STSPIN32G4 and PC13-PC14-PC15 for LED Indicators

[MLange]

Good morning,

In regards to the current limit placed on the PC13-PC15 pins of the STM32G431VBx3 that is used as the basis of the STSPIN32G4 (and whose datasheet is referred to multiple times in the STSPIN's), the '431 datasheet has the usual low-power domain caveat of "Since the switch only sinks a limited amount of current (3 mA) [...] These GPIOs must not be used as current sources (for instance to drive an LED)" (DS12589 Rev 6, pg. 60). The "Must Not" language seems rather definitive, and yet both of the demonstration boards go ahead and use at least one of the pins for an LED.

The EVLSPIN32G4-ACT schematic (as partially shown below) uses PC13 and PC14 with LEDs. The EVSPIN32G4 board implements the same circuit on PC13 but not PC14.

The yellow LED is listed in both BOMs as Wurth 150060YS75000, with 20mA test current and 2V typical forward voltage. The green LED is listed as Wurth 150060GS75000 at 20mA and 3.2V (typ). With the 320Ω worth of resistors in place, the yellow LED should be at a (rule-bending) 4.1mA, and with a 330Ω resistor, the Green LED should be at 0.3mA.

Does the STSPIN32G4 not have the same restriction on current sourcing through pins PC13-15, or is it just a matter of being very careful with current and/or implementing a PWM scheme so that the average dissipation across the power switch is lowered?

[STOne-32 · Accepted solution]

Dear @MLange ,

Welcome in our STCommunity and thank you for spotting attention to that warning . Indeed PC13, PC14, PC15 pins / GPIO are special as put in backup domain for LSE - low speed clock to keep running on VBAT mode and so these pins are powered by a special internal switch that is limited to sustain 3mA in worst case conditions. A particular care should be applied to PWM to not exceed 3mA sunk on the green and yellow LEDs on top of the serial resistances 330 and 320 Ohm . We will inform our colleagues internally for more visibility on the warning at documentation of our STSPIN package .

thanks again .

STOne-32