[arcadien]

New sub-project tinyDecoders/: an ATtiny84A receiver that decodes the frames tinySensors send and shows them on a 128×32 SSD1306 OLED. The tinySensor firmware is unchanged, apart from two radio IDs (see below).

Contents

• Decoders (pure logic, host-tested by loopback through the repo encoders via OrdersToPulses, TECH-DECODE-001):
  • X10 RFXmeter (FUNC-DECODE-001)
  • Lacrosse WS7000 frame and fields (FUNC-DECODE-002/003)
  • Oregon v3 THN132N / THGN123N / BTHR918 frame and fields (FUNC-DECODE-004/005)
• Display
  • SourceTable: last record of up to 6 sources (FUNC-DECODE-006)
  • FormatPage: two 16-character lines with a protocol+channel label and the age of the last frame (UI-DECODE-001)
  • Spleen 8x16 font, BSD-2-Clause (UI-DECODE-002)
  • Receiver loop: one page every 4 s, redrawing one changed character per poll (FUNC-DECODE-007)
• Hardware
  • ReceiverHal build-time alias with a test double (TECH-DECODE-002)
  • Timer1 capture on PA7 (PLAT-DECODE-001)
  • SSD1306 over TinyI2C, no framebuffer (PLAT-DECODE-002)
• Build
  • attiny84a env at 8 MHz; a post-build check fails the build past 384 B static RAM (CONF-DECODE-001)
  • CI and pre-commit cover tinyDecoders/ (CONF-DECODE-002)
• Config: S_04 moves to Oregon channel 3 (CONF-BUILD-003) and robot to Lacrosse address FOUR (CONF-BUILD-004), so labels don't collide. Both sensors must be reflashed, and existing receivers will see them under new IDs.

Status

• Native tests pass: tinyDecoders 42/42, firmware 82/82.
• attiny84a build: flash 6314 / 8192 B, static RAM 323 / 384 B.
• Not yet tested on hardware: radio capture, display, stack depth, and receiver noise.

🤖 Generated with Claude Code