PI4ZLB

Amateur radio license exams typically have a question about the bandwidths taken up by various modulation types. The concept behind the question is pretty obvious — as guardians of the spectrum, operators really should know how much space each emission type occupies. As a result, the budding ham is left knowing that continuous wave (CW) signals take up a mere 150 Hertz of precious bandwidth.

But is that really the case? And what does the bandwidth of a CW signal even mean, anyway? To understand that, we turn to [Alan (W2AEW)] and his in-depth look at CW bandwidth. But first, one needs to see that CW signals are a bit special. To send Morse code, the transmitter is not generating a tone for the dits and dahs and modulating a carrier wave, rather, the “naked” carrier is just being turned on and off by the operator using the transmitter’s keyer. The audio tone you hear results from mixing the carrier wave with the output of a separate oscillator in the receiver to create a beat frequency in the audio range.

That seems to suggest that CW signals occupy zero bandwidth since no information is modulated onto the carrier. But as [Alan] explains, the action of keying the transmitter imposes a low-frequency square wave on the carrier, so the occupied bandwidth of the signal depends on how fast the operator is sending, as well as the RF rise and fall time. His demonstration starts with a signal generator modulating a 14 MHz RF signal with a simple square wave at a 50% duty cycle. By controlling the keying frequency, he mimics different code speeds from 15 to 40 words per minute, and his fancy scope measures the occupied bandwidth at each speed. He’s also able to change the rise and fall time of the square wave, which turns out to have a huge effect on bandwidth; the faster the rise-fall, the larger the bandwidth.

It’s a surprising result given the stock “150 Hertz” answer on the license exam; in fact, none of the scenarios [Allen] tested came close to that canonical figure. It’s another great example of the subtle but important details of radio that [Alan] specializes in explaining.

Ham radio is having a bit of a resurgence these days, likely due to awards programs like Parks on the Air (POTA) and Summits on the Air (SOTA), which encourage amateur radio operators to head outside and “activate” at various parks and mountaintops. For semi-mobile operations like this, a low-power radio is often used, as well as other portable gear including antennas. In the VHF/UHF world, the J-pole is a commonly used antenna as well, and this roll-up tunable J-pole antenna is among the most versatile we’ve seen.

The antenna uses mostly common household parts which keeps the cost down tremendously. The structure of the antenna is replacement webbing for old lawn chairs, and the conductive elements for the antenna are made out of metallic HVAC tape which is fixed onto the chair webbing after being cut to shape. The only specialized parts needed for this is a 3D printed bracket which not only holds the hookup for the coax cable feeding the antenna, but is also capable of sliding up and down the lower section of the “J” to allow the antenna to be easily tuned.

As long as you have access to a 3D printer, this antenna is exceptionally portable and pretty easy to make as well. Although VHF and UHF aren’t too popular for POTA and SOTA, portable equipment like this for the higher frequency bands is still handy to have around when traveling or operating remotely. With the antenna situation sorted out, a DIY radio that can make use of it might be in order as well.

Generally, the biggest problem a new ham radio operator will come across when starting out on the high frequency (HF) bands is finding physical space for the antennas. For a quick example, a dipole antenna for the 20 m band will need around 10 m of wire, and the lower frequencies like 80 m need about four times as much linear space. But if you’re willing to trade a large space requirement for a high voltage hazard instead, a magnetic loop antenna might be just the ticket.

Loop antennas like these are typically used only for receiving, but in a pinch they can be used to transmit as well. To tune the antennas, which are much shorter than a standard vertical or dipole, a capacitor is soldered onto the ends, which electrically lengthens the antenna. [OM0ET] is using two loops of coax cable for the antenna, with each end soldered to one half of a dual variable capacitor which allows this antenna to tune from the 30 m bands to the 10 m bands, although he is using it mostly for WSPR on 20 m. His project also includes the use of an openWSPR module, meaning that he doesn’t have to dedicate an entire computer to run this mode.

The main downsides of antennas like these is that they are not omnidirectional, are not particularly good at transmitting, and develop a significantly high voltage across the capacitor as this similar mag loop antenna project demonstrated. But for those with extreme limitations on space or who, like [OM0ET] want a simple, small setup for running low-power applications like WSPR they can really excel. In fact, WSPR is a great mode for getting on the air at an absolute minimum of cost.

