Looking for ways to protect the Pinecil V2 against bad quality supplies

[CarlosGS]

My Pinecil V2 broke after using a new power supply with higher power (140W).

• The issue is mentioned in the wiki, which identifies the cause is the PD chip sometimes just stops working: https://wiki.pine64.org/wiki/Pinecil:_Test,_Repair,_Issues#Not_Powering_Up_Anymore
• The linked Reddit thread, some have been broken with as low as 23V: https://www.reddit.com/r/PINE64official/comments/q8d08a/pinecil_does_not_power_on/
• Also this video cut mentions two irons which PD chip failed with a new PSU: https://www.youtube.com/watch?v=VULQXBKzniQ&t=195s

At first I thought the cause was that the provided voltage goes up to 28V #2003, but Ralim kindly pointed out that electrically it was already taken into consideration.
Apparently in the chat there have been more cases attributed to interference from a lower grade power supply.

Checking the schematic brings up a 25V rated capacitor tied to VBUS, but from #1390 it seems the cap was already upgraded to a higher rating.

New hypothesis:
From the schematic a possible candidate are the I2C lines that go into the USB cable without any TVS or capacitors. This is the main difference I can see from the reference implementation suggested by the PD chip manufacturer.

• But those I2C lines are also connected to the main microcontroller and to the accelerometer, so it is odd that only the PD chip dies.
• And I want to understand why is it necessary to desolder the PD chip to allow powering up from the DC jack, whether it shorts out other pins that prevent boot, etc.

If anyone has more insights on this please share and any input will be very helpful :)

[CarlosGS]

Ok, in my case it seems the PD chip was not the cause after all! Instead, the main CPU got fried and now has VCC_3V3 shorted to ground.

Here is the 3V3 regulator (U5) getting quite warm:

And the main microcontroller internally shorted to ground:

From the Pinecil V2 schematic (https://files.pine64.org/doc/Pinecil/Pinecil_schematic_v2.0_20220608.pdf) we see there are a lot of pins that directly connect to the USB port.

While there are ESD protection diodes in place, in order to protect the microcontroller we would need to add clamping diodes that keep GPIOs always below 3.3V even when bad power supplies peak above that.

But we would first need to know exactly what happened when the CPU got fried (what voltages we had, was it a spike or sustained?). I am not going to risk plugging the other irons, but let me know if you have more suggestions to test this.

The moral of the story: Do not plug the iron to random power supplies, stick to the one that works :)