Open question #2 in docs/architecture.md §11. Shares the bench rig with #1.
Problem. GPIO interrupts under the Arduino framework can jitter by tens of µs or worse under load. The design assumes hardware capture (RMT/MCPWM on ESP32-S3) reduces capture jitter to noise — decision D13, status bench validation pending.
Test plan
- Feed the rig's optocoupler output to an ESP32-S3 DevKit (BOM 2.1).
- Compare capture methods: main-loop polling (baseline), GPIO interrupt, RMT capture, MCPWM capture.
- Each with WiFi/BT disabled (design state) and enabled (worst case), against the logic analyzer's ground truth.
Pass criteria. Capture jitter ≪ the 1 ms system target — expected single-digit µs with hardware capture. Document actual numbers per method.
Gates D13 and the MCU choice.
Open question #2 in
docs/architecture.md§11. Shares the bench rig with #1.Problem. GPIO interrupts under the Arduino framework can jitter by tens of µs or worse under load. The design assumes hardware capture (RMT/MCPWM on ESP32-S3) reduces capture jitter to noise — decision D13, status bench validation pending.
Test plan
Pass criteria. Capture jitter ≪ the 1 ms system target — expected single-digit µs with hardware capture. Document actual numbers per method.
Gates D13 and the MCU choice.