https://mapfumo.github.io/tags/range-testing/Recent content in Range-Testing on Antony MapfumoHugo -- gohugo.ioen-usTue, 28 Apr 2026 12:23:09 +1000https://mapfumo.github.io/posts/2026/april/i-walked-921-metres-with-lorawan-node-here-is-what-the-data-showed/Tue, 28 Apr 2026 12:23:09 +1000https://mapfumo.github.io/posts/2026/april/i-walked-921-metres-with-lorawan-node-here-is-what-the-data-showed/<h2 id="the-setup">The Setup</h2> <p>I built a two-node LoRaWAN sensor network on STM32WL55JC1 microcontrollers running Rust/Embassy firmware — the full story is in <a href="https://mapfumo.github.io/posts/2026/april/building-a-lorawan-sensor-network-in-rust-stm32wl55-embassy-and-the-bugs-that-taught-the-most/">Part 1</a>. One of the two nodes, lora-2, carries no sensor. Its only job is to send a confirmed uplink every ~10 seconds, display RSSI, SNR, and DR on its OLED, and let the gateway&rsquo;s InfluxDB record tell the story later.</p> <p>You might wonder why I didn&rsquo;t just use lora-1 — the sensor node — for this test. Both nodes send a 4-byte payload, so packet size is identical and makes no difference to range. What matters is transmission behaviour: lora-2 sends every uplink confirmed, so a missed ACK is detected within one cycle. lora-1 only confirms every fifth uplink, meaning gateway loss takes much longer to register. lora-2 also transmits every ~10 seconds versus lora-1&rsquo;s ~30 seconds — three times the data points in InfluxDB, three times the resolution when overlaying against GPS waypoints later. Purpose-built beats repurposed.</p>