Show HN: We chased a weather balloon across Montana and never found it
Show HN: The Montana Weather Balloon Odyssey
By Andrew M. Published: August 19th, 2026 | Reading Time: ~16 minutes
In June 2025, I partnered with New England Sci-Tech via Apex to launch StratoSpore, my inaugural venture into high-altitude ballooning. That project served as a testing ground for using algae as a biosensor to monitor altitude, providing a wealth of educational data.
Driven by a desire to return to the stratosphere, I teamed up with Sam Flynn to develop a more robust, flight-ready payload.
🎯 Project Objectives
Based on the lessons learned from the previous year, I established a specific set of goals for the new payload:
- Establish redundancy for all tracking systems.
- Transmit images to the ground with actual discernible detail.
- Upgrade to a more dependable GPS module.
- Enhance radio functionality (potentially leveraging my amateur radio license).
- Optimize telemetry via more efficient data packing.
Last year's images were a mere 18x10 pixels. My aim for this documentation is to detail the implementation of these upgrades and share the learning curve to assist future explorers.
🧪 The Experiments
While StratoSpore focused on UV exposure and algae fluorescence, our current payload, UpLink, pursued two primary objectives:
- Thermal Analysis: Comparing the insulation efficacy of foaming PLA versus standard non-foaming PLA.
- Visual Telemetry: Sending high-resolution imagery across a radio link.
📦 Testing Payload Insulation
The industry standard for high-altitude enclosures is typically styrofoam. While effective, it has distinct drawbacks:
| Feature | Styrofoam Enclosures | Custom 3D Printed Enclosures |
|---|---|---|
| Insulation | Excellent | Variable (depends on material/infill) |
| Sizing | Predefined/Rigid | Fully Customizable |
| Layout | Inefficient | Optimized for specific hardware |
| Cost | Relatively Expensive | Low (cost of filament) |
Our total launch mass—including the electronics, enclosure, flight line, and parachute—was . In ballooning, minimizing weight is critical because:
Sam utilized Fusion360 to design a bespoke enclosure. This design featured integrated standoffs, standardized mounting points, and dedicated slots for temperature sensors and cameras. To test insulation, we attached small capsules to the lid, sealing the internal sensors with hot glue to prevent ambient air from skewing the results.
The Challenge of Foaming PLA
We experimented with a specialized filament that contains microscopic air bubbles which expand during printing. While it is lighter than standard PLA, it is notoriously difficult to handle.
"Printing with foaming PLA is a nightmare unless you have a very specific profile."
To make it viable, I implemented the following Orca Slicer settings:
print_profile:
speed: "Reduced by ~7x"
cooling: "Disabled or heavily limited"
bed_temp: "Increased for better adhesion"
acceleration: "Disabled (prevents wall strength variance)"
maintenance: "Must be thoroughly dried"
(You can download my full profile for Orca Slicer if you wish to experiment.)
📊 The Results
Our hypothesis suggested that the microscopic bubbles in foaming PLA would significantly outperform standard PLA in thermal insulation. However, the data was inconclusive.
Observations:
- There was no consistent winner; foaming PLA performed better in some instances and worse in others.
- Conclusion: Filament composition is less important than infill density and wall count.
Future Iterations:
I suspect that multiline infill might be the answer, as insulation relies on air pockets to impede heat transfer. Thicker walls between these pockets could be the key.
Despite the inconclusive material test, the custom enclosure itself worked. The insulated sensors remained warmer than the outside air. We also discovered via simulation that painting the enclosure black (rather than white) increased internal heat due to solar radiation.
Regardless of the insulation results, foaming PLA remains the gold standard for the weight-to-strength ratio, and I will continue using it for future flights.
🖼️ Image Transmission
Capturing the view from is a rite of passage in ballooning. However, my previous attempt was a failure; I sent images that were pixels.
The Problem:
Radio links are slow and limited by a protocol maximum of 255 bytes per transmission. I attempted to solve this with an excessively complex compression algorithm, which ironically rendered the images useless.
The Evolution of the Solution:
- SSTV (Slow Scan Television): An analog method. Verdict: Too power-hungry () and poor quality.
- SSDV (Slow Scan Digital Video): A packetized digital alternative. Verdict: Ideal. If the ground station misses a few packets, the image can still be reconstructed.
I modified a project by Philip Hero to implement this.
Visual Comparison: Below is an example of the low-resolution output from the previous attempt:
