STM32C5 - PB6 - electrical short with VDD

[mechatron]

Hi,

 

I am currently evaluating STM32C532KBU6, where I noticed the PB6 gets shorted (~20 ohms) with VDD of the MCU, as soon as the MCU is powered. When not powered, I am measuring ~4 Mega ohms between PB6 and VDD. I have not set any Option Bytes and did not flash any firmware as yet. The controller is consuming <10mA and no other abnormality observed. 

 

I checked the reference manual as well as the internal PU/PD architecture. So far, I could not find any answer.

The PB6 is not connected to any external circuitry and only to a test point. Any leads will help me here!

 

Thanks!

[TDK]

Likely, it’s in the bootloader and PB6 is in use as AN2606 shows. The bootloader will be entered when the flash memory is empty.

Measuring resistance on a pin in use leads to invalid results.

[mechatron]

Hi ​@TDK , Thanks for the response.

I do not see any mention about PU or PD for PB6 in bootloader mode. AN2606 still does not explain the issue I am facing.

For PA15, PA13 and PB4, I am able to confirm the PU to VDD, as mentioned in the datasheet. Similarly, for PA14, I can measure pull-down resistance to VSS. The values were ~32-37 kOhms. Hence, I will not outrightly say that measuring resistance on a pin in use will lead to invalid results. It gives me an indicator, if not the final result.

[TDK]

I do not see any mention about PU or PD for PB6 in bootloader mode. AN2606 still does not explain the issue I am facing.

It shows here that the pin is put into output mode. FDCAN_TX output pin is idle high, hence the “short” to VDD. What’s really happening is the MCU is sinking the injected current from the multimeter. The response from the circuit is not due to resistance.

[mƎALLEm]

Hello,

How did you measure that resistance on PB6 with MCU powered-on?

What about other IOs with the same measurement procedure?

[mechatron]

Hi ​@mƎALLEm , Thank you for your response.

I measured using a standard multimeter. I was able to confirm the PU/PD on PA15, PA13, PA14 and PB4, as mentioned in the datasheet.

 

[mƎALLEm]

I measured using a standard multimeter

Sorry I didn’t understand how did you measure the pin resistance with a standard multimeter! Was it in Ohmmeter mode while MCU is in powered on?

[rich.g.williams]

​@mechatron  “My question was about the electrical short with VDD (or in other words, active HIGH state) of the PB6. I care less about the resistance measurement and more about why such a state exists”

What is being explained is that in your configuration (are you using a NUCLEO PCB or probably in Bootstrap) then the bootstrap firmware runs at start up and configures PB6 pin as FDCAN2 in transmission mode (and used in alternate push-pull, no pull mode).

The ST32C532 after reset enters pattern 19 (which gives from AN2606):-

1. Boot0(bit) = 1 and BootSel(bit) = 0

2. Boot0(bit) = 0 and BootSel(bit) = 0, and user flash empty

3. BootSel(bit) = 1 and Boot pin = 1

so if your ST32C532 flash is empty it has to, must, enter bootstrap 

Now if it enters bootstrap it will configure PA6 as an output pin pulling to VDD.

Don’t try to measure resistance with power on - measure pin voltage instead and it will likely be VDD.

If you want to prove this, power on the ST32C532 and hold it in reset - you should then find that all pins are floating (are inputs) while in reset.

Be careful not to short any pins - any shorts to voltages over 4 Volts is likely to destroy most 3.3V MCU’s

Hope that helps, please let us know if this helps.

 

[mechatron]

Hi ​@rich.g.williams , I tried a similar approach, to confirm whether the voltage levels hold during the MCU reset state. I pulled the NRST pin low, which inturn made the PB6 logic LOW as well. I need to verify this across other relevant pins as well, but your explanation could very well be the answer I am looking for. Thank you!

[rich.g.williams]

Yes well done! Almost certainly is your explanation, there are a few configs as output done by the bootloader:-