Hello,
I built two PCBs, one with the STM32WL33KCV6TR (QFN32 version) and one with the STM32WL33CCV6TR (QFN48) version. Both schematics are identical (48MHz + 32768Hz crystal connected, SMPS used, decoupling capacitors on power pads, otherwise nothing is connected to the chips). On both versions I flashed the PWR_DEEPSTOP_RTC example. On the *CC* the original version from the example library, on the *KC* I used cubeMX to change the .isc file to the *KC* version. Afterwards I used a diff tool to make sure both project folders are identical (apart from the KC/CC difference).
My expectation would be that both PCBs draw the same amount of current.
However during "active" time in the software example the CC current draw is 4.3mA vs 3.0mA of the KC version.
During deepstop the CC version draws 1.3mA while the KC version draws the expected 1.2µA.
These measurements have been done at 3.3V.
Here are more measurements.
| | active | deepstop |
| STM32WL33KCV6TR @ 3.3V | 4.3mA | 1.3mA |
| STM32WL33CCV6TR @ 3.3V | 3.0mA | 1.3µA |
| STM32WL33KCV6TR @ 1.8V | 4.3mA | 740µA |
| STM32WL33CCV6TR @ 1.8V | 3.7mA | 1.2µA |
All tests have been done with cubeIDE 1.8.1 on MacOS with FW 1.1.0.
Especially during deepstop the power consumption of the KC variant is far too high.
Is there any register setting I need to change for the KC variant to reduce the power consumption or anything else that needs to be different between these two versions?
Thanks!
