We have a custom `STM32WL33CCV6` (`VFQFPN32`) board that should reach about `4 µA` in Deepstop, similar to the NUCLEO-WL33, but instead it drops only to roughly `450–500 µA` and appears to restart on Deepstop entry. `NRST` does not glitch low, and the behavior is unchanged with only `3.3V` and `GND` connected. We also tested direct `HAL_PWR_EnterDEEPSTOPMode()` with wake sources disabled, and it still appears to reboot rather than stay in Deepstop.
We reviewed ST errata `ES0612` and applied the documented `VFQFPN32` workaround from section `2.2.9` for nonbonded GPIOs causing increased Run-mode current by forcing the listed nonbonded pins to pulldown at startup. That workaround had no effect on the Deepstop issue. The custom board’s `VFBSD / VLXSD / VDDSD / VDDRF` schematic and BOM appear close to the ST reference, so we are asking whether there are known WL33 low-power / SMPS / Deepstop pitfalls beyond errata `2.2.9` that could explain a persistent `~500 µA` floor or an apparent reset/restart on Deepstop entry instead of the expected `~4 µA`.
It almost seems like deepstop is resetting the pins to their reset value or something??
