uv254-experiments

note: experimenting with new button -- the GQ12H-10/S; mounting hole size 12mm; action momentary

TODO for main absorbance firmware (src/main.cpp)

• Increase LED warm-up and discard first post-warm-up sample. During LOD characterization (2026-06-29) the first ADC sample taken right after the existing 1.5 s warm-up was ~10 % low compared to subsequent samples; the LED clearly hadn't reached steady state. By the 2nd sample (1 s later) the reading was stable. Mitigation applied in lod_test/ and worth porting back into the main firmware:
  • Bump LED_WARMUP_MS from 1500 to 3000.
  • Take one analogRead after the warm-up timer fires and discard it before starting the averaging window.
  • This costs ~1.5 s extra per CALIBRATE/COMPUTE press, but removes a systematic first-sample bias that otherwise dominates the absorbance noise.