Home >Company News>Company News
Customer Case Study — Turning a $20 LoRa Board into a 9 Year Battery-Powered Sensor Node

How a community maker turned our $20 "EoRa Pi" board into a professional-grade, ultra-low-power HVAC monitoring system — and what it proves about the future of wireless IoT design

.
hvac

At a Glance

  • Customer: William Lucid (IoT Maker / Systems Developer)

  • Product: EoRa-S3-900TB ("EoRa Pi") — ESP32-S3 + SX1262 LoRa Development Board

  • Application: Ultra-low-power HVAC monitoring system (outdoor BME280 sensor node)

  • Challenge: Build a long-life outdoor sensor node that wakes only when needed and sips current in deep sleep.

  • Solution: Three event-driven nodes built on the EoRa Pi, using LoRa Wake-on-Radio (WOR) triggered by Channel Activity Detection (CAD).

  • Results: ~38.19 µA average deep-sleep current, ~107.7 months of theoretical battery life — and a 25.6% power reduction through CAD optimization.

The Challenge

Designing a wireless sensor node for long-term outdoor deployment usually means fighting a constant battle with the power budget. Periodic wakeups drain batteries fast, and getting a LoRa radio to wake a sleeping MCU reliably is notoriously difficult. The maker needed a platform that could stay in deep sleep almost all the time and spring to life only when a specific event occurred.

The Board: EoRa-S3-900TB ("EoRa Pi")

The EoRa Pi is our integrated development board designed to lower the barrier to professional LoRa design. A fully integrated ESP32-S3 + SX1262 LoRa board, it features:

  • ESP32-S3 MCU + SX1262 LoRa radio

  • OLED display and microSD slot

  • Li-ion charger with dual LDOs

  • JST battery connector and USB-C

  • Full GPIO breakout

  • Under $20 USD, shipping worldwide from EbyteIoT.com

What makes it special for battery-powered projects is its remarkably low deep-sleep current — ~25 µA when unplugged from USB, with peripherals off and the DC-DC regulator mode active. That gives engineers a ready-made foundation for long-life, field-deployed IoT nodes

The Solution: A Three-Node Ultra-Low-Power HVAC Monitor

The maker chose the EoRa Pi for its out-of-the-box ~25 µA deep-sleep current and complete GPIO breakout, and built Heating System Monitor IV — an event-driven, ultra-low-power monitoring system with three cooperating nodes:

NodeHardwareRole
Blower NodeESP32-S3 Super MiniDetects blower ON/OFF via MPU6050 vibration sensor, sends ESP-NOW packets
Receiver HubEoRa-S3-900TBReceives ESP-NOW, triggers LoRa Wake-on-Radio (WOR), merges data, reads indoor BME280, logs to Google Sheets
Outside NodeEoRa-S3-900TB~38.19 µA deep sleep, wakes on LoRa WOR, reads BME280, sends ESP-NOW, returns to sleep

The architecture eliminates periodic wakeups entirely — the system wakes only when the blower changes state, spending 99.9% of its time in deep sleep.

Engineering Highlights

1. Wake-on-Radio via DIO1 Rerouting
The factory wiring connects the SX1262's DIO1 to GPIO33 — but GPIO33 is not RTC-wake-capable, so it can't trigger the ESP32-S3 from deep sleep. The maker rerouted DIO1 to GPIO16 (a true RTC-wake-capable pin), enabling the full WOR wake path: blower OFF → Hub sends WOR packet → SX1262 detects preamble → DIO1 asserts → EXT0 wakes the node.

2. Channel Activity Detection (CAD) Optimization
The project migrated from autoDutyCycle to CAD interrupt as the pre-check before the WOR receive window, reducing average current from 51.30 µA to 38.19 µA — a 25.6% reduction and a 34.3% improvement in theoretical battery life on the same 3000 mAh LiPo.

3. Professional Power Profiling with Nordic PPK2
The maker validated every operating mode using the Nordic Power Profiler Kit 2, capturing true deep-sleep current, WOR wake spikes, ESP-NOW bursts, and return-to-sleep curves — real, reproducible data for the open-source community

The Results: Battery Life That Speaks for Itself

Validated end-to-end with a Nordic PPK2 power profiler, the system delivered:

MetricValue
Deep sleep average current~38.19 µA (99.9% of the time in sleep)
Theoretical battery life~107.7 months (~8.9 years) on a 3,000 mAh LiPo
Realistic (derated) battery life~62–90 months
Power reduction via CAD25.6% lower consumption

A notable finding: LiPo self-discharge — not deep-sleep current — turned out to be the dominant factor in the runtime budget. It's a valuable lesson for anyone designing battery-powered IoT, and a reminder that great hardware and great power design go hand in hand

Why This Project Is a Win for EBYTE

This build proves something we've believed all along: you don't need expensive hardware to achieve professional-grade IoT design. By pairing the low-power, fully integrated EoRa Pi with careful engineering, a community maker created a long-life, event-driven LoRa sensor node — exactly the kind of reliable, cost-effective solution we help customers build every day.

"This project shows how careful engineering can turn a less-than-$20 board into a highly efficient, long-life sensor node." — William Lucid, Project Author

At EBYTE, we support this ecosystem with:

  • Complete product portfolio — from core boards and single-board computers to LoRa, WiFi, 4G modules and industrial gateways

  • Full development resources — documentation, SDKs, schematics and technical support (7×12h online, remote assistance, on-site service)

  • Global reach — over 100 countries, 5,000+ enterprise partners, and service centers across Germany, the US and Brazil


Promotion de printemps

🌟 Build Something Great — Start with EBYTE

Inspired by this project? The EoRa-S3-900TB is ready for your next ultra-low-power LoRa design.

How to Order:👉Submit Your Request 

  • Want your project featured? Share what you build with EBYTE hardware — we'd love to tell your story.

  • Door-to-Door / DDP service to the US, Canada, and Europe — shop on RobotShop.

  • How to Order: Submit your request on our site.

  • Free Samples for University StudentsUniversity-Sponsorship Applications  

Ready to make wireless communication simpler? The EoRa Pi is a great place to start.



Source / Original Project: "38.19 µA Avg Outside BME280 Node — HVAC System Monitor IV" by William Lucid, published on Hackster.io (July 31, 2026).
Original URL: https://www.hackster.io/AB9NQ/38-19-a-avg-outside-bme280-node-hvac-system-monitor-e4e0a9


This article is based on the maker's open-source project, Project details and credits belong to the original author.