How to Build a 10 Meter Beacon
by Tom Sevart N2UHC

I'm not sure when the idea to put a beacon station on the air first came to me, but it was probably not long after the QRP To The Field event campout with the Four State QRP group in April of 2003. Walt, AG5P had given me a couple of old junk CB's he had brought along. Walt knew of my affinity for old CB's when it comes to scrounging parts for homebrew projects. My first QRP rig, a Small Wonder 20+, was placed into an old CB case, so Walt knew I'd probably get good use out of the ones he had picked up at a hamfest and generously donated to me.

At first, I think my plans for one of the radios was to convert it for use as a 10 meter CW rig. Just key the transmitter in unmodulated AM and build & install a BFO so that I could receive CW. Eventually, though, I thought it would be a better idea to put a beacon on the air. Jay, KØETC, operates a 6 meter beacon in Joplin, but as far as I knew there were no 10 meter beacons around. I had often listened for beacons on 10 meters to see if the band was open. This is the main important reason beacon stations exist, to provide an indicator of a band opening. By noting what stations are heard, one can determine what locations one can have QSO's to when the band is open. Not only that, but it's fun to DX beacons and collect QSL's. This, I thought, would be a good use for an old CB, since it should be an easy matter to convert a radio from 27 MHz to 28 MHz. The idea of using the unmodulated AM carrier to send CW would still work. I just wouldn't have to worry about the BFO since I was only concerned with building a transmitter.

The NorCal keyer kit, completed & ready to be installed.

The controller would be the easy part. It was a no-brainer what I would use. I had built a CW keyer using the same chip found in the NorCal keyer kits. The chip has a beacon function which repeats two of the three 40-character memories. This would work great as a beacon controller.

I checked out the two CB's that Walt gave to me. The first was an old Johnson Messenger 123A 23 channel CB. It is a non-synthesized crystal-controlled radio that mixes 10 crystals to get the CB frequencies. It checked out OK and seemed to worked just fine, though some of the channels didn't seem to work if I wiggled the channel knob just right. However, the other channels worked so it would be a fine candidate as a beacon transmitter. The second radio was a Midland Model 13-830 23 channel radio with a synthesized VFO that also checked out OK. I did find out later that it also had a slight problem with the channel selector knob, in that if I wiggled it just right on some channels, the frequency would change to a different channel. For example, if you were on, say, channel 12 and moved the knob just right, it would be on channel 6. Since I would only be using it on one frequency, this didn't concern me too much.

Inside the Johnson Messenger 123A, two reinstalled crystals can be seen on the rotary switch at right.

I opened up the Johnson and took a look at the crystals. Luckily they were all marked with the frequencies stamped into their cases. Some of the crystals were in the 32.700 to 32.950 range, while the others were in the 5 to 6 MHz. range. By listing and charting all the frequencies, I determined that the radio subtracted the 5 MHz. crystals from the 32 MHz. crystals to give the right channel frequency. The 6 MHz. crystals were subtracted from the 32 MHz. crystals which showed to be 455 kHz lower than the right frequency. It was obvious that the 6 MHz. crystals were the receive crystals and the 5 MHz. crystals the transmit crystals. It was important to map out which crystals were used for each channel in order to know where to put the replacement crystals.

Digging through my stock of crystals, most removed from other junked CB's, I found that I had some in the 37 MHz. range. If I could mix one with a 9 MHz. crystal, it should give me a frequency in the 28 MHz. range. I did find a few in the 9 MHz. range, and once I charted them against the 37 MHz. crystals, found that I had numerous combinations that ended in frequencies in the 10 meter band. I carefully made a note on paper of the positions of each crystal in the Johnson, then removed them all. I then soldered in a 37 and a 9 MHz. crystal in the position for channel 1. I was almost ready to go.

I had realized that I was facing a strange problem. Upon removing the microphone from the radio, which was hard wired to the PC board, I found that the Johnson has an odd PTT system. Whereas most CB's will ground the PTT pin to activate the PTT, the Johnson had 12V running through both points where the PTT wires connected, and neither were connected to ground. Grounding one side with a screwdriver caused a short circuit, so I knew that was the hot side. However, I couldn't readily connect the other PC trace to the ground of the keyer board, since when I tried that the 12V on the hot side of the PTT went straight to ground and the radio didn't key. This wouldn't work if I was going to put the keyer board inside the radio and power it from the radio's 12V. A possible solution was to use a separate power supply such as a wall wart for the keyer, but that really didn't appeal to me. To make matters worse, it was impossible to find a schematic for that radio on the internet. It was at a builder's meeting at Joe, WØMQY's house that Bill, WBØLXZ came up with a simple solution using an optical isolator chip. I would run the output & ground to one side of the chip, while the two radio PTT connections would be connected to the other side. It appeared that my problem was solved.

