glued it in a hole using a hot glue glu. Then I glued in a similar manner by placing it first speaker of the material, and then taping gently glue in several places, thus preventing any possible shifting. I thought at the same time not to push him too much and provide him with sufficient clearance.

First start of the receiver

I decided in the next step to start a trial of the receiving device in the set of synthesis. Antek has a 2-generator - VFO (VXO) and BFO. My version has a built Antek intermediate frequency filter circuit with a value of 8.6 MHz in the form of four crystals, each 8.6 MHz. Before connecting to the synthesis of the main board first, I removed all the elements of the two generators, leaving the input VFO capacitor C18 (1n) and resistors R17 and R18 (the 470R). The position marked similarly left BFO C67 (1n) and R19 (220R) and R20 (1k). The board left the stabilizer system US7 (78L05) for use later.

Coaxial cables connected the VFO output to input DDS1 of synthesis, and place the output BFO DDS2. Ultimately, I intend to connect an additional transformer RF Isolation of 1:1 according to the authors of the synthesis of DDS. For now, it was not necessary. Connected a 12V power input of US8 (also desoldering). Potentiometer mounted volume. The speaker used a small speaker with neodymium power 0.1 W and a diameter of 27mm. HF antenna plugged in (19.5 m dipole) and turned on the power. I measured the voltage on systems US1, US2, US3, US4, and US5. The speaker could be heard quiet noise and disturbance. Transgressed the BFO tuning, after all, is the basis of SSB. First on the menu I chose the setting of IF LSB. The default value of 9.000000000 I set my middle or 8.667000000 - as stated in the documentation of Antek. I turned on the 80m and LSB. It was immediately hear the correspondents, but no clear modulation was too shallow. I raised sensitivity delicately turning the separator L4/L5. At the moment I left the default input filter installed. Filtering system for the bands I'm going to design later. After setting the highest sensitivity experimentally changed the LSB IF frequency up and down. Finally, set the BFO in the LSB mode the value 8.662500000, at which the receiving station was clean and legible, and the same modulation seemed to be deep enough. USB mode I determined the frequency of the BFO with a value of 8.66500000

First start the transmitter

First I connected the power supply immediately 13.8 V to insert the transceiver. I turned on the trial PTT in order to measure voltage, but without signal modulation. I checked the voltage at R41 (about 0.9 V), R36 (about 1, 4V) and R31 (0.15 V) to establish the quiescent currents transistors. There were no major deviations. Mounting potentiometer R33 I established in the middle position. I connected the antenna to the input RF power meter with the load. Instead, you can connect the two resistors 100R 2W each. However, I immediately wanted to control the output power using a power meter. For I determined the frequency synthesis 3.714 MHz and LSB. The microphone input signal plugged LF sinusoidal with a frequency of 1.5 kHz. Instead, you can connect any signal such as the telegraph key. I turned on the PTT. I immediately noticed a strong heating of the IRF530 transistor. The power meter gave no indication. I turned off the power and checked the correctness transformers connecting the transmitter. It turned out that in both incorrectly connected the winding transformers, resulting in suppression of the signal. Dismantled transformer connections and improved soldering them back. The radiator is mounted transistor IRF530 the transceiver package. Przylutowałem everything again. I turned on the PTT. It's obvious it was an indication of transmitter output power of 0.5 W. In addition, I turned on the IC-718 transceiver to eavesdropping. With the PTT Antek gently hang around R33 thereby increasing power. And so I reached up to about 1W. Above this value was triggering the transmitter. With this setting, the TRX IC 1W-718 was heard clear signal LF Leaving such a setting tuned coil L10 and L11 for maximum power at which the radio next to it was strongest signal is heard. Finally got the power of the order 1.5 W Tentatively lowered to 12V power supply and turned on the transmitter. It turned out that when the voltage of power does not exceed 0.8 W. At 13.8 V power supply has increased to 1.5 W. Finally, I connected the meter to the load instead of My W8010 dipole antenna. I turned on the broadcast. On the radio next to the antenna is not connected could be heard clear signal LF Test completed successfully. The transmitter was initially launched for the 80m band, but the target transmitting power will be increased.

A switch on the band 80m, 40m, 20m, 10m and 6m

The intention to construct this device was equipped with Antek the opportunity to work in more than one band. Normally, my version was running the default Antek 80m. So I decided to add more bandwidth. To make the most of space in the package to the radio, I decided to leave for the 80m band filter original. To design them only for the band switches.

