Kiln controller

Kiln Controller

A MicroPython-based lumber drying kiln controller for a passive solar kiln in the Georgian Bay area of Ontario. Designed for drying hardwood lumber (maple and beech, 0.5--1 inch thick) using multi-stage schedules derived from USDA Forest Products Laboratory references (FPL-GTR-57 and FPL-GTR-118).

The kiln is approximately 5 ft x 2 ft x 3 ft, heated passively by solar gain with a 500W ceramic PTC backup heater for shoulder-season use. A Raspberry Pi Pico 2W controls all kiln hardware. Telemetry is transmitted over LoRa (~200m) to a Raspberry Pi 4 at the cottage, which stores data in SQLite and serves a REST API for a Kivy mobile app.

Hardware

Schematic and custom PCB

KiCad 9.0 project with schematic and custom PCB layout is in the PCB folder.

Firmware (Pico 2W -- MicroPython)

All hardware drivers live in lib/. Every module is class-based, accepts pin numbers as constructor arguments, and follows the pattern established in lib/exhaust.py.

Drying schedules are stored as JSON in schedules/ and loaded from the SD card at runtime. Schedules are included for hard maple and beech at 0.5 in and 1 in thickness.

Deploying to the Pico

python update_lib.py

This copies main.py, config.py, and all lib/ files to the Pico in one step. The MicroPython SPI SD card driver must be deployed as sdcard_driver.py:

mpremote cp sdcard_driver.py :sdcard_driver.py

Running tests

mpremote run test_modules.py

Tests all 12 modules in sequence and reports pass/fail. Does not start WiFi or the control loop.

Cottage-side architecture (Pi4)

Pico --> SPI1 --> Ra-02 ~~LoRa~~ Ra-02 --> SPI0 --> Pi4 kiln_server
                                                 --> SQLite
                                                 --> REST API :8080 --> Kivy app
                                                 --> ntfy.sh --> phone

The kiln_server/ Python package (in progress) runs as a systemd service on the Pi4. It receives LoRa packets from the Pico, writes telemetry and alerts to SQLite, serves a REST API for the mobile app, and pushes fault alerts via ntfy.sh.

The Pico also runs a WiFi AP and exposes its own REST API (port 80) for direct local control when within range.

Drying schedules

Schedules follow USDA FPL kiln-drying recommendations:

• Hard maple (T3-C2): schedules/maple_05in.json, schedules/maple_1in.json
• Beech (T4-C3): schedules/beech_05in.json, schedules/beech_1in.json

Each schedule has 9 stages: 7 drying stages + equalizing + conditioning. Stage advance is automatic (MC% + minimum time) for drying stages, and time-only for equalizing/conditioning. Manual advance is available via the REST API.

Status

Pico firmware is at the integration stage. All 12 lib/ modules are complete and hardware-tested. main.py is implemented and pending full hardware integration test. The Pi4 kiln_server daemon and Kivy mobile app are next.