Another problem I had found was that the Johnson's output power had dropped for some reason. When I first checked it out it was putting out a full 4 watts, but at some point it developed a fault and was only putting out a few milliwatts at best. If the final transistor was blown, that would be the end of the Johnson. Bill again came to the rescue. We determined that a transformer had gone bad and wasn't supplying full voltage to the final transistor. I simply wired a jumper in and bypassed the transformer, and the radio was back up to 4 watts.

Internal view of the Midland 13-830 showing the 9.585 crystal installed in the transmit slot.

In the meantime, while we were trying to solve that problem, we had also taken a look at the Midland 13-830. This radio had three crystals, one for the oscillator, one for transmit, and the other for receive. Unfortunately these crystals didn't have the frequencies marked on them. We had to use a frequency counter held close to each crystal to determine the frequency. We did find that the oscillator crystal was a 10.240 crystal. That number was familiar as I had pulled several out of old CB's. We found that the transmit crystal resonated at 10.692, and figured up that the receive crystal should be in the 11 MHz. range. Bill theorized that if we switched the transmit & receive crystals, it should make the radio operate in the 28 MHz. range. However, for one reason or another, it didn't work. The radio had absolutely no output when we tried that. We then thought about replacing the 10.240 crystal with an 11 MHz. crystal which would increase the frequency. Ideally, an 11.240 crystal would have worked great, but neither of us had one. Bill had several 11 MHz. crystals which did raise the frequency, however they only got us to around 27.500 instead of our goal of 28.200 to 28.300. The search for the right crystal would have to continue. What I didn't know at the time was that I already had one.

Some time later, Jay invited me to his QTH to work on getting the Johnson operational on 10 meters. Earlier I had tried a combination of a 37 MHz. and 9 MHz. crystals, but it didn't seem to work as there was no output. Jay reminded me that the 37 MHz. crystal may have actually been a frequency that was a fundamental of the 37 MHz. frequency, and would not cause the LO circuit to oscillate. He then suggested that I use an original 32 MHz. crystal and find a 4 MHz. crystal that would put the radio in the beacon band. He was searching for crystals on the Mouser Electronics website and found one that resonated at 4.4334, which when mixed with the 32.7 MHz. crystal would give a frequency of 28.2666, right in the middle of the 10 meter beacon band. Jay had wanted to put in a 4.000 MHz. crystal to test his theory, which would put the radio temporarily on 28.700 MHz. I soldered in the crystals and fired up the radio, but it didn't work! We messed around with it for a while, then I realized that I had soldered one of the crystals to the wrong post on the rotary switch! It was getting time for me to go, so I decided to work on it later. But I felt that we had had a real breakthrough.

View of the NorCal keyer installed inside the Midland case. The flat speaker can be seen attached to the bottom half of the radio's case.

An even bigger breakthrough was to come. At another meeting at Joe's, Bill suggested that we try replacing the 10 MHz. transmit crystal in the Midland with an 11 MHz. crystal. We did, which actually lowered the frequency. It became clear that instead of an 11 MHz. crystal, we needed to use a lower frequency crystal. Luckily I had an ample source of 9 MHz. crystals which, ironically enough, were the ones I had planned to use in the Johnson. I selected a crystal and soldered it in the transmit slot. It raised the frequency but was still a bit low. I finally selected a 9.585 crystal which did the trick. That crystal gave the radio 23 different 10 meter frequencies from 28.072 to 28.462, several of which ended up in the beacon band. I had an operational transmitter for 10 meters, and the Johnson was reverently placed back into the junk box.

The next step was to tune up the radio by turning the slugs in the coils. We borrowed a watt meter from Joe and connected a dummy load. I was expecting to get a full 4 watts out, but we could only get 2 watts out, despite all the tweaking we did. I later found on W5JO's website that when a CB is modified for 10 meters, 2 watts is about all that one will be able to get without completely replacing the coils. This made me feel better. At any rate, I would not have to install a small CPU cooling fan like I had at first planned. The 2 watts & 50% duty cycle only makes the transistor's heat sink warm but not hot.

The final disaster was yet to come. I had tried installing the keyer board, but I could not get the keyer to operate from 12V. It worked fine from a 9V battery, but whenever the higher voltage was applied, there was no response. The circuit I was using included a 78L05 voltage regulator, which apparently will handle 9V OK but not 12. I was back to the possible dilemma of having to use a separate power supply for the keyer circuitry. I had tried using several resistors in series with the +12V supply line in order to try to pull the voltage down a bit, but it still didn't work. At one point something must have gone wrong since the keyer chip didn't want to respond at all, not even with 9V. I just got a steady keydown from the output. The chip was fried.