KAntek design in a nutshell is that we can distinguish three stages of filters: low pass filter symmetric transceiver, receive filter with a bandpass filter and a transmit bandpass similar to the previous one. The first filter I decided performed in a simplified form of the original solution, leaving the two coils and three capacitors and matching their respective values ​​for each band, respectively. The second filter - like the first receiver was modified according to each band. The third filter due to the inclusion of the LC circuit in the loop of the collector of the transistor in the first gain stage, required a repetition of the each band to fine tune the transmitter in each band. The purpose of this filter is to pass the signal from the mixer difference being BFO and VFO signal and filter out the rest of the signals.
In my simple diagram shows that it is simply duplicated all three systems, including filters them using transmitters based on the selected band. This selection can be made, including low (below approximately 1V) at the base of transistors controlling relays. The values ​​of LC elements in different filters using the chose a simple formula and you should not hold onto them. The output circuit of the receiver filter, attached to the input mixer is also included in the key input signal transistor PTT to prevent the ingress of signals from the transmitter to the input mixer.
The entire system is controlled by a BCD decoder to one of 10 in circuit 74LS145. At the entrance of the system turned on the output of the processor synthesis DDS.

Designing a printed circuit board with a relatively small dimensions (135x85mm) for items containing 30 transmitters turned out to I not an easy task. PCB has been designed to allow its placement on the PCB of Antek.

Printer-friendly version: bandswitch.ZIP
List of components: BOM.ZIP

For those interested, enter the dimensions of the deployment of three additional holes in the PCB AVT-2310 for fixing the plate filters.

Tile also has 5 holes 6mm to allow for tuning coils for 80m and adjust potentiometers located on the bottom plate. In addition, these holes were to be brought some shielded wire from the lower part of the filter switches. Size of the tile was also taken into account the presence of PCBs DDS synthesis and the possibility of getting into the radio transmitter. You can opt out of any of the bands and simpler PCB design with less parts, but I stubbornly decided to turn 5 bands including 50 MHz! I designed a double-sided PCB, every available space on both sides filled the ground for better shielding.

PCB made ​​by myself without metallization of holes (which incidentally would be very handy), a technique similar to that used it in performance of the keyboard plate for the synthesis of DDS. I used the method of iron, digesting both plates in installments by placing it on the surface of ferric chloride first one way and then another. Most drill holes drilled 0.6 mm. Holes in the stand of U10, relays and filters 7x7 drill drilled 0.8 mm holes drilled earlier thickening.
After drilling the holes slightly further honed surface with sandpaper plates with a density of 1000 to eliminate edge troublesome holes. At the end of PCB painted with a solution of rosin in a spirit by painting both surfaces two times PCB and leaving to dry for four nails driven into the board.
PCB Assembly, I started to solder all the guides. I made them just soldered on both sides of the pieces ends of the resistors and cutting off at the surface of the PCB.

When assembling some difficulty was putting all the relays in a situation where some of the tips should be solder from the TOP. Installation of the relay started from those that were difficult to reach the tip of the side TOP. After that side of the So only the soldered ends of which were to lead other than weight and was not free. Soldering iron was applied gently as close as it was thought possible, and at the same time not to damage the relay cover. Deploying relays in the project I tried to allow the solder is in order. During the assembly carefully thought not to miss any of them. Any relay can be forgotten patch BOTTOM side of the plate.

7x7 coils each coil of solder by placing approximately 0.5 mm above the PCB and slightly "grabbing" tin the tip, which had to be soldered also from the TOP keeping the rules similar to the soldering relays. Before soldering each coil, first took off with her ​​body, then solder the necessary ends of the side of TOP, BOTTOM later, and the final overlap back coil body by pressing it and soldered to ground. As the plate was painted rosin, tin easily grasped quickly the solder pads.
During installation, the relays and coils successively also soldered capacitors between coils to a later time does not have problems with access to them.

Trim the signal wires and solder to coaxial inserts it took me an hour. However, preparation for soldering the pins on the board It took me a second transceiver that. It is not too complicated, but tedious. For the prepared PCB from the transceiver and switch brazed DDS synthesizer lines. To control the synthesis of brazed pins ribbon cable from the control input of the switch.
PCB switch plate placed over the transceiver with 4 screws 3mm, blocking the PCB with insulating nuts and washers. Here I used the heatsink pad TO92 transistor.

First start of the receiving switch

Preconditioned system finally hooked up to 12V. At the beginning I tried to switch bands. When pressed on the keyboard buttons switch the band there was no reaction. I checked the voltage at the relay. Liquidated break on one track. I checked again layout. This time we could hear the relay switching.
Then I checked with my simple individual broadband frequency generator set. To this end I connected generator only one end directly to the NE612 mixer input at the point marked as the C66 (at this point is now to connect the relay PZ1A / O). switching another band generator tuned by checking the correct reception. Finally, I made sure that the receiver is ready to tune in the individual bands.

