SV1OAN's Blog

Greek radio amateur station, operating as SV9/SV1OAN from Heraklion, Crete

Season finale

It always feels like the season ends at the last weekend of July, with the IOTA Contest. August is for relaxing, thinking over the past months and planning for after September – when it starts to cool down enough for the brain to start functioning again. In Greece, the summer is hot and it is common to take your vacations during its last month. So here I am, next to the beach, thinking over all the excitement ham radio had to offer in the last year: building useful gadgets, chasing awards, and upgrading my equipment.

Running SZ9ERK

Moreover, this year’s IOTA participation was unique in another extent: with a few fellow hams we ran the contest from the local radio club with the club’s callsign (SZ9ERK). This took several months to organize and prepare, and besides proving a very effective team-building exercise, it gave us all the satisfying experience of running a contest with better equipment than we had at home. I was actively involved in the arrangements, and I am thankful to all the people that helped to make it happen. I certainly look forward to teaming up for more contests in the future.

The coming season may also find me with new antennas. I have collected all the necessary hardware and will hopefully soon replace my trusty dual wire dipoles (20 and 15 meters) with a rotating three-band dipole (20, 15, and 10 meters) and an inverted-v for 40 meters. What keeps the project back is that I have to put new cables up to the roof to connect everything, and am still missing a rigid mast to support both antennas.

So, it looks like Crete is becoming my permanent home. It’s been two years here and since I am thinking of installing a “proper” antenna system, I decided to request an address change in my license as well. This means that my callsign prefix will become SV9. The new season brings new equipment, new antennas, and a new callsign. The time may be right for a new blog as well.



I was reluctant to sell my Yaesu FT-897D and give up the all-band, all-mode convenience. However, as has been stated by many, while it was pretty good at doing all things, it did not excel in any one (except maybe SSB operation at VHF/UHF which has proved to be of very limited use to me). I also had a lot of good moments with the radio and had learnt to use it pretty well during the last 6 years. But it was time for an upgrade, and it was necessary to let go of old equipment to cover the cost of something new. Luckily, within almost a week I managed to sell both my (practically unused) Yaesu VX-8E handheld and the FT-897D along with all the accessories I had collected over the years: the SSB filter, an LDG AT-897Plus tuner, and a SignaLink USB.


I wanted an HF-oriented radio, capable of 100 Watts, with an integrated tuner, a USB port to connect to the computer, and priced under 1500 €. This narrowed down the selection to one model from each of the “big three” manufacturers: the Yaesu FT-991, the Kenwood TS-590S, and Icom’s brand-new IC-7300. The Yaesu was getting bad reviews all over the Internet, with reports of early units blowing their finals. It felt like a gamble and not a significant upgrade from the FT-897D. So, it was between the Kenwood and the Icom. The Kenwood had already been many years in action and part of a very prestigious list – the equipment used by the world’s top contesters. On the other hand, the Icom had all the latest technology, raving reviews, and a panadapter. The latter, and the fact that I had to stretch the budget a couple hundred euros more for the Kenwood, were enough to make me place the order for the Icom.

The IC-7300 is overall a great radio. I will not give a full review, as I don’t have the experience or the equipment to do so, but in comparison to other radios I have used, it sounds good, the noise reduction, auto-notch and filtering work great, using the menus to navigate through configuration parameters is extremely easy, and all knobs and buttons feel like being in the right place. It has a tuner, a USB port, and a voice keyer built-in, so no accessories needed. Also, the touch-screen is a pleasure to use and a game-changer. Ten minutes after using the radio I accidentally found myself touching the callsigns showing up on the bandscope window of my laptop screen, expecting for the logger to switch the radio automatically. Well… not yet.

But, still, none of these things feel as important as the panadapter. I had read that once you get to use a panadapter, you can never really go back. It’s true. All other features are nice to have and their absence would be a nuisance, but it’s the panadapter that makes the difference. Because I can now literally see the signals instead of just hearing them. I can check any band in a glimpse – tell immediately if there is propagation, were the activity is, spot a weak signal, etc. I still like to use the dial, but now I now where I am heading to when I turn it around. Using your sight to interface with a radio is a big paradigm shift.

