Protect the MCU without making hardware changes.

[Wadeva]

Hello everyone,

I’m facing an issue described in this link: My STM32 device is overheating and has malfunctioned.

My STM32G070CBT MCU has been damaged, specifically from the PC6 pin, which wasn’t in use. Unfortunately, the pin is connected to another device that is shorted to the chassis, and this is a finished product that has already been released. Here’s what I cannot do:

• I cannot modify the PCB design.
• I cannot add protection diodes.
• I cannot cut the cable.

Given these constraints, is there a way to protect the MCU without hardware changes? For instance, could setting PC6 to pull-down mode or output mode help mitigate the issue?

Thanks!

[Wadeva]

(moved from previous thread)

Now, my challenge is different. Since the design is already completed and the PCBs have been manufactured, it’s unlikely that we can make changes to the layout at this stage. Is there a way I can protect this unprotected pin to safeguard the MCU from ESD without redesigning the PCB? Alternatively, is there a way to reduce the risk of the device being damaged?

[Andrew Neil]

---

@Wadeva wrote:

Now, my challenge is different. 

---

Please start a new thread for a new question.

Provide a link here, so that people can find it.

You can provide a link in the new thread back to here for reference.

You're going to need to give more details about this "unprotected" pin - show schematic, etc.

[LCE]

> unprotected pin that is connected to the chassis ground of the device

>> Since the design is already completed and the PCBs have been manufactured,
>> it’s unlikely that we can make changes to the layout at this stage.

Bad idea with that BIG design error! I wonder how that one got through any design checks.

You could still test the device's ESD behavior.
If the chassis is perfectly grounded and chassis potential is the same as STM32's local GND, it might still work...

But as it is already having problems without ESD... or have you solved the heat problem by changing GPIO mode of that pin?

[AScha.3]

On existing boards: your only chance is a simple but good, low impedance connection to board/cpu GND.

Maybe a solder joint to ground plane , direct at the point, where the "signal line" comes to the board.

[Andrew Neil]

---

@Wadeva wrote:

PC6 pin, which wasn’t in use. Unfortunately, the pin is connected to another device that is shorted to the chassis, 

---

If PC6 is configured as an input (which is the default state), that shouldn't matter.

So make sure that PC6 is configured as an input - or just left unconfigured.

[Wadeva]

Hello Andrew,

This pin indeed unconfigured. 

[Andrew Neil]

Then pulling it to ground would cause no problem.

[Wadeva]

Hello, I’m a bit confused. If I configure PC6 with an internal pull-down, will that resolve the issue? Or is leaving the pin unconfigured sufficient for protection?

To clarify: the product has already been distributed, and I can only make changes via an OTA update. If there’s a way to improve safety by adjusting the configuration of PC6, what would you recommend? Or, as you mentioned, is it better to leave the pin unconfigured?

This leads to another question: could a different pin be causing the problem? I will check PCB design a little bit further then.

Thanks for the help. 

[Wadeva]

Hello LCE,

We conducted tests in the past, but unfortunately, they were done incorrectly, so we didn’t notice the issue until malfunctioning problems began. While we don’t have time to perform another ESD test right now, we plan to do so in the future.

Although we haven’t changed any pin configurations and no further devices have failed, it feels like a ticking time bomb—we don’t know when the problem might occur again.