What you could firstly do is measure the resistance between 3.3V and GND with the board off!!!
There are twos solutions :
- Short circut, 0 Ohms, you're voltage regulator is dead closed, unless you really know what to do and how (changing the power regulator) the raspberry is dead.
- No shorting, In this configuration the regulator died open, you could try to power it with a good 3.3V power supply. (next part)
If and only if there are no short :
In order to do so find a 3.3V power suppply which can provide easily more than 1 or 2 Amp (linear regulator even good will becomes hot very hot !). A laboratory power supply may be a great idea to set the voltage and the maximum current to see what is happening !
Connect ground to ground and 3.3V from your power supply to the 3.3V pin of the GPIO. And try to boot this way…
For sceptics : Indeed there are multiple voltage on the board
- 5V externally provided
- 3.3V internally generated
- 1.2V internally generated
- VDD CORE
What could happen is that manipulating the GPIO you fried the 3.3V output, but it's likely nothing happened to the 1.2V which is a separate voltage regulator in the same chip but another regulator.
So in short one of the regulator may be fried (the 3.3V) the other one generating 1.2V is likely to be OK.
And as the 1.2V regulator takes his voltage from the 5V having the 3.3V fried open is likely to have no impact on the 1.2V.
The chip is a XR77004