A week ago, my new radio made its brief contest debut during the IARU HF World Championship. Sweeping through the stations one-by-one was never easier, because I could now see their imprints side-by-side on the screen. I just moved from one to the next. After a few hours on the air, I did not miss my old radio. I felt excited. Highly recommended.

Across the finishing line

I got all excited when I heard the Australian station question which “sierra victor” was calling him. I gave my callsign a few times and waited. I was touching the 100 DXCC entities finishing line… The VK station repeated slowly: “SV9/SV1OAN, you are 5-4 in Australia.” Fireworks! The line was cut, and as its sides fell to the ground, I was running around the house celebrating.

DXCC entities worked and confirmed

It all started about 11 months ago, when I decided to pursue a DXCC award as SV9/SV1OAN. A persistent goal has been a great driving force. It urged me to set up the logger software on my computer “properly” (with connectivity to the radio, the cluster, LoTW, and eQSL), try out different antennas (just wire ones), participate in contests, do some CQing, and listen around to chase stations from missing DXCCs.

My logbook shows 946 QSOs in 74 days of activity in total. I could have done this faster I guess, if I joined all major contests, if I realized earlier that each HF band favors propagation from different countries, and if I had more hours on the radio, trying to find the callers before they hit the cluster and generate huge pile-ups.

The QSLing process is not fast, so the DXCC award will take months to follow. I am at 85 confirmed entities right now and still have to send out some direct QSL cards.

What comes next? Digital modes, a better antenna, and a new radio cross my mind. I have also been thinking of learning CW. Should I go for more DXCC entities? I think the entity-chasing safari is never over, but I can participate more casually. Like yesterday afternoon, when I heard a booming signal from Guadeloupe. I grabbed the microphone and called a few times. 101! Oh, the sweet feeling of the “new one”…

The hunt continues

My first award safari trophy is hanging on the wall. I applied for the CQ WPX Award early December last year, right after getting the last needed QSO confirmation. The award manager was very helpful in guiding me through the process, as I needed a combination of eQSL, LoTW and paper card credits to reach 300 prefixes. About a month later, the award was issued, and about three months since I filed the application, I found it in my mailbox.

Shack on March 20, 2016

The hunt continues. I have not done any DXing in the last months, but a few direct QSL replies have raised the number of confirmed DXCC entities to 75 (out of 86 logged). Last week, I took down the 20 meter dipole, which I am intending to replace with a c-pole or some other type of vertical. I hope that the change in polarization will help me reach out to other parts of the world.

Update: My short participation to the CQ World Wide WPX Contest was surprisingly fruitful. I easily logged a couple of new African countries with the 20 meter c-pole on Saturday morning. However, when I came back in the afternoon, I realized that there was nothing new on 20 meters. Instead, on 15 I could hear a whole lot of countries from South America. Next morning, I replaced the c-pole with a wire vertical for 15 meters, with just one radial aiming east. Then, on Sunday afternoon, I logged 8 new countries from Central and South America! Amazing. With an additional African contact a few days later, I am now at 97 DXCC entities (81 confirmed).

The 10 € TNC

Mobilinkd make a nice Bluetooth-enabled TNC. At their site, one can also find the firmware for a simple, Arduino-based TNC, along with detailed instructions on how to interface the Arduino board with a radio. What is even more interesting, is that the Arduino Nano board, used in the example implementation, can be found at extremely cheap prices. While an original one costs over 30 € in Greece, I found a shop selling a clone for 5.5 €! If I had all the other parts required for interfacing with the radio, this project would cost me even less than 10 €.

Arduino Nano TNC shield

I first sketched out a simple layout for the adapter board (the “shield” in Arduino nomenclature), and then soldered all the components on perforated board. The Arduino Nano clone uses the CH340G chip for serial communication, which required a driver for Mac OS X. To flash the firmware, I used the avrdude binary included in the Arduino application and the following commands (adjust the serial port and firmware location if you use these):

cd /Applications/
./avrdude -v \
          -C ../etc/avrdude.conf \
          -c arduino \
          -p m328p \
          -b 57600 \
          -P /dev/cu.wchusbserialfd120 \
          -U ~/Downloads/mobilinkd-473-arduino.hex

The first test was with the TNC connected to my Yaesu FT-897D’s data port. On the other side of the TNC, I connected the TP-Link TL-MR3020 running Aprx, as described here. I could send packets, but not receive a thing.