When Japan’s SLIM lunar lander made a rather unconventional touch-down on the lunar surface, it had already disgorged two small lunar excursion vehicles from its innards: LEV-1 and LEV-2. Of these, the LEV-1 is not only capable of direct to Earth transmission, but it also has been assigned its own amateur radio license: JS1YMG, which makes it the first Ham radio station on the Moon. LEV-1 receives data from LEV-2, which is transmitted to Earth using its 1 Watt UHF circular polarization antenna as Morse code at 437.410 MHz. Although the data format hasn’t been published, [Daniel Estévez] (EA4GPZ) has been sleuthing around to figure it out.

Using captures from the 25 meter radiotelescope at Dwingeloo in the Netherlands, [Daniel] set to work deciphering what he knew to be telemetry data following a CCSDS standard. After some mix-and-matching he found that the encoding matched PCM/PSK/PM with a symbol rate of 64 baud and 2048 kHz subcarrier. The residual carrier is modulated in amplitude with Morse code, but initially this Morse code made no sense.

Fortunately a few fellow Hams pitched in and figured out that the amplitude signs for the Morse code were inverted. By inverting the amplitude, suddenly the Morse code looked a lot more clear, with the LEV-1’s call sign and what looked like hexadecimal data following it. Each of the frames is also followed by a CRC-16, which should make it possible to start decoding the data transmitted in each frame.

As far as timekeeping goes, there’s nothing more accurate and precise than an atomic clock. Unfortunately, we can’t all have blocks of cesium in our basements, so various agencies around the world have maintained radio stations which, combined with an on-site atomic clock, send out timekeeping signals over the air. In the United States, this is the WWVB station located in Colorado which is generally receivable anywhere in the US but can be hard to hear on the East Coast. That’s why [JonMackey], who lives in northern New Hampshire, built this WWVB simulator.

Normally, clocks built to synchronize with the WWVB station include a small radio antenna to receive the 60 kHz signal and the 1-bit-per-second data transmission which is then decoded and used to update the time shown on the clock. Most of these clocks have internal (but much less precise) timekeeping circuitry to keep themselves going if they lose this signal, but [JonMackey] can go several days without his clocks hearing it. To make up for that he built a small transmitter that generates the proper timekeeping code for his clocks. The system is based on an STM32 which receives its time from GPS and broadcasts it on the correct frequency so that these clocks can get updates.

The small radio transmitter is built using one of the pins on the STM32 using PWM to get its frequency exactly at 60 kHz, which then can have the data modulated onto it. The radiating area is much less than a meter, so this isn’t likely to upset any neighbors, NIST, or the FCC, and the clocks need to be right beside it to update. Part of the reason why range is so limited is that very low frequency (VLF) radios typically require enormous antennas to be useful, so if you want to listen to more than timekeeping standards you’ll need a little bit of gear.

There are plenty of hobbies around with huge price tags, and ham radio can certainly be one of them. Experienced hams might have radios that cost thousands of dollars, with huge, steerable antennas on masts that can be similarly priced. But there’s also a side to the hobby that throws all of this out of the window in favor of the simplest, lowest-cost radios and antennas that still can get the job done. Software-defined radio (SDR) turned this practice up to 11 as well, and this radio module uses almost nothing more than a microcontroller to get on the air.

The design uses the capabilities of the Raspberry Pi Pico to handle almost all of the radio’s capabilities. The RF oscillator is driven by one of the Pico’s programmable I/O (PIO) pins, which takes some load off of the processor. For AM and SSB, where amplitude needs to be controlled as well, a PWM signal is generated on another PIO which is then mixed with the RF oscillator using an analog multiplexer. The design also includes a microphone with a preamplifier which can be fed into a third PIO; alternatively it can receive audio from a computer via the USB interface. More processor resources are needed when generating phase-modulated signals like RF, but the Pico is still quite capable of doing all of these tasks without jitter larger than a clock cycle.

Of course this only outputs a signal with a few milliwatts of power, so for making any useful radio contacts with this circuit an amplifier is almost certainly needed. With the heavy lifting done by the Pico, though, the amplifier doesn’t need to be complicated or expensive. While the design is simple and low-cost, it’s not the simplest radio possible. This transmitter sends out radio waves using only a single transistor but you will be limited to Morse code only.