There’s an interesting development in amateur ballooning: using so-called superpressure balloons, which float high in the atmosphere indefinitely rather than simply going up and up and then popping like a normal weather balloon. Superpressure balloons can last for months and travel long distances, potentially circumnavigating the globe, all the while reporting their position.
You might imagine that an undertaking like this would be immensely difficult and cost thousands of dollars. In fact, you can build and launch such a balloon for about the cost of a fancy dinner out. You just have to think small! That’s why amateur balloonists call them pico balloons.
The payload of a pico balloon is so light (between 12 to 30 grams) that you can use a large Mylar party balloon filled with helium to lift it. They’re also inexpensive; that’s important because you won’t get your payload back. And because such diminutive payloads don’t pose a danger to aircraft, they aren’t subject to the many rules and restrictions on free-floating balloons that carry more mass.
The essential advances that made pico ballooning possible were figuring out how to track a balloon no matter where in the world it might be and how to power such tiny payloads. A lot of folks worked on these challenges and came up with good solutions that aren’t hard or expensive to reproduce.
What is WSPR?
Amazingly, the global tracking of the balloon’s telemetry is done without satellites. Instead, pico balloonists take advantage of an amateur-radio network called WSPR (Weak Signal Propagation Reporter), a protocol developed by a rather famous ham-radio enthusiast—Joseph Hooton Taylor Jr., one of the two scientists awarded the 1993 Nobel Prize in Physics for discovering binary pulsars.
A Raspberry Pi Pico microcontroller [top left] is soldered directly to a daughterboard consisting of a high-frequency transmitter and a GPS module [bottom left], which are all powered by solar panels [right].James Provost
WSPR was designed to monitor signal-propagation conditions for different radio bands—useful information if you’re a ham trying to make distant contacts. WSPR can also record low-power balloon-telemetry signals. WSPR is very low bandwidth—less than 10 bits per minute—but it does the job. A worldwide network of radio amateurs receives these WSPR signals and reports them publicly over the internet, which gives picoballoonists a way to track their flights. You need at least a general-class ham-radio license to launch a pico balloon, as one is required to transmit on the bands used for long-distance telemetry.
The pico balloon payload I chose to build is based on the aptly named US $4 Raspberry Pi Pico board, with a solder-on daughterboard that contains a GPS receiver and transmitter. The folks who developed this daughterboard and associated software (to create what they call the Jetpack WSPR Tracker) have done a fantastic job of making their work easy to reproduce.
You could, in principle, power the Jetpack tracker with batteries, but in practice it would be impossible to keep them warm in the stratosphere, where average temperatures can be as low as –51 °C. Instead, the tracker runs off two lightweight solar modules. At night, it gracefully powers down. When the sun rises high enough in the morning, the tracker powers up and starts transmitting again.
My first pico balloon made it only halfway across the Atlantic before going silent.
I had five Jetpack boards custom-manufactured in China for just $39. The cost nearly doubled after adding shipping and tariff charges. Still that’s really cheap, even when you add the cost of the Raspberry Pi ($4), the party balloon ($10 for two), the helium ($10 at my local supermarket), and the two solar modules ($7 each).
The biggest sticking point I had with the Jetpack design was the liberties it takes with spurious emissions from its transmitter. Federal Communications Commission (FCC) regulations call for spurious emissions to be at least 43 decibels below the power of the transmitted signal. But my transmitter had strong unwanted emissions at odd harmonics of the fundamental frequency. (That’s because the transmitter is a Si5351A temperature-controlled oscillator, which outputs a square wave, not a sinusoid.) Taking measurements, I could see that the third harmonic at 42 megahertz was only 25 dB quieter than the 14-MHz fundamental of my WSPR signal’s frequency.
As of press time, the WSPR network had tracked my balloon from the Eastern United States to the Mediterranean coast. James Provost
In practical terms, this shouldn’t create any noticeable interference, given that this transmitter puts out milliwatts at most and floats miles away from the nearest receiver. Still, I wanted to be fully compliant with FCC regulations, so I added traps to the antenna—simple circuit elements that hams use to allow a single antenna to work on multiple bands by altering how the antenna resonates at different frequencies. Each trap was made of a small inductor (four 5-millimeter-diameter loops of No. 32 magnet wire) in parallel with a 220-picofarad capacitor. I tuned them with the help of a NanoVNA signal analyzer by stretching the loops apart slightly. I attached the traps directly to the tracker board, so that they quashed the spurious 42-MHz emissions at the source. That worked well and added only 0.3 grams of weight.
With my payload complete, I partially filled my balloon with helium. You want the balloon to hold just a little more gas than it takes to lift the payload off the ground. This will give the helium room to expand as the balloon climbs to its final altitude.
My first pico balloon, launched from a park near my home in North Carolina, made it only halfway across the Atlantic before going silent. My second went up and was never heard from again. The third was indeed the charm. It crossed the Iberian Peninsula and at the time of this writing is somewhere over the Mediterranean at an altitude of nearly 12 kilometers. With any luck, it might go on to orbit the planet.
I’m a little puzzled about the balloons’ telemetry messages received on the WSPR network, as they have been few and far between. My best guess is that power from the horizontal solar panels I’m using is marginal, with the winter sun being so low in the sky. That’s something I should have thought about before launching the first balloon just 24 hours after the winter solstice!
This article appears in the February 2026 print issue as “Long-Duration Amateur Ballooning.”
Instagram Engineering
Shopify Engineering