I searched around the Internet and found that the output level of the radio’s data port is 100 mVolts peak-to-peak. That was clearly not enough for the TNC. (Mobilinkd also provide a tool to monitor the input level.) I either needed some kind of amplifier, or to connect the input to the headphone or speaker output. I changed cables to interface with the radio’s front headphone and microphone sockets. This time the TNC would hear incoming packets as well! I set Aprx to act as an IGate and set off on a car ride with my VX-8GE and a small external magnet antenna, to see if it could track me far enough. And it did – up to the edge of the city.

The Mobilinkd TNC supports 1200-baud AFSK only. A quick look at it’s source code reveals it is based on BeRTOS. A similar BeRTOS-based TNC is investigated by KI4MCW here, while a comprehensive list of Arduino-based TNCs and various implementations is provided by M1GEO here. With an Arduino, you can also build a MicroModem. M0PZT has used an Arduino UNO clone to build the Mobilinkd TNC and has made a video showing it working at his site.

Smaller than a raspberry

The cheap embedded Linux device craze did not start with the Raspberry Pi. It dates back to the Linksys WRT54G and NSLU2. More than 10 years ago, these two devices cost less than $100, ran Linux, and were supported by a large community of users. They were perfect for hundreds of applications – and it turns out that they still are.

TP-Link TL-MR3020

Until a few weeks ago, I thought that everyone was using a Raspberry Pi for generic embedded applications nowadays, when I stumbled upon the PirateBox project. Of course they are supporting the Raspberry Pi as well, but one of their main platforms is the TP-Link TL-MR3020 (and its sibling, the TL-MR3040). This modern-day incarnation of the WRT54G is smaller and cheaper than the Raspberry Pi. For about 35 € you get Ethernet and wireless interfaces, a USB port, an enclosure, a power supply, and enough RAM to run OpenWrt. Even better? The computer store across the street had one available in stock.

I got it without thinking too much. I had to connect the OpenTracker USB to the USB port and run some APRS application on it. Running an IGate seemed a nice idea and I was not the first one to think so. A quick search revealed that KI6PSP, YP0NXX, DK7XE, and PA0ESH have done it already. However, this could be the smallest setup. A handheld radio, the modem, and the router would still make a very small package. As a bonus, the modem could be powered by the USB port of the router (if I had the TL-MR3040 everything could run off batteries).

The OpenTracker USB was recognized by the cdc_acm driver (part of the kmod-usb-acm package), which created a /dev/ttyACM0 device (as mentioned here). Next, I needed Aprx – an IGate designed for embedded devices. There was no “official” package readily available for OpenWrt 15.05, so I made one. You can find it here, and in case you want to create one yourself, here are the necessary files. PA0ESH also has Aprx packages for more OpenWrt versions here. After installation, I connected the modem to my Yaesu FT-60E and configured Aprx to forward all received packets to one of the APRS-IS Tier 2 servers. I sent a beacon from my VX-8GE and it instantly showed up in Success!

Award safari

The summer QSO frenzy and some casual DX chasing have yielded 81 DXCC entities in my Cretan log, 62 of which are currently confirmed via LoTW. Raising the count seems each time even more difficult. I can hear and work almost every European, as well as close-by Asian or African station, but North America and Oceania are impossible to reach. I see the stations in the cluster, but their signal is too week – if audible. It feels that I am now limited by my dipole antenna, regrettably placed at a relatively low height. I guess propagation does not help me either. I can hear some extremely strong stations beaming directly to Europe, but the competition in the pile-ups is so strong, it is hopeless to even try.

QSL cards

On the other hand, I realized that LoTW can also be used for the CQ WPX award. That looked like an interesting parallel path towards the DXCC and another incentive to get on the air and call CQ. I did that a few times, although with mixed results. While my prefix collection grew, it made no difference to my DXCC entities count.

