Guide

How WSPR balloon trackers encode telemetry, how protocols differ, and what to expect from a high-altitude float mission.

Types of Trackers

ZachTek — WSPR2

Simplest tracker · No channels · Altitude in power fields

ZachTek is named after the ZachTek WSPR-TX Pico, a popular commercial transmitter kit. It is by far the most widely used protocol in the DIY balloon tracking community — practically every home-built tracker you encounter uses this approach.

Technically, ZachTek does not define a new radio protocol. It reuses WSPR2 — a standard weak-signal propagation mode that has nothing to do with balloons and was never designed for them. What ZachTek actually defines is a single convention: how to encode altitude into the WSPR power fields. Everything else — the two-minute slots, the callsign, the locator — is plain unmodified WSPR2.

What ZachTek adds to WSPR2

power1Reported dBm of Frame 1 → coarse altitude component (× 300 m)
power2Reported dBm of Frame 2 → fine altitude component (× 20 m)

All other fields (callsign, locator, timing) are standard WSPR2 — no modifications.

Two-frame structure

Frame 1 — Basiceven 2-min slot
CallsignSP0LND
LocatorKO02
Power30 dBm
+
Frame 2 — Telemetryodd 2-min slot
CallsignSP0LND/6
LocatorKO02ab
Power20 dBm
Decoded altitude = 9 000 + 400 = 9 400 m

Altitude formula

altitude = power1 × 300 + power2 × 20  (meters)

WSPR power levels are discrete (0, 3, 7, 10 … 57, 60 dBm). The smallest possible step in the fine component is 3 dBm, giving a resolution of 3 × 20 = 60 m. When both fields report 60 dBm, the altitude is considered invalid.

U4B — Extended Telemetry

Rich telemetry · Channel-based · No suffix

U4B is a dedicated WSPR balloon telemetry protocol that packs a rich set of data into the standard WSPR callsign and locator fields. The most widely known tracker implementing it is Traquito.

Unlike ZachTek, U4B encodes far more than just altitude. All data is packed into the character values of the callsign and locator fields through multi-step arithmetic — the result looks like a normal WSPR transmission but the callsign in Frame 2 is not a real amateur call.

Position

6-char locator

Altitude

~20 m res.

Temperature

°C

Battery

Voltage (V)

GPS

Lock + sat count

Speed

km/h horiz.

Two-frame structure

Frame 1 — Basicstandard WSPR
CallsignSP0LND
LocatorKO02
Powernormal dBm
+
Frame 2 — Encodedpayload ID 1 & 3
Callsign0A2B3C
LocatorXX99
Powerencoded dBm
Frame 2 callsign and locator are not real WSPR identifiers — they are encoded telemetry data

Channel system

U4B uses a channel system to prevent collisions between multiple balloons operating on the same band. Each channel defines a specific pair of WSPR time slots and an encoded callsign prefix for Frame 2.

Important: only one balloon per channel can operate at a time. If two balloons share a channel, their encoded Frame 2 transmissions will collide — both will decode incorrectly or not at all.
Note: U4B does not support callsign suffixes (e.g. /6). No suffix can be carried in the encoded Frame 2.

Balloon Floating

How Balloon Floating Works

Physics, payload weight, precision & luck

Unlike weather sondes that ascend and burst, ZachTek and Traquito balloons are designed to float — finding an equilibrium altitude and drifting with upper-atmosphere winds for days, weeks, or even months.

Typical flight timeline

nightnightday 2day 3+highlowaltitudelaunch~3h~12h~36h3+ daysAscent~1 m/s alwaysCatch floatup, then dip downmay drop herenight — may dropless risk nowStable floatsurvival chance ↑↑

Ascent rate is typically ~1 m/s regardless of free lift. Survival probability rises sharply after the first 72 hours.
Diagram is illustrative only and does not reflect actual flight data.

The biggest failure point: catching float

The most critical moment of any flight is catching float — the phase during ascent when the balloon slows down and settles into its equilibrium altitude. If the envelope has any weakness, a micro-leak, or was slightly overfilled, this is usually when it reveals itself. The transition from ascent to float puts significant stress on the latex, and if it's going to fail, it most often happens right here.