Undaunted, I took my homebrew keyer which had the exact same circuitry and a chip that wasn't fried, and connected it with clip leads to the beacon. I programmed the keyer and let it operate the beacon until I came up with another solution. I had been emailing Doug, KI6DS, about the NorCal keyer kits, and he told me that they operate fine on 12V. They obviously have a different voltage regulator than the one I had. I ordered a couple of them from Doug, and after a nice long delay by the Post Office, Doug finally got my order & sent the kits to me.

N2UHC beacon, completed & on the air.


The keyer kit went together quickly and was easily installed. I drilled a hole in the side of the radio's chassis and used a spacer to connect the small keyer board. On the front face of the Midland, I had epoxied a piece of PC board over the hole where the microphone connected. I then drilled a hole in the middle of it and installed a pushbutton switch. I had removed the squelch pot from the radio and in the hole in the front panel installed a second pushbutton. I cut the traces to the volume pot and connected a small, flat speaker in series with it. I had previously removed the CB's speaker and put the small speaker in its place. I then connected it to the keyer so that I would be able to hear the keyer's programming and sidetone. One of the pushbuttons was connected to the S1 spot on the board, while the other pushbutton was connected to the dit side of the keyer board. This allows me to run the menu and start or stop the beacon. To start it, I push the program button and hold it until I hear "O?" in CW. I press it once more and get "BE," for the beacon mode. I then press the other button which starts the beacon. To stop the beacon, I simply press the dit button again. To finish it off, I installed a 3.5mm stereo jack in the back panel and connected it to the dit & dah sides of the keyer. This allows me to change the programming of the beacon by plugging in a set of iambic paddles.

The beacon's antenna consists of a vertical dipole made from 1/2" electrical conduit. I used a wooden dowel as a spacer and fed the antenna directly from the coax. I had seen this plan in QST several years earlier, and had a friend who built one that worked great. I cut the antenna down to get it resonant, then glued PVC caps on each end and coated the wooden dowel with epoxy to weatherproof it. I then mounted it on a 2X4 that was stuck in the ground and connected to our privacy fence. The bottom part of the antenna rests on a short length of 4X4.

Schematic of the battery backup circuit.

The beacon is powered by a modified computer power supply that puts out 2A at 12V. I added battery backup to the beacon by connecting a 12V gel cell battery in parallel with the power supply, along with a diode to prevent current backflow when the power supply is turned off. The battery is charged through a 100 Ohm 10W resistor, which prevents overcharging by supplying a 120 mA trickle charge. A second diode bypasses this resistor when the power supply goes off and the battery powers the beacon. The battery backup system provides flawless changeover when the power supply is turned off & back on again.

One thing I have been thinking about was moving the beacon's antenna. It is only about 15 feet or so away from my VHF/UHF antenna mast, which supports one end of my HF dipole. When I had the beacon connected to my CW keyer, I noticed a great deal of noise on the HF bands. However I did learn that most of it was due to the keyer being connected to a wall wart power supply, and once I installed the NorCal keyer in the beacon that problem went away. However it still tends to interfere with my scanner somewhat, so relocating the beacon antenna should help out. I had the idea of putting it in a corner of our fence on the back of the property, but I would have to have a 200' long run of coax. There would be a great deal of loss, and with only 2 watts would not be a good idea. The other alternative would be to put the beacon out there near the antenna, and run a long power cord out to it. This would be more feasible, but the long run of electrical cord would have a resistance of around 3 Ohms. This would draw down the supplied voltage a bit. It could work, however I do enjoy having the beacon on my desk where I can easily turn it on or off. I have to turn it off when using 10 meters since it does wipe out reception, a problem that won't go away by locating the beacon's antenna a few feet further away from my other HF antenna. It's a simple matter to start or stop it when it's located on the desk, but if it were located outside I'd have to walk all the way out there to stop & then restart it.

I had found several different 10 meter beacon lists on the internet, and studied them closely to find a frequency that wasn't being used heavily. I settled on 28.232 as there was only one other beacon on that frequency. It was also one of the frequencies available to me from converting the radio from 27 MHz. to 28 MHz. The beacon has been active since 6 August 2003 and, aside from locals, was first heard in Texas. Later reports were received from Montana, Georgia, New York, and California. Incidentally, the Montana ham who heard my 10 meter beacon also heard the KØETC beacon on 6 meters, according to his website. My first DX report came from New Brunswick, Canada, and an interesting exchange of QSL's was made with HP1AC in Panama, who sent me a QSL for reporting his beacon, then sent me a report for hearing my beacon. What makes it interesting is that HP1AC/B is on 28.231.4, just 600 Hz down from my beacon.

In the future, I may have plans to build a QRPp beacon from an old Archer Space Patrol toy walkie talkie. It is a crystal-controlled radio that operates on CB channel 14 and probably puts out around 200-300 mW. I do plan to convert another CB into a 10 or maybe 15 meter CW transceiver. I'm always ready for another challenge.

Tom Sevart N2UHC

October 26, 2003