The CQ WPX award proved also a good motivation to register at eQSL, as eQSLs count towards CQ awards. RUMlog integrates nicely with both LoTW and eQSL, so sending and receiving eQSLs was just a matter of setting the username, password, some options, and selecting the appropriate menu items. Now, from the log’s 731 entries, 232 are confirmed via LoTW and 198 via eQSL. Total unique confirmations either way: 321. Total prefixes confirmed: 254 out of 456. More than half, and just 46 left for the award! (I have written a script to count the WPXs.)

In the meantime, I have prepared my SV9/SV1OAN QSL card. Electronic QSLs are fast and easy, but I love the personal touch of the paper-based exchange. Paper cards will also be used for direct confirmations – sometimes a necessity. I use card labels printed by the logging program instead of writing out each card by hand, but for each QSO, I first search online for the QSLing details. It took me several days to check the latest batch of 431 QSOs with, print the labels, stick them on the cards, stamp them, organize by QSL bureau, and ship them over to RAAG. To simplify sorting, I use a 10×15 cm index card box and A-Z guides.

With the SV9/SV1OAN cards, I have also prepared and will send out the QSLs for SX3AM. Again, better late than never. I have already replied to the direct requests.

The award safari continues. And there is a lot more work to it than making QSOs.

Update: It turns out that my confirmed prefix count is wrong (off by about 10). eQSL QSLs are valid only if both users have verified their identity (both have the “Authenticity Guaranteed” certification). The script has been updated.

In the shadow of giants

This year, I decided to be brave enough and run the RSGB IOTA Contest on my own. After all, I was already on an island and had the experience of J48A and SX3AM. This was not my original intention, but the economic situation here in Greece finally rendered any friends’ plans to come to Crete very difficult to implement.

20 m dipole antenna

I was not going QRP this time. I had my FT-897D ready, pumping 100 Watts to the 20 meter dipole. That would suggest making more contacts than SX3AM. On the other hand, I was alone, not expecting to handle 24 hours in a row on the radio, and without low band antennas, that were a missing necessity to the J48A team as well.

Two weeks before RSGB’s IOTA Contest, was IARU’s HF World Championship. I thought it would be a good idea to try it out, to see what QSO count I could accomplish, and to pass the equipment through a proper test.

On July 11th, a little before 12:00 GMT, I was already on the radio, waiting for the contest to start. The 20 m band was pretty active, but not really crowded, so I was hoping to find a nice spot to call CQ. Amazingly, the second it started, there was a station calling every 2 KHz – and in some cases even closer! I quickly dumped the idea of calling CQ and chased the headquarter stations of nearby countries: Poland, the Czech Republic, Croatia, etc. However, my rate was not really good, ranging close to 1 contact every 3-4 minutes in the first 3 hours.

Eventually, I tried calling CQ. I had to go towards the edge of the band to find clear space. The radio’s DSP filters would not help much. Neither did the amplified signals of fellow competitors that would splatter multiple KHz around their center frequency. Some stations did answer my call, but it was not until I was announced in the cluster that I got a small pile-up, resulting in an hour’s worth of contacts in several minutes.

After about 5:30 hours, I took a break. I was very tired and my ears where buzzing. I managed to sit 3 more hours on the radio, before calling it a day. The next morning I was welcomed with a light headache – I assume I owed it to the initial long operating streak of the previous day. After that was settled, I “searched and pounced” for a couple of hours, before calling CQ again. This time the results were better. At about 9:00 GMT, I guess the big guns were in other bands and the propagation was low, so it was easier for close-by stations to hear me and reply. The QSO rate was initially pretty much the same as when scouting around the band. Once more, as soon as I hit the cluster, I got a pile-up. But it was too late. The band opened up again and the giants were back. Stations would pop up so close that I couldn’t stand their high-pitched noise and it was evident that no one could hear me anymore. With small breaks, I managed to make contact #250 just 2 minutes before the end. RUMped recorded almost 11 hours of operating, 512 QSO points, and 61 multipliers, for a total score of 31,232 points.

I didn’t expect to do much more in the next contest. Indeed. The moment it started, it was easily recognizable that participation was much larger. More giants around and I was in their shadow, struggling to be heard. Another problem was that this time I had much more things to say for each QSO. I had to give the island number, in addition to my lengthy, non contest-friendly callsign. I took the challenge a bit lighter, trying not to push myself. Again, on the morning of the second day, although earlier this time (at 5:25 GMT), other stations started replying to my call and some minutes later I was in the cluster. During a little bit over an hour, I managed to log about 60 QSOs – nearly a third of my overall count. In total, I did 170 QSOs in about 7 hours, which gave me 1,440 points and 34 multipliers. My final claimed score of 48,960 points was very close to last year’s.

