For the protection diodes to work for large overvoltage conditions you need to limit the current with a series resistance. There is usually an absolute maximum input current specification in the datasheet.You have less to worry about when it is an input. If the input is high impedance, these can tolerate a lot of overvoltage because the input diodes clamp the signal.
Many microcontrollers allow it, so it probably varies by manufacturer. I have even seen zero-crossing detectors straight from 120V into the analog inputs through a high impedance.These diodes are never supposed to pass any current. In fact I have seen Microchip's document mentioned that current through these protection diodes are out of spec. of the PIC. They don't even put a max. current value for them in the datasheet.
At least this is my understanding. Higher input voltages can be dealt with using clamp diodes and current limiting resistors.
That note warns against passing *sustained* currents through the diodes. ESDI have located the document that warns against doing so for PICs by Microchip. Funny enough some of their application notes are doing just that.
But the document is quite recent so maybe it doesn't apply retrospectively.
Thread starter | Similar threads | Forum | Replies | Date |
---|---|---|---|---|
S | 2 ESP32's GPIOs to same header pins isolation | General Electronics Chat | 10 | |
Modeling 8266 GPIOs in LTSpice? | PCB Layout , EDA & Simulations | 0 | ||
STM32 Nucleo GPIOs, 5V Tolerant? | Microcontrollers | 1 | ||
PhotoGate control with MCU GPIOs. | Microcontrollers | 6 | ||
Wake up ESP8266 with multiple GPIOs | Microcontrollers | 7 |