A slow leak at float means the balloon will gradually lose lift and start descending — sometimes so slowly it's barely noticeable on the tracking map. During the day the sun heats the gas, expanding it and generating extra lift that masks the leak — the balloon can still appear to hold altitude. Then night comes: no sun, the gas cools and contracts, the lift it provides shrinks below the payload weight, and the balloon descends slowly. Sometimes, if it lands in an open area, the tracker will continue transmitting WSPR from the ground — a fortunate ending that at least confirms where it came down.

Why do balloons come down?

20km18km16km14km12km10km8kmCirrus cloudsice crystals, ~16km#1 cause of descentonly 2 km gapBalloonfloat ~14km4kmgapCommercial aircraft10–12 km, below floatlow risk

The overwhelming majority of descents — ~99.9% of cases — come down to one thing: clouds. Not equipment failure, not bad luck, just clouds. High-altitude cirrus and similar formations routinely reach 14, 15, even 16 km, and a balloon drifting at float altitude will eventually fly straight through one. Ice crystals and moisture coat the envelope, adding weight and disrupting the delicate lift balance until the balloon can no longer stay up.

A common question is whether aircraft pose a risk. They don't — commercial planes cruise at 10–12 km, well below typical float altitude, and military or business jets rarely climb higher. At those altitudes a balloon is statistically far more likely to encounter a cloud than any aircraft. And even in the unlikely scenario where paths cross, the jet would never actually touch the balloon — the wake turbulence alone would scatter it long before any physical contact.

Only after clouds comes everything else: equipment failures (solar panel delamination, cold-induced battery cutoff, solder joint fracture from thermal cycling), envelope degradation from UV exposure slowly embrittling the latex, and gradual micro-leaks. These are real failure modes, but they are secondary — if a balloon avoids clouds long enough, it can fly for weeks.

Surviving the first days

The most dangerous period is the first 24–72 hours. The envelope faces intense UV radiation, repeated temperature cycling (scorching heat in sunlight, extreme cold in shadow), and the pressure stress of the initial float equilibrium settling. Most failures happen here. Once a balloon survives this phase, it becomes significantly more stable and can continue flying indefinitely.

Payload weight matters

Total payload weight (tracker + battery + antenna + balloon envelope) is one of the most critical variables. A payload of around ~10 g typically floats at approximately ~14 km altitude. Lighter means higher and more stable. A single gram can shift the float ceiling by several kilometers.

A precision hobby

Float ballooning is genuinely precise work. There is no room for mistakes in the areas you can control — a wrong helium fill ratio, a gram too much weight, a poor antenna solder joint, or even the wrong knot can end a flight before it properly begins. Every detail matters and shortcuts tend to have consequences at altitude.

Luck plays a big role

And yet — balloon flying has a humbling habit of rewarding beginners and punishing veterans in equal measure. The clumsiest launcher can stumble into a record-breaking float while the most meticulous builder watches their craft descend in the first hour. Once the technical basics are covered, luck becomes the dominant factor. Skill sets the floor, fortune determines the ceiling.

You can follow your balloon's journey in real time on Amateur Sondehub — this service automatically forwards your WSPR spots there.

Other Hardware

What about radiosondes like M20 / RS41?

Weather balloon sondes repurposed for amateur use

Radiosondes such as the Meteomodem M20 and Vaisala RS41 are mass-produced weather balloon instruments that are regularly recovered by hobbyists after official flights. Their onboard GPS, microcontroller, and radio hardware make them an attractive starting point — and people do launch them. Theoretically, under the right conditions, a radiosonde could even achieve a float.

In practice, however, radiosondes are simply too heavy for a meaningful float mission. They were designed to ascend, transmit once, and be discarded — not to float for weeks at altitude. The full assembled unit is far outside the weight budget that makes low-altitude float viable (~10 g total payload).

Better approach

If you have recovered radiosondes, the most practical use is to harvest their PCB components — the GPS module, oscillator, or MCU — and integrate them into a purpose-built lightweight tracker. Launching a radiosonde as-is for a float attempt is rarely worth it when dedicated trackers like ZachTek Pico or Traquito exist specifically for this purpose and weigh a fraction of the mass.

This page is a draft

The content here is work in progress and may be incomplete or inaccurate. If you have suggestions, corrections, or anything you'd like to see added, feel free to reach out.