The summer contesting season ended with a very casual participation in the European HF Championship, on August the 1st.

I certainly wouldn’t expect to reach the 1,338 QSOs of J48A. That was a team of 5, working 24 hours on 4 bands and both modes. I was one person, working less than 12 hours on 1 band and SSB only. My initial thoughts were that I would certainly need more and better antennas to be more competitive. I would also like a radio with better filtering capabilities to avoid the fatigue. I even started thinking of buying an amplifier (take a look at this talk for an explanation on how dBs relate to contest score). However, given my QSO rate at both contests, a 24 hour run would yield about 540-580 QSOs (assuming of course the same band conditions throughout the period). So, I surely need to build my operating skills as well. To become a giant, I must grow stronger, not only get larger. After all, isn’t that what contesting is all about?

Thanks to all stations that participated. The giants are there to raise the bar and make this even more interesting and fun to all – big or small.

On a side note, I now have 73 DXCC entries in my SV9/SV1OAN log, 48 of which confirmed. I am nearly halfway to my goal.

6 down, 94 to go

I have been off the radio waves for several months till recently, as I now got back transmitting as SV9/SV1OAN. Last August, I found a new job, which required moving to Crete.

Shack on July 1, 2015

I left my soldering iron and any ongoing electronics projects behind, but took all my radios with me, which now include an FT-897D, an FT-817ND, and two handhelds (a VX-8GE and an FT-60E). I did not know beforehand if I would have the opportunity to set up a base station, or would need the flexibility of a light-weight, portable radio for occasional activation of my callsign from SV9-land (IOTA reference number EU-015).

Settling in the new house and getting up to speed with the new work environment required a lengthy period of adjustment, even in such a familiar setting, as I have already lived in Heraklion for 5 years as a student. It was just a few weeks ago, when I finally got to draw two runs of RG-58 to the roof of the building and erect both a small V/U vertical antenna (a Diamond Χ-50) and a 20 m dipole. I am glad to say that I have already made my first 50 QSOs from 29 DXCC entities, 6 of which confirmed via LoTW. As a new challenge, I promised myself to go for a DXCC award as SV9/SV1OAN. It will be my first.

Pirates of Agia Mariani

This June, as soon as the first summer heat wave hit us, I started thinking of the IOTA Contest. After last year’s exciting experience in contesting, I began contemplating something even more adventurous: a DXpedition to an uninhabited, rocky islet, similar in style to the great ham radio excursions organized in far away islands of the Pacific, albeit in a much smaller scale. Fortunately, Greece is full of such islets, so I just had to find one that was accessible, and a team of brave, fellow operators that would eagerly participate in a DX crusade.



Panos, SV1COX, was thrilled with the idea. We started looking for an appropriate destination and somehow decided to journey to Agia Mariani, a very small island at the southwestern edge of Peloponnese. Agia Mariani is part of IOTA EU-158, which also seems to be the most wanted group of Greek islands. Moreover, we found a video on YouTube, showing that the island had a dock for small boats, from which a path lead through the bushes to the church of Agia Marina. Built on the islet’s highest point, the church seemed recently renovated, with an adjacent tiled shelter providing a place to camp and operate in the shade of the burning sun. Ideal.

I took up the task of issuing the special callsign, SX3AM, and arranging our transport to and from the islet. Panos was put in charge of organizing the equipment and preparing the antennas. We decided that we should arrive on target early Saturday morning, so it would be best to sleep in Methoni the previous night. We would probably sail from Methoni anyway.

Another major decision was to run the contest in the QRP category. Operating at 100 Watts, would require carrying over big radios and power supplies, plus a generator and gas to power them. The generator would require resting every 4-6 hours, meaning we would also need batteries. A lot of stuff to handle, too much weight to move around, and – most importantly – too many things to go wrong. Lower power meant just ultra-small radios and a few batteries in our backpacks, which would already have to include clothes, food, and water.

