Input protection

[Andrew Mowry]

I have a case where a maximum of 1.4V will be applied to an GPIO pin immediately before the MCU powers up (the voltage enables a boost converter, which then powers up the NRF52840). I am wondering if a series silicon diode would be adequate to protect the GPIO during this time? Perhaps combined with a series resistor? I believe technically a series diode would limit current to the sub-uA level rather than limiting voltage. Is that okay, or do I strictly need to limit the voltage to 0.3V? 

[ketiljo]

You can use a 10k series resistor here. This will limit the current in the ESD protection diodes and clamp the voltage to VDD + 0.3 V.

[Andrew Mowry]

Great, thank you!

[hmolesworth]

You may find a reset is affected where the pin provides phantom power to the MCU and reset circuit via the schottky clamp diode if power is removed from the MCU after being applied. At low VCC this is tiny, few 10's nA, and so the 10k won't help. No worries if MCU power is not removed and restored by the boost converter being truned on, then off, then sometime later back on. I have experienced this.

I would suggest a test to illustrate this if you have time. The workaround is often to use a buffer/level shifter powered by the MCU VCC.

[Andrew Mowry]

Thank you-- is that the same issue that you mentioned here:?

 Protection diodes on NRF52840: clamping current 

I'm not sure I totally understand the implications of this, but that other thread you make it sound bad :). Is the issue basically that the MCU won't be able to restart if the power is momentarily interrupted for some reason?

[hmolesworth]

Quite so; although of course it might work sometimes then occasionally get locked up. An other alternative is to use a crowbar or a boost converter which guarantees output discharge when disabled.