Tuning the receiver started from the lowest band, which is 80m. Connected a generator with AM modulated signal to the antenna input by setting the generator and the synthesis frequency of 3.700 MHz. I connected to the antenna input of the antenna, although there would be enough to run a simple 50 ohm resistor. Gently turning the coil L5 through a hole in the plate I tried to switch to "catch" the signal from the generator. Unfortunately, silence. So I started to turn touch the probe with different degrees of the generator filter. In the end it has proved necessary desoldering short-circuiting transistor T140 signal to ground during the reception. Do not know why the transistor polarized at once to lead. The signal appeared. Then unplugged the generator and tuned coils L4 / 5 and L3 to the signal from the antenna directly correspondents HF.

Another band 40m, 20m, 10m and 6m running the Easy in the same way by joining the generator to the input and trying to tune the frequency. Just managed to get one after the other reception band 80m, 40m, 20m, 10m and 6m .
Sensitivity to the bandwidth 14MHz, 28MHz and 50MHz is a little lower, but I do not mind listening to it in a nice band. In addition, I decided to use the input amplifier system based on MAR6 - A simple amplifier circuit soldered directly between point PZ1A / 0, and the antenna filter. But it is not necessary, because the receiver on all bands is behaving itself.
When starting receiver were necessary to correct the resistors R17 and R18 on the board Antek to get the VFO signal from the synthesis of approximately 300 mV. Likewise, I did for BFO changing R19 and R20 to 470omów.

The initial start of the broadcast switch

Transmitter tuning started from 80m band. Connected a 12V power supply, and PTT switch around 2kHz signal for the modulator input. I connected to the antenna input RF Power Meter In addition to the IC-718 turned on and set both radios to the frequency of 3.700 MHz. I turned on the PTT. There was no signal, so I checked the correctness of assembly. I removed break on one of the relays. After the amendment the signal appeared. It turned out that after my modified output filter power output rose to more than 6W ! More precise tuning of the transmitter and the sustainability of the modulator I decided to do at the end of the device is fitted in the housing.

Switched to a bandwidth 40m/LSB. I set the frequency of 7.050 MHz and the transmission checked. The signal appeared immediately, so there is no short circuit or interruption of the plate. Adjust slightly coils L105 and L106 have set the highest level of signal. However, the output power from a transmitter in the 600mW was 7MHz band.

The next band I switched mode to USB transceiver. For the 14MHz band transmitter output power was 200 mW. Continue for approximately 60mW 28MHz bandwidth and 50MHz for about 5 mW.
Transmitter tuning on the higher bands, caused in turn triggering a power amplifier in the 80m band, which is characteristic of the transistor amplifier IRF520. Measured the capacity of the individual steps controls for each band, measured at the input power C47_1. Amounted to 170uW, 130uW, 120uW, 61uW and 42uW to 50MHz bandwidth. It follows that this system can be used Band switch for controlling a better amplifier.
In this device, however, is planned to launch a more powerful transmitter.

Final assembly of the device in the housing

I proceeded to install the transceiver in the housing. First, I put the main PCB to the housing Antek. Then screwed potentiometer and the pulser. Soldered to a shielded wire in a potentiometer. Then I started using a keyboard tray pads TO92, cutting them to be equal. Then I placed the DDS synthesizer. Soldered the wires from the pulser synthesizer.
At the end of the band put a switch module before setting caps on the screws M3 distance. Soldered wires in the screen to the microphone jack and speaker. Finally, I turned on the machine again and corrected the reception band. In the final I drove a little transmitter power setting the 80m band power about 6.5 W, at which the system is not aroused.
I turned the cabinet, provided the cover. The next step was to adjust the microphone amplifier with additional compression system condenser microphone to my ICOM HM-36, which requires a good fit.

As I made a simple compressor microphone amplifier system with compressor system dynamics based on LM324 operational amplifiers. I used the finished project here AVT2454 (Compressor Dynamic SSB). I made a classic compressor circuit board thermal transfer method. Submission of took me about 2 hours. PCB mounted on the bar, aluminum sheet having a cut arms which are bent downwards. They drilled two holes 3 mm. PCB M3 screws screwed into the blade and the whole secured by means of one screw M3 of the band switch module.
For electret microphone to work, it was necessary to bring supply voltage of 7-8V on the mic. For this purpose, IC 7808 stabilizer is screwed into one of the screws display. 8V voltage stabilizer plugged into the microphone through the 4k7 resistor.
Finally adjusted the compression ratio of the system. It quickly became clear that for the 80m band compression had to be very small. Increasing the excitation caused compression of the compression system and ultimately the formation of echo (duck). On the higher bands there was no problem.

Wojciech Matuszyk SQ9GL