A few weeks later, the callsign was approved and we made an open call for participation. Many friends liked the idea and expressed their intention to join us, but we needed confident answers, in order to know how many beds to book and what kind of boat to hire. Takis, SV1IZF, responded quickly and became part of the team. He also agreed to solve the problem of powering our equipment. He would bring over many amp-hours worth of batteries, fully charged and double-checked.

So, on the afternoon of the 25th of July, I put my bags in the car and drove over to pick up the others. Half an hour before midnight, we arrived at Methoni. A quick bite and back to the hotel to get some sleep. Methoni was very quiet, and luckily the bursty winds of the previous week had calmed down. I talked to our captain and he would wait for us at the port at 9 AM.

Next morning, we stopped at the local bakery for food supplies and headed to the port. We had a lot of bags to carry. Right on time, Mr. Christos picked us up and we departed for a 40-minutes long sail to Agia Mariani. Upon reaching the island, I was amazed by the dense vegetation. The path leading to the church was very narrow and its sides were decorated by giant spiders, that watched us peacefully going up and down four times to bring up all the equipment to the shelter. Takis brought along two 70 Ah batteries. Very big to handle, but necessary to ease our anxiety of running out of power.

An hour before the contest’s start, everything was ready. We had four dipoles – one for every band, except 80 m, and an additional C-Pole antenna for 20 m. Our radio setup included an Icom IC-703 and a Yaesu FT-817ND, with an Elecraft T1 tuner. We also had a spare Yaesu FT-817, connected to a 2 m Moxon antenna, for local communications.

Exactly at noon UTC, I started calling on 20 m. A few responses. Actually nothing, compared to J49A’s pile-ups. Soon, the band got crowded and there was a station to be heard in nearly every KHz. I began feeling desperate. Over an hour later, and we did not even have 10 QSOs recorded. I imagined QRP being a bit hard, but not so difficult.

We started chasing other stations to gather some points. Every time I got to operate, I started by calling CQ for 10 minutes, in case anyone heard us. Once, I was lucky enough to create a small pile-up and log about 20 contacts in 15 minutes. All other attempts yielded a couple of replies at most.

It was time for CW. Panos took over and nearly doubled our contacts in an hour. Me and Takis tried to get some multipliers on SSB. Every QSO was an adventure. Again, a completely different experience than last year’s. You had to work very hard to get each point. It was not a matter of speed and accuracy to handle tens of stations calling at the same time. You needed persistence and patience.

Nightfall brought changes in propagation and in the sounds around us. Panos’s concentration was often shattered by squeaks coming from the nearby bushes. Suddenly, we realized that the church’s front yard and surroundings were filled with rats. The island’s residents were not daunted by our noise and did not even seem to care when we flashed them with our spotlights. We had camped in the pirate’s lair, and soon the pirates started raiding on our supplies. They must have smelled the freshly baked bread and the juicy fruit.

We quickly changed roles. One had to work the contest and the others had to protect the camp. In the calm between battles, I managed to get a few hours of sleep in our tent, guarded by the others. The morning light, our mightiest ally, drove the rodents back to their hideouts. When I woke up, I found the others exhausted. “An unbelievable night,” they said. I believed them.

I made some more contacts before midday. They were even harder. Most bands were closed. We decided to leave a bit earlier, as operating in these conditions seemed pointless, we had a lot of things to pack and carry back to the dock, plus a long drive back to Athens. Indeed, right on the contest’s end we were all ready, waiting for the boat to take us to Methoni. As a bonus, we spared a few minutes for a quick, refreshing swim in the crystal-clear waters of Agia Mariani, before waving farewell to its inhabitants.

We made 180 QSOs, as show in the table below, of which only 2 invalid for the contest. We logged 21 different island groups and got 38 multipliers. We claimed 50,160 points.

40 m 11 64
20 m 49 47
15 m 8 1
Total 68 112

Many thanks to SV1COX and SV1IZF and to all stations contacted. It was a remarkable journey, from which we learned a lot. Next time, it will be even easier. Our battles with the pirates of Agia Mariani did not let us down, but have already turned into stories of courage and endurance; properties necessary for operating with very low power amongst giants, in the heat of one of the world’s biggest contests.