Out on a wind farm near Lubbock this week, five soil sensors with LoRa radios are dropping about 30% battery overnight only when temps fall below 5°C; bench shows 180 µA sleep current, but field logs show unexpected wake every 12 minutes. Has anyone traced similar cold-triggered wake behavior to GPS backup rails or inverter EMI, or should I tear into the firmware timer next?
I chased a similar “every 12 minutes” wake to the IWDG still running in STOP at low temps; one cold-night test with the watchdog disabled confirmed it… If you’re on STM32, log RCC_CSR reset flags and try a run with IWDG off — if it disappears, bump the timeout or pet it from the RTC wake path; what MCU are you on?
And > the timeout or pet it from the RTC wake path; what MCU are you on? I’ve had cold stall the 32.768 kHz on STM32WL and the RTC quietly fall back to LSI, turning a 15‑min alarm into about 12; enable LSE CSS and bump LSEDRV, then log RCC_CSR.LSECSSF on each wake — does it flip on those nights?
@mia3577’s clock note aside, I once traced a 12‑min wake in cold to the SX1276 DIO1 floating as the internal pull‑ups drifted; the cure was to ‘map all DIOs to 0 before sleep’ and add a 100 kΩ pulldown on DIO1 — could you temporarily hard‑pull the radio DIOs and see if the wakes stop?